Progress in Brain Research Volume 161 Neurotrauma: New Insights into Pathology and Treatment

PROGRESS IN BRAIN RESEARCH VOLUME 161 NEUROTRAUMA: NEW INSIGHTS INTO PATHOLOGY AND TREATMENT Other volumes in PROGRES...

Author: John T. Weber | Andrew I.R. Maas

93 downloads 683 Views 12MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

PROGRESS IN BRAIN RESEARCH VOLUME 161 NEUROTRAUMA: NEW INSIGHTS INTO PATHOLOGY AND TREATMENT

Other volumes in PROGRESS IN BRAIN RESEARCH 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: Transmission Revisited, by L.F. Agnati, K. Fuxe, C. Nicholson and E. Sykova´ (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 0-44450332-3. Volume 127: Neural Transplantation II. Novel Cell Therapies for CNS Disorders, by S.B. Dunnett and A. Bjo¨rklund (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. Sandku¨hler, B. Bromm and G.F. Gebhart (Eds.) – 2000, ISBN 0-44450509-1. 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 Lo´pez 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 0-444-50586-5. Volume 135: Do Seizures Damage the Brain, by A. Pitka¨nen 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. Volume 139: Vasopressin and Oxytocin: From Genes to Clinical Applications, by D. Poulain, S. Oliet and D. Theodosis (Eds.) – 2002, ISBN 0-444-50982-8. Volume 140: The Brain’s Eye, by J. Hyo¨na¨, D.P. Munoz, W. Heide and R. Radach (Eds.) – 2002, ISBN 0-444-51097-4. Volume 141: Gonadotropin-Releasing Hormone: Molecules and Receptors, by I.S. Parhar (Ed.) – 2002, ISBN 0-444-50979-8. Volume 142: Neural Control of Space Coding, and Action Production, by C. Prablanc, D. Pe´lisson and Y. Rossetti (Eds.) – 2003, ISBN 0-444-509771. Volume 143: Brain Mechanisms for the Integration of Posture and Movement, by S. Mori, D.G. Stuart and M. Wiesendanger (Eds.) – 2004, ISBN 0-444-513892. Volume 144: The Roots of Visual Awareness, by C.A. Heywood, A.D. Milner and C. Blakemore (Eds.) – 2004, ISBN 0-444-50978-X. Volume 145: Acetylcholine in the Cerebral Cortex, by L. Descarries, K. Krnjevic´ and M. Steriade (Eds.) – 2004, ISBN 0-444-51125-3. Volume 146: NGF and Related Molecules in Health and Disease, by L. Aloe and L. Calza` (Eds.) – 2004, ISBN 0-444-51472-4. Volume 147: Development, Dynamics and Pathology of Neuronal Networks: From Molecules to Functional Circuits, by J. Van Pelt, M. Kamermans, C.N. Levelt, A. Van Ooyen, G.J.A. Ramakers and P.R. Roelfsema (Eds.) – 2005, ISBN 0-444-51663-8. Volume 148: Creating Coordination in the Cerebellum, by C.I. De Zeeuw and F. Cicirata (Eds.) – 2005, ISBN 0-444-51754-5. Volume 149: Cortical Function: A View from the Thalamus, by V.A. Casagrande, R.W. Guillery and S.M. Sherman (Eds.) – 2005, ISBN 0-444-51679-4. Volume 150: The Boundaries of Consciousness: Neurobiology and Neuropathology, by Steven Laureys (Ed.) – 2005, ISBN 0-444-51851-7. Volume 151: Neuroanatomy of the Oculomotor System, by J.A. Bu¨ttner-Ennever (Ed.) – 2006, ISBN 0-444-51696-4. Volume 152: Autonomic Dysfunction after Spinal Cord Injury, by L.C. Weaver and C. Polosa (Eds.) – 2006, ISBN 0-444-51925-4. Volume 153: Hypothalamic Integration of Energy Metabolism, by A. Kalsbeek, E. Fliers, M.A. Hofman, D.F. Swaab, E.J.W. Van Someren and R. M. Buijs (Eds.) – 2006, ISBN 978-0-444-52261-0. Volume 154: Visual Perception, Part 1, Fundamentals of Vision: Low and Mid-Level Processes in Perception, by S. Martinez-Conde, S.L. Macknik, L.M. Martinez, J.M. Alonso and P.U. Tse (Eds.) – 2006, ISBN 978-0-444-52966-4. Volume 155: Visual Perception, Part 2, Fundamentals of Awareness, Multi-Sensory Integration and High-Order Perception, by S. Martinez-Conde, S.L. Macknik, L.M. Martinez, J.M. Alonso and P.U. Tse (Eds.) – 2006, ISBN 978-0-444-51927-6. Volume 156: Understanding Emotions, by S. Anders, G. Ende, M. Junghofer, J. Kissler and D. Wildgruber (Eds.) – 2006, ISBN 978-0444-52182-8. Volume 157: Reprogramming of the Brain, by A.R. Møller (Ed.) – 2006, ISBN 978-0-444-51602-2. Volume 158: Functional Genomics and Proteomics in the Clinical Neurosciences, by S.E. Hemby and S. Bahn (Eds.) – 2006, ISBN 9780-444-51853-8. Volume 159: Event-Related Dynamics of Brain Oscillations, by C. Neuper and W. Klimesch (Eds.) – 2006, ISBN 978-0-444-52183-5 Volume 160: GABA and the Basal Ganglia: From Molecules to Systems, by J.M. Tepper, E.D. Abercrombie and J.P. Bolam (Eds.) – 2007, ISBN 978-0-444-52184-2

PROGRESS IN BRAIN RESEARCH

VOLUME 161

NEUROTRAUMA: NEW INSIGHTS INTO PATHOLOGY AND TREATMENT EDITED BY JOHN T. WEBER Department of Neuroscience, Erasmus Medical Centre, Rotterdam, The Netherlands and School of Pharmacy, Memorial University of Newfoundland, St. John’s, NL, Canada

ANDREW I.R. MAAS Department of Neurosurgery, Erasmus Medical Centre, Rotterdam, The Netherlands

AMSTERDAM – BOSTON – HEIDELBERG – LONDON – NEW YORK – OXFORD PARIS – SAN DIEGO – SAN FRANCISCO – SINGAPORE – SYDNEY – TOKYO 2007

Elsevier Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands Linacre House, Jordan Hill, Oxford OX2 8DP, UK

First edition 2007 Copyright r 2007 Elsevier B.V. All rights reserved No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; e-mail: [email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://www.elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material 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 drug dosages should be made Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing in Publication Data Neurotrauma: new insights into pathology and treatment. – (Progress in brain research; v. 161) 1. Nervous system – Wounds and injuries I. Weber, John II. Maas, Andrew I.R. 617.40 8044 ISBN-13: 9780444530172

ISBN: 978-0-444-53017-2 (this volume) ISSN: 0079-6123 (Series)

For information on all Elsevier publications visit our website at books.elsevier.com

Printed and bound in The Netherlands 07 08 09 10 11

10 9 8 7 6 5 4 3 2 1

List of Contributors

J. Ai, St. Michael’s Hospital, Trauma Research, Toronto, ON M5B 1W8, Canada G.J. Amelink, Rudolf Magnus Institute of Neuroscience, Department of Neurosurgery, University Medical Center Utrecht, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands A.J. Baker, St. Michael’s Hospital, Trauma Research, Toronto, ON M5B 1W8, Canada D.C. Baptiste, Division of Cellular and Molecular Biology, Toronto Western Research Institute and Krembil Neuroscience Centre, Toronto Western Hospital, 12th Floor Room 407, McLaughlin Pavilion, 399 Bathurst Street, Toronto, ON M5 T 2S8, Canada M.D. Baumann, Department of Chemical Engineering and Applied Chemistry, University of Toronto, Terrence Donnelly Centre for Cellular and Biomolecular Research, 164 College Street, Toronto, ON M5S 3G9, Canada S.J. Bernstein, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadelphia, PA 19104-6392, USA N. Biasca, Clinic of Orthopaedic, Sports Medicine and Traumatology, Department of Surgery, Spital Oberengadin, CH-7503 Samedan/St. Moritz, Switzerland O. Bloch, Department of Neurological Surgery, University of California, San Francisco, CA, USA and Brain and Spinal Injury Center, University of California, 1001 Potrero Avenue, Room 101, San Francisco, CA 94110, USA H.M. Bramlett, The Miami Project to Cure Paralysis, Department of Neurological Surgery, University of Miami Miller School of Medicine, 1095 NW 14th Terrace, Miami, FL 33136, USA M. Buchfelder, Department of Neurosurgery, Friedrich-Alexander-University, Erlangen-Nuremberg, Schwabachanlage 6, D-91054 Erlangen, Germany A. Buki, Department of Neurosurgery, Medical Faculty, Pe´cs University, Re´t St. 2, Pe´cs, H-7624, Hungary P. Bukovics, Department of Neurosurgery, University Medical School, Pe´cs University, Re´t St. 2, Pe´cs, H-7624, Hungary M.R. Bullock, Department of Neurosurgery, Virginia Commonwealth University Medical Center, Richmond, VA, USA A.S. Cohen, Department of Pediatrics, University of Pennsylvania School of Medicine and Division of Neurology, Children’s Hospital of Philadelphia, Philadelphia, PA 19104-4318, USA D.K. Cullen, Neural Injury Biomechanics and Repair Laboratory, Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, 313 Ferst Drive, Atlanta, GA 30332-0535, USA E. Czeiter, Department of Neurosurgery, University Medical School, Pe´cs University, Re´t St. 2, Pe´cs, H-7624, Hungary W.D. Dietrich, The Miami Project to Cure Paralysis, Department of Neurological Surgery, University of Miami Miller School of Medicine, 1095 NW 14th Terrace, Miami, FL 33136, USA T. Doczi, Department of Neurosurgery, University Medical School, Pe´cs University, Re´t St. 2, Pe´cs, H-7624, Hungary J.J. Donkin, Discipline of Pathology, Level 3, Medical School North, The University of Adelaide, Adelaide, SA 5005, Australia v

vi

A.-C. Duhaime, Pediatric Neurosurgery, Children’s Hospital at Dartmouth, Dartmouth Hitchcock Medical Center, Lebanon, NH 03756, USA S. Durham, Pediatric Neurosurgery, Children’s Hospital at Dartmouth, Dartmouth Hitchcock Medical Center, Lebanon, NH 03756, USA O. Farkas, Department of Anatomy and Neurobiology, Medical College of Virginia Campus, Virginia Commonwealth University, 1101 E. Marshall St., 12-048 P.O. Box 980709, Richmond, VA 23298, USA M.G. Fehlings, Division of Cellular and Molecular Biology, Toronto Western Research Institute and Krembil Neuroscience Centre, Toronto Western Research Institute, 4th ﬂoor West Wing Room 449, 399 Bathurst Street, Toronto, ON M5 T 2S8, Canada C.L. Floyd, Department of Physical Medicine and Rehabilitation, Center for Glial Biology in Medicine, 547 Spain Rehabilitation Center, University of Alabama at Birmingham, Birmingham, AL 35249, USA B.F. Fuller, Department of Psychiatry, Center for Neuroproteomics and Biomarkers Research at the McKnight Brain Institute of the University of Florida, PO Box 100256, Gainesville, FL 32610, USA D.M. Geddes-Klein, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadelphia, PA 19104-6392, USA P.B. Goforth, Department of Pharmacology and Toxicology, VCU School of Medicine, Virginia Commonwealth University, Richmond, VA, USA M.S. Grady, Department of Neurosurgery, University of Pennsylvania School of Medicine, Philadelphia, PA, USA B.C. Hains, Department of Neurology and Center for Neuroscience and Regeneration Research, LCI-707, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06510, USA and Rehabilitation Research Center, VA Connecticut Healthcare System, West Haven, CT 06516, USA I.K. Haitsma, Department of Neurosurgery, Erasmus Medical Center, ’s-Gravendijkwal 230, 3015 CE Rotterdam, The Netherlands I. Hassan, Discipline of Pathology, Level 3, Medical School North, The University of Adelaide, Adelaide, SA 5005, Australia R.L. Hayes, Department of Neuroscience, Center for Traumatic Brain Injury Studies at the McKnight Brain Institute of the University of Florida, PO Box 100244, Gainesville, FL 32610, USA F. Hesse, Department of Neurosurgery, Friedrich-Alexander-University, Erlangen-Nuremberg, Schwabachanlage 6, D-91054 Erlangen, Germany M. Horowitz, Laboratory of Environmental Physiology, Faculty of Dental Medicine, Hebrew University of Jerusalem, Jerusalem 91120, Israel P. Jendelova, Institute of Experimental Medicine ASCR, Videnska 1083, 14220 Prague 4, Czech Republic C. Kang, Department of Chemical Engineering and Applied Chemistry, University of Toronto, Terrence Donnelly Centre for Cellular and Biomolecular Research, 164 College Street, Toronto, ON M5S 3G9, Canada Y. Katayama, Department of Neurological Surgery, Nihon University School of Medicine, 30-1 Oyaguchi Kamimachi, Itabashi-ku, Tokyo 173-8610, Japan T. Kawamata, Department of Neurological Surgery, Nihon University School of Medicine, 30-1 Oyaguchi Kamimachi, Itabashi-ku, Tokyo 173-8610, Japan A. Kleindienst, Department of Neurosurgery, Friedrich-Alexander-University, Erlangen-Nuremberg, Schwabachanlage 6, D-91054 Erlangen, Germany F.H. Kobeissy, Department of Psychiatry, Center for Neuroproteomics and Biomarkers Research at the McKnight Brain Institute of the University of Florida, PO Box 100256, Gainesville, FL 32610, USA J.D. Kocsis, Yale University School of Medicine, Neuroscience Research Center (127A), VA Connecticut Health Care System, West Haven, CT 06516, USA E.J.O. Kompanje, Department of Intensive Care, Erasmus MC University Medical Center Rotterdam, ’s-Gravendijkwal 230, P.O. Box 2040, 3015 CE Rotterdam, The Netherlands

vii

E. Kovesdi, Department of Neurosurgery, University Medical School, Pe´cs University, Re´t St. 2, Pe´cs, H-7624, Hungary K.L. Lankford, Rehabilitation Research Center, Veterans Affairs Connecticut Healthcare System, West Haven, CT 06516, USA M.C. LaPlaca, Neural Injury Biomechanics and Repair Laboratory, Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, 313 Ferst Drive, Atlanta, GA 30332-0535, USA B. Li, Research Institute of Surgery, Daping Hospital, Third Military Medical University, Chongqing, P.R. of China M.C. Liu, Department of Neuroscience, Center for Traumatic Brain Injury Studies at the McKnight Brain Institute of the University of Florida, PO Box 100244, Gainesville, FL 32610, USA N. Luka´cˇova´, Institute of Neurobiology, Slovak Academy of Sciences, Sˇolte´sovej 4, 040 01 Kosˇ ice, Slovak Republic B.G. Lyeth, Department of Neurological Surgery, University of California, 1515 Newton Court, One Shields Avenue, Davis, CA 95816, USA A.I.R. Maas, Department of Neurosurgery, Erasmus Medical Center, ’s-Gravendijkwal 230, 3015 CE Rotterdam, The Netherlands G.T. Manley, Department of Neurosurgery, University of California, 1001 Potrero Avenue, Room 101, San Francisco, CA 94110, USA J. Marsˇ ala, Institute of Neurobiology, Slovak Academy of Sciences, Sˇolte´sovej 4, 040 01 Kosˇ ice, Slovak Republic W.L. Maxwell, Institute of Biomedical and Life Sciences, University of Glasgow, Glasgow G12 8QQ, UK D.F. Meaney, Departments of Bioengineering and Neurosurgery, University of Pennsylvania, 3320 Smith Walk, Philadelphia, PA 19104-6392, USA M. Mesﬁn, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadelphia, PA 19104-6392, USA W.J. Miller, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadelphia, PA 19104-6392, USA J.P. Muizelaar, Department of Neurosurgery, University of California at Davis Medical Center, 4860 Y Street, Suite 3740, Sacramento, CA 95817, USA M.W. Oli, Banyan Biomarkers Inc., 12085 Research Drive, Alachua, FL 32615, USA J. Orenda´cˇova´, Institute of Neurobiology, Slovak Academy of Sciences, Sˇolte´sovej 4, 040 01 Kosˇ ice, Slovak Republic A.K. Ottens, Department of Psychiatry, Center for Neuroproteomics and Biomarkers Research at the McKnight Brain Institute of the University of Florida, PO Box 100256, Gainesville, FL 32610, USA J. Pal, Neurobiology Research Group of the Hungarian Academy of Sciences, University Medical School, Pe´cs University, Re´t St. 2, Pe´cs, H-7624, Hungary E. Park, St. Michael’s Hospital, Trauma Research, Toronto, ON M5B 1W8, Canada B.J. Pﬁster, Department of Biomedical Engineering, New Jersey Institute of Technology, Newark, NJ, USA N. Plesnila, Laboratory of Experimental Neurosurgery, Department of Neurosurgery and Institute for Surgical Research, University of Munich Medical Center, Grosshadern, Marchioninistr 15, 81377 Munich, Germany P. Poon, Department of Chemical Engineering and Applied Chemistry, University of Toronto, Donnelly Center for Cellular and Biomolecular Research, Toronto, ON M5S 3E1, Canada J.T. Povlishock, Department of Anatomy and Neurobiology, Medical College of Virginia Campus, Virginia Commonwealth University, 1101 E. Marshall St., 12-048 Sanger Hall, P.O. Box 980709, Richmond, VA 23298, USA

viii

G.R. Prado, Neural Injury Biomechanics and Repair Laboratory, Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, 313 Ferst Drive, Atlanta, GA 30332-0535, USA C. Radtke, Department of Plastic, Hand and Reconstructive Surgery, Hannover Medical School, Hannover, Germany D. Reglodi, Department of Anatomy, University Medical School, Pe´cs University, Szigeti u. 12, Pe´cs, H-7624, Hungary P. Reilly, Department of Neurosurgery, Royal Adelaide Hospital, Adelaide, SA 5000, Australia M. Sasaki, Department of Neurology and Center for Neuroscience and Regeneration Research, Yale University School of Medicine, New Haven, CT 06510, USA L.S. Satin, Department of Pharmacology and Toxicology, VCU School of Medicine, Virginia Commonwealth University, Richmond, VA, USA E. Schwarzbach, Department of Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, PA, USA N.A. Shein, Department of Pharmacology, School of Pharmacy, Hebrew University of Jerusalem, Jerusalem 91120, Israel E. Shohami, Department of Pharmacology, School of Pharmacy, Hebrew University of Jerusalem, Jerusalem, 91120, Israel M.S. Shoichet, Department of Chemical Engineering and Applied Chemistry, University of Toronto, Terrence Donnelly Centre for Cellular and Biomolecular Research, 160 College Street, Room 514, Toronto, ON M5S 3E1, Canada C.M. Simon, Neural Injury Biomechanics and Repair Laboratory, Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, 313 Ferst Drive, Atlanta, GA 30332-0535, USA P. Singh, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadelphia, PA 19104-6392, USA F.J.A. Slieker, Department of Neurosurgery, Erasmus MC University Medical Center Rotterdam, ’s-Gravendijkwal 230, P.O. Box 2040, 3000 CA Rotterdam, The Netherlands J.M. Spaethling, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadelphia, PA 19104-6392, USA D.G. Stein, Department of Emergency Medicine, Emory University School of Medicine, Brain Research Laboratory, Suite 5100, 1365B Clifton Road N.E., Atlanta, GA 30322, USA N. Stocchetti, Ospedale Policlinico IRCCS, Milan University, Milan, Italy E. Sykova, Institute of Experimental Medicine ASCR, EU centre of Excellence, Videnska 1083, 14220 Prague 4, Czech Republic D. Szellar, Department of Neurosurgery, University Medical School, Pe´cs University, Re´t St. 2, Pe´cs, H-7624, Hungary A. Tamas, Department of Anatomy, University Medical School, Pe´cs University, Szigeti u. 12, Pe´cs, H-7624, Hungary C.H. Tator, Krembil Neuroscience Centre, Toronto Western Research Institute and Department of Surgery, University of Toronto, 399 Bathurst Street, Toronto, ON M5 T 2S8, Canada E. Thornton, Discipline of Pathology, Level 3, Medical School North, The University of Adelaide, Adelaide, SA 5005, Australia R.J. Turner, Discipline of Pathology, Level 3, Medical School North, The University of Adelaide, Adelaide, SA 5005, Australia C. Van Den Heuvel, Discipline of Pathology, Level 3, Medical School North, The University of Adelaide, Adelaide, SA 5005, Australia

ix

I. Vanicky´, Institute of Neurobiology, Slovak Academy of Sciences, Sˇolte´sovej 4, 040 01 Kosˇ ice, Slovak Republic B.H. Verweij, Rudolf Magnus Institute of Neuroscience, Department of Neurosurgery, University Medical Center Utrech, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands R. Vink, Discipline of Pathology, Level 3, Medical School North, The University of Adelaide, Adelaide, SA 5005, Australia C.R. Von Reyn, Department of Bioengineering, University of Pennsylvania, 3320 Smith Walk, Philadelphia, PA 19104-6392, USA K.K.W. Wang, Department of Psychiatry, Center for Neuroproteomics and Biomarkers Research at the McKnight Brain Institute of the University of Florida, PO Box 100256, Gainesville, FL 32610, USA S.G. Waxman, Department of Neurology and Center for Neuroscience and Regeneration Research, LCI707, Yale School of Medicine, 333 Cedar Street, New Haven, CT 06510, USA and Rehabilitation Research Center, VA Connecticut Healthcare System, West Haven, CT 06516, USA J.T. Weber, Department of Neuroscience, Erasmus Medical Centre, Rotterdam, The Netherlands and School of Pharmacy, Memorial University of Newfoundland, Health Sciences Centre, 300 Prince Philip Drive, St. John’s, NL A1B 3V6, Canada X. Yao, Department of Neurological Surgery, University of California, San Francisco, CA, USA and Brain and Spinal Injury Center, 1001 Potrero Avenue, Room 101, San Francisco, CA 94110, USA Z. Zador, Department of Neurological Surgery, University of California, San Francisco, CA, USA and Brain and Spinal Injury Center, 1001 Potrero Avenue, Room 101, San Francisco, CA 94110, USA

This page intentionally left blank

Preface

This current volume of Progress in Brain Research, entitled ‘‘Neurotrauma: New insights into pathology and treatment’’ is a compilation of chapters on major topics which were discussed at the 8th International Neurotrauma Symposium held in Rotterdam, The Netherlands in May of 2006. Neurotrauma, which encompasses both traumatic brain injury (TBI) and spinal cord injury (SCI), is the leading cause of death and disability in young adults, and the incidence in older patients is increasing. As such, neurotrauma is a ﬁeld in medicine with one of the highest unmet needs. Concentrated, focused and multidisciplinary efforts are required to combat this important disease. Therefore, one of the major goals of the symposium was to stimulate cross-talk between basic researchers and clinicians, as well as between individuals in the ﬁelds of TBI and SCI. We feel that TBI and SCI researchers have much to learn from one another in their approach to treating these disorders. In addition, researchers have an obligation to design and conduct clinically relevant research, just as clinicians have an obligation to learn about the latest approaches to treating neurotrauma. Exciting ﬁndings from basic research open opportunities for improving treatment results, and this volume represents a unique and comprehensive overview of the latest ﬁndings and insights on translational research in neurotrauma. Experts who delivered presentations at the symposium have submitted the majority of chapters in this volume. In general, these individuals were selected because the topics that they, and their research groups, are currently pursuing are pertinent for understanding the pathology and treatment of neurotrauma. We are grateful also to other experts, who were not in attendance in Rotterdam, for writing chapters in speciﬁc areas of neurotrauma, which we felt were needed to make this a more comprehensive volume. The result, we think, is a report on some of the most relevant and up-to-date topics in the neurotrauma ﬁeld. Also, we are proud that the author base is a true representation of the international neurotrauma society (INTS), with contributors from Western and Eastern Europe, The United States, Canada, Asia and Australia. The global problem of neurotrauma can best be served by a worldwide representation of neurotrauma specialists combining and exchanging expertise. We are extremely grateful to all of the participants, assistants and sponsors who made the 8th International Neurotrauma Symposium a resounding success, and are especially gracious to all of the authors who have taken the time to contribute to this timely volume. In addition, we would like to apologize to some individuals or groups who may feel excluded, which was primarily due to space limitations. This exclusion is by no means intentional, and we have no doubt that new and additional topics, which are not covered in the current volume, will soon be represented elsewhere on the international stage. John T. Weber Andrew I.R. Maas

xi

This page intentionally left blank

Contents

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

v

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

xi

Section I. Introduction 1.

The impact of neurotrauma on society: an international perspective P. Reilly (Adelaide, South Australia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3

Section II. Biomechanics of Injury 2.

3.

CNS injury biomechanics and experimental models M.C. LaPlaca, C.M. Simon, G.R. Prado and D.K. Cullen (Atlanta, GA, USA). . . .

13

Linking impact to cellular and molecular sequelae of CNS injury: modelling in vivo complexity with in vitro simplicity J.M. Spaethling, D.M. Geddes-Klein, W.J. Miller, C.R. von Reyn, P. Singh, M. Mesﬁn, S.J. Bernstein and D.F. Meaney (Philadelphia, PA, USA) . . . . . . . . . . . . . . . . . . .

27

Section III. Pathological Mechanisms of Injury 4.

5.

6.

7.

Cellular and subcellular change evoked by diffuse traumatic brain injury: a complex web of change extending far beyond focal damage O. Farkas and J.T. Povlishock (Richmond, VA, USA) . . . . . . . . . . . . . . . . . . . . .

43

Astroglia: Important mediators of traumatic brain injury C.L. Floyd and B.G. Lyeth (Davis, CA, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . .

61

Rescuing neurons and glia: is inhibition of apoptosis useful? E. Kovesdi, E. Czeiter, A. Tamas, D. Reglodi, D. Szellar, J. Pal, P. Bukovics, T. Doczi and A. Buki (Pe´cs, Hungary) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

81

Substance P in traumatic brain injury J.J. Donkin, R.J. Turner, I. Hassan and R. Vink (Adelaide, South Australia) . . . . .

97

xiii

xiv

8.

9.

10.

11.

12.

13.

Current concepts of cerebral oxygen transport and energy metabolism after severe traumatic brain injury B.H. Verweij, G.J. Amelink and J.P. Muizelaar (Utrecht, The Netherlands and Sacramento, CA, USA). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

111

Progressive damage after brain and spinal cord injury: pathomechanisms and treatment strategies H.M. Bramlett and W.D. Dietrich (Miami, FL, USA) . . . . . . . . . . . . . . . . . . . . . . 125 Injury-induced alterations in CNS electrophysiology A.S. Cohen, B. J. Pﬁster, E. Schwarzbach, M.S. Grady, P.B. Goforth and L.S. Satin (Philadelphia, PA, Newark, NJ and Richmond, VA, USA) . . . . . . . . . .

143

Traumatic injury of the spinal cord and nitric oxide J. Mars˘ ala, J. Orenda´cˇova´, N. Luka´cˇova´ and I. Vanicky´ (Kosˇ ice, Slovak Republic)

171

Aquaporins: role in cerebral edema and brain water balance Z. Zador, O. Bloch, X. Yao and G.T. Manley (San Francisco, CA, USA) . . . . . . .

185

Sodium channel expression and the molecular pathophysiology of pain after SCI B.C. Hains and S.G. Waxman (West Haven, CT, USA) . . . . . . . . . . . . . . . . . . . .

195

Section IV. Novel Aspects of Clinical Research in CNS Injury 14.

15.

16.

17.

Monitoring cerebral oxygenation in traumatic brain injury I.K. Haitsma and A.I.R. Maas (Rotterdam, The Netherlands) . . . . . . . . . . . . . . .

207

Update on the treatment of spinal cord injury D.C. Baptiste and M.G. Fehlings (Toronto, ON, Canada) . . . . . . . . . . . . . . . . . .

217

Cerebral contusion: a role for lesion progression T. Kawamata and Y. Katayama (Tokyo, Japan) . . . . . . . . . . . . . . . . . . . . . . . . .

235

Ethical implications of time frames in a randomized controlled trial in acute severe traumatic brain injury E.J.O. Kompanje, A.I.R. Maas, F.J.A. Slieker and N. Stocchetti (Rotterdam, The Netherlands and Milan, Italy). . . . . . . . . . . . . . . . . . . . . . . . . .

243

Section V. Emerging Topics in CNS Trauma 18.

19.

Experimental models of repetitive brain injuries J.T. Weber (Rotterdam, The Netherlands) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

253

Minor traumatic brain injury in sports: a review in order to prevent neurological sequelae N. Biasca and W.L. Maxwell (Samedan/St. Moritz, Switzerland and Glasgow, UK) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

263

xv

20.

21.

22.

23.

24.

25.

Traumatic brain injury in infants: the phenomenon of subdural hemorrhage with hemispheric hypodensity (‘‘Big Black Brain’’) A.-C. Duhaime and S. Durham (Lebanon, NH, USA) . . . . . . . . . . . . . . . . . . . . .

293

Traumatic brain injury and Alzheimer’s disease: a review C. Van Den Heuvel, E. Thornton and R. Vink (Adelaide, Australia). . . . . . . . . . .

303

The neurotrophic protein S100B: value as a marker of brain damage and possible therapeutic implications A. Kleindienst, F. Hesse, M.R. Bullock and M. Buchfelder (Erlangen-Nuremberg, Germany and Richmond, VA, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

317

Cerebellar injury: clinical relevance and potential in traumatic brain injury research E. Park, J. Ai and A.J. Baker (Toronto, ON, Canada) . . . . . . . . . . . . . . . . . . . . .

327

Sex differences in brain damage and recovery of function: experimental and clinical ﬁndings D.G. Stein (Atlanta, GA, USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

339

Heat acclimation: a unique model of physiologically mediated global preconditioning against traumatic brain injury N.A. Shein, M. Horowitz and E. Shohami (Jerusalem, Israel). . . . . . . . . . . . . . . .

353

Section VI. The Future of Neurotrauma: Developing Novel Treatment Strategies 26.

In vivo tracking of stem cells in brain and spinal cord injury E. Sykova and P. Jendelova (Prague, Czech Republic) . . . . . . . . . . . . . . . . . . . . .

367

27.

Intrathecal drug delivery strategy is safe and efﬁcacious for localized delivery to the spinal cord M.S. Shoichet, C.H. Tator, P. Poon, C. Kang and M.D. Baumann (Toronto, ON, Canada). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385

28.

Decompression craniectomy after traumatic brain injury: recent experimental results N. Plesnila (Munich, Germany) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

393

Novel neuroproteomic approaches to studying traumatic brain injury A.K. Ottens, F.H. Kobeissy, B.F. Fuller, M.C. Liu, M.W. Oli, R.L. Hayes and K.K.W. Wang (Gainesville and Alachua, FL, USA) . . . . . . . . . .

401

29.

30.

Remyelination of the injured spinal cord M. Sasaki, B. Li, K.L. Lankford, C. Radtke and J.D. Kocsis (New Haven, CT, USA, Chongqing, People’s Republic of China and Hannover, Germany) . . . . . . . . . . . . 419

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

435

This page intentionally left blank

SECTION I

Introduction

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 1

The impact of neurotrauma on society: an international perspective Peter Reilly Department of Neurosurgery, Royal Adelaide Hospital, Adelaide, South Australia 5000

Abstract: Neurotrauma, in many countries and particularly in the younger age group kills more people than AIDS or cancer but unlike these diseases the causes are known and it is preventable. The costs to communities in terms of suffering and economics are enormous. The common causes are road trafﬁc accidents, falls and violence. Neurotrauma affects particularly the developing world where it consumes already over stretched health resources. In the developed world steps to reduce the incidence of neurotrauma and to treat the victims have had some effect nevertheless it stills remains an endemic problem which does not receive the public awareness or the political support it deserves. For the victims there is general agreement on the principles of clinical management but often difﬁculties in applying early and effective care in countries with the greatest need because of shortage of facilities and expertise. To reduce the overall burden of neurotrauma demands actions which extend from the political to basic patient care. There have been remarkable advances in the understanding of acute brain and spinal cord injury and encouraging possibilities for effective neuroprotection, repair and regeneration but in the broader context prevention of neurotrauma is the urgent imperative. In this endeavour the neuroscientist has knowledge which informs and encourages policy makers to take the steps necessary to reduce injury. These steps require political will and community support for hard decisions which impact on the way people conduct their daily lives. The WHO predicts that unless there are changes in present policies and if there are no additional road safety countermeasures put in place, there will be a major increase in road trafﬁc fatalities over the next 20 years and beyond (World Health Organisation. (2004). www.who.int). Keywords: epidemiology; prevention; brain and spinal injury statistics ﬁrearms that are easily available; for poor industrial safety standards and for sports such as boxing which aim to inﬂict neurotrauma or other sports which regularly do so. Some accidents are unavoidable and unforeseeable. Much of neurotrauma is all too predictable and avoidable. In this introduction to the Proceedings of the 8th International Neurotrauma Symposium I wish to consider some global aspects of neurotrauma and the role of the International Neurotrauma Society (INTS) in the overall effort to reduce the impact of neurotrauma on societies.

Introduction It is a sad paradox that one of the greatest health problems of this age is largely avoidable — neurotrauma of the brain and spinal cord is caused by injury inﬂicted by misadventure, violence or carelessness. The principle causes of neurotrauma reﬂect choices by individuals and by society for fast, freely available and personal transport; for

Corresponding author. Tel.: +61-8-8222-5232;

Fax: +61-8-8222-5668; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61001-7

3

4

Neurotrauma as a global problem The global incidence of traumatic brain injury (TBI) is generally reported as 200 in 100,000 with a mortality of 20 per 100,000. Incidences reported from different countries range from 91 to 430 per 100,000 with mortalities of 9–89 per 100,000 (Fearnside and Simpson, 2005). Despite awareness of the causes and of the economic and human costs of neurotrauma the incidence remains depressingly high and in developing countries continues to rise. One of the difﬁculties in describing accurately the global impact of neurotrauma is the variable way in which statistics are recorded. Even in countries with well-developed data collection systems, information on the full impact of head and spinal cord injury is difﬁcult to obtain. Mortality ﬁgures are often based on A&E admissions or hospital separations and do not include pre hospital deaths which account for 50% of road trauma deaths. The incidence of spinal cord injury ranged from 14.5 to 57.8 cases per million in a review of reported studies; the low ﬁgure of 14.5 from Australia did not include deaths at the scene of trauma (Ackery et al., 2004). Death and serious injury can be measured with reasonable accuracy but the greater numbers of mild and moderate head injuries are less likely to be encompassed, yet they represent a major component of the total burden of neurotrauma. Changing practices in hospital admissions also affect comparisons between studies. The greater use of CT scanning has lead to fewer hospital admissions, increasing the proportion of severe head injury admissions and reducing the mild and moderate head injury admissions (Thurman and Guerrero, 1999). A true measure of incidence and prevalence of neurotrauma would require that all head and spinal cord injuries be included. In fact there is no fully comprehensive study of the incidence of TBI in a deﬁned population (Fearnside and Simpson, 2005). Similar problems affect spinal injury statistics (Ackery et al., 2004). Despite these difﬁculties and limitations accurate data are essential in order to develop prevention and management strategies.

These wide ranges no doubt reﬂect a combination of actual differences, varying deﬁnitions of injury, different populations of inclusion and sampling errors (Bruns and Hauser, 2003). A number of studies indicate important regional variations. A higher mortality for patients injured in rural areas compared with urban areas has been recorded in Australia and Taiwan (Chiu et al., 1995, 1997; Hillier et al., 1997). Injury and mortality rates are bimodal affecting particularly the under 25 and over 65 years. In the UK the TBI incidence of 10:100,000 represents 1% of all deaths but 20% of deaths in age range 5–45 years. As in other developed countries, TBI is the common cause of death in this age range.

The causes of neurotrauma The main causes of neurotrauma, transport accidents, falls and gunshot wounds, reﬂect societal behaviour. From the beginning motorised transport has inﬂicted injury to a degree which would undoubtedly have shocked its inventors. By the 1930s the motor car ‘‘had emerged as the most persistent killer in the western world’’ (Gilbert, 1997). It is estimated that by 1997 25 million people had died on the roads (Odero et al., 1997). In 2002 the global death rate from road trafﬁc accidents was 1.2 million and between 20 and 50 million people are estimated to be injured or disabled each year (World Health Organisation, 2000, 2004). Of road trauma deaths it is estimated that 1/3–1/2 are due to brain injury. Vehicular accidents are also the major cause of spinal cord injury worldwide (Ackery et al., 2004). The cost of road trafﬁc injuries to society is estimated as 2% of a country’s gross domestic product (World Health Organisation, 2004). Global ﬁgures encompass important regional variations and trends (Soderlund and Zwi, 1995) (Table 1). Ninety percent of road trafﬁc accident deaths occur in low and middle income countries and at nearly twice the population adjusted rate of high-income countries. Road trafﬁc fatalities per 100,000 population in the year 2000 ranged from 28.3 in the African region to 5.9 in Great Britain

5 Table 1. Mechanisms of TBI

Australia (Tate et al., 1998) India (Gururaj, 2002) China (Wang et al., 1986) USA (CDC, 1999)

RTA

Falls

Violence

Sport

40 45–60 32 49

21 20–30 22 26

8 10–20 1 17

25 15

Occupation

24

Note: Numbers are percentages of trauma victims in each of the respective categories. RTA, road trafﬁc accidents.

Fig. 1. Road trafﬁc fatality trends in three high-income countries (Australia, United Kingdom, United States of America). Sources: Transport Safety Bureau, Australia; Department of Transport, United Kingdom; Fatality Analysis Reporting System, United States of America (World Health Organisation, 2004).

(World Health Organisation, 2004). Furthermore TBI from road trafﬁc accidents has been falling in high-income countries for some years. In Australia there has been an average yearly decline in TBI of 5%, most due to an 8% decline in vehicle occupancy injuries (O’Connor and Cripps, 1999). Between 1968 and 1983 road trafﬁc accident mortality declined by more than 20% in Europe (Fig. 1) In contrast it increased by more than 150% in Asian countries and more than 200% in African countries (Soderlund, 1995) (Fig. 2). Trafﬁc type has important inﬂuences on injury patterns. In Beijing one third of trafﬁc deaths occur among bicyclists. In India pedestrians account for 15–35% of injuries, 2-wheeler occupants 20–40% and bicyclists 5–15% (Gururaj, 2002). Taiwan had a high rate of TBI from road trafﬁc accidents reported in 1991 to be 89 per 100,000

(Chiu et al., 1995). This was considered to be related to the rapidly increasing use of motor cycles, the most dangerous form of road transport in current use anywhere in the world, and an increasing component of trafﬁc in developing countries (World Health Organisation, 2004). Other causes of neurotrauma also show marked regional variation and changing trends. In the USA ﬁrearm injuries exceeded road trafﬁc accidents for the ﬁrst time in 1990 and this trend contrasted with a fall in road trafﬁc accident deaths (Sosin et al., 1995). Other trends in injury causation have been documented in several countries. A review of head injury mortality from 1987 to 2000 in Sweden showed a constant rate of injury over this period, however a fall in transport-related injury was balanced by an increase in falls below the age of 25 and over the age of 65 (Kleiven et al., 2003). In the UK

6

Fig. 2. Global and regional road fatality trends, 1987–1995. Data are displayed according to regional classiﬁcation of TRL Ltd., United Kingdom. The World Health Organisation, 2004 document from which this comes states that reproduced with permission of the authors namely, Jacobs G, Aeron-Thomas A, Astrop A. Estimating Global Road Fatalities. Crowthorne, Transport Research Laboratory, 2000 (TRL Report 445).

there were regional differences between the incidence of road trafﬁc accident versus falls and of alcohol-related accidents (Kay and Teasdale, 2001). Surveys from different countries also indicate variation in the causation of spinal cord injury between countries and over time. In most countries vehicular accidents are the leading cause of spinal cord injury in the younger age group but, as with head injury, this ﬁgure is falling while the incidence of injury due to falls in the older age group is increasing (Cripps, 2006). Falls and violence are more prevalent in developing countries and in the lower economic strata of developed countries (Ackery et al., 2004).

The community costs Most studies of TBI focus on severe injury. Ninety percent of patients presenting to A&E with head injury in the UK had minor, 5% moderate and 5% severe head injury. For every severe head injury there were 17 mild or moderate injuries (Kay and Teasdale, 2001). Most patients who suffer from minor TBI make uneventful recoveries but many suffer continuing, disabling symptoms. WHO

ﬁgures indicate a possible treatment rate of moderate head injury of 100–300 per 100,000 but the actual rate may be greater than 600 per 100,000 (Cassidy et al., 2004). Thornhill et al. (2000) reported that following mild head injury 51% had continuing symptoms and after moderate head injury 54%. Following mild head injury (GCS 13–15) 79% show persistent headache, 59% memory problems and 34% remained unemployed. Guerrero et al. (2000) found that at 1 year 15% still had symptoms. This represents an enormous and often under treated group of potential disabilities. A pathophysiological basis has been identiﬁed for continuing symptoms (Blumbergs et al., 1994; Bigler, 2001). The ensuing neuropsychological effects and the value of psychological therapy have been documented (Ponsford et al., 2000; Carroll et al., 2004; Cassidy et al., 2004; Ponsford, 2005).

Providing care Since the 1970s routine intracranial pressure recording, the CT scan and Intensive Care Units have focused treatment on severe head injury. The general principles of management of brain and

7

spinal injury may be agreed upon but there are major difﬁculties in delivering care to patients in developing countries through shortage of neurosurgeons, a lack of trained personal for primary care, particularly in rural areas, basic facilities and transport systems for trauma victims. These pressures are often compounded by the need to address many other health problems with limited resources. In such situations general surgeons may need to be trained in acute neurosurgery. In Australia distance imposes special needs and the general surgeons in Darwin, a city of 70,000 and 2000 km from a neurosurgical centre, have managed acute neurosurgery for many years with support by neurosurgeons and teleradiology (Treacy et al., 2005). To meet these needs protocols for on site management of neurotrauma by non neurosurgeons have been developed (Neurosurgical Society of Australasia Guidelines, 2000).

What can be done? Neurotrauma is a major public health challenge. Acknowledging that most TBI and spinal cord injury is avoidable clearly emphasises the need for strategies to identify causes and prevent injury. Motor vehicle accident rates have fallen in many countries in response to public awareness campaigns and legislation (McDermott et al., 1996). Each new legislated measure, for random alcohol testing, speed cameras, seat belts or helmets, tends to be followed by a modest fall in accident rates (Campbell, 1987; MacLennan et al., 2004). In response to the high number of motorcyclerelated head injuries in Taiwan a concerted effort was made to introduce motorcycle helmets from 1997. In the ﬁrst year motorcycle-related head injuries fell by 33%. There was a better than 90% compliance rate with helmet wearing which was mandated by law (Chiu et al., 2000a, b). Motorcycle helmets have now been introduced via legislation in many countries including Australia. In the US the repeal of helmet legislation in several states, provided a human experiment unlikely to be approved by an ethics committee. In Arkansas repeal of the state helmet law in 1997 led to a fall in helmet usage from 97 to 52% over

9 months. There was a marked increase in crash scene fatalities and in admissions of non helmeted crash survivors. It was concluded that helmet usage decreased fatalities by 20–40% (Bledsoe et al., 2002). Other studies have supported the beneﬁt of motor cycle helmets in reducing head and facial injury (Wagle et al., 1993; Liu et al., 2003). Bicycle helmets have been mandatory in Australia since 1990 and several studies have indicated their beneﬁt (Attewell et al., 2001; Thompson et al., 1999). There is great potential beneﬁt for helmets in developing countries where bicycle and increasingly motor bikes are a major cause of TBI. Societal attitudes are clearly key to any injury prevention campaign (Weinstein, 1987). Cars are undoubtedly safer now than they were decades ago. Standard safety ratings are accepted by most manufacturers, but they can be safer. It is often remarked that safety does not sell compared with appearance and performance. There appears to be an acceptance of the risk of neurotrauma even though there must be few families who do not have a victim of neurotrauma of some degree. Those at highest risk have the least perception of risk (Gilk et al., 1999).

Community support for research Funding levels are a barometer of the value a society puts on a particular ﬁeld of research. The Brain Injury Association of America estimates that TBI in the US costs the community $48.3 billion per year, $31.7 billion in hospitalisation costs and $16.6 billion from fatal TBI. Spinal cord injury has been estimated to cost the US $9.7 billion each year (Berkowitz, 1998). In 2006 the US National Institutes of Health, the major government funding body, allocated $86 million for TBI research and $87 million for spinal cord injury research. Whether these are appropriate amounts can be argued. In Australia the principle commonwealth funded body, the National Health and Medical Research Council provided $600,000 towards neurotrauma research in 2005–2006 from a total budget of $60 million, that is to say 1% of the research budget.

8

Researchers in neurotrauma in some states have been more successful in gaining substantial grants from the state government third party insurers. One of the more innovative funding initiatives occured in Western Australia for 5 years ﬁnes raised by speed cameras amounting to $500,000 per year were allocated to trauma research. In the US and Australia the funds directed towards neurotrauma research appear small compared with the societal costs and in comparison with the funds directed to other major diseases. Spinal research has proﬁted greatly from patient advocacy. A visit by Christopher Reeves to Sydney in 1999 lead to the NSW Premier’s Initiative research fund for spinal research. There are now moves to establish national platforms to coordinate research and develop funds for spinal injury research — important given the need to support highquality research in a relatively small population. The challenge Neurotrauma affects all communities, all age groups and walks of life. Despite understanding of the causes and being able to treat victims better it is an increasing global problem with important regional variations. There is a lack of adequate comprehensive data of the incidence of neurotrauma in any community, most of all in those communities which have the major burden of neurotrauma. Neurotrauma is a particular burden on the developing countries which have the least capacity to manage it. Reducing the incidence of neurotrauma can only occur by greater political action and the engagement of societies in order to see neurotrauma as avoidable. The role of the INTS and neuroscience The INTS comprises a worldwide body of experts in neurotrauma from clinical, non clinical and basic science backgrounds. Many are actively involved in acute neurotrauma care or working with victims of the delayed effects of traumatic brain or spinal cord injury or in injury prevention. The INTS aims to encourage research in all aspects of neurotrauma by

providing an international forum for basic, preclinical and clinical research. It aims to encourage young neuroscientists to pursue their interests and in this way to foster research around the globe. Scientists have a role in neurotrauma prevention and treatment in every level. Scientiﬁc evidence underpins effective political and social action. The neuroscientist can play a vital role in informing and stimulating policy makers towards action that will reduce the incidence of neurotrauma. The INTS plays a highly signiﬁcant role in encouraging the worldwide search for better treatments for victims of neurotrauma. It is important that the INTS liaises with other neurotrauma organisations such as the WFNS, National Neurotrauma Societies and regional consortia so that the international neuroscience community can use its wide experience and expertise to the best effect.

References Ackery, A., Tator, C. and Krassioukov, A. (2004) A global perspective on spinal cord injury epidemiology. J. Neurotrauma, 21: 1355–1370. Attewell, R.G., Glase, K. and McFadden, M. (2001) Bicycle helmet efﬁcacy: a meta-analysis. Accid. Anal. Prev., 33: 345–352. Berkowitz, M., O’Leary, P., Kruse, D. and Harvey, C. (1998) Spinal cord injury: an analysis of medical and social costs. Demos Medical Publishing Inc., New York. Bigler, E. (2001) The lesion(s) in traumatic brain injury: implications for clinical neuropsychology. Arch. Clin. Neuropsychol., 16: 95–131. Bledsoe, G.H., Schexnayder, S.M., Carey, M.J., Dobbins, W.N., Gibson, W.D., Hindman, J.W., Collins, T., Wallace, B.H., Cone, J.B. and Ferrer, T.J. (2002) The negative impact of the repeal of the Arkansas motorcycle helmet law. J. Trauma, 53: 1078–1087. Blumbergs, P.C., Scott, G. and Manavis, J. (1994) Staining of amyloid precursor protein to study axonal injury in mild head injury. Lancet, 344: 1055–1056. Bruns Jr., J. and Hauser, W.A. (2003) The epidemiology of traumatic brain injury: a review. Epilepsia, 44(Suppl 10): 2–10. Campbell, B.J. (1987) Safety belt injury reduction related to crash severity and front seated position. J. Trauma, 27: 733–739. Carroll, L.J., Cassidy, J.D. and Holm, L. (2004) Methodological issues and research recommendations for mild traumatic brain injury: the WHO Collaborating Centre Task Force on Mild Traumatic brain injury. J. Rehabil. Med., 43(Suppl): 113–135.

9 Cassidy, J.D., Carroll, L.J. and Peloso, P.M. (2004) Incidence, risk factors and prevention of mild traumatic brain injury: results of the WHO Collaborating Centre Task Force on Mild Traumatic brain injury. J. Rehabil. Med., 43(Suppl): 26–28. Center for Disease Control and Prevention (CDC). (1999) Traumatic brain injury in the United States. A Report to Congress Current Data on Traumatic Brain Injury and Mortality. Chiu, W.T., Ho, Y.S. and Lee, Y.S. (2000a) Sharp decline of injury mortality rate in a developing country. Am. J. Public Health, 90(5): 793–796. Chiu, W.T., Hung, C.C., Le, L.S., Lin, L.S., Shih, C.J. and LaPorte, R.E. (1997) Head injury in urban and rural populations in a developing country. J. Clin. Neurosci., 4: 469–472. Chiu, W.T., Hung, C.C. and Shih, C.J. (1995) Epidemiology of head injury in rural Taiwan: a four year survey. J. Clin. Neurosci., 2: 210–215. Chiu, W.T., Kuo, C.Y., Hung, C.C. and Chen, M. (2000b) The effect of the Taiwan motorcycle helmet use law on head injuries. Am. J. Public Health, 5(90): 793–796. Cripps, R.A. (2006) Spinal Cord Injury, Australia 2003–04. Australian Institute of Health and Welfare, Canberra. Fearnside, M.R. and Simpson, D.A. (2005) Epidemiology. In: Reilly P.L. and Bullock R.B. (Eds.), Head Injury: Pathophysiology and Management (2nd ed). Hodder Arnold, London, pp. 3–25. Gilbert, M. (1997) A History of the Twentieth Century, Vol. 1: 1900–1933. Harper Collins, London, 41p. Gilk, D.C., Kronenfeld, J.J., Jackson, K. and Zhang, W. (1999) Comparison of trafﬁc accident and chronic disease risk perceptions. Am. J. Health. Behav., 23: 198–209. Guerrero, J.L., Thurman, D.J. and Sniezek, J.E. (2000) Emergency department visits associated with traumatic brain injury: United States, 1995–1996. Brain Inj., 14: 181–186. Gururaj, G. (2002) Epidemiology of traumatic brain injuries: Indian scenario. Neurol. Res., 24: 24–28. Hillier, S., Hillier, J. and Metzer, J. (1997) Epidemiology of traumatic brain injury in South Australia. Brain Inj., 11: 649–659. Kay, A. and Teasdale, G. (2001) Head injury in the United Kingdom. World J. Surg., 25: 1210–1220. Kleiven, S., Peloso, P.M. and von Holst, H. (2003) The epidemiology of head injuries in Sweden from 1987 to 2000. Inj. Control Saf. Promot., 10: 173–180. Liu, B., Norton, R., Blows, S. and Lo, S.K. (2003) Helmets for preventing injury in motorcycle riders. The Cochrane Database of Systemic Reviews, Issue 4. MacLennan, P.A., McGwin, G., Metzger, J., Moran, S.G. and Rue III, L.W. (2004) Risk of injury for occupants of motor vehicle collisions from unbelted occupants. Inj. Prevent., 10: 363–367. McDermott, F.T., Cordner, S.M. and Tremayne, A.B. (1996) Evaluation of the medical management and preventability of

death in 137 road trafﬁc fatalities in Victoria, Australia: an overview. J. Trauma, 40: 520–535. Neurosurgical Society of Australasia Guidelines. (2000) The Management of Acute Neurotrauma in Rural and Remote Locations. A Set of Guidelines for the Care of Head and Spinal Injuries. The Neurosurgical Society of Australasia, Royal Australasian College of Surgeons, Melbourne. O’Connor, P.J. and Cripps, R.A. (1999) Traumatic Brain Injury (TBI) Surveillance Issues. AIHW National Injury Surveillance Unit, Flinders University Research Centre for Injury Studies, Adelaide. Odero, W., Garner, P. and Zwi, A. (1997) Road trafﬁc injuries in developing countries: a comprehensive review of epidemiological studies. Trop. Med. Int. Health, 2: 445–460. Ponsford, J. (2005) Rehabilitation interventions after mild head injury. Curr. Opin. Neurol., 18: 692–697. Ponsford, J., Wilmott, C., Rothwell, A., Cameron, P., Kelly, A.M., Nelms, R., Curran, C. and Ng, K.T. (2000) Factors inﬂuencing outcomes following mild traumatic brain injury in adults. J. Int. Neuropsychol. Soc., 6: 568–597. Soderlund, N. and Zwi, A.B. (1995) Trafﬁc related mortality in industrialized and less developed countries. World Health Organisation Bull., 73: 175–182. Sosin, D.M., Sniezek, J.E. and Waxweiler, R.J. (1995) Trends in death associated with traumatic brain injury, 1979 through 1992. JAMA, 273: 1778–1780. Tate, R.L., McDonald, S. and Lulham, J.M. (1998) Incidence of hospital-treated traumatic brain injury in an Australian community. Aust. N.Z. J. Public Health, 22: 419–423. Thompson, D.C., Rivara, F.P. and Thompson, R. (1999) Helmets for preventing head and facial injuries in bicyclists. The Cochrane Database of Systemic Reviews, Issue 4. Thornhill, S., Teasdale, G.M., Murray, G.D., McEwen, J., Roy, C.W. and Penny, K.I. (2000) Disability in young people and adults one year after head injury: prospective cohort study. BMJ, 320: 1631–1635. Thurman, D. and Guerrero, J. (1999) Trends in hospitalization associated with traumatic brain injury. JAMA, 282(10): 954–957. Treacy, P.J., Reilly, P. and Brophy, B.P. (2005) Emergency neurosurgery by general surgeons at a remote major hospital. ANZ J. Surg., 75: 852–857. Wagle, V.G., Perkins, C. and Vallera, A. (1993) Is helmet use beneﬁcial to motorcyclists? J. Trauma, 34: 120–122. Wang, C.C., Schoenberg, B.S., Li, S., Yang, Y., Cheng, X. and Bolis, L. (1986) Brain injury due to head trauma. Epidemiology in urban areas of the People’s Republic of China. Arch. Neurol., 43: 570–572. Weinstein, N.D. (1987) Unrealistic optimism about susceptibility to health problems: conclusions from a community-wide sample. J. Behav. Med., 10: 481–500. World Health Organisation. (2000) Injury: leading cause of the global burden of disease. www.who.int World Health Organisation. (2004) World report on road trafﬁc injury prevention. www.who.int

SECTION II

Biomechanics of Injury

This page intentionally left blank

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 2

CNS injury biomechanics and experimental models M.C. LaPlaca, C.M. Simon, G.R. Prado and D.K. Cullen Neural Injury Biomechanics and Repair Laboratory, Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, 313 Ferst Dr., Atlanta, GA 30332-0535, USA

Abstract: Traumatic brain injury (TBI) and traumatic spinal cord injury (SCI) are acquired when an external physical insult causes damage to the central nervous system (CNS). Functional disabilities resulting from CNS trauma are dependent upon the mode, severity, and anatomical location of the mechanical impact as well as the mechanical properties of the tissue. Although the biomechanical insult is the initiating factor in the pathophysiology of CNS trauma, the anatomical loading distribution and the resulting cellular responses are currently not well understood. For example, the primary response phase includes events such as increased membrane permeability to ions and other molecules, which may initiate complex signaling cascades that account for the prolonged damage and dysfunction. Correlation of insult parameters with cellular changes and subsequent deficits may lead to reﬁned tolerance criteria and facilitate the development of improved protective gear. In addition, advancements in the understanding of injury biomechanics are essential for the development and interpretation of experimental studies at both the in vitro and in vivo levels and may lead to the development of new treatment approaches by determining injury mechanisms across the temporal spectrum of the injury response. Here we discuss basic concepts relevant to the biomechanics of CNS trauma, injury models used to experimentally simulate TBI and SCI, and novel multilevel approaches for improving the current understanding of primary damage mechanisms. Keywords: traumatic brain injury; traumatic spiral cord injury; neurotrauma; biomechanics; membrane permeability; ﬁnite element analysis; injury tolerance criteria mechanical input that has exceeded structural limits of cells and tissue. Primary damage is characterized by nonspeciﬁc cell loss as well as sublethal injury, which activates a cascade of secondary responses leading to prolonged cell death, network dysfunction, and system level changes (Fig. 2). Although the mechanical impact is the initiating event in traumatic CNS injury, the relationship between biomechanical inputs and the downstream pathological effects are not well understood. Investigation of relevant loading parameters and the resulting cell and tissue responses in a variety of model systems is imperative for deciphering injury-induced pathophysiological mechanisms and developing experimental models that hold ﬁdelity to the human clinical situation.

Introduction Traumatic brain injury (TBI) and spinal cord injury (SCI) result in a range of deficits depending on the insult severity and the anatomical region(s) affected. In traumatic central nervous injury (CNS) injury, a mechanical impact (caused by motor vehicle accidents, gunshot wounds, blows to the head or spine, etc.) induces a mechanical response at the cell and tissue level that ultimately causes a pathophysiological injury response (as shown in Fig. 1). In the acute phase of injury, primary damage occurs as a direct result of a Corresponding author.

E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61002-9

13

14

Mechanical Input

Mechanical Response

Injury Response

Fig. 1. Steps in CNS trauma. Traumatic brain and spinal cord injuries result from mechanical loading to the tissue. Pathophysiological events are initiated by the mechanical tissue response to impact.

insult

secondary response

lifespan

primary response

Fig. 2. Temporal aspects of injury. Mechanical loading causes an acute primary phase followed by a prolonged secondary phase. The primary response is characterized by nonspeciﬁc cell loss, which initiates a cascade of complex secondary events such as inﬂammation, excitotoxicity, and neurodegeneration.

A notable application of the study of biomechanics in CNS trauma is the determination of accurate tissue tolerances. Tissue tolerances are deﬁned as the point at which structural and/or physiological failure occurs. An improved understanding of injury biomechanics and the resulting brain and spinal cord responses will ultimately facilitate the development of improved protective gear (e.g., helmets and seat belts). Determination of tolerance criteria requires information about the forces and deformations that lead to failure, but the mechanical parameters (i.e., magnitude and rate of force and deformation) are only partially understood. Tissue response and tolerance criteria for humans are largely based on cadaveric studies, but may not accurately represent the properties of living tissue. Basic cell and animal studies, in which a deﬁned mechanical insult can be applied to live cells in culture or in an intact animal, have an advantage for the determination of tissue tolerance and may lead to the reﬁnement of human tolerance criteria. These tolerance criteria must be model-independent and represent inherent system properties. We will discuss basic biomechanical concepts as they relate to traumatic brain and spinal cord injuries and present experimental models that have been developed and characterized in an attempt to mimic the forces and deformations occurring in human CNS trauma. The mechanisms by which the mechanical response to a traumatic insult leads

to dysfunction are complex, yet can be simpliﬁed using controlled cellular injury models that account for deformation magnitude and rate. Biomechanically relevant in vitro TBI models, used in combination with animal studies and computer simulations, may lead to improved cellular and tissue injury tolerance criteria as well as a more complete understanding of the relationship between the biomechanical input and pathophysiological changes. This multilevel approach will be discussed with respect to selection of experimental models, development of mechanistically driven treatment strategies, and future research priorities.

Basic biomechanics Biomechanics is the study of forces and physical responses in stationary (static) and moving (dynamic) biological systems. A system (in the case of traumatic CNS injury — the brain or spinal cord) reacts in a speciﬁc way when a force, or load, is placed on it. These external loads may result in initial damage or lead to delayed damage. The point at which loading causes tissue damage is the threshold (or the tolerance) of the system and is dependent on the type and duration of the load. The basic terms, or descriptors, that biomechanicians use to describe applied loads are force and stress and the resulting responses are deformations and strains.

15

Force is deﬁned as the action of one body (a physical entity in the system, such as a windshield) on another (as a result of an impact), which will cause acceleration of the second body (e.g., the head) unless acted upon by an equal and opposite action counteracting the effect of the ﬁrst body. The unit is a Newton (N); 1 N is the force that will give 1 kg an acceleration of 1 m/s2 (English unit is pound-force, lbf). When forces are generated in tissue, deformation may ensue depending on the material properties and the nature of the force itself. Deformation is deﬁned as the change in shape of a body undergoing a force. A rigid body, for example, would experience extremely small deformations, while biological tissue (usually referred to as deformable or nonrigid) can often undergo substantially large deformations. Stress is another term frequently used in biomechanical analysis and refers to the distribution of force relative to the area on which it acts. Normal stresses (designated by the Greek letter sigma (s)) act perpendicular to the surface, while shear stresses (designated by the Greek letter tau (t)) act tangential to the surface. The unit is the Pascal (Pa); 1 Pa ¼ 1 N/m2. A given force acting on a small surface produces greater stress than the same force acting over a larger surface. In other words, the amount of mechanical stress created by a force is dependent on the size of the area over which the force is applied. The resulting strain that occurs relates the deformed state of the body to the undeformed state and is unitless. Extensional strain is the change in length divided by the original length (designated by the Greek letter epsilon) (e ¼ Dl/lo) and can be further classiﬁed as being in tension (positive strain) or compression (negative strain). Extensional strain results from stresses generated from linear (or translational) loads. Shear strain, often resulting from rotational loads, is also the change in length divided by the original length (designated by the Greek letter gamma) (g ¼ Dl/lo). Brain tissue is thought to be more sensitive to shear strain than extensional strain (Holbourn, 1943). Therefore loading that involves rotation of the head has been thought to result in more severe injuries, although this assumption has recently been questioned (King et al., 2003). The relationships between stress

and strain are referred to as constitutive relationships and the resulting equations are used to deﬁne behavior of the tissue (or the mechanical response). The basic mechanics terms deﬁned above are valuable in describing the conditions that lead to injuries, although several factors surround biomechanical analysis of damage prediction. Mechanical conditions can be referred to as the insult parameters and the result as the injury (Fig. 1). Two broad categories of insults can be deﬁned as static and dynamic loading, with dynamic loading being the most common. The mechanical response to insult is the tissue deformation or strain and will initiate the ensuing pathological events. The insult parameters and the mechanical response will dictate the types of injury (focal and/or diffuse). We will consider the categories of insults, the mechanical response to traumatic insult, the types of injuries produced, as well as two overlapping response phases (primary and secondary) in light of the biomechanical ﬁdelity of experimental models used to simulate these conditions.

Traumatic mechanical insults Loads are described as direct (e.g., physical contact between the head and another object) or indirect (e.g., as the result of motion of the head). In indirect loading, acceleration of the second body (e.g., the head) can act analogously to applied forces. Loads can be translational (linear), rotational, or angular (a combination of translational and rotational). The type of force and the direction, or plane, of loading, will also affect the resulting mechanical response in the tissue. The extent and severity of deformation increases with increasing force, and this relationship is nonlinear. In other words, the increase in tissue damage may be greater than the proportional increase in force. Static loading is a very slowly applied direct load. Usually there are no deficits until there is substantial tissue deformation. These loading conditions are relatively rare and often occur in human entrapment situations (e.g., earthquakes). Dynamic loading, on the other hand, can occur quite

16

rapidly (under 1 s, often o50 ms) and is the most common cause of TBI and SCI. Dynamic loading can further be broken down into impact loading (direct loading where an impact occurs with an object hitting the head or the head hitting an object) or impulsive loading (indirect loading where no contact occurs). Impact loading can be either focal or diffuse, depending on the magnitude of the force and area of impact. Although pure impact would involve contact with no head movement, impact loading is usually a combination of contact forces — from the impact itself — and inertial forces — from the motion of the head and the brain within the skull. It is important to consider the size, mass, and hardness of the impacting object as well as the surface area and velocity at which contact occurs. For example, impact with smaller objects (i.e., o2 in. in diameter) results in high local stress concentrations and therefore is associated with a greater risk for more local and severe damage and is more likely to result in tissue penetration. Impulsive loading is due to inertial forces alone and leads to diffuse brain injuries. Models of impulsive loading include angular acceleration of the head, yet many of the models utilized for impact loading are designed to deliver a rapid bulk insult that has inertial components. Ultimately, the response is dictated by the mechanical response of the tissue or cells. Loads, in particular rotational inputs to the brain, however, do not linearly scale between humans and animal, as the mass of the brain is much smaller. In fact, to produce an equivalent rotational load in a rodent brain as in a human brain the angular acceleration would need to be approximately two orders of magnitude higher. This anatomical complexity introduces difﬁculty in directly linking pathological consequences to the biomechanical input. In addition to these constraints in animal modeling, the regional stresses and strains have yet to be well characterized. Future investigations to determine the relationships between biomechanical parameters and cellular responses will require a detailed spatial characterization of local cellular stresses and strains in animal models of CNS trauma.

Mechanical response to traumatic insult A traumatic insult to brain or spinal cord will lead to a mechanical response of the tissue that is dependent on the mode, severity, and anatomical location of the impact as well as the mechanical properties of the tissue. The mechanical properties of a tissue vary from individual to individual, as well as with age and previous injuries or disease (Prange and Margulies, 2002). In addition, cellular orientation and tissue composition varies among anatomical regions of the brain and spinal cord, creating nonuniform (or heterogeneous) mechanical properties that directly affect structural and functional tolerances as well as the load distribution throughout the tissue upon mechanical loading. Because of the properties of soft tissues, like brain and spinal cord, both the rate and the duration of the insult will also inﬂuence the response. Loads that are applied quickly may incur more damage due to the material properties of CNS tissue. When loads are applied at a high rate, the tissue cannot absorb (or reduce) the force fast enough and can fail both structurally and functionally. In contrast, slowly applied loads give the tissue ‘‘time’’ to reduce the force and generally result in less damage. For short durations of force, much of the effects of the force are reduced. As the duration of force increases, less reduction occurs and therefore less force is needed to produce tissue deformation. These behaviors are deﬁned by a mechanical property termed viscoelasticity.

Types of traumatic CNS injury Focal injuries result from direct loading and can often occur without widespread, or diffuse, damage. Focal injuries are typically induced when an object penetrates the skull or vertebral column as a result of a motor vehicle accident, gunshot wound, or a blow. As a result, macroscopically visible damage is typically visible at the site of impact, and the clinical symptoms are often very speciﬁc to the area that is directly injured. Focal injuries to the brain include epidural hematomas and skull fracture (with or without brain damage). When

17

there is osteal or dural compromise, this is often termed open head injury in the clinical setting. Contact loading can also result in coup (at the site of impact) and contra-coup (away from the site of impact) contusions to the brain, involving both cellular and vascular components. Focal injuries account for one-half of all severe head injuries, but two-third of all deaths in this group (Thurman and Guerrero, 1999; Adekoya et al., 2002). SCI is most commonly caused by fracture and dislocation of the spinal column, resulting in a focal injury. The mechanical impact causes displacement of bone fragments, intervertebral discs, or ligaments, resulting in transient compression or contusion of spinal cord tissue. Spinal cord is compressed at the site of impact that causes the surrounding tissue to lengthen in the longitudinal direction. Tissue near the center of the spinal cord is most vulnerable, suggesting that the mechanical loads are highest in this anatomical region. Large myelinated axons in the surrounding white matter are also highly susceptible to mechanical damage, due to stress concentrations at the nodes of Ranvier (Maxwell, 1996). As in TBI, the rate, magnitude, and duration of the biomechanical insult can dictate the injury response and may affect functional outcome. Slow stretching of the spinal cord results in very little tissue damage. In fact, increasing the length of the spinal cord up to twice the original length results in very little damage if the elongation is applied slowly (Shi and Whitebone, 2006). However, biomechanical inputs applied rapidly or for an extended duration (longer than 20–30 min) may surpass tissue thresholds and result in irreversible damage. Diffuse injuries are most often caused by inertial loading, which describes the motion of objects. The acceleration (velocity change divided by change in time) is an important parameter in determining tissue response. Higher accelerations correspond to higher forces (force equals mass times acceleration, Newton’s second law). This must be taken into account when establishing thresholds for tissue damage. Because of the complex head-neck dynamics, the brain can undergo high acceleration when subjected to an external load and therefore TBI often manifests as a diffuse

injury. When the acceleration is translational, injuries tend to be localized to a smaller area. Rotational acceleration, on the other hand, can lead to large strains deep within the brain, resulting in diffuse axonal injury (DAI) (Gennarelli et al., 1982). Most injuries seen clinically are a combination of translational and rotational accelerations (referred to as angular acceleration). Diffuse injuries are thought to occur as a result of not only the acceleration portion of loading, but also from the deceleration portion of the insult, creating very fast moving, uneven load distributions (Margulies et al., 1990). Diffuse strains can lead to differential movement of the skull relative to the brain, causing parasagittal bridging vein injury, as well as intracerebral hemorrhage. Diffuse injury to the brain tends to lead to widespread dysfunction, making these injuries the most prevalent cause of persistent neurological disability. Clinically, diffuse injury is often seen in closed head injury and arises most often from motor vehicle accidents.

Experimental modeling of traumatic CNS injury Experimental models of CNS injury have been invaluable in the investigation of pathological mechanisms and treatment strategies. However, due to the variable nature of clinical traumatic CNS injury (e.g., inconsistencies in the anatomical location of impact and the magnitude and duration of loading), experimental models must simplify the human condition in order to create a reproducible injury that can be utilized for controlled experimental testing. Although relevance to the clinic may be sacriﬁced, these simpliﬁcations allow the assessment of various outcome measures at the cellular, tissue, and organism level in response to deﬁned bulk loading parameters. In vivo animal models preserve much of the complexity associated with human traumatic CNS injury while allowing the investigator to experimentally manipulate certain parameters (e.g., treatment variables, time of sacriﬁce) that are not possible in humans. In the study of injury biomechanics, in vivo models provide a more complete representation of the human brain and spinal cord

18

because they more closely mimic the material properties and anatomical architecture. Therefore, the load distribution and structural failure in animal models are expected to be similar to human injury when clinically relevant biomechanical loading parameters are applied in a scale-appropriate manner. In vivo models commonly used in TBI and SCI research have been used to experimentally represent aspects of the biomechanics of CNS trauma. Direct loading has been mimicked using contusion, weight drop, ﬂuid percussion, or compression injuries. Contusion or weight drop involves brief, rapid loading of CNS tissue using a piston or a weight dropped from various heights (Dixon et al., 1991; Anderson and Stokes, 1992; Marmarou et al., 1994; Young, 2002; Scheff et al., 2003). These models are designed to deliver a rapid bulk insult that has both impact and inertial components. Compression injury is also used to experimentally replicate mechanical loads applied to spinal cords over long durations (e.g., due to abnormal, prolonged twisting of the spine during an automobile accident) (Rivlin and Tator, 1978; Dolan and Tator, 1979). Inertial loading experienced during TBI is modeled with ﬂuid percussion injury (Dixon et al., 1987; McIntosh et al., 1989; Thibault et al., 1992) (which has components of impact loads as well) and angular acceleration of the head (Gennarelli et al., 1982; Smith et al., 1997), which results in characteristic pathophysiological changes such as DAI. In vitro TBI models offer several advantages over whole animal models, including control over cellular components and real-time measurement of acute responses. Neural cultures and tissue explants have been subjected to compression, tension, or shear to experimentally mimic aspects of CNS trauma (see Morrison et al., 1998b). Models include deformable membranes that are stretched biaxially (Ellis et al., 1995; Cargill and Thibault, 1996; Geddes and Cargill, 2001; Morrison et al., 1998a) or uniaxially (Pﬁster et al., 2003; Lusardi et al., 2004) to transfer strain to attached cells, some with the capability of deforming neurites aligned longitudinally to the strain ﬁeld (Galbraith and Thibault, 1993; Smith et al., 1999). These in vitro models allow for isolation of speciﬁc

biomechanical parameters (e.g., deformation mode, rate, and magnitude), allowing for systematic assessment of cellular responses to deﬁned inputs. The recent development of a three-dimensional (3D) model in which neural cells are cultured in a hydrogel offers an intermediate degree of complexity, as bulk deformation of the culture results in heterogeneous strain ﬁelds at the cellular level depending on the orientation of the cell within the matrix (LaPlaca et al., 2005; Cullen and LaPlaca, 2006).

Tolerance criteria for CNS injury To date, several cellular tolerance criteria have been established to describe the contribution of both acceleration and pulse duration for a speciﬁc head injury (e.g., skull fracture, concussion), including the Wayne State Tolerance Criteria (Lissner et al., 1960), the Gadd Severity Index (Gadd, 1966), and the Head Injury Criterion (Versace, 1971). The basic overlying principle is that short pulses of high acceleration can produce injury, while lower accelerations require longer pulses to produce injury. These criteria have contributed to the development of a fundamental foundation; however, the tolerance stipulations have been based on cadaver or primate data in which the measure of injury did not consider damage at the cellular level. The efforts at the National Highway Trafﬁc & Safety Administration (NHTSA) have produced models of the head in the SIMon project. The predictive capability of SIMon and other computational models hinge on adoption of rational and experimentally veriﬁed thresholds for damage. Because different regions of CNS tissue have different cellular orientations and tissue composition, resulting in nonuniform (or heterogeneous) mechanical properties, structural and functional tolerances of the brain and spinal cord differ depending on the region affected. More complex and realistic computer models have been developed to provide more accurate information relevant to the biomechanics of injury (e.g., Zhang et al., 2004). Iterative veriﬁcation of these models is imperative to their successful application. Measurement of brain tissue strain

19

during a dynamic mechanical event is exceedingly difﬁcult in the intact animal or postmortem human subject (Hardy et al., 2003). Consequently, it has proven challenging to determine quantitative tolerances to be used for the damage measures embedded in computer models. Current efforts have utilized existing experimental data and scaling relationships to empirically derive thresholds to predict physiological outcome in animal experiments (Takhounts et al., 2003). Therefore, experimental models that enable the correlation of strain and acute injury could potentially determine detailed cellular tolerances.

Response phases of traumatic CNS injury Acute cellular response The initial damage that is a direct result of loading to the brain is deﬁned as the primary phase of injury. Biomechanicians study this phase in order to determine tissue tolerances to mechanical loading because the effects of the mechanical insult can be more easily isolated from biochemical events occurring in the secondary or more chronic phase. Our understanding of tolerances at the cellular level is vital to developing better safety equipment and understanding mechanotransduction in the pathological range. At the time of the insult there may be a varying amount of primary damage that results from the physical force itself. This includes compromised skin, bony fractures, tissue tearing, cellular rupture, and reorientation of the tissue components. If a deformation threshold is surpassed, these structural failures result and can severely compromise brain function. Due to the heterogeneity of CNS tissue, it is likely that loads and deformations experienced by cells in various anatomical regions are not consistent and cannot be accurately estimated by simplistic models assuming homogeneity. Certain anatomical regions may be subjected to more severe loading during impact because of differences in the material properties in that particular location (due to variations in cellular orientation, myelination, etc.). Anatomical regions experiencing larger strains would therefore be expected to be

more susceptible to primary damage caused by the mechanical insult itself. Although identiﬁcation of these regions would allow more accurate correlations between the mechanical input and pathophysiological responses, very little is currently known about local cellular strains in animal models of CNS trauma, mainly due to limitations in detection techniques. One approach for addressing these technical limitations is the development of more sensitive methods for the detection of mechanically induced damage. Although detection of structural failures can be relatively obvious in some instances (such as the presence of large focal lesions), more subtle damage may also be present and can provide a unique opportunity for assessment of local cellular strains after trauma. Visualization of the anatomical localization of this mechanical damage can provide a more sensitive measure of the load distribution throughout the tissue. We and others have investigated nonspeciﬁc plasma membrane damage as an indicator of mechanical damage in various models of TBI and SCI (Pettus et al., 1994; LaPlaca et al., 1997; Shi and Borgens, 2000; Geddes et al., 2003; Farkas et al., 2006). This type of cellular damage occurs as a direct result of mechanical loading, creating rips or tears in the plasma membrane at regions of high local strain. We have utilized Lucifer yellow as an indicator of acute biophysical membrane failure after TBI and SCI. Lucifer yellow is normally membraneimpermeable; therefore, cellular presence of this molecule can be used to detect plasma membrane compromise. In these experiments, Lucifer yellow was injected intrathecally 3 h prior to brain or spinal cord contusion, and animals were sacriﬁced 10 min after injury (a schematic of the injury devices are illustrated in Fig. 3). Histological evidence demonstrated heterogeneous uptake of the permeability marker in various anatomical locations (as shown in Fig. 4), indicating that the distribution of mechanical loading in CNS tissue is complex and not well understood. Although we have focused on acute membrane damage as an indicator of the load distribution throughout the brain and spinal cord, others have explored membrane compromise as an initiator of downstream pathological events. Cell membrane damage can

20 A. Infinite Horizons spinal cord contusion device

B. Cortical contusion impact device

Displacement sensor and motor controller

LVDT

Impactor tip Impactor tip

Load cell Stereotactic frame

Air tank

Spinal clamps

Control Box

500 µm

Fig. 3. In vivo contusion injury devices. Injury devices are used to experimentally deliver prescribed injury parameters to the exposed brain or spinal cord. For example, the Inﬁnite Horizons spinal cord contusion device (A) allows the user to select an impact force for injury, while the controlled cortical impact device (B) utilizes a pneumatic system to injure the brain at a deﬁned tissue displacement.

B. SCI

A. TBI

100 µm 500 µm

500 µm

Fig. 4. Acute cellular permeability following TBI and SCI. In the acute phase of traumatic injury, the plasma membrane becomes damaged due to local cellular strains that exceed structural thresholds. Lucifer yellow uptake in the injured brain (A) and spinal cord (B) demonstrates a heterogeneous distribution of membrane failure, suggesting that loading is not evenly distributed throughout the CNS parenchyma.

lead to abnormal ion movement across the membrane, resulting in pathophysiological changes such as conduction block, neuroﬁlament compaction, and impaired axonal transport (Pettus et al., 1994; Shi and Pryor, 2002). Thus, mechanical loading may directly result in pathophysiological changes. Experimental evidence has demonstrated that the extent of membrane compromise is dependent on the magnitude and rate of strain (LaPlaca et al., 1997; Geddes et al., 2003; Shi and Whitebone, 2006). In addition, others have suggested that the mode of injury may play a critical role in dictating the extent of mechanically induced cell membrane damage (Geddes-Klein et al., 2006). After TBI, membrane disruption has been shown to occur

after focal injury in a contusion model (Fig. 4) as well as diffuse loading after impact acceleration injury (Farkas et al., 2006), with patterns of marker uptake speciﬁc to the mode of impact. Because there is a correlation between injury severity and membrane compromise, permeability markers can therefore be used as an indicator of the extent of local cellular loading parameters. For example, experiments conducted in our laboratory have demonstrated more extensive permeability marker uptake in speciﬁc hippocampal regions after contusion injury, suggesting that local cellular loading is more severe in certain anatomical locations. These data may explain the preferential cell death seen in these regions in the subacute and chronic phases, as mechanical damage during the initial

21

3-D Neural Co-Cultures

piston

confocal microscope 3-D Cell Compression Device

3-D Cell Shearing Device

undeformed

Strain

Trapezoidal Input undeformed

0.50 0.25 10

deformed

20 30 Time (ms)

40 deformed

Fig. 5. In vitro injury devices for shear and compression injuries. Neural cells cultured in a 3-D conﬁguration were subjected to either shear or compression injury with a prescribed strain magnitude and rate. This experimental model provides control over bulk material deformation, while local strains may vary due to cellular orientation within the matrix.

impact in these regions may make the cells more susceptible to death and/or dysfunction during the secondary phase of injury. Although in vivo models can provide a more anatomically accurate representation of the structural and functional damage associated with human CNS injury, in vitro models allow for more thorough investigation of tissue tolerances because biomechanical insult parameters can be more precisely controlled and manipulated. In a recent study, the effects of both shear and compression modes of impact were investigated (Fig. 5). This is an example of how strain estimations derived from ﬁnite element analysis (FEA) can be applied to simpliﬁed culture environments to isolate components of the heterogeneous mechanical response (Fig. 6). Brieﬂy, mixed cultures consisting of neurons and astrocytes were plated in a 3-D matrix and subjected to either shear or compressive loading (0.50 strain at strain rates of 1, 10, or 30 s1). Both types of loading resulted in significant increases in membrane permeability in a strain rate dependent manner, with no differences in the density or percentage of permeabilized cells based on mode of deformation. However, the degree of

permeability marker uptake per permeabilized cell, potentially a gauge of local cellular strain/stress concentrations, was greater following shear deformation (Fig. 7). Interestingly, the density of dead cells was also significantly greater following shear deformation (5–7 fold increase) compared to compression (2-fold increase), suggesting that there is a correlation between the degree of membrane permeability and the extent of cell death. This study agrees with previous work demonstrating that shear deformation is the primary mode of tissue failure (Holbourn, 1943; Sahay et al., 1992). Although this study evaluated cellular responses based on different modes of bulk deformation, local cellular strains are heterogeneous, and may be a function of cell orientation with respect to the bulk strain ﬁeld (amongst other factors) (LaPlaca et al., 2005; Cullen and LaPlaca, 2006). We have demonstrated that neuronal response to loading depends on cell orientation, and hence local cellular strain, where maximal neurite loss occurred at shear-dominated strain regimes (LaPlaca et al., 2005). Ongoing in vitro studies are aimed at deﬁning the biomechanical parameters (deformation mode, rate, and magnitude) that lead to structural

22

C A

B

Finite element modeling of strain propagation following a focal insult (controlled cortical impact) in a rat. in vivo simulations

in vitro

A. Shear-dominated shear B. Compression-dominated compression C Unloaded region static Fig. 6. Finite element model (FEM) simulations and corresponding isolation of loading components in vitro. Traumatic loading to the brain results in the generation of complex, heterogeneous strain patterns at tissue and cell levels. Heterogeneity in the cellular response to traumatic loading may be due to several factors, including mode of deformation, cell population, and cell orientation. Neural cell tolerances to traumatic loading may therefore be elucidated based on these parameters.

23

Static Control

1 s-1 rate

10 s-1 rate

30 s-1 rate

Compression

Shear

Fig. 7. Acute cellular permeability increases in vitro depend on mode of bulk loading. Representative confocal reconstructions of calcein+ cells following static control conditions or mechanical loading (0.50 strain at 1, 10, or 30 s1 strain rate). Calcein, a normally cell-impermeant molecule, was added to the extracellular space prior to loading but becomes intracellularly sequestered following loading. Reconstructions from 50 mm thick z-stacks are shown here.

failure at the cellular level. Models of neural trauma that represent the related biomechanics and pathophysiology are important for the elucidation of cellular tolerances and the development of mechanistically driven intervention strategies.

during a traumatic insult. The response (whether cellular or whole organism) can better represent the clinical setting and therefore potential treatments can be evaluated in a more relevant setting. Future directions

Secondary response Primary damage initiates a cascade of secondary responses, leading to cell death, network dysfunction, and system-level changes (Fig. 8). While there is no absolute time when primary damage evolves into delayed effects, the secondary phase of injury can be deﬁned as any consequence of the primary insult. This may be in the acute (minutes to hours) period or in a more delayed fashion (days to months) and is dependent on the severity of the initial insult, as well as the health and age of the individual. There is a role for biomechanics in determining injury mechanisms in both the primary and secondary phases of the injury response by utilizing laboratory models that best mimic the forces/stresses and deformations/strains that occur

Determination of tolerance criteria for traumatic CNS injury will likely require a multilevel approach that incorporates both existing data and new knowledge from animal and cellular studies with more reﬁned computer modeling. Computer modeling in the form of FEA can provide estimates of the mechanical response of tissue to a large range of traumatic insult parameters, allowing parametric analysis. These models need to contain anatomical detail (for both human and animals) and corresponding mechanical property data to maintain the highest possible ﬁdelity. In addition, they should be able to simulate large, high rate deformations for both impact and inertial insult conditions. These estimated strain and stress patterns should be veriﬁed with in situ measurements when possible. This

24 MECHANICAL INSULT

MEMBRANE STRAIN

DEPOLARIZATION abnormal cell signaling

energy deficits

abnormal gene expression

MEMBRANE PERMEABILITY

NON-SPECIFIC ION FLUX (e.g., Ca2+, Na+, K+) loss of structural integrity

proteolytic degradation

PERSISTENT DYSFUNCTION OR DEATH Fig. 8. Simpliﬁed schematic of injury cascades initiated by mechanical loading. Mechanical injury may directly initiate downstream pathophysiological events, but the cause-and-effect relationship has not been thoroughly explored. Plasma membrane damage is hypothesized to trigger cell death or dysfunction through the inability to regulate ion ﬂux.

represents an experimental challenge and is worthy of consideration with new advances in nano- and micro-fabrication techniques, which permit electromechanical sensors to be instrumented. Animal models provide an opportunity to study the acute phase of injury and therefore can be correlated with estimated strain patterns in order to improve our current understanding of mechanotransduction. In addition, parallel long-term studies of delayed cell death and functional outcome can provide correlative data to acute responses. Furthermore, the cell response can be studied under very controlled conditions, and in vitro models of traumatic injury can be used to isolate elements of the mechanical response and reﬁne our understanding of cellular tolerances. Altogether, these data (with known temporal responses) can be applied to human models of traumatic injury (with unknown temporal responses) and tolerance criteria for humans extracted and predicted for speciﬁc scenarios. Conclusion Given the tremendous consequences that TBI and SCI have on society, it is important to better understand the biomechanical circumstances as they relate to the physiological and clinical implications. Biomechanics can play a role in improving preventative measures such as safety design in automobiles and sports equipment, as well as

highway and road safety by determining loading thresholds to the soft tissue of the brain and spinal cord. In addition to preventative strategies, biomechanics plays an important role in experimental modeling which, in turn, is vital to the development and application of mechanistically inspired pharmaceutical agents. By applying consistent and clinically relevant mechanical parameters (e.g., shear strain applied at high rates) to isolated neural cells or animal tissue, the response to mechanical disturbances can be assessed. The strain response is dependent on the tissue heterogeneity, namely the region-speciﬁc material properties and tissue orientation, therefore making elucidation of the cellular-level response to mechanical-trauma complex. The correlation of the injury response with strain enables detailed cellular tolerances that can be used to predict human injury criteria using FEA. In addition to cellular-level investigations, biomechanical models can be utilized at the animal level to achieve preclinical testing settings. Taken together, multilevel investigations can be used to eventually decrease the incidence of traumatic CNS injury and improve clinical outcomes. Acknowledgements We acknowledge Liying Zhang and King Yang from Wayne State University for the FEA computer simulations. Partial funding for the results

25

presented was provided by NSF (BES-0093830) and by Cooperative Agreement No. DTNH2201-H-07551 from the U.S. Department of Transportation — National Highway Trafﬁc Safety Administration to the University of Alabama at Birmingham, Southern Consortium for Injury Biomechanics.

References Adekoya, N., Thurman, D.J., White, D.D. and Webb, K.W. (2002) Surveillance for traumatic brain injury deaths — United States, 1989–1998. MMWR Surveill. Summ., 51: 1–14. Anderson, T.E. and Stokes, B.T. (1992) Experimental models for spinal cord injury research: physical and physiological considerations. J. Neurotrauma, 9(Suppl 1): S135–S142. Cargill, R.S. and Thibault, L.E. (1996) Acute alterations in [Ca2+]i in NG108-15 cells subjected to high strain rate deformation and chemical hypoxia: an in vitro model for neural trauma. J. Neurotrauma, 13: 395–407. Cullen, D.K. and LaPlaca, M.C. (2006) Neuronal response to high rate shear deformation depends on heterogeneity of the local strain ﬁeld. J. Neurotrauma, 23: 1304–1319. Dixon, C.E., Clifton, G.L., Lighthall, J.W., Yaghmai, A.A. and Hayes, R.L. (1991) A controlled cortical impact model of traumatic brain injury in the rat. J. Neurosci. Methods, 39: 253–262. Dixon, C.E., Lyeth, B.G., Povlishock, J.T., Findling, R.L., Hamm, R.J., Marmarou, A., Young, H.F. and Hayes, R.L. (1987) A ﬂuid percussion model of experimental brain injury in the rat. J. Neurosurg., 67: 110–119. Dolan, E.J. and Tator, C.H. (1979) A new method for testing the force of clips for aneurysms or experimental spinal cord compression. J. Neurosurg., 51: 229–233. Ellis, E.F., McKinney, J.S., Willoughby, K.A., Liang, S. and Povlishock, J.T. (1995) A new model for rapid stretch-induced injury of cells in culture: characterization of the model using astrocytes. J. Neurotrauma, 12: 325–339. Farkas, O., Lifshitz, J. and Povlishock, J.T. (2006) Mechanoporation induced by diffuse traumatic brain injury: an irreversible or reversible response to injury? J. Neurosci., 26: 3130–3140. Gadd, C.W. (1966) Use of a weighted impulse criteria for estimating injury hazard. In: 10th Stapp Car Crash Conference 1966, pp. 164–174. Galbraith, J. and Thibault, L.E. (1993) Mechanically induced depolarizations in the squid giant axon. J. Biomech. Eng., 115: 13–22. Geddes, D.M. and Cargill II., R.S. (2001) An in vitro model of neural trauma: device characterization and calcium response to mechanical stretch. J. Biomech. Eng., 123: 247–255. Geddes, D.M., Cargill II., R.S. and LaPlaca, M.C. (2003) Mechanical stretch to neurons results in a strain rate and

magnitude-dependent increase in plasma membrane permeability. J. Neurotrauma, 20: 1039–1049. Geddes-Klein, D.M., Schiffman, K.B. and Meaney, D.F. (2006) Mechanisms and consequences of neuronal stretch injury in vitro differ with the model of trauma. J. Neurotrauma, 23: 193–204. Gennarelli, T.A., Thibault, L.E., Adams, J.H., Graham, D.I., Thompson, C.J. and Marcincin, R.P. (1982) Diffuse axonal injury and traumatic coma in the primate. Ann. Neurol., 12: 564–574. Hardy, W.N., Foster, C., Mason, M., Yang, K.H., King, A.I. and Tashman, S. (2003) Investigation of head injury mechanisms using neutral density technology and high-speed biplanar X-ray. Stapp Car Crash J., 45: 337–368. Holbourn, A.H. (1943) Mechanics of head injuries. Lancet, 2: 438–441. King, A.I., Yang, K.H., Zhang, L., Hardy, W. and Viano, D. (2003) Is head injury caused by linear or angular acceleration? In: 2003 IRCOBI Conference. Lisbon, Portugal, pp. 1–12. LaPlaca, M.C., Cullen, D.K., McLoughlin, J.J. and Cargill II., R.S. (2005) High rate shear strain of three-dimensional neural cell cultures: a new in vitro traumatic brain injury model. J. Biomech., 38: 1093–1105. LaPlaca, M.C., Lee, V.M. and Thibault, L.E. (1997) An in vitro model of traumatic neuronal injury: loading rate-dependent changes in acute cytosolic calcium and lactate dehydrogenase release. J. Neurotrauma, 14: 355–368. Lissner, H.R., Lebow, M. and Evans, F.G. (1960) Experimental studies on the relation between acceleration and intracranial pressure changes in man. Surg. Gynecol. Obstet., 111: 329–338. Lusardi, T.A., Rangan, J., Sun, D., Smith, D.H. and Meaney, D.F. (2004) A device to study the initiation and propagation of calcium transients in cultured neurons after mechanical stretch. Ann. Biomed. Eng., 32: 1546–1558. Margulies, S.S., Thibault, L.E. and Gennarelli, T.A. (1990) Physical model simulations of brain injury in the primate. J. Biomech., 23: 823–836. Marmarou, A., Foda, M.A., van den Brink, W., Campbell, J., Kita, H. and Demetriadou, K. (1994) A new model of diffuse brain injury in rats. Part I: pathophysiology and biomechanics. J. Neurosurg., 80: 291–300. Maxwell, W.L. (1996) Histopathological changes at central nodes of Ranvier after stretch-injury. Microsc. Res. Tech., 34: 522–535. McIntosh, T.K., Vink, R., Noble, L., Yamakami, I., Fernyak, S., Soares, H. and Faden, A.L. (1989) Traumatic brain injury in the rat: characterization of a lateral ﬂuid percussion model. Neuroscience, 28: 233–244. Morrison III., B., Meaney, D.F. and McIntosh, T.K. (1998a) Mechanical characterization of an in vitro device designed to quantitatively injure living brain tissue. Ann. Biomed. Eng., 26: 381–390. Morrison III., B., Saatman, K.E., Meaney, D.F. and McIntosh, T.K. (1998b) In vitro central nervous system models of mechanically induced trauma: a review. J. Neurotrauma, 15: 911–928.

26 Pettus, E.H., Christman, C.W., Giebel, M.L. and Povlishock, J.T. (1994) Traumatically induced altered membrane permeability: its relationship to traumatically induced reactive axonal change. J. Neurotrauma, 11: 507–522. Pﬁster, B.J., Weihs, T.P., Betenbaugh, M. and Bao, G. (2003) An in vitro uniaxial stretch model for axonal injury. Ann. Biomed. Eng., 31: 589–598. Prange, M.T. and Margulies, S.S. (2002) Regional, directional, and age-depedent properties of the brain undergoing large deformation. J. Biomed. Eng., 124: 244–252. Rivlin, A.S. and Tator, C.H. (1978) Effect of duration of acute spinal cord compression in a new acute cord injury model in the rat. Surg. Neurol., 10: 38–43. Sahay, K.B., Mehrotra, R., Sachdeva, U. and Banerji, A.K. (1992) Elastomechanical characterization of brain tissues. J. Biomech., 25: 319–326. Scheff, S.W., Rabchevsky, A.G., Fugaccia, I., Main, J.A. and Lumpp Jr., J.E. (2003) Experimental modeling of spinal cord injury: characterization of a force-deﬁned injury device. J. Neurotrauma, 20: 179–193. Shi, R. and Borgens, R.B. (2000) Anatomical repair of nerve membranes in crushed mammalian spinal cord with polyethylene glycol. J. Neurocytol., 29: 633–643. Shi, R. and Pryor, J.D. (2002) Pathological changes of isolated spinal cord axons in response to mechanical stretch. Neuroscience, 110: 765–777. Shi, R. and Whitebone, J. (2006) Conduction deficits and membrane disruption of spinal cord axons as a function of

magnitude and rate of strain. J. Neurophysiol., 95: 3384–3390. Smith, D.H., Chen, X.H., Xu, B.N., McIntosh, T.K., Gennarelli, T.A. and Meaney, D.F. (1997) Characterization of diffuse axonal pathology and selective hippocampal damage following inertial brain trauma in the pig. J. Neuropathol. Exp. Neurol., 56: 822–834. Smith, D.H., Wolf, J.A., Lusardi, T.A., Lee, V.M. and Meaney, D.F. (1999) High tolerance and delayed elastic response of cultured axons to dynamic stretch injury. J. Neurosci., 19: 4263–4269. Takhounts, E.G., Eppinger, R.H., Campbell, J.Q., Tannous, R.E. and Power, E.D. (2003) On the development of the SIMon ﬁnite element head model. Stapp Car Crash J., 47: 107–133. Thibault, L.E., Meaney, D.F., Anderson, B.J. and Marmarou, A. (1992) Biomechanical aspects of a ﬂuid percussion model of brain injury. J. Neurotrauma, 9: 311–322. Thurman, D. and Guerrero, J. (1999) Trends in hospitalization associated with traumatic brain injury. JAMA, 282: 954–957. Versace, J. (1971) A review of the severity index. In: Proceedings of the 15th Stapp Car Crash Conference, Society of Automotive Engineers, New York, pp. 771–796. Young, W. (2002) Spinal cord contusion models. Prog. Brain Res., 137: 231–255. Zhang, L., Yang, K.H. and King, A.I. (2004) A proposed injury threshold for mild traumatic brain injury. J. Biomech. Eng., 126: 226–236.

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 3

Linking impact to cellular and molecular sequelae of CNS injury: Modeling in vivo complexity with in vitro simplicity Jennifer M. Spaethling, Donna M. Geddes-Klein, William J. Miller, Catherine R. von Reyn, Pallab Singh, Mahlet Mesﬁn, Steven J. Bernstein and David F. Meaney Departments of Bioengineering and Neurosurgery, University of Pennsylvania, 3320 Smith Walk, Philadelphia, PA, 19104-6392, USA

Abstract: Traumatic brain injury (TBI) represents one of most common disorders to the central nervous system (CNS). Despite signiﬁcant efforts, though, an effective clinical treatment for TBI is not yet available. The complexity of human TBI is modeled with a broad group of experimental models, with each model matching some aspect of the human condition. In the past 15 years, these in vivo models were complemented with a group of in vitro models, with these in vitro models allowing investigators to more precisely identify the mechanism(s) of TBI, the different intracellular events that occur in acute period following injury, and the possible treatment of this injury in vitro. In this paper, we review the available in vitro models to study TBI, discuss their biomechanical basis for human TBI, and review the ﬁndings from these in vitro models. Finally, we synthesize the current knowledge and point out possible future directions for this group of models, especially in the effort toward developing new therapies for the traumatically brain injured patient. Keywords: traumatic brain injury; biomechanics; in vitro models

Sattin, 2005). In the elderly, TBI is only eclipsed by cardiovascular disease and cancer as a major cause of death. Clinically, brain injuries are categorized broadly as focal or diffuse (Gennarelli et al., 1982a). Focal injuries are readily observable lesions that appear on standard CT or MRI scans, and include injuries to the vasculature (epidural, subdural hematoma), the microvasculature (cerebral contusions), and visible tears in the brain parenchyma (intracerebral hemorrhage). Diffuse injuries are diagnosed when no visible lesions are present using conventional imaging, yet the patient has clear neurological impairment. A major substrate of diffuse brain injuries is diffuse axonal injury (DAI), but this category of injury also includes diffuse brain swelling

Introduction The enormous consequences of traumatic brain injury (TBI) continue in society, despite the rapid advance of technology to reduce the severity of injuries and new approaches in trauma patient care. Even with these changes, traumatic brain injuries remain the leading cause of death in people less than 45 years old (Thurman et al., 1999). The incidence rate is equally startling — the number of people hospitalized each year for traumatic brain injuries exceed those diagnosed with multiple sclerosis, breast cancer, and spinal cord injury combined (Langlois and Corresponding author. Tel.: +1 215 573 3155; Fax: +1 215

573 2071; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61003-0

27

28

and brain edema (Graham et al., 1995; Povlishock and Katz, 2005). Generations of investigators directed their efforts toward understanding the mechanisms of TBI. From these efforts, researchers and clinicians recognize that TBI for a given patient is not captured well with only a ‘diffuse’ or ‘focal’ description. Rather, many different injuries are grouped with the broad clinical subtypes and leads to a tremendous diversity of injuries in the patient population. Injury mechanisms share a similar diversity — the mechanisms of injury for one speciﬁc traumatic injury may not apply universally to other injury types, and vice versa. As a ﬁeld, we often use a reductionist approach to understand individual components of clinical TBI. Models to study the in vivo complexity of TBI exist in different species, along different length scales that span from the single cell to the whole organism. The purpose of this review is to provide a current synthesis of the ﬁndings from models intended to study TBI in vitro, with an emphasis on discussing how these models relate to experimental efforts using animal models of TBI. In turn, we will summarize the ﬁndings from these models and point out new areas of opportunity where these models may prove invaluable in developing new treatments for TBI.

Biomechanical mechanisms that cause TBI in vivo A discussion of the linkages between in vitro and in vivo experimental TBI studies would be incomplete without a brief review of the underlying biomechanical mechanisms that cause TBI. One major aim of in vitro models is to replicate the underlying physical forces that the tissue experiences during traumatic injury. However, the forces experienced by the tissue during even a single head impact can vary greatly with impact direction, force, and the impacting surface. A person designing an in vitro model must ask — which of these forces to the tissue are important for causing injury? In addition, one asks — how does one control the inherent variability of these forces? Pressures developed within the brain during injury were considered one of the primary

mechanisms for causing immediate impairment after brain injury, with the ﬁrst modern studies dating back over six decades (Denny-Brown and Russel, 1941; Denny-Brown, 1945). These studies were soon followed by a number of in vivo experimental models that simulated these pressures to cause a concussive insult in animals (Stalhammar and Olsson, 1975; Sullivan et al., 1976). The intracranial pressure patterns that are generated in the human brain during impact is known (Nahum et al., 1977) and is predicted accurately with computational models (e.g., Zhang et al., 2004). Therefore, one can draw a clear relationship between the pressures generated in the in vivo models and within the brain during TBI. The effect of transient pressure changes on tissue function, though, is less clear. Of the possible mechanisms, there is only consensus that pressure cause damages is when the pressure drops below the cavitation threshold for tissue, thereby causing immediate (primary) tissue damage (Nusholtz et al., 1995). Much less agreement exists on whether high positive pressures will cause impairment, or an additional physical injury mechanism is needed. Tissue deformation during impact is the second major physical mechanism explaining the primary patterns of injury in the brain. Although the brain shape under pressure loading largely does not change, the compliant properties of the tissue make the brain susceptible to shearing deformations during impact. Until recently, we knew little about the properties of the brain during situations causing injury. Recent information suggests that the brain softens as it deforms (Prange and Margulies, 2002; Coats and Margulies, 2006), an intriguing material property that may lead to unexpected patterns of tissue damage. Although there is wide recognition that brain deformation/stress is a major mechanism leading to cellular damage, exactly how the stresses are transferred from the tissue to the cellular structures within the material are not well known. Estimates of the tissue strain needed to cause injury are now available (Shreiber et al., 1997; Bain and Meaney, 2000; Zhu et al., 2006), but there remains considerable discussion about how these strains should be modeled in either in vivo or in vitro models, and if these

29

models can truly capture the complete simulation of injury.

Reproducing injury mechanisms with in vivo TBI models Much like other diseases and disorders, an investigator faces several challenges when modeling human TBI in an animal model. The clinical mix of focal and diffuse injuries conﬂicts with the need to develop a consistent, reliable, and repeatable model in the laboratory. Inevitably, the investigator chooses which components of clinical TBI to study in the laboratory. Here, we brieﬂy categorize the different experimental models and describe their uses. We use this description as background material for the in vitro models discussed in the next section. For nonpenetrating injuries, an early area of study was to model the effects of pressure loading on the brain, as the concussive effects of pressure were reported in the early literature (DennyBrown, 1945). The effect of pressure led to the so called ‘percussion concussion’ models (Gennarelli, 1994) that used a pulse of either air or ﬂuid on the exposed cortex to cause neurological impairment. The most common model in current use is the ﬂuid percussion model, available in many species and used in different conﬁgurations (Stalhammar and Olsson, 1975; Sullivan et al., 1976; Dixon et al., 1988; McIntosh et al., 1989a). Early work with ﬂuid percussion showed how the pressure applied to the cortex was dissipated throughout the brain and the spinal canal (Stalhammar and Olsson, 1975). Later work showed that this pressure also caused a movement of intracranial tissue, leading to pressure and strain throughout the brain (Thibault et al., 1992). Possibly due to the complex mechanics of this model, different variations of the percussion model can lead to forms of injury that resemble components of human TBI. A more recent technique to study brain injury in vivo is to directly deform the brain with a solid indentor, often referred to as the cortical impact model (Lighthall, 1988). The advantages of the model include a highly quantiﬁed impact condition, an ability to easily scale the impact condition

across species (Smith et al., 1995), and modifying the model to injure different areas of the cortex. Recent modiﬁcations of the model used a direct skull impact, leading to a closed head impact model that could be considered closer to the clinical condition. With the precise control of the model, the cortical impact model is frequently the choice when studying TBI in transgenic animals. From an injury mechanism standpoint, the cortical model is qualitatively similar to the ﬂuid percussion technique — apply a mechanical input locally to the cortex and study the progressive pattern of injury throughout the brain. A ﬁnal method for TBI models is using acceleration to injure the brain, reproducing a common mechanical loading that causes TBI in humans. Studies show the necessary acceleration to cause injury increases quickly with decreasing brain mass (Ommaya et al., 1967); therefore, the acceleration method is most readily used in large animal species that include the nonhuman primate and the miniature pig (Gennarelli et al., 1982b; Meaney et al., 1995). Although models using acceleration input in small animals (e.g., rat, ferret) appear in the literature (Marmarou et al., 1994; Xiao-Sheng et al., 2000; Gutierrez et al., 2001), it is difﬁcult to separate the effects of the acceleration from the possible shape changes in the skull caused by the relatively large force needed to create these accelerations. A combination of the acceleration input and the higher proportion of white matter in the gyrencephalic brains of the nonhuman primate and pig makes these models ideal for studying injury to the white matter.

Broad categorization of in vivo models — what do they reproduce? The different in vivo models described above are generally designed to simulate the mechanisms of injury that occur in humans. Results from in vivo models are sometimes difﬁcult to interpret, though, as measurements of intracellular signaling, single cell function, and the role of different cell types are difﬁcult. Moreover, the effect of mechanical and hypoxic injury are not easily separable with in vivo models, but are easily divided with in vitro

30

approaches. The use of in vitro models to simulate TBI began several decades ago and has signiﬁcantly evolved in complexity (Morrison et al., 1998b). In this section, we describe the in vitro models that are currently in use to study TBI, and provide an in vivo correlate for these in vitro methods.

Scratch model/laceration The most direct method for mechanical injury is to tear or lacerate cultures with a stylus or punch. The direct mechanical disruption of cultures is one of the earliest methods to study the progression of injury, beginning with tissue chunks (Epstein, 1971) but soon moving to mixed cultures of neurons and glia (Tecoma et al., 1989; Regan and Choi, 1994; Regan and Panter, 1995). The scratch method is well suited for high throughput drug screening, and has been used by Faden and colleagues to examine both inhibitors and antisense oligonucleotide treatment (Faden et al., 1997; Mukhin et al., 1997, 1998). This technology also scales easily to slice culture tissue (Sieg et al., 1999), and has suggested factors that inﬂuence the vulnerability or survivability of neurons in different regions of the brain. The direct mechanical injury remains in use, now with an emphasis that includes the release of molecules that could be considered biomarkers of the injury (Yang et al., 2006) and the potential factors that can regulate the migratory activity of glial cells to the injury site (Barral-Moran et al., 2003). The technique best correlates to the primary tissue tearing that can occur in different brain regions following severe head injury, a penetrating ballistic injury, or the local tissue damage that occurs with depressed skull fractures. An interesting variation of the tearing model is using a laser to focally disrupt or transect the processes of neurons in culture (Gross et al., 1983). The relative position of the laser cut can be controlled and can be close to or distant from the neuronal soma, yielding distinct differences in cell fate (Lucas, 1987). The distance from the lesion to the soma also changes the subsequent ultrastructure response, as well as the electrophysiological properties of the cell (Lucas et al., 1985). Moreover, the

cuts can lead to gene expression changes over time (Raghupathi et al., 1998). This remains probably the most precise model to study distal and proximal effects of transection. The model best corresponds to the physical disruption that can occur to some neuronal processes following injury (Maxwell et al., 1997), especially at the severe levels. The most important utility of the model, though, may be the ability to distinguish the role of local and remote signaling on cell survival/death following injury (Singleton et al., 2002). Weight drop/compression An additional and straightforward technique to use on cultures is to mechanically compress the cultures with a weight, akin to the weight drop method developed initially by Allen (1911) to study spinal cord injury in vivo. Indeed, one of the ﬁrst in vitro models for central nervous system (CNS) injury used this technique of spinal cord cultures (Balentine et al., 1988). The technique is well suited to organotypic cultures that have a deﬁned thickness and more realistic 3D architecture, and can be used to study the effects of both mechanical injury and a superimposed hypoxic injury (Adamchik et al., 2000). The order of the injuries can be changed, so that the mechanical injury can be considered the secondary injury, or vice versa. The technique to compress the tissue construct can also change; a dropped weight can be replaced by a rolling stainless steel bar, or a composite foam indentor over a region of the culture (Adamchik et al., 2000). Recently, this type of model showed the potential spreading depression that can occur from mechanical injury (Church and Andrew, 2005). One primary disadvantage of this technique, though, is drawing the direct in vivo correlate for this method. Crush injuries to the brain parenchyma are rare, and are complicated by overlying skull fracture. Cell/substrate stretch model Based on the number of completed studies, the most commonly used in vitro technique to study the consequences of TBI in vitro is the cell stretch

31

(or substrate deformation) model. These models replicate the magnitude and rate of tissue deformation that occurs in vivo during injury (Meaney et al., 1995), but often simplify the multiple oscillations of tissue deformation into a single, transient stretch insult. One early feature of the model was using a design where the cultured cells, plated to an elastic substrate, were simultaneously deformed in two directions. Although ﬁrst used in astrocytes, this technique was rapidly tested in many different cell types of the CNS including neuron and neuronlike cell lines, endothelial cells, and, more recently, microglia (Cargill and Thibault, 1996; Ellis et al., 1995; McKinney et al., 1996; Rzigalinski et al., 1997). A variation of the initial stretch models is now available, where the cultures are stretched only in one direction (Lusardi et al., 2004). The stretch can be conﬁned to just cultured axons to model diffuse axonal injury (Smith et al., 1999). Methods to stretch dissociated cultures transfer easily to studying the effects of stretch on tissue slice cultures (Morrison et al., 1998a), where the transfer between the substrate stretch and the resulting tissue stretch are deﬁned (Cater et al., 2006). More recent work shows the cell stretch technique can scale to study the effect of more complex, 3D strain patterns on the morphology and viability of cells in tissue constructs (LaPlaca et al., 2005). With the increasing number of controlled mechanical manipulations available on cultures and tissue constructs, it is now possible to identify if CNS cells are uniquely vulnerable to a certain type of mechanical deformation (Geddes-Klein et al., 2006a), if the effects of injury are cumulative (Slemmer et al., 2002), or if the mechanism(s) of injury will change if the cells experience stretch, compression, or ﬂuid shear deformation (LaPlaca et al., 1997).

Mechanoactivation — the ﬁrst therapeutic target The increasing diversity of models used to study TBI in vitro leads one to a common question — what processes do these forces activate rapidly following injury, and how do these ‘mechanoactivated’ signals lead to the pathological changes observed hours to days following the initial injury?

Identifying these early changes in CNS cells provides initial therapeutic targets. Knowing how the mechanoactivated targets lead to subsequent intracellular events and/or consequences will naturally generate new therapeutic targets, is perhaps as important. In this section, we review the current knowledge on the early changes that occur following mechanical injury and their relative utility as a target for reversing the effects of the injury.

NMDAR are the most commonly studied mechanoactivated target Alterations in ionic homeostasis have been observed across nearly all in vitro and in vivo models of TBI and therefore provide a natural starting point for identifying mechanoactivated receptors. Due to the role of glutamate receptors in both physiologic (learning and memory) and pathologic (stroke and epilepsy) conditions, N-methyl D-aspartate receptors (NMDAR) are among the most widely studied receptors responsible for the cytosolic calcium overload (Gagliardi, 2000; Arundine and Tymianski, 2004). Multiple models of TBI have shown that the bulk of the loss of ionic homeostasis can be attributed to activation of NMDARs. The connectivity of NMDARs to the actin cytoskeleton may provide a potential mechanism as to how these receptors are activated by stretch (Geddes-Klein et al., 2006b). One unique feature of mechanical injury to neurons is the mechanically induced reduction in the normal magnesium block of the NMDAR (Zhang et al., 1996). Altering the magnesium block of NMDARs changes the relative inﬂuence of NMDAR even under normal neurotransmission signaling and provides a therapeutic target that has been found to be somewhat effective in vivo (McIntosh et al., 1989a, b). It is worth noting that mechanically induced changes in the NMDAR properties are not observed across all models of TBI, including mixed cultures (Faden et al., 2001), or when the cultures are subject to a sublethal stretch (Arundine et al., 2003, 2004). Additionally, the mechanically induced activation or modulation of NMDARs leads to the propagation of intracellular calcium into adjacent, uninjured cells (Lusardi et al., 2004). While

32

NMDAR antagonists have not been effective in the clinic, recent data suggests that targeting the location of NMDAR (synaptic vs. extrasynaptic) may differentiate between the protective and excitotoxic processes of NMDARs (DeRidder et al., 2006).

Are other glutamate receptors involved? The roles of other glutamate receptors are also beginning to generate serious attention in treating traumatic brain injuries. a-Amino-3-hydroxy-5methyl-4-isoxazolepropionic acid receptors (AMPAR) undergo a unique transformation, losing their rapid desensitization property following mechanical injury (Goforth et al., 1999). The loss in desensitization is persistent for at least 24 h following injury, and can be avoided if the NMDARs are inhibited prior to mechanical injury (Goforth et al., 2004). The impact of this loss in desensitization of the AMPARs on neuronal function is not yet well described, though. As AMPARs are centrally involved in fundamental processes such as synaptic plasticity and metaplasticity, even a transient alteration in the normal desensitizing properties of the AMPAR can lead to immediate and long term changes in neuronal network behavior. Although there is evidence that inhibition of metabotropic glutamate receptors (mGluRs) will provide some protection against the effects of mechanical injury, it is not known if these changes are directly caused by mechanical modulation of the mGluRs, or if these protective effects are from simply inhibiting the activation of these receptors from enhanced glutamate levels following injury (Faden et al., 2001; Movsesyan et al., 2001; Movsesyan and Faden, 2006). One interesting recent theory is the role of potential physical coupling between the group I mGluR receptors, IP3, and phospholipase C (PLC) (Floyd et al., 2004). The role of different mGluR subtypes also reveals a complex interplay between mGluRs after injury, where the inhibition of one subtype (type III) can eliminate the protection offered by an antagonist directed against a second subtype (type II) (Movsesyan and Faden, 2006). Given the role of mGluRs in either enhancing or modulating synaptic efﬁcacy (Gubellini et al., 2004; Simonyi et al.,

2005), these mechanically initiated changes may have long lasting effects on synaptic physiology following injury. Voltage-gated sodium channels can contribute to regional changes in neuronal axons Voltage-gated sodium channels have been shown to indirectly contribute to shifts in intracellular calcium following mechanical injury. The relative role of the voltage-gated sodium channel only appears when the axonal segment of the neuron is deformed, potentially because the NMDAR and AMPAR mediated changes dominate for dendritic processes and the neuronal soma. Moreover, it is worth noting, that little in vitro work exists for myelinated axons due to the difﬁculties of creating myelinated cultures in vitro, even though it has been shown that myelinated axons respond differently to stretch than myelinated axon (Reeves et al., 2005). Studies show the sodium channel activation following mechanical injury leads to a dramatic and sudden rise in axoplasmic calcium, but not from calcium entering directly through the sodium channel (Wolf et al., 2001). Rather, the source of increased axoplasmic calcium is through voltage-gated calcium channels and through reversal of the sodium calcium exchanger. These changes in the sodium channel occur simultaneously with a larger, but reversible change in the morphology of axons subjected to mechanical injury (Smith et al., 1999). Over time, these sodium channels are the targets of proteolysis, which can lead to a sustained change in neuronal activity (Iwata et al., 2004). Changes in mechanical permeability One ﬁnal consequence of mechanical force is an immediate, but transient, change in the plasma membrane permeability termed ‘mechanoporation’. In a series of studies, both the relative size and duration of these transient pores in the membrane were estimated following neuronal stretch (Geddes and Cargill, 2001; Geddes et al., 2003). These effects are also measured following traumatic injury in vivo, although these changes appear to occur over a

33

longer time period than in vitro preparations (Pettus et al., 1994; Stone et al., 2004; Farkas et al., 2006). It is not clear, though, how these changes in permeability can change for different neuronal regions (axon, dendrite, soma), or if this process is limited to one or more cell types. The effect of these transient pores can be minimized by resealing the membrane with surfactants (Serbest et al., 2006). Even though these changes may be brief, recent evidence suggests this mechanism may be capable of speciﬁcally stimulating different components of the mitogenactivated protein kinase (MAPK) cascade, and can therefore play a role in the ensuing cell death that occurs after injury (Serbest et al., 2006). However, the role of permeability increases in viability seems to be inconsistent across models, as some studies reveal no change in membrane resistance or permeability to a ﬂuorescent dye (Tavalin et al., 1995; Zhang et al., 1996; Smith et al., 1999). Mechanoactivation cascade — the next target Even though the early events of mechanical injury in vitro are becoming clear, these efforts only begin to establish the complex cascade of events that will inﬂuence neuronal or glial survival after TBI. The initial mechanoactivated receptors and ensuing loss of ionic homeostasis activates a cascade of cellular processes that ultimately result in loss of function or cell death. Understanding these cascades will be essential to developing novel therapeutic targets that are admissible in a clinically relevant therapeutic window. We term these next events as the ‘mechanoactivation cascade’. These cascades can occur rapidly following the initial mechanoactivation step, or can progress more slowly over time. In this section, we group these cascades according to the cell types studied to date in the CNS. Neuronal cascades The changes in neurons following mechanical injury include alterations in the electrophysiological properties, organelle function, receptor proﬁle, and intracellular calcium buffering. Some of these changes lead to neuronal vulnerability, while others will transiently impair function. Early after

injury, there is a delayed, but persistent depolarization that is linked to the alteration in the electrogenic sodium/potassium exchanger (Tavalin et al., 1995, 1997). The depolarization is triggered with a mild glutamate stimulation, and can be attenuated by restoring intracellular ATP levels. Findings from several groups show that mechanical injury leads to an enhanced response to glutamate up to 24 h following injury (Weber et al., 1999; Arundine et al., 2004; Geddes-Klein et al., 2006a), in turn leading to an enhanced vulnerability to glutamate excitotoxicity (Arundine et al., 2003). Although most reports focus on the role of the NMDAR as the underlying factor for this enhanced glutamate response, this is not the only ligand-gated receptor that changes following injury. The loss in desensitization of the AMPAR appears in parallel with an enhancement of AMPA currents after injury, as does the enhancement of g-aminobutyric acid (GABA) currents. Both the AMPA and GABA current changes are linked to the NMDAR, as inhibiting the NMDA prior to injury eliminates the enhancement of both currents (Goforth et al., 2004; Kao et al., 2004). These changes may also contribute to the changes in neural network activity that appears following mechanical injury (Prado et al., 2005), although these changes remain to be explored in detail. These changes are likely among the initiating factors that contribute to failure to induce LTP in in vivo models and ultimately result in learning and memory deﬁcits. Neuronal mitochondria are also affected, with a reduction in the mitochondria membrane potential that can persist and which is dependent on the presence of surrounding glia (Ahmed et al., 2000, 2002). A series of reports show that the normal regulation of intracellular calcium stores are altered soon after mechanical injury, and this change in capacitive calcium inﬂux can alter the homeostatic mechanisms for calcium induced calcium release (Weber et al., 2001; Chen et al., 2004; Weber, 2004). Cysteine protease activation is dependent on the severity of the injury, and can be affected by NMDAR activation inhibition/activation (Pike et al., 2000; DeRidder et al., 2006). Inhibiting the activation of one cysteine protease, caspase-3, will have a short therapeutic window but also reveals

34

some crosstalk among the classic apoptotic and necrotic pathways (Knoblach et al., 2004). Gene expression changes correlate with the mechanical input applied to the culture (Morrison et al., 2000); some of these changes reﬂect changes observed in single neurons following TBI in vivo (O’Dell et al., 2000). The role of MAPKs in vitro are similar to the role observed in vivo — extracellular signal regulated kinase (ERK) inhibition may be a potential target (Mori et al., 2002), but the timing of the therapy needs to be better deﬁned (Dash et al., 2002).

within astrocytes, but also mediate damage in adjacent neurons (Lamb et al., 1997). Changes in mitochondria membrane potential and ATP levels occur only transiently in astrocytes after injury, but the presence of the astrocytes in cultures change the time course of neuronal mitochondria changes (Ahmed et al., 2000). The presence of astrocytes also appears to inﬂuence the release of matrix metalloproteinases following injury in vitro, a factor that may signiﬁcantly affect the regeneration phase after injury (Wang et al., 2002). Gene expression changes can be altered in the presence or absence of glia (Katano et al., 1999), indicating that the crossover effect extends to the genomic level.

Astrocytic cascades The role of astrocyte-based changes can inﬂuence not only glial reactivity in the injured brain, but also neuronal activity through the coupling of astrocytes and neurons at the synapse. Intracellular calcium signaling in astrocytes after injury is coupled to the presence of extracellular calcium, and the absence of extracellular calcium will inﬂuence astrocytic death (Rzigalinski et al., 1997). The complex intracellular regulation of calcium also changes, as stimulation of mGluRs is less coupled to IP3 mediated calcium release and is regulated, in part, by PLC (Rzigalinski et al., 1998; Floyd et al., 2001). Changes in intracellular sodium occur in parallel with these calcium changes and are linked with glutamate uptake in astrocytes, but may also be loosely coupled to intracellular calcium changes through changes in the sodium calcium exchanger (Floyd et al., 2005). These changes in ion homeostasis lead to alterations in MAPK signaling (Neary et al., 2003), some of which can be linked to the eventual glial reactivity phase that forms an important part of the neuropathology found in human TBI. Perhaps more than studies using cultured neurons, there are now several reports focusing on the ‘crossover’ effect of mechanically injured astrocytes on other CNS cell types. For example, injured astrocytes generate isoprostanes that can inﬂuence vasoconstriction and reduce blood ﬂow after trauma (Hoffman et al., 2000). Generation of reactive oxygen species from mechanically injured astrocytes can affect not only metabolic changes

What therapies have emerged? A primary advantage of in vitro models is the ability to quickly test the efﬁcacy of new or existing compounds in reducing the effects of mechanical injury. Most work to date, though, has focused on the mechanisms that contribute to either early changes in intracellular signaling or cell function. Surprisingly few studies have moved this work toward studying therapies that may be used for in vivo TBI treatment. Early work showed the effectiveness of NMDA antagonists (Regan and Panter, 1995), free radical scavengers (Lamb et al., 1997; Shah et al., 1997), and mGluR activation or inhibition (Allen et al., 1999; Movsesyan et al., 2001). More recent work includes the development of new peptides to reduce cell death after injury, even if the compound is delivered hours following injury (Faden et al., 2003, 2004, 2005). These peptides offer a strategic advantage over receptor antagonists by offering more selective inhibition of intracellular signaling triggered by a target receptor. Customized peptide designs can interfere with critical regulatory points in the mechanoactivation cascade, such as the point where NMDAR activation can lead to the formation of reactive oxygen species. Using a peptide to inhibit the linkage between the NMDAR and PSD-95, a prominent postsynaptic density protein, leads to an interruption in the ROS cascade and will lead to neuronal protection (Arundine et al., 2004). Delayed efﬁcacy is also possible by using subunit speciﬁc

35

antagonists for the NMDAR (DeRidder et al., 2006), perhaps due to the recent data showing the dual role for NMDAR in cell survival and apoptosis (Hardingham. and Bading, 2003). In vitro models are also useful in determining more speciﬁc biomarkers for both TBI diagnosis and treatment. A group of released factors have been found in the media following mechanical injury, with some contributing to the ensuing neuronal death. Injury using a stylus transection method is inhibited with NMDAR antagonists and aminosteroids (Regan and Panter, 1995), and the surrounding media can contain free radicals that, if applied to uninjured cultures, would cause apoptotic cell death within hours (Shah et al., 1997). A physical scratching injury also activates inﬂammatory mediated injury in neuronal cultures, as well as astrocytic migration (Fitch et al., 1999), MAPK activation, and the release of matrix metalloproteinases (MMPs) (Wang et al., 2002). Mechanical stretch injury will cause the synaptic release of zinc (Cho et al., 2003), an increase in s100b in the media (Willoughby et al., 2004). It is interesting to note that at least some of these molecules are being investigated as potential biomarkers for in vivo injury (Pineda et al., 2004).

Pointing toward the future Collectively, this work shows the extensive efforts completed in understanding how mechanical forces are transmitted to cells of the CNS, what processes are activated by these mechanical forces, and how this information reveals different intervention points for treating the effects of mechanical injury in the CNS. Motivated by an understanding of the physical forces that occur within the tissue during injury, there are now models to study the effect of both simple and more complex mechanical loading to cultures. These models are well controlled, can be easily extended to understand more realistic loading conditions, and are even approaching the very complex question of how these forces are transferred to cells in 3D tissue. One needs to consider if additional in vitro models are needed, or if the complex mechanical environment is suitably modeled with existing technology. Certainly, the

current work does not comprehensively address the initial mechanoactivation process for all cell types in the brain; notable exceptions are the oligodendrocytes, brain endothelial cells, and microglia. Even within a given cell type such as neurons, there are differences that appear among neurons from different brain regions. Moreover, there is little consensus information on how the culture platform — nearly pure or mixed cultures, tissue constructs, or organotypic slice cultures — contribute to the measured response; information to date show the potential synergistic interactions that appear between different cell types. These areas, as well as many others, can provide more insight into how these models will move toward representing the in vivo environment following injury. In closing, it is worth noting that the natural division between in vivo and in vitro models of injury is slowly disappearing. The culturing techniques are becoming more sophisticated, allowing one to construct functional tissues with highly precise and more ‘in vivo like’ architectures. The methods to injure these cultures is now developed enough to mechanically load these constructs with any desired loading proﬁle, mimicking the loading proﬁles occurring in vivo following injury. In comparison, rapid advances with in vivo imaging technology now allows one to use ﬂuorescent indicator dyes in vivo to monitor shifts ion homeostasis, and technology is already emerging to monitor the electrical activity in living, awake animals over time. As a result, technologies and measurement normally restricted to in vitro systems are now available for testing in the in vivo animal. The increasing convergence of these different approaches will likely make comparative approaches in the future more common, and will lead to a more rapid translation of in vitro ﬁndings to the in vivo setting. Ultimately, this will translate into the more rapid testing of therapeutics in vivo that originated from in vitro systems, and lead to more therapeutic options for the traumatically brain injured patient.

Acknowledgments Funds for this work were provided by grants from the National Institutes of Health (RO1 HD-41699

36

and NS-35712) and the Commonwealth of Pennsylvania.

References Adamchik, Y., Frantseva, M.V., Weisspapir, M., Carlen, P.L. and Perez Velazquez, J.L. (2000) Methods to induce primary and secondary traumatic damage in organotypic hippocampal slice cultures. Brain Res. Brain Res. Protoc., 5(2): 153–158. Ahmed, S.M., Rzigalinski, B.A., Willoughby, K.A., Sitterding, H.A. and Ellis, E.F. (2000) Stretch-induced injury alters mitochondrial membrane potential and cellular ATP in cultured astrocytes and neurons. J. Neurochem., 74(5): 1951–1960. Ahmed, S.M., Weber, J.T., Liang, S., Willoughby, K.A., Sitterding, H.A., Rzigalinski, B.A. and Ellis, E.F. (2002) NMDA receptor activation contributes to a portion of the decreased mitochondrial membrane potential and elevated intracellular free calcium in strain-injured neurons. J. Neurotrauma, 19(12): 1619–1629. Allen, A. (1911) Surgery of experimental lesion of spinal cord equivalent to crush injury of fracture dislocation of spinal column — a preliminary report. JAMA, 57(11): 878–890. Allen, J.W., Ivanova, S.A., Fan, L., Espey, M.G., Basile, A.S. and Faden, A.I. (1999) Group II metabotropic glutamate receptor activation attenuates traumatic neuronal injury and improves neurological recovery after traumatic brain injury. J. Pharmacol. Exp. Ther., 290(1): 112–120. Arundine, M., Aarts, M., Lau, A. and Tymianski, M. (2004) Vulnerability of central neurons to secondary insults after in vitro mechanical stretch. J. Neurosci., 24(37): 8106–8123. Arundine, M., Chopra, G.K., Wrong, A., Lei, S., Aarts, M.M., MacDonald, J.F. and Tymianski, M. (2003) Enhanced vulnerability to NMDA toxicity in sublethal traumatic neuronal injury in vitro. J. Neurotrauma, 20(12): 1377–1395. Arundine, M. and Tymianski, M. (2004) Molecular mechanisms of glutamate-dependent neurodegeneration in ischemia and traumatic brain injury. Cell Mol. Life Sci., 61(6): 657–668. Bain, A.C. and Meaney, D.F. (2000) Tissue-level thresholds for axonal injury in an experimental model of CNS white matter injury. J. Biomech. Eng., 122: 615–622. Balentine, J.D., Greene, W.B. and Bornstein, M. (1988) In vitro spinal cord trauma. Lab. Invest., 58(1): 93–99. Barral-Moran, M.J., Calaora, V., Vutskits, L., Wang, C., Zhang, H., Durbec, P., Rougon, G. and Kiss, J.Z. (2003) Oligodendrocyte progenitor migration in response to injury of glial monolayers requires the polysialic neural celladhesion molecule. J. Neurosci. Res., 72(6): 679–690. Cargill II, R.S. and Thibault, L.E. (1996) Acute alterations in [Ca2+]i in NG108-15 cells subjected to high strain rate deformation and chemical hypoxia: an in vitro model for neural trauma. J. Neurotrauma, 13(7): 395–407. Cater, H.L., Sundstrom, L.E. and Morrison III, B. (2006) Temporal development of hippocampal cell death is

dependent on tissue strain but not strain rate. J. Biomech., 39(15): 2810–2818. Chen, T., Willoughby, K.A. and Ellis, E.F. (2004) Group I metabotropic receptor antagonism blocks depletion of calcium stores and reduces potentiated capacitative calcium entry in strain-injured neurons and astrocytes. J. Neurotrauma, 21(3): 271–281. Cho, I.H., Im, J.Y., Kim, D., Kim, K.S., Lee, J.K. and Han, P.L. (2003) Protective effects of extracellular glutathione against Zn2+-induced cell death in vitro and in vivo. J. Neurosci. Res., 74(5): 736–743. Church, A.J. and Andrew, R.D. (2005) Spreading depression expands traumatic injury in neocortical brain slices. J. Neurotrauma, 22(2): 277–290. Coats, B. and Margulies, S.S. (2006) Material properties of porcine parietal cortex. J. Biomech., 39(13): 2521–2525. Dash, P.K., Mach, S.A. and Moore, A.N. (2002) The role of extracellular signal-regulated kinase in cognitive and motor deﬁcits following experimental traumatic brain injury. Neuroscience, 114(3): 755–767. Denny-Brown, D. (1945) Cerebral concussion. Physiol. Rev., 25: 296–325. Denny-Brown, D. and Russel, W.R. (1941) Experimental cerebral concussion. Brain, 64(2–3): 92–164. DeRidder, M.N., Simon, M.J., Siman, R., Auberson, Y.P., Raghupathi, R. and Meaney, D.F. (2006) Traumatic mechanical injury to the hippocampus in vitro causes regional caspase-3 and calpain activation that is inﬂuenced by NMDA receptor subunit composition. Neurobiol. Dis., 22(1): 165–176. Dixon, C.E., Lighthall, J.W. and Anderson, T.E. (1988) Physiologic, histopathologic, and cineradiographic characterization of a new ﬂuid-percussion model of experimental brain injury in the rat. J. Neurotrauma, 5(2): 91–104. Ellis, E.F., McKinney, J.S., Willoughby, K.A., Liang, S. and Povlishock, J.T. (1995) A new model for rapid stretchinduced injury of cells in culture: characterization of the model using astrocytes. J. Neurotrauma, 12(3): 325–339. Epstein, M.H. (1971) Relative susceptibility of elements of the cerebral cortex to mechanical trauma in the rat. J. Neurosurg., 35(5): 517–522. Faden, A.I., Fox, G.B., Di, X., Knoblach, S.M., Cernak, I., Mullins, P., Nikolaeva, M. and Kozikowski, A.P. (2003) Neuroprotective and nootropic actions of a novel cyclized dipeptide after controlled cortical impact injury in mice. J. Cereb. Blood Flow Metab., 23(3): 355–363. Faden, A.I., Ivanova, S.A., Yakovlev, A.G. and Mukhin, A.G. (1997) Neuroprotective effects of group III mGluR in traumatic neuronal injury. J. Neurotrauma, 14(12): 885–895. Faden, A.I., Knoblach, S.M., Movsesyan, V.A. and Cernak, I. (2004) Novel small peptides with neuroprotective and nootropic properties. J. Alzheimers Dis., 6(Suppl 6): S93–S97. Faden, A.I., Movsesyan, V.A., Knoblach, S.M., Ahmed, F. and Cernak, I. (2005) Neuroprotective effects of novel small peptides in vitro and after brain injury. Neuropharmacology, 49(3): 410–424. Faden, A.I., O’Leary, D.M., Fan, L., Bao, W., Mullins, P.G. and Movsesyan, V.A. (2001) Selective blockade of the

37 mGluR1 receptor reduces traumatic neuronal injury in vitro and improves outcome after brain trauma. Exp. Neurol., 167(2): 435–444. Farkas, O., Lifshitz, J. and Povlishock, J.T. (2006) Mechanoporation induced by diffuse traumatic brain injury: an irreversible or reversible response to injury? J. Neurosci., 26(12): 3130–3140. Fitch, M.T., Doller, C., Combs, C.K., Landreth, G.E. and Silver, J. (1999) Cellular and molecular mechanisms of glial scarring and progressive cavitation: in vivo and in vitro analysis of inﬂammation-induced secondary injury after CNS trauma. J. Neurosci., 19(19): 8182–8198. Floyd, C.L., Gorin, F.A. and Lyeth, B.G. (2005) Mechanical strain injury increases intracellular sodium and reverses Na+/ Ca2+ exchange in cortical astrocytes. Glia, 51(1): 35–46. Floyd, C.L., Rzigalinski, B.A., Sitterding, H.A., Willoughby, K.A. and Ellis, E.F. (2004) Antagonism of group I metabotropic glutamate receptors and PLC attenuates increases in inositol trisphosphate and reduces reactive gliosis in straininjured astrocytes. J. Neurotrauma, 21(2): 205–216. Floyd, C.L., Rzigalinski, B.A., Weber, J.T., Sitterding, H.A., Willoughby, K.A. and Ellis, E.F. (2001) Traumatic injury of cultured astrocytes alters inositol (1,4,5)-trisphosphatemediated signaling. Glia, 33(1): 12–23. Gagliardi, R.J. (2000) Neuroprotection, excitotoxicity and NMDA antagonists. Arq. Neuropsiquiatr., 58(2B): 583–588. Geddes, D.M. and Cargill II, R.S. (2001) An in vitro model of neural trauma: device characterization and calcium response to mechanical stretch. J. Biomech. Eng., 123(3): 247–255. Geddes, D.M., LaPlaca, M.C. and Cargill II, R.S. (2003) Susceptibility of hippocampal neurons to mechanically induced injury. Exp. Neurol., 184(1): 420–427. Geddes-Klein, D.M., Schiffman, K.B. and Meaney, D.F. (2006a) Mechanisms and consequences of neuronal stretch injury in vitro differ with the model of trauma. J. Neurotrauma, 23(2): 193–204. Geddes-Klein, D.M., Serbest, G., Mesﬁn, M.N., Cohen, A.S. and Meaney, D.F. (2006b) Pharmacologically induced calcium oscillations protect neurons from increases in cytosolic calcium after trauma. J. Neurochem., 97(2): 462–474. Gennarelli, T.A. (1994) Animate models of human head injury. J. Neurotrauma, 11(4): 357–368. Gennarelli, T.A., Spielman, G.M., Langﬁtt, T.W., Gildenberg, P.L., Harrington, T., Jane, J.A., Marshall, L.F., Miller, J.D. and Pitts, L.H. (1982a) Inﬂuence of the type of intracranial lesion on outcome from severe head injury. J. Neurosurg., 56: 26–32. Gennarelli, T.A., Thibault, L.E., Adams, J.H., Graham, D.I., Thompson, C.J. and Marcincin, R.P. (1982b) Diffuse axonal injury and traumatic coma in the primate. Ann. Neurol., 12: 564–574. Goforth, P.B., Ellis, E.F. and Satin, L.S. (1999) Enhancement of AMPA-mediated current after traumatic injury in cortical neurons. J. Neurosci., 19(17): 7367–7374. Goforth, P.B., Ellis, E.F. and Satin, L.S. (2004) Mechanical injury modulates AMPA receptor kinetics via an NMDA receptor-dependent pathway. J. Neurotrauma, 21(6): 719–732.

Graham, D.I., Adams, J.H., Nicoll, J.A., Maxwell, W.L. and Gennarelli, T.A. (1995) The nature, distribution and causes of traumatic brain injury. Brain Pathol., 5(4): 397–406. Gross, G.W., Lucas, J.H. and Higgins, M.L. (1983) Laser microbeam surgery: ultrastructural changes associated with neurite transection in culture. J. Neurosci., 3(10): 1979–1993. Gubellini, P., Pisani, A., Centonze, D., Bernardi, G. and Calabresi, P. (2004) Metabotropic glutamate receptors and striatal synaptic plasticity: implications for neurological diseases. Prog. Neurobiol., 74(5): 271–300. Gutierrez, E., Huang, Y., Haglid, K., Bao, F., Hansson, H.A., Hamberger, A. and Viano, D. (2001) A new model for diffuse brain injury by rotational acceleration: I model, gross appearance, and astrocytosis. J. Neurotrauma, 18(3): 247–257. Hardingham, G.E. and Bading, H. (2003) The Yin and Yang of NMDA receptor signalling. Trends Neurosci., 26(2): 81–89. Hoffman, S.W., Rzigalinski, B.A., Willoughby, K.A. and Ellis, E.F. (2000) Astrocytes generate isoprostanes in response to trauma or oxygen radicals. J. Neurotrauma, 17(5): 415–420. Iwata, A., Stys, P.K., Wolf, J.A., Chen, X.H., Taylor, A.G., Meaney, D.F. and Smith, D.H. (2004) Traumatic axonal injury induces proteolytic cleavage of the voltage-gated sodium channels modulated by tetrodotoxin and protease inhibitors. J. Neurosci., 24(19): 4605–4613. Kao, C.Q., Goforth, P.B., Ellis, E.F. and Satin, L.S. (2004) Potentiation of GABA(A) currents after mechanical injury of cortical neurons. J. Neurotrauma, 21(3): 259–270. Katano, H., Fujita, K., Kato, T., Asai, K., Kawamura, Y., Masago, A. and Yamada, K. (1999) Traumatic injury in vitro induces IEG mRNAs in cultured glial cells, suppressed by coculture with neurons. Neuroreport, 10(12): 2439–2448. Knoblach, S.M., Alroy, D.A., Nikolaeva, M., Cernak, I., Stoica, B.A. and Faden, A.I. (2004) Caspase inhibitor z-DEVD-fmk attenuates calpain and necrotic cell death in vitro and after traumatic brain injury. J. Cereb. Blood Flow Metab., 24(10): 1119–1132. Lamb, R.G., Harper, C.C., McKinney, J.S., Rzigalinski, B.A. and Ellis, E.F. (1997) Alterations in phosphatidylcholine metabolism of stretch-injured cultured rat astrocytes. J. Neurochem., 68(5): 1904–1910. Langlois, J.A. and Sattin, R.W. (2005) Traumatic brain injury in the United States: research and programs of the Centers for Disease Control and Prevention (CDC). J. Head Trauma Rehabil., 20(3): 187–188. LaPlaca, M.C., Cullen, D.K., McLoughlin, J.J. and Cargill II, R.S. (2005) High rate shear strain of three-dimensional neural cell cultures: a new in vitro traumatic brain injury model. J. Biomech., 38(5): 1093–1105. LaPlaca, M.C., Lee, V.M. and Thibault, L.E. (1997) An in vitro model of traumatic neuronal injury: loading rate-dependent changes in acute cytosolic calcium and lactate dehydrogenase release. J. Neurotrauma, 14(6): 355–368. Lighthall, J.W. (1988) Controlled cortical impact: a new experimental brain injury model. J. Neurotrauma, 5: 1–15. Lucas, J.H. (1987) Proximal segment retraction increases the probability of nerve cell survival after dendrite transection. Brain Res., 425(2): 384–387.

38 Lucas, J.H., Gross, G.W., Emery, D.G. and Gardner, C.R. (1985) Neuronal survival or death after dendrite transection close to the perikaryon: correlation with electrophysiologic, morphologic, and ultrastructural changes. Cent. Nerv. Syst. Trauma, 2(4): 231–255. Lusardi, T.A., Rangan, J., Sun, D., Smith, D.H. and Meaney, D.F. (2004) A device to study the initiation and propagation of calcium transients in cultured neurons after mechanical stretch. Ann. Biomed. Eng., 32(11): 1546–1558. Marmarou, A., Foda, M.A., van den Brink, W., Campbell, J., Kita, H. and Demetriadou, K. (1994) A new model of diffuse brain injury in rats. Part I: Pathophysiology and biomechanics. J. Neurosurg., 80(2): 291–300. Maxwell, W.L., Povlishock, J.T. and Graham, D.L. (1997) A mechanistic analysis of nondisruptive axonal injury: a review. J. Neurotrauma, 14(7): 419–440. McIntosh, T.K., Vink, R., Noble, L., Yamakami, I., Fernyak, S., Soares, H. and Faden, A.L. (1989a) Traumatic brain injury in the rat: characterization of a lateral ﬂuid-percussion model. Neuroscience, 28(1): 233–244. McIntosh, T.K., Vink, R., Yamakami, I. and Faden, A.I. (1989b) Magnesium protects against neurological deﬁcit after brain injury. Brain Res., 482(2): 252–260. McKinney, J.S., Willoughby, K.A., Liang, S. and Ellis, E.F. (1996) Stretch-induced injury of cultured neuronal, glial, and endothelial cells. Effect of polyethylene glycol-conjugated superoxide dismutase. Stroke, 27(5): 934–940. Meaney, D.F., Smith, D.H., Shreiber, D.I., Bain, A.C., Miller, R.T., Ross, D.T. and Gennarelli, T.A. (1995) Biomechanical analysis of experimental diffuse axonal injury. J. Neurotrauma, 12(4): 689–694. Mori, T., Wang, X., Jung, J.C., Sumii, T., Singhal, A.B., Fini, M.E., Dixon, C.E., Alessandrini, A. and Lo, E.H. (2002) Mitogen-activated protein kinase inhibition in traumatic brain injury: in vitro and in vivo effects. J. Cereb. Blood Flow Metab., 22(4): 444–452. Morrison III, B., Eberwine, J.H., Meaney, D.F. and McIntosh, T.K. (2000) Traumatic injury induces differential expression of cell death genes in organotypic brain slice cultures determined by complementary DNA array hybridization. Neuroscience, 96(1): 131–139. Morrison III, B., Meaney, D.F. and McIntosh, T.K. (1998a) Mechanical characterization of an in vitro device designed to quantitatively injure living brain tissue. Ann. Biomed. Eng., 26(3): 381–390. Morrison III, B., Saatman, K.E., Meaney, D.F. and McIntosh, T.K. (1998b) In vitro central nervous system models of mechanically induced trauma: a review. J. Neurotrauma, 15(11): 911–928. Movsesyan, V.A. and Faden, A.I. (2006) Neuroprotective effects of selective group II mGluR activation in brain trauma and traumatic neuronal injury. J. Neurotrauma, 23(2): 117–127. Movsesyan, V.A., O’Leary, D.M., Fan, L., Bao, W., Mullins, P.G., Knoblach, S.M. and Faden, A.I. (2001) mGluR5 antagonists 2-methyl-6-(phenylethynyl)-pyridine and (E)-2methyl-6-(2-phenylethenyl)-pyridine reduce traumatic neuronal

injury in vitro and in vivo by antagonizing N-methyl-D-aspartate receptors. J. Pharmacol. Exp. Ther., 296(1): 41–47. Mukhin, A.G., Ivanova, S.A., Allen, J.W. and Faden, A.I. (1998) Mechanical injury to neuronal/glial cultures in microplates: role of NMDA receptors and pH in secondary neuronal cell death. J. Neurosci. Res., 51(6): 748–758. Mukhin, A.G., Ivanova, S.A., Knoblach, S.M. and Faden, A.I. (1997) New in vitro model of traumatic neuronal injury: evaluation of secondary injury and glutamate receptor-mediated neurotoxicity. J. Neurotrauma, 14(9): 651–663. Nahum, A.M., Smith, R. and Ward, C.C. (1977) Intracranial pressure dynamics during head impact. In: 21st Annual Stapp Car Crash Conference, Society of Automotive Engineers. San Diego, CA. Neary, J.T., Kang, Y., Willoughby, K.A. and Ellis, E.F. (2003) Activation of extracellular signal-regulated kinase by stretchinduced injury in astrocytes involves extracellular ATP and P2 purinergic receptors. J. Neurosci., 23(6): 2348–2356. Nusholtz, G.S., Wylie, E.B. and Glascoe, L.G. (1995) Internal cavitation in simple head impact model. J. Neurotrauma, 12(4): 707–714. O’Dell, D.M., Raghupathi, R., Crino, P.B., Eberwine, J.H. and McIntosh, T.K. (2000) Traumatic brain injury alters the molecular ﬁngerprint of TUNEL-positive cortical neurons in vivo: a single-cell analysis. J. Neurosci., 20(13): 4821–4828. Ommaya, A.K., Yarnell, P., Hirsch, A.E. and Harris, E.H. (1967) Scaling of experimental data on cerebral concussion in sub-human primates to concussion threshold for man. In: 13th Stapp Car Crash Conference, Anaheim, CA, Society of Automotive Engineers. Pettus, E.H., Christman, C.W., Giebel, M.L. and Povlishock, J.T. (1994) Traumatically induced altered membrane permeability: its relationship to traumatically induced reactive axonal change. J. Neurotrauma, 11(5): 507–522. Pike, B.R., Zhao, X., Newcomb, J.K., Glenn, C.C., Anderson, D.K. and Hayes, R.L. (2000) Stretch injury causes calpain and caspase-3 activation and necrotic and apoptotic cell death in septo-hippocampal cell cultures. J. Neurotrauma, 17(4): 283–298. Pineda, J.A., Wang, K.K. and Hayes, R.L. (2004) Biomarkers of proteolytic damage following traumatic brain injury. Brain Pathol., 14(2): 202–209. Povlishock, J.T. and Katz, D.I. (2005) Update of neuropathology and neurological recovery after traumatic brain injury. J. Head Trauma Rehabil., 20(1): 76–94. Prado, G.R., Ross, J.D., DeWeerth, S.P. and LaPlaca, M.C. (2005) Mechanical trauma induces immediate changes in neuronal network activity. J. Neural Eng., 2(4): 148–158. Prange, M.T. and Margulies, S.S. (2002) Regional, directional, and age-dependent properties of the brain undergoing large deformation. J. Biomech. Eng., 124(2): 244–252. Raghupathi, R., Grants, I., Rosenberg, L.J., McIntosh, T.K. and Lucas, J.H. (1998) Increased jun immunoreactivity in an in vitro model of mammalian spinal neuron physical injury. J. Neurotrauma, 15(7): 555–561. Reeves, T.M., Phillips, L.L. and Povlishock, J.T. (2005) Myelinated and unmyelinated axons of the corpus callosum

39 differ in vulnerability and functional recovery following traumatic brain injury. Exp. Neurol., 196(1): 126–137. Regan, R.F. and Choi, D.W. (1994) The effect of NMDA, AMPA/kainate, and calcium channel antagonists on traumatic cortical neuronal injury in culture. Brain Res., 633(1–2): 236–242. Regan, R.F. and Panter, S.S. (1995) Traumatic neuronal injury in cortical cell culture is attenuated by 21-aminosteroids. Brain Res., 682(1–2): 144–150. Rzigalinski, B.A., Liang, S., McKinney, J.S., Willoughby, K.A. and Ellis, E.F. (1997) Effect of Ca2+ on in vitro astrocyte injury. J. Neurochem., 68(1): 289–296. Rzigalinski, B.A., Weber, J.T., Willoughby, K.A. and Ellis, E.F. (1998) Intracellular free calcium dynamics in stretchinjured astrocytes. J. Neurochem., 70(6): 2377–2385. Serbest, G., Horwitz, J., Jost, M. and Barbee, K. (2006) Mechanisms of cell death and neuroprotection by poloxamer 188 after mechanical trauma. FASEB J., 20(2): 308–310. Shah, P.T., Yoon, K.W., Xu, X.M. and Broder, L.D. (1997) Apoptosis mediates cell death following traumatic injury in rat hippocampal neurons. Neuroscience, 79(4): 999–1004. Shreiber, D., Bain, A.C. and Meaney, D.F. (1997) In vivo thresholds for mechanical injury to the blood brain barrier. J. Passenger Cars, 106: 3792–3806. Sieg, F., Wahle, P. and Pape, H.C. (1999) Cellular reactivity to mechanical axonal injury in an organotypic in vitro model of neurotrauma. J. Neurotrauma, 16(12): 1197–1213. Simonyi, A., Schachtman, T.R. and Christoffersen, G.R. (2005) The role of metabotropic glutamate receptor 5 in learning and memory processes. Drug News Perspect., 18(6): 353–361. Singleton, R.H., Zhu, J., Stone, J.R. and Povlishock, J.T. (2002) Traumatically induced axotomy adjacent to the soma does not result in acute neuronal death. J. Neurosci., 22(3): 791–802. Slemmer, J.E., Matser, E.J., De Zeeuw, C.I. and Weber, J.T. (2002) Repeated mild injury causes cumulative damage to hippocampal cells. Brain, 125(Pt 12): 2699–2709. Smith, D.H., Soares, H.D., Pierce, J.S., Pearlman, K., Saatman, K.E., Meaney, D.F., Dixon, C.E. and McIntosh, T.K. (1995) A model of parasagittal controlled cortical impact in the mouse: cognitive and histopathologic effects. J. Neurotrauma, 12(2): 169–178. Smith, D.H., Wolf, J.A., Lusardi, T.A., Lee, V.M. and Meaney, D.F. (1999) High tolerance and delayed elastic response of cultured axons to dynamic stretch injury. J. Neurosci., 19(11): 4263–4269. Stalhammar, D. and Olsson, Y. (1975) Experimental brain damage from ﬂuid pressures due to impact acceleration. Acta Neurol. Scand., 52: 38–55. Stone, J.R., Okonkwo, D.O., Dialo, A.O., Rubin, D.G., Mutlu, L.K., Povlishock, J.T. and Helm, G.A. (2004) Impaired axonal transport and altered axolemmal permeability occur in distinct populations of damaged axons following traumatic brain injury. Exp. Neurol., 190(1): 59–69. Sullivan, H.G., Martinez, J., Becker, D.P., Miller, J.D., Grifﬁth, R. and Wist, A.O. (1976) Fluid-percussion model of mechanical brain injury in the cat. J. Neurosurg., 45(5): 521–534.

Tavalin, S.J., Ellis, E.F. and Satin, L.S. (1995) Mechanical perturbation of cultured cortical neurons reveals a stretchinduced delayed depolarization. J. Neurophysiol., 74(6): 2767–2773. Tavalin, S.J., Ellis, E.F. and Satin, L.S. (1997) Inhibition of the electrogenic Na pump underlies delayed depolarization of cortical neurons after mechanical injury or glutamate. J. Neurophysiol., 77(2): 632–638. Tecoma, E.S., Monyer, H., Goldberg, M.P. and Choi, D.W. (1989) Traumatic neuronal injury in vitro is attenuated by NMDA antagonists. Neuron, 2(6): 1541–1545. Thibault, L.E., Meaney, D.F., Anderson, B.J. and Marmarou, A. (1992) Biomechanical aspects of a ﬂuid percussion model of brain injuiry. J. Neurotrauma, 9(4): 311–322. Thurman, D.J., Alverson, C., Dunn, K.A., Guerrero, J. and Sniezek, J.E. (1999) Traumatic brain injury in the United States: a public health perspective. J. Head Trauma Rehabil., 14(6): 602–615. Wang, X., Mori, T., Jung, J.C., Fini, M.E. and Lo, E.H. (2002) Secretion of matrix metalloproteinase-2 and -9 after mechanical trauma injury in rat cortical cultures and involvement of MAP kinase. J. Neurotrauma, 19(5): 615–625. Weber, J.T. (2004) Calcium homeostasis following traumatic neuronal injury. Curr. Neurovasc. Res., 1(2): 151–171. Weber, J.T., Rzigalinski, B.A. and Ellis, E.F. (2001) Traumatic injury of cortical neurons causes changes in intracellular calcium stores and capacitative calcium inﬂux. J. Biol. Chem., 276(3): 1800–1807. Weber, J.T., Rzigalinski, B.A., Willoughby, K.A., Moore, S.F. and Ellis, E.F. (1999) Alterations in calcium-mediated signal transduction after traumatic injury of cortical neurons. Cell Calcium, 26(6): 289–299. Willoughby, K.A., Kleindienst, A., Muller, C., Chen, T., Muir, J.K. and Ellis, E.F. (2004) S100B protein is released by in vitro trauma and reduces delayed neuronal injury. J. Neurochem., 91(6): 1284–1291. Wolf, J.A., Stys, P.K., Lusardi, T., Meaney, D. and Smith, D.H. (2001) Traumatic axonal injury induces calcium inﬂux modulated by tetrodotoxin-sensitive sodium channels. J. Neurosci., 21(6): 1923–1930. Xiao-Sheng, H., Sheng-Yu, Y., Xiang, Z., Zhou, F. and Jianning, Z. (2000) Diffuse axonal injury due to lateral head rotation in a rat model. J. Neurosurg., 93(4): 626–633. Yang, X.F., Cao, F., Pan, D.S., Liu, W.G., Hu, W.W., Zheng, X.J., Zhao, X.Q. and Lu, S.T. (2006) Establishment of a mechanical injury model of rat hippocampal neurons in vitro. Chin. J. Traumatol., 9(1): 29–33. Zhang, L., Rzigalinski, B.A., Ellis, E.F. and Satin, L.S. (1996) Reduction of voltage-dependent Mg2+ blockade of NMDA current in mechanically injured neurons. Science, 274(5294): 1921–1923. Zhang, L., Yang, K.H. and King, A.I. (2004) A proposed injury threshold for mild traumatic brain injury. J. Biomech. Eng., 126(2): 226–236. Zhu, Q., Prange, M. and Margulies, S. (2006) Predicting unconsciousness from a pediatric brain injury threshold. Dev. Neurosci., 28(4–5): 388–395.

SECTION III

Pathological Mechanisms of Injury

This page intentionally left blank

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 4

Cellular and subcellular change evoked by diffuse traumatic brain injury: a complex web of change extending far beyond focal damage Orsolya Farkas and John T. Povlishock Department of Anatomy and Neurobiology, Medical College of Virginia Campus, Virginia Commonwealth University, P.O. Box 980709, Richmond, VA 23298, USA

Abstract: Until recently, our understanding of the cellular and subcellular changes evoked by diffuse traumatic brain injury has been framed in the context of primary focal injury. In this regard, the ensuing cell death cascades were linked to contusional-mediated changes associated with frank hemorrhage and ischemia, and these were assumed to contribute to the observed apoptotic and necrotic neuronal death. Little consideration was given to the potential that other non-contusional cell death cascades could have been triggered by the diffuse mechanical forces of injury. While the importance of these classical, contusion-related apoptotic and necrotic cell death cascades cannot be discounted with diffuse injury, more recent information suggests that the mechanical force of injury itself can diffusely porate the neuronal plasmalemma and its axolemmal membranes, evoking other forms of cellular response that can contribute to cell injury or death. In this regard, the duration of the membrane alteration appears to be a dependent factor, with enduring membrane change, potentially leading to irreversible damage, whereas more transient membrane perturbation can be followed by cell membrane resealing associated with recovery and/or adaptive change. With more enduring mechanical membrane perturbation, it appears that some of the traditional death cascades involving the activation of cysteine proteases are at work. Equally important, non-traditional pathways involving the lysosomal dependent release of hydrolytic enzymes may also be players in the ensuing neuronal death. These mechanically related factors that directly impact upon the neuronal somata may also be inﬂuenced by concomitant and/or secondary axotomy-mediated responses. This axonal injury, although once thought to involve a singular intraaxonal response to injury, is now known to be more complex, reﬂecting differential responses to injuries of varying severity. Moreover, it now appears that ﬁber size and type may also inﬂuence the axon’s reaction to injury. In sum, this review explicates the complexity of the cellular and subcellular responses evoked by diffuse traumatic brain injury in both the neuronal somata and its axonal appendages. This review further illustrates that our once simplistic views framed by evidence based upon contusional and/or ischemic change do not fully explain the complex repertoire of change evoked by diffuse traumatic brain injury. Keywords: diffuse traumatic brain injury; neuronal injury; axonal injury; necrosis; apoptosis

Corresponding author. E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61004-2

43

44

Introduction Over the last 40 years, our understanding of the complex pathobiology of traumatic brain injury (TBI) has improved signiﬁcantly. From both the clinical and basic science perspective, now most consider the pathobiology of TBI in the context of focal as well as diffuse change (Povlishock and Katz, 2005). It is also recognized in the clinical setting that focal injuries are typically embedded within concomitant diffuse pathologies, with the diffuse pathologies being the major determinants of the adverse outcomes associated with TBI (Povlishock and Katz, 2005). Despite this recognition of the importance of diffuse change, a disproportionate number of basic science studies have focused on focal injuries, such as contusional change without a parallel consideration of any diffuse changes that accompany the injury. In this review, we attempt to address what is known as well as what is controversial in our understanding of diffuse change within the traumatically injured brain parenchyma, examining both the cellular and subcellular responses evoked by diffuse traumatic brain injury (DTBI) to the neurons and their axonal extensions. In this review, we attempt also to consider new experimental directions for this important area of scientiﬁc inquiry.

Neuronal damage associated with DTBI Neuronal damage and death associated with several types of brain injury including TBI in humans and animals have been widely studied and, as such, have been the focus of several major reviews (Adams et al., 1980; Cervos-Navarro and Lafuente, 1991; Kotapka et al., 1992, 1993; Ross et al., 1993; Kermer et al., 1999; Raghupathi, 2004; Yakovlev and Faden, 2004). The majority of these studies and reviews, however, have focused on focal injuries, addressing neuronal death in localized contusional or pericontusional regions. Because all these studies provided excellent, detailed descriptions of the necrotic and apoptotic neuronal change associated with focal TBI, the present review follows a different track, focusing primarily on the under

appreciated phenomenon of diffuse traumatic neuronal change.

Morphological characteristics of neuronal injury after DTBI As noted above, unlike the wide body of literature on neuronal death after focal TBI, there are relatively few descriptions of neuronal injury and death occurring remote from contusional regions (Dietrich et al., 1994a, b; Colicos et al., 1996; Hicks et al., 1996), or after DTBI (Smith et al., 1997; Runnerstam et al., 2001; Cernak et al., 2002). These studies, similar to those focusing on focal TBI, describe two distinct forms of cell death, namely necrosis and apoptosis, within the neocortex, hippocampus and diencephalon. With diffuse injury, damaged/dying neurons have been described not only in the neocortex in or near contusions, but also in regions remote from the impact/contusional site, such as the C1, C2, C3 layers and the dentate hilus of hippocampus and thalamus. Additional neuronal damage/death has been described scattered in the caudate/putamen and the inferior and superior colliculi, consistent with the diffuse nature of the injury. Like previous descriptions of necrosis, these damaged neurons were dark and shrunken with Nissl staining (Dietrich et al., 1994b; Hicks et al., 1996) or eosinophilic with acid fuchsin staining (Cortez et al., 1989; Dietrich et al., 1994a; Hicks et al., 1996), revealing distorted proﬁles and vacuolization. At the ultrastructural level, such neurons demonstrated increased cytoplasmic and nuclear electron density, swollen mitochondria, vacuolated cytoplasm, plasma membrane disruption, pyknotic nuclei and perisomatic glial swelling (Dietrich et al., 1994b). Importantly, neuronal necrosis has also been observed after DTBI in the absence of any focal lesion, such as contusion or hematoma formation. Speciﬁcally, neuronal necrosis bilaterally scattered in the neocortex (Singleton et al., 2002; Singleton and Povlishock, 2004; Farkas et al., 2006) as well as in the cerebellum, and the C1 and C3 pyramidal layers and dentate hilus of the hippocampus has been described (Kotapka et al., 1991; Smith et al., 1997; Singleton et al., 2002; Singleton

45

and Povlishock, 2004). While this limited literature supports the occurrence necrotic cell death with DTBI, to date, the mechanisms underlying this necrosis are not well understood. The presence of necrotic neurons remote from or in the absence of contusion suggests that the pathogenesis of this necrosis does not follow from any overt destructive or ischemic processes comparable to those occurring within contusional foci. Unfortunately, our current beliefs regarding the pathogenesis of DTBI-induced necrotic change have been biased by the ischemic literature, wherein disproportionate emphasis has been placed on membrane pump failure and subsequent ionic dysregulation and calpain-mediated proteolysis. It is likely, however, that other potentially important factors and mechanisms associated with direct mechanical perturbation and its sequelae may be at work in the pathogenesis of this diffuse necrotic neuronal death and, as such, merit consideration (vide infra). Like necrosis, scattered apoptosis has also been described following DTBI. Although underappreciated and originally thought to take place only in ‘‘physiologic’’ programmed neuronal death processes and/or slowly progressive neurodegenerative diseases, its role in the pathobiology of DTBI has become more apparent. Rink et al. (1995) ﬁrst identiﬁed this process in TBI, positing that it played a role in delayed neuronal damage in contrast to necrosis that was linked to both acute and delayed neuronal death. Neurons with apoptotic morphology have been described in contusional foci, but more commonly have been identiﬁed either in the absence of contusions or in foci remote from contusional change, such as the hippocampus and/or diencephalon (Colicos and Dash, 1996; Colicos et al., 1996; Pravdenkova et al., 1996; Conti et al., 1998; Fox et al., 1998; Newcomb et al., 1999; Lin et al., 2001; Runnerstam et al., 2001; Cernak et al., 2002, 2004; Raghupathi et al., 2002). At ultrastructural level, these apoptotic neurons demonstrated cytoplasmic condensation, nuclear pyknosis, chromatin condensation, cell rounding, membrane blebbing, cytoskeletal collapse and, ultimately, disintegration of the cell into small fragments forming apoptotic bodies. Similar to our understanding of necrosis, the pathogenesis of apoptosis occurring after DTBI is not well

appreciated. It is believed that the balance between the anti- and pro-apoptotic signals determines if the injured neuron dies or survives. At present, apoptotic cell death is considered to play a role in delayed neuronal death occurring several hours to weeks after DTBI, when the dominance of proapoptotic factors in the presence of a persistent energy supply results in the activation of cysteine proteases, such as caspases that are regulators and effectors of apoptotic cell death.

Mechanisms eliciting neuronal injury and death after DTBI Similar to their morphological differences, apoptosis and necrosis have also been distinguished by the differences in their pathological mechanisms. As noted, apoptosis is an active process requiring energy while necrosis is a passive event that results from energy failure and consequent loss of ionic homeostasis with the ultimate recruitment of an inﬂammatory response. Although both processes were once considered distinct, recent evidence demonstrates that necrosis and apoptosis can evolve in parallel in the same injured tissue. Further, it has been recently suggested based upon the cellular microenvironment and energy supply that a switch between necrosis and apoptosis can occur within an individual cell (Kermer et al., 1999; Kitanaka and Kuchino, 1999; Nicotera et al., 1999; Raghupathi, 2004). This hybrid form of cell death, sometimes referred as aponecrosis, reﬂects the fact that these cells show the morphological characteristics of both types of cell death (Formigli et al., 2000). To date, multiple reviews have considered those cellular, subcellular and pathophysiological cascades that may contribute to the above described necrotic and apoptotic neuronal cell death cascades (Kermer et al., 1999; Raghupathi et al., 2000; Zipfel et al., 2000; Keane et al., 2001; Raghupathi, 2004; Yakovlev and Faden, 2004). Accordingly, it is unnecessaryto revisit these details. Rather, we provide here only a summary of these well-characterized processes, while providing, in subsequent passages, more detail on other not well appreciated mechanisms related to the pathogenesis of DTBI.

46

Neuroexcitation and calcium dysregulation Calcium dysregulation has been linked to both necrotic and apoptotic processes, although different intracellular cascades are involved. To date, it is known that the traumatic episode can be associated with generalized neuroexcitotoxicity, involving excitatory amino acid (EAA) release that can trigger several subcellular alterations via Na+ and Ca2+ inﬂux from the extracellular space. Using microdialysis, the EAA glutamate has been found to be elevated after TBI both in animals and humans. While this glutamate elevation initially was described after focal TBI (Faden et al., 1989; Katayama et al., 1990; Palmer et al., 1993), recent studies also showed the involvement of EAA release after DTBI (Runnerstam et al., 2001; Goda et al., 2002; Fei et al., 2005). The subsequent massive Ca2+ inﬂux via glutamate/NMDA channels triggers further Ca2+ release from intracellular stores and thereby dramatically elevates free intracellular Ca2+. This increased cytosolic Ca2+ can lead to activation of proteases, such as calpains, phosphatases, protein kinases or nitric oxide synthase, all which have been linked in various scenarios to either necrotic or apoptotic cell death (Kermer et al., 1999).

Free radicals Concurrent with and inﬂuenced by the glutamate NMDA receptor activation and Ca2+ dysregulation described above, the generation of oxygen free radicals can also occur after DTBI. Either via the generation of nitric oxide, the accelerated metabolism of arachidonate acid via multiple pathways or the generation of superoxide anions via an evoked inﬂammatory response, destructive radical species can be generated with lethal results for the brain (Kontos and Povlishock, 1986; Povlishock and Kontos, 1992). With more localized radical generation and Ca2+ dysregulation, this creates the permissive environment for mitochondrial damage, cytochrome C release and apoptotic cell death. With more generalized radical generation, the oxygen radicals participate in highly destructive processes including lipid

peroxidation, protein nitrosylation and DNA degradation (Braughler and Hall, 1992).

Calpains and caspases As noted above, increased intracellular Ca2+ can activate proteases, such as calpain. Calpain, a member of cysteine proteases, and its activation are well known to be associated with several different types of brain injury, including DTBI (Kupina et al., 2001, 2002; Buki et al., 2003; Farkas et al., 2006). It has several different isotypes, including m-calpain and m-calpain that are ubiquitous in the central nervous system. Although calpains have several different substrates within the cell, such as cytoskeletal elements, neuroﬁlaments, protein kinase C, calmodulin binding proteins and transcription factors (for review, see Wang (2000); Huang and Wang (2001)), one of the most widely investigated of the calpain substrates is spectrin. Spectrin is the primary component of the neuronal cytoskeleton and its cleavage by calpain results in speciﬁc and stable breakdown products, such as 150 and 145 kDa spectrin breakdown products (SBDPs). The presence of calpain-speciﬁc SBDPs as a result of calpain-mediated spectrin proteolysis (CMSP) has been routinely observed both after experimental (Kampﬂ et al., 1996; Saatman et al., 1996; Pike et al., 1998, 2001; Farkas et al., 2006) and human TBI (McCracken et al., 1999; Pineda et al., 2004; Farkas et al., 2005), and now is being used as one of the markers of Ca2+-induced cell injury and death. Excessive Ca2+ inﬂux and the concomitant calpain activation lead to the proteolytic degradation of intracellular proteins and membranes. The ensuing cytoskeletal and membrane damage can destroy cell integrity, cause increased membrane permeability, compromise the transport of essential cell products, induce aberrant signaling cascades and ﬁnally lead to cell death (Kermer et al., 1999). Although calpain activation can lead to either necrotic or apoptotic change, most concur that in TBI it is primarily responsible for necrotic cell death (Wang, 2000; Raghupathi, 2004). Like calpains, caspases are also members of the cysteine protease family, with these proteases thought to play a major role in apoptotic cell

47

death (Eldadah and Faden, 2000; Wang, 2000; Raghupathi, 2004). Caspase-3, the common apoptosis effector, is activated via two major pathways, either the mitochondrial intrinsic or the death receptor, extrinsic pathways. Brieﬂy, the apoptotic stimuli can result in the permeabilization of the mitochondrial membrane via different pathways, subsequently triggering the release of cytocrome-c into the cytoplasm. Cytocrome-c binds to apoptosis activating protein-1 (Apaf-1) and the complex activates caspase-9, which then activates caspase-3 via cleaving its proenzyme form. During the extrinsic pathways, apoptotic stimuli activate the death receptors, such as TNFa1-receptor or Fas receptor. The receptors recruit adapter proteins and form a complex, which then induces the autolytic activation of caspase-8, which activates caspase-3 (for review, see Wang (2000); Yakovlev and Faden (2004)). Caspase-3, similar to calpain, cleaves spectrin in a speciﬁc manner resulting in a 120 kDa SBDP that can be used for the speciﬁc detection of caspase-3 activity and apoptosis after TBI. The activation of caspase-3 has been observed in widespread regions both after focal and DTBI (Pike et al., 1998, 2001; Beer et al., 2000; Clark et al., 2000; Keane et al., 2001; Yakovlev et al., 2001; Cernak et al., 2002), as well as in human head injury (Clark et al., 1999; Pineda et al., 2004; Farkas et al., 2005).

The Bcl-2 oncogene family Integral to the apoptotic death cascade is the Bcl-2 proto-oncogene family, which is one of the key regulators of apoptosis. This family consists of members with anti-apoptotic (Bcl-2 and BclXL) and pro-apoptotic (Bax, Bad, Bid, BclXS) activity. Typically, following apoptotic stimuli, Bax translocates from the cytoplasm to the mitochondrial membrane resulting in cytocrome-c release, which by binding with Apaf-1 activates caspases. Bax is also suggested to directly activate caspase-3. In contrast, anti-apoptotic Bcl-2 and BclXL can prevent the mitochondrial permeabilization by inactivating Bax via heterodimerization. Bcl-2 and BclXL can also inhibit caspase-3 activation (Merry and Korsmeyer, 1997; Yang et al., 1997).

Pro-apoptotic Bax overexpression has been demonstrated after DTBI (Cernak et al., 2002). Concomitant decrease in the expression of the anti-apoptotic Bcl-2 and BclXL has also been reported (Felderhoff-Mueser et al., 2002). However, some studies have demonstrated increased Bcl-2 expression after DTBI (Cernak et al., 2002). Overall, it has been suggested that the shift between pro- and anti-apoptotic oncogene expression regulates the fate of the injured neuron, resulting in cell death if the expression of pro-apoptotic proteins exceeds the expression of anti-apoptotic proteins, while promoting cell survival if the anti-apoptotic Bcl protein expression is upregulated over the proapoptotic members of the Bcl-family (Merry and Korsmeyer, 1997).

Role of hydrolytic enzymes In the previous passages, attention has been focused on those pathways that have received emphasis in terms of the apoptotic/necrotic pathways occurring with TBI. While it is clear that calcium dysregulation and cysteine protease activation are integral to traumatically induced neuronal death, it is of note that other potentially important pathways and processes may be operant, although, they have been give little attention in the context of TBI. Foremost among these are the calcium-mediated activation of lysosomal hydrolytic pathways that have received considerable attention in non-TBI-related neuronal death. Further, it is well known that lysosomes and their enzymes are involved in different cell death pathways in various non-CNS systems as reviewed by Kroemer and Jaattela (2005) and Yamashima (2004). Multiple damaging insults are known to result in the disruption of the lysosomal membrane leading to the release of several lysosomal enzymes, such as cathepsins. It has been posited that low levels of stress result in limited lysosomal membrane rupture and cathepsin release that in turn elicit apoptosis. In contrast, high levels of stress can cause generalized lysosomal membrane rupture resulting in necrosis (Brunk et al., 1997). In the later scenario, targeted Ca2+-mediated calpain activation and the likely concomitant production of free radicals are posited to destroy the lysosomal

48

membrane leading to the leakage of damaging hydrolytic lysosomal enzymes into the cytoplasm to cause the digestion of the cell’s structural proteins (Yamashima, 2000; Yamashima et al., 2003). Despite the consistent involvement of lysosomes and cathepsin in cell death in multiple organs, their contribution to neuronal death following DTBI has not been rigorously evaluated and obviously requires attention. Mechanically related factors While the previous passages brieﬂy reviewed some of the neurochemical cascades triggered by the forces of injury, until recently there has not been any parallel consideration that the mechanical force of injury itself could directly perturb the neuronal cell membrane or its appendages leading to mechanically induced neuronal death cascades. In this regard, we comment here on recently published data evaluating the potential damaging neuronal somatic consequences of traumatically induced axotomy and/or the direct disruption of the neuronal plasmalemma. Axotomy-related changes Neuronal somatic changes occurring adjacent to sites of traumatic axonal injury (TAI) have been recognized previously (Van den Heuvel et al., 1998, 1999; Cernak et al., 2004); yet, the relation of these events to any subsequent injury cascades has not been appreciated. Recently, using immunohistochemical techniques recognizing amyloid precursor protein (APP), a well accepted marker of TAI, neuronal somata directly linked to TAI were identiﬁed following DTBI in the cerebral cortex, hippocampus and thalamus (Singleton et al., 2002). These studies showed that contrary to contemporary thought, TAI, even in immediate proximity to the neuronal soma, did not result in acute neuronal death. Rather, neurons sustaining TAI revealed signs of reactive change, including the loss and degranulation of the rough endoplasmic reticulum, disaggregation of polysomes and dispersal of the Golgi apparatus without any cytoskeletal or mitochondrial alterations (Singleton

et al., 2002). These TAI-linked neurons also revealed a transient suppression of protein synthesis. Taken together, these transient reactive changes suggest a potential neuronal attempt at reorganization and repair, rather than the initiation of any prenecrotic/apoptotic change.

Mechanoporation and the fate of neurons with membrane disruption Using extracellular tracer infusion techniques, multiple in vitro and in vivo studies have demonstrated that immediately following DTBI, other non-axotomized neurons can take up either high molecular weight tracers, such as horseradish peroxidase or different molecular weight dextrans or other molecules normally excluded from the neuronal cytoplasm by the intact cell membrane. This immediate tracer uptake suggests that the mechanical force of the injury itself evoked neuronal cell membrane disruption (mechanoporation), that then most likely allowed the inﬂux of damaging ions through the compromised cell membrane (Geddes et al., 2003; Prado and LaPlaca, 2004; Prado et al., 2004, 2005; Singleton and Povlishock, 2004). Following these initial descriptions of neuronal membrane perturbation, subsequent studies revealed a rather heterogeneous neuronal response to this disruption (Singleton and Povlishock, 2004; Farkas et al., 2006). Some neurons sustaining cell membrane disruption revealed Fluoro Jade positivity and ultrastructural evidence of overt neuronal necrosis. In contrast, other neurons sustaining membrane disruption did not reveal comparable signs of overt cell damage. Further, at the same time and in the same brain foci, different populations of injured neurons did not reveal evidence of membrane disruption, yet, they demonstrated either perisomatic axonal injury resulting in increased neuronal somatic APP-positivity, or the non-axotomy-related induction of heat shock protein expression (Singleton and Povlishock, 2004), and/or CMSP (Farkas et al., 2006, vide infra). In part, because of these divergent neuronal responses, some of which were suggestive of cell death whereas others suggested recovery, the

49

potential for post-traumatic neuronal cell membrane resealing and recovery became a consideration. Via the administration of different extracellular tracers at varied times post-injury, in vitro studies have revealed that the majority of neurons sustaining mechanoporation could reseal their disrupted membranes in the ﬁrst minutes post-TBI (Geddes et al., 2003; Prado and LaPlaca, 2004; Prado et al., 2005). In contrast, in vivo studies did not conﬁrm these observations. Rather, in vivo studies revealed that more than 50% of the tracer-containing cortical neurons contained both pre- and post-injury administered tracers at several hours post-DTBI, suggesting enduring membrane permeability (Fig. 1) (Farkas et al., 2006). Resealing was observed after DTBI, yet, to a lesser extent than suggested in vitro (Fig. 2) (Farkas et al., 2006). To further complicate this issue, the recent ﬁnding of neurons demonstrating enduring membrane permeability without evidence of overt cell damage points to the occurrence of potentially delayed membrane resealing, rather than the rapid resealing/repair observed in vitro (Farkas et al., 2006). Further complexity was also associated with the ﬁnding that other neurons only revealed post-injury tracer uptake without evidence of the pre-injury tracer, suggesting the potential for delayed membrane disruption (Fig. 3) (Farkas et al., 2006). Although the underlying mechanism(s) of this delayed membrane disruption is (are) unknown, the possibility exists that the sustained, elevated intracranial pressure occurring after injury and persisting for several hours post-TBI (Farkas et al., 2006) could contribute to this phenomenon. Statistical analysis of the neuronal populations demonstrating different tracer uptake over time post-DTBI has revealed signiﬁcant tracer distribution/redistribution consistent with the complex neuronal membrane changes described above. Speciﬁcally, between 4 and 8 h post-injury, the proportion of neurons with delayed membrane perturbation was signiﬁcantly different from the proportion of resealed neurons as well as the proportion of those neurons with enduring membrane perturbation. The change in the proportion of neurons with resealed and enduring membrane perturbation over time was not signiﬁcant. This redistribution of membrane perturbation types between 4 and 8 h post-injury may be the result of resealed neurons

reopening their membranes to now manifest enduring damage and/or additional neurons suffering delayed membrane perturbation (Fig. 4) (Farkas et al., 2006). Collectively, these in vivo studies clearly conﬁrm the existence of DTBI-injured neuronal membrane perturbation, while further illustrating the complex pathobiology associated with DTBI. This complexity was further highlighted by recent in vitro studies that explored the potential relationship of the above described altered membrane permeability to CMSP positing that CMSP co-existed with increased membrane permeability (Pike et al., 2000; Liu and Schnellmann, 2003; Liu et al., 2004). In contrast, with in vivo DTBI, the majority of neurons sustaining either membrane disruption and/or demonstrating necrotic change did not reveal CMSP, suggesting that membrane disruption itself did not lead to calpain activation. Conversely, with DTBI, neurons demonstrating CMSP did not reveal membrane disruption, and, as the majority of these CMSP-positive neurons revealed only limited ultrastructural damage, this suggested that CMSP was not associated with neuronal death after DTBI (Fig. 5) (Farkas et al., 2006). Taken together with other studies (Saatman et al., 1996; Brana et al., 1999), these ﬁndings emphasize the need for caution in interpreting the occurrence of CMSP and its overall implications for the neuronal injury and death associated with DTBI.

Axonal damage associated with DTBI In addition to diffuse neuronal perturbation and death described above, diffuse axonal injury is also a distinguishing feature of DTBI, occurring across the spectrum of brain injury ranging from mild through severe. Historically, the histological identiﬁcation of diffuse axonal injury was based upon the use of silver salts to detect, within the ﬁrst days of injury, grossly swollen axonal bulbs that lacked continuity with their downstream axonal partners (Strich, 1956). This suggested their mechanical transsection resulting in axonal retraction and axoplasmic pooling at the site of disconnection (Strich, 1956; Adams, 1982). More contemporary studies, however, have shown that in large part, this premise of transsection and retraction

50

Fig. 1. This ﬁgure illustrates the phenomenon of mechanically induced neuronal membrane disruption and its consequences for the neuron. Via double labeled confocal images, panels A–C show neurons ﬂooding with both dextrans with evidence of concomitant cellular injury, reﬂected in their irregular, distorted proﬁles and vacuolization (arrows). In those cells showing the most severe damage, note that the dextrans are also typically found within the nucleus (arrowhead). Note that other double-labeled neurons (D) demonstrate little or no pathological damage and that despite homogenous tracer uptake no nuclear accumulation or vacuolization occurs (arrows). Scale bars: 100 mm. Panel E illustrates three tracer ﬂooded neurons, conﬁrmed by routine ﬂuorescent microscopy and followed via EM. The most severely damaged neuron (asterisk) demonstrates increased electron density, organelle vacuolization (arrows) and perisomatic glial ensheathment (arrowheads). The two other cells (double and triple asterisks) demonstrate little or no pathological change. Note that the surrounding neuropil demonstrates little overt pathologic change consistent with the confocal observations. Scale bar: 5 mm.

is not correct. Rather, it has been shown in multiple animal studies, as well as limited human investigations, that the forces of injury diffusely alter focal axonal segments. This results in a local

impairment of axonal transport, with progressive local axonal swelling followed by detachment over a post-traumatic course ranging from several hours up to a day (Povlishock and Jenkins, 1995). Given

51

Fig. 2. This ﬁgure illustrates the ﬁnding that although mechanoporation can occur, some neurons reseal their perturbated membranes with increased post-traumatic survival. Here confocal images of pre-injury infused dextrans (A), post-injury-infused dextrans (B) and their overlay (C) demonstrate some cortical neurons ﬂooding with the pre-injury infused dextran alone without concomitant ﬂooding with the post-injury administrated tracer (arrows), all of which is suggestive of cell membrane closure/recovery. Note that one doubleﬂooded neuron demonstrating severe damage (arrowhead) and one double-labeled neuron demonstrating less severe pathology (double arrowhead) are also shown. Scale bar: 50 mm. Panel D illustrates a routine ﬂuorescent image that reveals a single labeled cell. In panel E, the same neuron is visualized through the use of antibodies to the ﬂuorophore and then carried to the EM level (F–G). Note that neurons ﬂooding with the pre-injury dextran alone do not show overt pathological damage. Immunoreactive products (anti-Alexa Fluor IR) are labeled with arrows in panel G, which is an enlargement of the area, blocked out in panel F. Scale bar 2 mm. (Adapted with permission from Farkas et al. (2006); Copyright 2006 by the Society for Neuroscience).

the fact that these reactive axonal changes were found in scattered axons related to other intact axons and their vascular elements, this precluded the potential for direct mechanical renting, suggesting that more subtle intraaxonal changes were at work in the pathogenesis of this progressive axonal change leading to disconnection.

Cellular and subcellular factors related to the initiating pathogenesis of DTBI-associated axonal injury Appreciating that the above described cascades of axonal perturbation leading to impaired transport and disconnection evolved over a relatively long

52

Fig. 3. In contrast to the phenomenon of resealing shown in Fig. 2, this ﬁgure provides evidence of delayed neuronal plasmalemmal opening. Here confocal images of pre-injury infused dextran (A), post-injury-infused dextran (B) and their overlay (C) demonstrate scattered neurons ﬂooding with the post-injury infused dextran alone (arrows), among with double-ﬂooded neurons (arrowheads). Scale bar: 100 mm. (Adapted with permission from Farkas et al. (2006); Copyright 2006 by the Society for Neuroscience).

post-traumatic course, emphasis has been placed upon identifying the initiating intraaxonal cellular and subcellular factors both to understand better the pathobiology of this axonal injury and to develop therapies targeting these cellular and subcellular changes. While immediate physical transsection of the axon cylinder has been ruled out, the potential for focal disturbances in the axolemma leading to local ionic dysregulation was evaluated in the experimental setting using extracellular tracers normally excluded by the intact/unaltered axolemma. Through this approach, our laboratory showed that discreet axonal foci scattered throughout the injured brain revealed evidence of altered focal axolemmal permeability to various normally excluded extracellular tracers (Pettus and Povlishock, 1996; Povlishock and Pettus, 1996). Moving on the premise that this alteration in axolemmal permeability would be accompanied by local calcium dysregulation, we probed the same axonal segments with antibodies targeting calcium-activated, CMSP. In these studies, CMSP was observed initially in the subaxolemmal domain followed by its activation in the axon’s core, with a particular predilection for the mitochondria that appeared swollen with disrupted cristae (Buki et al., 1999). Because such mitochondrial damage appeared consistent with local calcium overloading and caspase activation,

our laboratory probed these same segments with antibodies targeting cytochrome C and caspasemediated spectrin proteolysis (Buki et al., 2000). Via this approach, we demonstrated that the damaged mitochondria released cytochrome C that, in turn, activated caspase-mediated spectrin degradation (Buki et al., 2000; Buki and Povlishock, 2006). Collectively, both cysteine proteases, calpain and caspase were recognized to participate in the degradation of the axonal cytoskeleton associated with concomitant neuroﬁlament side-arm cleavage, neuroﬁlament compaction and microtubular loss (Buki and Povlishock, 2006). Because of these local cytoskeletal abnormalities, these compacted axonal segments could also be routinely identiﬁed by use of antibodies targeting altered neuroﬁlament subunits (RMO14) that, in our hands, provided a surrogate marker for identifying such sites of injury (Povlishock et al., 1997). Based on these ﬁndings in our lab, as well as others, it was assumed that these intraaxonal changes universally led to an upstream impairment of axonal transport leading to swelling and disconnection (Maxwell et al., 1997). However, continued investigation failed to establish a routine correlation between these two axonal events, leading us to question whether neuroﬁlament compaction and axonal swelling, consistently occurred within the same axonal segment. More recent studies, using multiple strategies to qualitatively and

53

Fig. 4. This ﬁgure provides a semiquantitative, graphic representation of those different forms and prognosis of membrane perturbation following DTBI. Here the distribution of neurons with resealed (gray column), enduring (white column) or delayed membrane perturbation (black column) are shown after DTBIinduced membrane perturbation. As membranes remain closed, open or reseal in response to injury, tracer availability determines the type of membrane perturbation assigned categorically to each neuron. The pre-injury administration of Alexa Fluor 488 and post-injury administration of Texas Red conjugated dextrans permit the evaluation of neuronal membrane perturbation distribution at 4 and 8 h after injury. Across time points, the proportion of neurons with delayed membrane perturbation is signiﬁcantly different from the proportion of resealed neurons and the proportion of enduring membrane perturbation (w2, po0.016). The change in the proportion of neurons with resealed and enduring membrane perturbation over time was not signiﬁcant (w2, p ¼ 0.19). The redistribution of the types of membrane perturbation between 4 and 8 h post-injury may result from resealed neurons reopening (gray arrow) to become neurons with enduring damage and/or additional neurons (black arrow) suffering delayed membrane perturbation.

quantitatively assess axonal numbers as well as the spatial relationship of axons showing neuroﬁlament compaction and impaired transport have now convincingly shown that the above described sites of altered axonal permeability, cysteine protease activation and cytoskeletal collapse do not routinely correlate with sites of impaired axonal transport (Stone et al., 2004; Marmarou et al., 2005). This suggested that the neuroﬁlament-compacted axonal segments and swollen axonal segments demonstrating impaired axonal transport most likely represent two different populations of injured axons responding to the traumatic episode in different fashions. For those axonal segments showing

focally altered axolemmal permeability it appears, as noted previously, that local calcium dysregulation with the activation of the cysteine proteases, causes local degradation of the axonal cytoskeleton, leading to axonal failure and disconnection (Stone et al., 2004; Marmarou et al., 2005). Why these sites of axonal injury do not reveal impaired axonal transport and axonal swelling is unclear, yet, it is conceivable that the suprathreshold calcium uptake occurring at these sites most likely converts anterograde to retrograde transport, thereby precluding the development of reactive axonal swelling (Sahenk and Lasek, 1988; Martz et al., 1989). The validity of the above is supported by the use of various therapeutic strategies targeting calpain inhibition that, as such, signiﬁcantly reduce the numbers of axonal proﬁles showing the above described cysteine protease activation and cytoskeletal collapse (Buki et al., 2003; Buki and Povlishock, 2006). In contrast, it appears that those axons showing impaired axonal transport and local swelling do not sustain any alteration in local axolemmal permeability or any activation of the cysteine proteases (Povlishock and Stone, 2001). Rather, it is posited that other mechanisms are at work and that these are linked to more subtle forms of calcium dysregulation. These potentially involve the activation of micromolar calpains to trigger the activation of calcineurin that, in turn, alters the microtubular network to disrupt local axonal transport kinetics and thereby elicit the swellings described above (Povlishock and Stone, 2001). Although limited direct evidence exists to support this pathway, the use of calcineurin antagonists, such as FK506, directly attenuate the numbers of axons showing impaired axonal transport and swelling while having no effect upon those axons showing neuroﬁlament compaction and disconnection (Marmarou and Povlishock, 2006). This supports the premise that calcineurin is integral to the pathogenesis of impaired axonal transport and swelling. Collectively, these studies illustrate the complexity of the pathogenesis of diffuse axonal injury, suggesting at least two differing types of initiating mechanisms, with the caveat that both populations of injured axons will not likely be amenable to one form of therapeutic intervention.

54

55

Other potential cellular and subcellular factors related to diffuse axonal injury In the above passages, we have focused upon the contemporary appreciation of the cellular and subcellular changes involved in diffuse axonal injury. Importantly, all these changes were believed to be ongoing only in myelinated nerve ﬁber populations with no involvement of unmyelinated axons that, in general, have received virtually no consideration either in the context of TBI (Jafari et al., 1998) or, for that matter, any other CNS disorder. Recently, however, this perception was changed by work conducted by Reeves et al. (2005) who, through the use of electrophysiological methods, provided compelling evidence for unmyelinated nerve ﬁber damage and dysfunction within the corpus callosum of traumatically brain-injured animals. Using an analysis of compound action potentials by examining two speciﬁc wave forms that can be independently related to either myelinated or unmyelinated axons populations, Reeves and colleagues showed signiﬁcant and sustained depression of the compound action potentials associated with the unmyelinated axon population. These electrophysiological studies were accompanied by routine morphological analysis using electron microscopy. Although these studies were preliminary, they suggested that the morphological progression of unmyelinated ﬁber change observed was quite dissimilar from that described above for myelinated axons. This potential difference is also partially supported by excellent in vitro studies that have examined non-myelinated neurites subjected to mechanical strains that ultimately led to local axonal beading and disconnection (Wolf et al., 2001; Iwata et al., 2004). In this pathology, however, no evidence of overt axolemmal change or

axolemmal disruption was discerned. Rather, the depolarization associated with this injury evoked sodium inﬂux, the activation of voltage gated calcium channels and the concomitant activation of sodium/calcium exchangers, all of which contributed to local intraaxonal calcium overloading (Wolf et al., 2001). These calcium-mediated changes, comparable with some of the changes described in myelinated axons, were linked with the activation of proteases. These, in turn, contributed to subsequent proteolysis of its NaCh subunit to promote a persistent elevation in intracellular calcium, fueling additional pathological changes through many of the pathways addressed above (Iwata et al., 2004). While these changes obviously remain to be conﬁrmed in vivo, they are intriguing and speak to yet another, different form of cellular and subcellular change that, in this case, involves a potential channelopathy as major player in the ensuing unmyelinated axonal perturbation.

Concluding comments In this brief review, we have attempted to critically evaluate current thought on the cellular and subcellular change evoked by DTBI at both the neuronal somatic and axonal fronts. Recognizing that our appreciation of cellular and subcellular change in the context of TBI has been framed primarily by our understanding of the pathobiology of contusional injury, this review cautions against moving on the assumption that these contusional cellular and subcellular cascades must transfer to those neuronal cell bodies and their appendages that sustain diffuse injury. Basic differences between focal and diffuse injury give rise to the possibility

Fig. 5. This ﬁgure illustrates the relation between CMSP and tracer uptake. Panel A shows triple labeled confocal image demonstrating dextran ﬂooded neurons as well as CMSP immunopositive neurons. Note that at both 4 and 8 h post-injury CMSP immunoreactive neurons were observed throughout the neocortex (arrows). Scale bar: 100 mm. In panels B–E confocal images of pre-injury dextran ﬂooding (B), post-injury dextran ﬂooding (C), CMSP immunopositivity (D) and their overlay (E) demonstrate neurons showing enduring membrane disruption reﬂected in their content of both tracers. Note that some neurons colocalize with CMSPimmunopositivity (arrow), whereas other neurons demonstrate tracer ﬂooding without CMSP (arrowheads). Also note that at these same time points, several CMSP immunoreactive neurons also could be identiﬁed without concomitant tracer ﬂooding (double arrowhead). Lastly note that in the same region, a neuron demonstrating only the initial tracer ﬂooding (big arrow), as well as a neuron revealing only secondary tracer ﬂooding (triple arrowhead) can also be seen. Scale bar: 50 mm. (Adapted with permission from Farkas et al. (2006); Copyright 2006 by the Society for Neuroscience).

56

for the involvement of other cell perturbation/ death cascades that may not be operant with focal injury. Speciﬁcally, the pathogenesis of diffuse injury highlights the potential for mechanically induced membrane perturbation to serve as a more signiﬁcant player in the ensuing cellular and subcellular change. The potential for transient versus more enduring mechanically induced membrane perturbation may also inﬂuence subsequent neuronal somatic and axonal perturbation and/or death. Additionally, the activation of other cell death cascades including, but not limited to, activation of lysosomal-mediated change through the release of hydrolytic enzymes must now be considered. Collectively, this review illustrates the complexity of the pathobiology of DTBI and the danger in assuming that a singular mechanistic view of cellular and subcellular change can translate to the diffusely injured brain.

Abbreviations CMSP DTBI SBDP TBI

calpain-mediated spectrin proteolysis diffuse traumatic brain injury spectrin breakdown product traumatic brain injury

Acknowledgments This work was supported in part by NIH grants NS045824, NS047463 and HD055813.

References Adams, J.H. (1982) Diffuse axonal injury in non-missile head injury. Injury, 13(5): 444–445. Adams, J.H., Graham, D.I., Scott, G., Parker, L.S. and Doyle, D. (1980) Brain damage in fatal non-missile head injury. J. Clin. Pathol., 33(12): 1132–1145. Beer, R., Franz, G., Srinivasan, A., Hayes, R.L., Pike, B.R., Newcomb, J.K., Zhao, X., Schmutzhard, E., Poewe, W. and Kampﬂ, A. (2000) Temporal proﬁle and cell subtype distribution of activated caspase-3 following experimental traumatic brain injury. J. Neurochem., 75(3): 1264–1273. Brana, C., Benham, C.D. and Sundstrom, L.E. (1999) Calpain activation and inhibition in organotypic rat hippocampal

slice cultures deprived of oxygen and glucose. Eur. J. Neurosci., 11(7): 2375–2384. Braughler, J.M. and Hall, E.D. (1992) Involvement of lipid peroxidation in CNS injury. J. Neurotrauma, 9(Suppl 1): S1–S7. Brunk, U.T., Dalen, H., Roberg, K. and Hellquist, H.B. (1997) Photo-oxidative disruption of lysosomal membranes causes apoptosis of cultured human ﬁbroblasts. Free Radic. Biol. Med., 23(4): 616–626. Buki, A., Farkas, O., Doczi, T. and Povlishock, J.T. (2003) Preinjury administration of the calpain inhibitor MDL-28170 attenuates traumatically induced axonal injury. J. Neurotrauma, 20(3): 261–268. Buki, A., Okonkwo, D.O., Wang, K.K. and Povlishock, J.T. (2000) Cytochrome c release and caspase activation in traumatic axonal injury. J. Neurosci., 20(8): 2825–2834. Buki, A. and Povlishock, J.T. (2006) All roads lead to disconnection? Traumatic axonal injury revisited. Acta Neurochir. (Wien.), 148(2): 181–193. Buki, A., Siman, R., Trojanowski, J.Q. and Povlishock, J.T. (1999) The role of calpain-mediated spectrin proteolysis in traumatically induced axonal injury. J. Neuropathol. Exp. Neurol., 58(4): 365–375. Cernak, I., Chapman, S., Hamlin, G. and Vink, R. (2002) Temporal characterisation of pro- and anti-apoptotic mechanisms following diffuse traumatic brain injury in rats. J. Clin. Neurosci., 9(5): 565–572. Cernak, I., Vink, R., Zapple, D.N., Cruz, M.I., Ahmed, F., Chang, T., Fricke, S.T. and Faden, A.I. (2004) The pathobiology of moderate diffuse traumatic brain injury as identiﬁed using a new experimental model of injury in rats. Neurobiol. Dis., 17(1): 29–43. Cervos-Navarro, J. and Lafuente, J.V. (1991) Traumatic brain injuries: structural changes. J. Neurol. Sci., 103(Suppl): S3–S14. Clark, R.S., Kochanek, P.M., Chen, M., Watkins, S.C., Marion, D.W., Chen, J., Hamilton, R.L., Loeffert, J.E. and Graham, S.H. (1999) Increases in Bcl-2 and cleavage of caspase-1 and caspase-3 in human brain after head injury. FASEB J., 13(8): 813–821. Clark, R.S., Kochanek, P.M., Watkins, S.C., Chen, M., Dixon, C.E., Seidberg, N.A., Melick, J., Loeffert, J.E., Nathaniel, P.D., Jin, K.L. and Graham, S.H. (2000) Caspase-3 mediated neuronal death after traumatic brain injury in rats. J. Neurochem., 74(2): 740–753. Colicos, M.A. and Dash, P.K. (1996) Apoptotic morphology of dentate gyrus granule cells following experimental cortical impact injury in rats: possible role in spatial memory deﬁcits. Brain Res., 739(1–2): 120–131. Colicos, M.A., Dixon, C.E. and Dash, P.K. (1996) Delayed, selective neuronal death following experimental cortical impact injury in rats: possible role in memory deﬁcits. Brain Res., 739(1–2): 111–119. Conti, A.C., Raghupathi, R., Trojanowski, J.Q. and McIntosh, T.K. (1998) Experimental brain injury induces regionally distinct apoptosis during the acute and delayed post-traumatic period. J. Neurosci., 18(15): 5663–5672.

57 Cortez, S.C., McIntosh, T.K. and Noble, L.J. (1989) Experimental ﬂuid percussion brain injury: vascular disruption and neuronal and glial alterations. Brain Res., 482(2): 271–282. Dietrich, W.D., Alonso, O., Busto, R., Globus, M.Y. and Ginsberg, M.D. (1994a) Post-traumatic brain hypothermia reduces histopathological damage following concussive brain injury in the rat. Acta Neuropathol. (Berl.), 87(3): 250–258. Dietrich, W.D., Alonso, O. and Halley, M. (1994b) Early microvascular and neuronal consequences of traumatic brain injury: a light and electron microscopic study in rats. J. Neurotrauma, 11(3): 289–301. Eldadah, B.A. and Faden, A.I. (2000) Caspase pathways, neuronal apoptosis, and CNS injury. J. Neurotrauma, 17(10): 811–829. Faden, A.I., Demediuk, P., Panter, S.S. and Vink, R. (1989) The role of excitatory amino acids and NMDA receptors in traumatic brain injury. Science, 244(4906): 798–800. Farkas, O., Lifshitz, J. and Povlishock, J.T. (2006) Mechanoporation induced by diffuse traumatic brain injury: an irreversible or reversible response to injury? J. Neurosci., 26(12): 3130–3140. Farkas, O., Polgar, B., Szekeres-Bartho, J., Doczi, T., Povlishock, J.T. and Buki, A. (2005) Spectrin breakdown products in the cerebrospinal ﬂuid in severe head injury: preliminary observations. Acta Neurochir. (Wien.), 147(8): 855–861. Fei, Z., Zhang, X., Jiang, X.F., Huang, W.D. and Bai, H.M. (2005) Altered expression patterns of metabotropic glutamate receptors in diffuse brain injury. Neurosci. Lett., 380(3): 280–283. Felderhoff-Mueser, U., Sifringer, M., Pesditschek, S., Kuckuck, H., Moysich, A., Bittigau, P. and Ikonomidou, C. (2002) Pathways leading to apoptotic neurodegeneration following trauma to the developing rat brain. Neurobiol. Dis., 11(2): 231–245. Formigli, L., Papucci, L., Tani, A., Schiavone, N., Tempestini, A., Orlandini, G.E., Capaccioli, S. and Orlandini, S.Z. (2000) Aponecrosis: morphological and biochemical exploration of a syncretic process of cell death sharing apoptosis and necrosis. J. Cell Physiol., 182(1): 41–49. Fox, G.B., Fan, L., LeVasseur, R.A. and Faden, A.I. (1998) Sustained sensory/motor and cognitive deﬁcits with neuronal apoptosis following controlled cortical impact brain injury in the mouse. J. Neurotrauma, 15(8): 599–614. Geddes, D.M., Cargill, R.S. and LaPlaca, M.C. (2003) Mechanical stretch to neurons results in a strain rate and magnitude-dependent increase in plasma membrane permeability. J. Neurotrauma, 20(10): 1039–1049. Goda, M., Isono, M., Fujiki, M. and Kobayashi, H. (2002) Both MK801 and NBQX reduce the neuronal damage after impactacceleration brain injury. J. Neurotrauma, 19(11): 1445–1456. Hicks, R., Soares, H., Smith, D. and McIntosh, T. (1996) Temporal and spatial characterization of neuronal injury following lateral ﬂuid-percussion brain injury in the rat. Acta Neuropathol. (Berl.), 91(3): 236–246. Huang, Y. and Wang, K.K. (2001) The calpain family and human disease. Trends Mol. Med., 7(8): 355–362.

Iwata, A., Stys, P.K., Wolf, J.A., Chen, X.H., Taylor, A.G., Meaney, D.F. and Smith, D.H. (2004) Traumatic axonal injury induces proteolytic cleavage of the voltage-gated sodium channels modulated by tetrodotoxin and protease inhibitors. J. Neurosci., 24(19): 4605–4613. Jafari, S.S., Nielson, M., Graham, D.I. and Maxwell, W.L. (1998) Axonal cytoskeletal changes after nondisruptive axonal injury. II. Intermediate sized axons. J. Neurotrauma, 15(11): 955–966. Kampﬂ, A., Posmantur, R., Nixon, R., Grynspan, F., Zhao, X., Liu, S.J., Newcomb, J.K., Clifton, G.L. and Hayes, R.L. (1996) mu-Calpain activation and calpain-mediated cytoskeletal proteolysis following traumatic brain injury. J. Neurochem., 67(4): 1575–1583. Katayama, Y., Becker, D.P., Tamura, T. and Hovda, D.A. (1990) Massive increases in extracellular potassium and the indiscriminate release of glutamate following concussive brain injury. J. Neurosurg., 73(6): 889–900. Keane, R.W., Kraydieh, S., Lotocki, G., Alonso, O.F., Aldana, P. and Dietrich, W.D. (2001) Apoptotic and antiapoptotic mechanisms after traumatic brain injury. J. Cereb. Blood Flow Metab., 21(10): 1189–1198. Kermer, P., Klocker, N. and Bahr, M. (1999) Neuronal death after brain injury. Models, mechanisms, and therapeutic strategies in vivo. Cell Tissue Res., 298(3): 383–395. Kitanaka, C. and Kuchino, Y. (1999) Caspase-independent programmed cell death with necrotic morphology. Cell Death Differ., 6(6): 508–515. Kontos, H.A. and Povlishock, J.T. (1986) Oxygen radicals in brain injury. Cent. Nerv. Syst. Trauma, 3(4): 257–263. Kotapka, M.J., Gennarelli, T.A., Graham, D.I., Adams, J.H., Thibault, L.E., Ross, D.T. and Ford, I. (1991) Selective vulnerability of hippocampal neurons in acceleration-induced experimental head injury. J. Neurotrauma, 8(4): 247–258. Kotapka, M.J., Graham, D.I., Adams, J.H., Doyle, D. and Gennarelli, T.A. (1993) Hippocampal damage in fatal paediatric head injury. Neuropathol. Appl. Neurobiol., 19(2): 128–133. Kotapka, M.J., Graham, D.I., Adams, J.H. and Gennarelli, T.A. (1992) Hippocampal pathology in fatal non-missile human head injury. Acta Neuropathol. (Berl.), 83(5): 530–534. Kroemer, G. and Jaattela, M. (2005) Lysosomes and autophagy in cell death control. Nat. Rev. Cancer, 5(11): 886–897. Kupina, N.C., Detloff, M.R., Dutta, S. and Hall, E.D. (2002) Neuroimmunophilin ligand V-10,367 is neuroprotective after 24-hour delayed administration in a mouse model of diffuse traumatic brain injury. J. Cereb. Blood Flow Metab., 22(10): 1212–1221. Kupina, N.C., Nath, R., Bernath, E.E., Inoue, J., Mitsuyoshi, A., Yuen, P.W., Wang, K.K. and Hall, E.D. (2001) The novel calpain inhibitor SJA6017 improves functional outcome after delayed administration in a mouse model of diffuse brain injury. J. Neurotrauma, 18(11): 1229–1240. Lin, X., Zhi, D. and Zhang, S. (2001) Inhibiting effect of moderate hypothermia on cell apoptosis after diffuse brain injury in rats. Chin. J. Traumatol., 4(1): 14–19.

58 Liu, X. and Schnellmann, R.G. (2003) Calpain mediates progressive plasma membrane permeability and proteolysis of cytoskeleton-associated paxillin, talin, and vinculin during renal cell death. J. Pharmacol. Exp. Ther., 304(1): 63–70. Liu, X., Van Vleet, T. and Schnellmann, R.G. (2004) The role of calpain in oncotic cell death. Annu. Rev. Pharmacol. Toxicol., 44: 349–370. Marmarou, C.R. and Povlishock, J.T. (2006) Administration of the immunophilin ligand FK506 differentially attenuates neuroﬁlament compaction and impaired axonal transport in injured axons following diffuse traumatic brain injury. Exp. Neurol., 197(2): 353–362. Marmarou, C.R., Walker, S.A., Davis, C.L. and Povlishock, J.T. (2005) Quantitative analysis of the relationship between intra-axonal neuroﬁlament compaction and impaired axonal transport following diffuse traumatic brain injury. J. Neurotrauma, 22(10): 1066–1080. Martz, D., Garner, J. and Lasek, R.J. (1989) Protein changes during anterograde-to-retrograde conversion of axonally transported vesicles. Brain Res., 476(1): 199–203. Maxwell, W.L., Povlishock, J.T. and Graham, D.L. (1997) A mechanistic analysis of nondisruptive axonal injury: a review. J. Neurotrauma, 14(7): 419–440. McCracken, E., Hunter, A.J., Patel, S., Graham, D.I. and Dewar, D. (1999) Calpain activation and cytoskeletal protein breakdown in the corpus callosum of head-injured patients. J. Neurotrauma, 16(9): 749–761. Merry, D.E. and Korsmeyer, S.J. (1997) Bcl-2 gene family in the nervous system. Annu. Rev. Neurosci., 20: 245–267. Newcomb, J.K., Zhao, X., Pike, B.R. and Hayes, R.L. (1999) Temporal proﬁle of apoptotic-like changes in neurons and astrocytes following controlled cortical impact injury in the rat. Exp. Neurol., 158(1): 76–88. Nicotera, P., Leist, M. and Manzo, L. (1999) Neuronal cell death: a demise with different shapes. Trends Pharmacol. Sci., 20(2): 46–51. Palmer, A.M., Marion, D.W., Botscheller, M.L., Swedlow, P.E., Styren, S.D. and DeKosky, S.T. (1993) Traumatic brain injury-induced excitotoxicity assessed in a controlled cortical impact model. J. Neurochem., 61(6): 2015–2024. Pettus, E.H. and Povlishock, J.T. (1996) Characterization of a distinct set of intra-axonal ultrastructural changes associated with traumatically induced alteration in axolemmal permeability. Brain Res., 722(1–2): 1–11. Pike, B.R., Flint, J., Dutta, S., Johnson, E., Wang, K.K. and Hayes, R.L. (2001) Accumulation of non-erythroid alpha II-spectrin and calpain-cleaved alpha II-spectrin breakdown products in cerebrospinal ﬂuid after traumatic brain injury in rats. J. Neurochem., 78(6): 1297–1306. Pike, B.R., Zhao, X., Newcomb, J.K., Glenn, C.C., Anderson, D.K. and Hayes, R.L. (2000) Stretch injury causes calpain and caspase-3 activation and necrotic and apoptotic cell death in septo-hippocampal cell cultures. J. Neurotrauma, 17(4): 283–298. Pike, B.R., Zhao, X., Newcomb, J.K., Posmantur, R.M., Wang, K.K. and Hayes, R.L. (1998) Regional calpain and

caspase-3 proteolysis of alpha-spectrin after traumatic brain injury. Neuroreport, 9(11): 2437–2442. Pineda, J.A., Wang, K.K. and Hayes, R.L. (2004) Biomarkers of proteolytic damage following traumatic brain injury. Brain Pathol., 14(2): 202–209. Povlishock, J.T. and Jenkins, L.W. (1995) Are the pathobiological changes evoked by traumatic brain injury immediate and irreversible? Brain Pathol., 5(4): 415–426. Povlishock, J.T. and Katz, D.I. (2005) Update of neuropathology and neurological recovery after traumatic brain injury. J. Head Trauma Rehabil., 20(1): 76–94. Povlishock, J.T. and Kontos, H.A. (1992) The role of oxygen radicals in the pathobiology of traumatic brain injury. Hum. Cell, 5(4): 345–353. Povlishock, J.T., Marmarou, A., McIntosh, T., Trojanowski, J.Q. and Moroi, J. (1997) Impact acceleration injury in the rat: evidence for focal axolemmal change and related neuroﬁlament sidearm alteration. J. Neuropathol. Exp. Neurol., 56(4): 347–359. Povlishock, J.T. and Pettus, E.H. (1996) Traumatically induced axonal damage: evidence for enduring changes in axolemmal permeability with associated cytoskeletal change. Acta Neurochir. Suppl. (Wien.), 66: 81–86. Povlishock, J.T. and Stone, J.R. (2001) Traumatic axonal injury. In: Miller L.P., Hayes R.L. and Newcomb J.K. (Eds.), Head Trauma: Basic, Preclinical and Clinical Directions. Wiley, pp. 281–302. Prado, G.R., Ross, J.D., DeWeerth, S.P. and LaPlaca, M.C. (2005) Mechanical trauma induces immediate changes in neuronal network activity. J. Neural Eng., 2(4): 148–158. Prado, G., Caltagirone, C.L. and LaPlaca, M.C. (2004) Neuronal plasma membrane disruption in the traumatically injured brain. J. Neurotrauma, 21: 1331. Prado, G. and LaPlaca, M.C. (2004) Neuronal plasma membrane is transiently disrupted by mechanical trauma: an insight into the mechanisms involved. J. Neurotrauma, 21: 1331. Pravdenkova, S.V., Basnakian, A.G., James, S.J. and Andersen, B.J. (1996) DNA fragmentation and nuclear endonuclease activity in rat brain after severe closed head injury. Brain Res., 729(2): 151–155. Raghupathi, R. (2004) Cell death mechanisms following traumatic brain injury. Brain Pathol., 14(2): 215–222. Raghupathi, R., Conti, A.C., Graham, D.I., Krajewski, S., Reed, J.C., Grady, M.S., Trojanowski, J.Q. and McIntosh, T.K. (2002) Mild traumatic brain injury induces apoptotic cell death in the cortex that is preceded by decreases in cellular Bcl-2 immunoreactivity. Neuroscience, 110(4): 605–616. Raghupathi, R., Graham, D.I. and McIntosh, T.K. (2000) Apoptosis after traumatic brain injury. J. Neurotrauma, 17(10): 927–938. Reeves, T.M., Phillips, L.L. and Povlishock, J.T. (2005) Myelinated and unmyelinated axons of the corpus callosum differ in vulnerability and functional recovery following traumatic brain injury. Exp. Neurol., 196(1): 126–137.

59 Rink, A., Fung, K.M., Trojanowski, J.Q., Lee, V.M., Neugebauer, E. and McIntosh, T.K. (1995) Evidence of apoptotic cell death after experimental traumatic brain injury in the rat. Am. J. Pathol., 147(6): 1575–1583. Ross, D.T., Graham, D.I. and Adams, J.H. (1993) Selective loss of neurons from the thalamic reticular nucleus following severe human head injury. J. Neurotrauma, 10(2): 151–165. Runnerstam, M., Bao, F., Huang, Y., Shi, J., Gutierrez, E., Hamberger, A., Hansson, H.A., Viano, D. and Haglid, K. (2001) A new model for diffuse brain injury by rotational acceleration: II. Effects on extracellular glutamate, intracranial pressure, and neuronal apoptosis. J. Neurotrauma, 18(3): 259–273. Saatman, K.E., Bozyczko-Coyne, D., Marcy, V., Siman, R. and McIntosh, T.K. (1996) Prolonged calpain-mediated spectrin breakdown occurs regionally following experimental brain injury in the rat. J. Neuropathol. Exp. Neurol., 55(7): 850–860. Sahenk, Z. and Lasek, R.J. (1988) Inhibition of proteolysis blocks anterograde-retrograde conversion of axonally transported vesicles. Brain Res., 460(1): 199–203. Singleton, R.H. and Povlishock, J.T. (2004) Identiﬁcation and characterization of heterogeneous neuronal injury and death in regions of diffuse brain injury: evidence for multiple independent injury phenotypes. J. Neurosci., 24(14): 3543–3553. Singleton, R.H., Zhu, J., Stone, J.R. and Povlishock, J.T. (2002) Traumatically induced axotomy adjacent to the soma does not result in acute neuronal death. J. Neurosci., 22(3): 791–802. Smith, D.H., Chen, X.H., Xu, B.N., McIntosh, T.K., Gennarelli, T.A. and Meaney, D.F. (1997) Characterization of diffuse axonal pathology and selective hippocampal damage following inertial brain trauma in the pig. J. Neuropathol. Exp. Neurol., 56(7): 822–834. Stone, J.R., Okonkwo, D.O., Dialo, A.O., Rubin, D.G., Mutlu, L.K., Povlishock, J.T. and Helm, G.A. (2004) Impaired axonal transport and altered axolemmal permeability occur in distinct populations of damaged axons following traumatic brain injury. Exp. Neurol., 190(1): 59–69. Strich, S.J. (1956) Diffuse degeneration of the cerebral white matter in severe dementia following head injury. J. Neurol. Neurosurg. Psychiatry, 19: 163–185.

Van den Heuvel, C., Blumbergs, P.C., Finnie, J.W., Manavis, J., Jones, N.R., Reilly, P.L. and Pereira, R.A. (1999) Upregulation of amyloid precursor protein messenger RNA in response to traumatic brain injury: an ovine head impact model. Exp. Neurol., 159(2): 441–450. Van den Heuvel, C., Lewis, S., Wong, M., Manavis, J., Finnie, J., Blumbergs, P.C., Jones, N. and Reilly, P. (1998) Diffuse neuronal perikaryon amyloid precursor protein immunoreactivity in a focal head impact model. Acta Neurochir. Suppl. (Wien.), 71: 209–211. Wang, K.K. (2000) Calpain and caspase: can you tell the difference? Trends Neurosci., 23(1): 20–26. Wolf, J.A., Stys, P.K., Lusardi, T., Meaney, D. and Smith, D.H. (2001) Traumatic axonal injury induces calcium inﬂux modulated by tetrodotoxin-sensitive sodium channels. J. Neurosci., 21(6): 1923–1930. Yakovlev, A.G. and Faden, A.I. (2004) Mechanisms of neural cell death: implications for development of neuroprotective treatment strategies. NeuroRx, 1(1): 5–16. Yakovlev, A.G., Ota, K., Wang, G., Movsesyan, V., Bao, W.L., Yoshihara, K. and Faden, A.I. (2001) Differential expression of apoptotic protease-activating factor-1 and caspase-3 genes and susceptibility to apoptosis during brain development and after traumatic brain injury. J. Neurosci., 21(19): 7439–7446. Yamashima, T. (2000) Implication of cysteine proteases calpain, cathepsin and caspase in ischemic neuronal death of primates. Prog. Neurobiol., 62(3): 273–295. Yamashima, T. (2004) Ca2+-dependent proteases in ischemic neuronal death: a conserved ‘calpain-cathepsin cascade’ from nematodes to primates. Cell Calcium, 36(3–4): 285–293. Yamashima, T., Tonchev, A.B., Tsukada, T., Saido, T.C., Imajoh-Ohmi, S., Momoi, T. and Kominami, E. (2003) Sustained calpain activation associated with lysosomal rupture executes necrosis of the postischemic CA1 neurons in primates. Hippocampus, 13(7): 791–800. Yang, J., Liu, X., Bhalla, K., Kim, C.N., Ibrado, A.M., Cai, J., Peng, T.I., Jones, D.P. and Wang, X. (1997) Prevention of apoptosis by Bcl-2: release of cytochrome c from mitochondria blocked. Science, 275(5303): 1129–1132. Zipfel, G.J., Babcock, D.J., Lee, J.M. and Choi, D.W. (2000) Neuronal apoptosis after CNS injury: the roles of glutamate and calcium. J. Neurotrauma, 17(10): 857–869.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 5

Astroglia: Important mediators of traumatic brain injury Candace L. Floyd1, and Bruce G. Lyeth2 1

Department of Physical Medicine and Rehabilitation, Center for Glial Biology in Medicine, 547 Spain Rehabilitation Center, University of Alabama at Birmingham, Birmingham, AL 35249, USA 2 Department of Neurological Surgery, University of California, 1515 Newton Court, One Shields Avenue, Davis, CA 95816, USA

Abstract: Traumatic brain injury (TBI) research to date has focused almost exclusively on the pathophysiology of injured neurons with very little attention paid to non-neuronal cells. However in the past decade, exciting discoveries have challenged this century-old view of passive glial cells and have led to a reinterpretation of the role of glial cells in central nervous system (CNS) biology and pathology. In this chapter we review several lines of evidence, indicating that glial cells, particularly astrocytes, are active partners to neurons in the brain, and summarize recent ﬁndings that detail the significance of astrocyte pathology in traumatic brain injury. Keywords: astrocyte; glutamate; sodium; calcium; acidosis; mechanical strain injury the significance of astrocyte pathology in traumatic brain injury.

Introduction Traumatic brain injury (TBI) research to date has focused almost exclusively on the pathophysiology of injured neurons with very little attention paid to non-neuronal cells. This near-exclusive focus on neuroprotection likely reﬂects the predominant paradigm of the neuroscience community as a whole, which characterized glial cells as a specialized type of connective tissue that merely provided support for the neurons. However in the past decade, exciting discoveries have challenged this century-old view of passive glial cells and have led to a reinterpretation of the role of glial cells in central nervous system (CNS) biology and pathology. In this chapter we review several lines of evidence, indicating that glial cells, particularly astrocytes, are active partners to neurons in the brain, and summarize recent ﬁndings which detail

Astrocytes in normal brain function: recasting an old ‘‘star’’ One line of evidence that astrocytes are more than supportive connective tissue comes from recent detailed analyses of the cellular morphology. A unique morphological feature of astrocytes is that nearly the entire cell surface is covered with processes that extend and become lamellae or ﬁlopodia as indicated by three-dimensional reconstruction of electron micrographs. Most lamellae and ﬁlopodia originate from processes, which, like the cell body, contain organelles and cytoskeletal elements; however, the cell surface extensions do not contain organelles and intermediate ﬁlaments themselves and are therefore not visualized with GFAP immunohistochemistry (Chao et al., 2002). Thus, GFAP immunohistochemistry provides only a

Corresponding author. E-mail: clﬂ[email protected] DOI: 10.1016/S0079-6123(06)61005-4

61

62

limited view of the total three-dimensional area and morphological complexity of a given astrocyte’s domain. In proprotoplasmic astrocytes of the cortex, cell surface extensions account for 50–60% of the cytoplasmic volume of the cell (the remainder is the cell body and processes) and cover 80% of the cell surface. Structurally, astrocytes are highly interdigitated such that neighboring processes often correspond to different astrocytic cell bodies. Also astrocyte processes frequently surround neuronal synapses; for example in the rat neocortex, 56% of all synapses are contacted by astrocytes (Chao et al., 2002). Large variations in the structural characteristics of glia-synaptic contacts exist such that common morphologies run the gambit from partial to entire encasement of a single axodendritic synapse or a triadic/serial synapse to complete coverage of complex synapses including centro-dendritic, centro-axonic, and multicenteric glomeruli synapses (Chao et al., 2002). The extensive astrocyte–synapse connections have led to the hypothesis that many synapses are actually of a tripartite nature, comprising pre- and post-synaptic neuronal elements and a third glial component (Haydon, 2001; Chao et al., 2002). A second line of evidence for the active role of astrocytes in brain function is that astrocytes express diverse neurotransmitter receptors that regulate the concentration of intercellular calcium ([Ca2+]i). Astrocytes express receptors for glutamate (Glaum et al., 1990; Holzwarth et al., 1994), acetylcholine (Kondou et al., 1994; Shao and McCarthy, 1995), ATP (Neary and Zhu, 1994), and histamine (Peakman and Hill, 1995). Additionally, astrocytes express voltage-gated Ca2+ channels similar to those in neurons (L- and N-type) and Ca2+ entry through these channels alters [Ca2+]i (MacVicar et al., 1991; D’Ascenzo et al., 2004). Activation of voltage-gated Ca2+ channels requires depolarization that has been shown in cultured astrocytes, organotypic cultures, and acutely isolated astrocytes following local changes in extracellular K+ concentration (Verkhratsky et al., 1998). Importantly, astrocytes respond to neuronal transmitter release by generation of calcium waves (Dani et al., 1992; Porter and McCarthy, 1996). Thus astrocytes have intact cyto-machinery to sense and respond to the

extracellular environment in ways that were traditionally the domain of neurons, the ‘‘excitable cells’’ of the CNS. Not only do astrocytes have the cyto-machinery of excitable cells, but also they use it. Astrocyte excitability comprises activation by external or internal signals followed by subsequent transfer of a speciﬁc message to nearby cells and has been termed ‘‘gliotransmission’’ (Bezzi and Volterra, 2001). The foundation of astrocyte excitability is calcium signaling, both intracellular and intercellular; and since astrocytes do not generate action potentials, imaging of calcium transients and oscillations has been more effective than traditional electrophysiology techniques in studying astrocyte excitation (Volterra and Meldolesi, 2005). Two well-documented forms of astrocyte excitation are (a) neuron-dependent excitation and (b) neuronindependent excitation. Neuron-dependent excitation, mainly through elevations in transmitter outside of the synaptic cleft, was originally described for glutamate transmission in hippocampus and cerebellum but now includes multiple brain circuits as well as multiple receptors systems including GABA, acetylcholine, dopamine, ATP, BDNF, etc. (Haydon, 2001). Neuron-independent, or spontaneous, excitation of astrocytes occurs mainly during development, but has also been demonstrated in adult preparations (Aguado et al., 2002) and may occur more frequently under pathological conditions such as CNS injury as discussed later in this chapter. Neuron-independent calcium oscillations principally involve release of calcium from intracellular stores, but calcium entry via voltage-gate channels may also play a role (Parri et al., 2001; Nett et al., 2002; Parri and Crunelli, 2003). Spontaneous excitation of astrocytes results in subsequent activation of neighboring cells, including astrocytes and neurons, which indicates that both cell types are sources of excitation and communication networks may be far more complex than an astrocytic response to neuronal excitation (Volterra et al., 2005). The cellular mechanisms of intercellular calcium signaling in astrocytes remain an area of active inquiry. Comprehensive reviews detailing ﬁndings leading to the current understanding of the mechanisms of astrocyte intercellular calcium waves

63

have recently been published (Haydon, 2001; Volterra et al., 2005), but the most prevalent working hypotheses for these mechanisms are brieﬂy described here. Intercellular calcium signaling in astrocyte comprises a calcium wave that is essentially a sequence of increases in intracellular calcium that spreads, like a wave, throughout a group of astrocytes. Astrocyte calcium waves were initially characterized in response to glutamate application or mechanical stimulation to single astrocytes in cell culture (Cornell-Bell et al., 1990; Cornell-Bell and Finkbeiner, 1991), and these ﬁndings have been corroborated in organotypic hippocampal slice cultures (Harris-White et al., 1998), Mu¨ller glia of acutely isolated retina (Newman and Zahs, 1998), and more recently in vivo (Nimmerjahn et al., 2004). In astrocytes, a ligand (such as glutamate) binds to a G-protein receptor coupled to phosphoinositide-speciﬁc phospholipase C (PLC) which cleaves phosphatidylinositol 4,5 bisphosphate (PIP2) yielding the second messenger inositol-1,4,5 trisphosphate (IP3). IP3 then binds to the IP3 receptor (IP3R) located on intracellular calcium stores, causing increased intracellular calcium. Elevated intracellular calcium in a single cell can translate into an intercellular calcium wave by at least two known mechanisms, working independently or in concert. First, an intercellular calcium wave can spread between astrocytes via gapjunction mediated metabolic coupling. In this model, IP3 diffuses between gap junctions and stimulates the release of intracellular calcium from adjacent astrocytes resulting in a wave of calcium. Additionally, an extracellular component may also be involved since studies show that calcium waves can travel between cells separated by a cell-free zone (Hassinger et al., 1996; Newman and Zahs, 1997) and that waves follow the direction of extracellular perfusion (Hassinger et al., 1996). Since waves of extracellular ATP can accompany intercellular calcium waves, the most likely candidate for the extracellular message for calcium wave propagation is ATP (Cotrina et al., 1998b, 2000; Guthrie et al., 1999). In this second model, astrocytes release ATP that then activates the purinergic receptor P2Y1, thereby increasing IP3 and producing a calcium signal in nearby cells. These two pathways may work together to produce

both shorter-range (gap-junction mediated) and longer-range (ATP-mediated) calcium waves (for review, see Haydon, 2001). One functional consequence of astrocyte calcium signaling is glutamate release (Parpura et al., 1994; Bezzi et al., 1998). Using live-cell ﬂuorescent imaging and cell culture, glutamate release from astrocyte was visualized by applying the cofactors NAD+ and the enzyme glutamate dehydrogenase in the bath so that any released glutamate is converted to a-ketogluatarate and the NAD+ is reduced to the ﬂuorescent NADH. Combination of this imaging technique with live-cell calcium imaging demonstrated that a wave of extracellular glutamate accompanies an intercellular calcium wave (Innocenti et al., 2000). Subsequent characterization of glutamate release from astrocytes has shown that this release is not only related to intracellular calcium increases, but is actually calcium dependent and is very likely regulated by exocytosis (Parpura and Haydon, 2000; Anlauf and Derouiche, 2005). In contrast to neurons where extracellular calcium is a main component of calcium-dependent exocytosis, both the IP3 and ryanodine-sensitive internal calcium stores seem to be key components in calcium-dependent glutamate release from astrocytes (Araque et al., 1999). In further support of the hypothesis that astrocytes exocytose glutamate, distinct vesicular compartments containing glutamate have been localized in astrocytes (Kreft et al., 2004); and astrocytes express key elements of vesicular exocytosis, SNARE-family proteins (Parpura et al., 1995; Crippa et al., 2006). However, astrocyte exocytosis of glutamate is slower (2 orders of magnitude) than neuronal counterparts (Kreft et al., 2004). The primary implication of astrocyte release of transmitters, such as glutamate, is modulation of synaptic transmission in nearby neurons and this modulation has been shown in several experimental preparations. In cell culture, astrocytes were shown to either evoke or depress synaptic activation and this activity was regulated by NMDA or metabotropic glutamate receptors (Araque et al., 1999). In hippocampal slice cultures, glutamate release from astrocytes augmented interneuron pyramidal synaptic connections (Kang et al., 1998; Liu et al., 2004a) and activated kainate

64

receptors on interneurons (Liu et al., 2004b). Functional consequences of astrocyte release of D-serine have also been demonstrated in that astrocyte-derived D-serine, through activation at the allosteric glycine-binding site, facilitated NMDA receptor activity (Mothet et al., 2005; Panatier et al., 2006). Also, as described above, astrocytes can release ATP. ATP can be converted to adenosine, which then can hyperpolarize adjacent neurons as has been shown in the retina (Newman, 2003). Taken together, these studies demonstrate that under physiological conditions astrocytes release transmitters that alter neuronal signaling. While the effects of CNS injury on intra- and intercellular calcium signaling as well as on astrocyte neuronal signaling have yet to be fully elucidated, a strong body of evidence suggests that astrocytes are not merely passive support cells but are active communication partners to neurons in the uninjured brain. Perhaps a better understanding of the effect of trauma on these elements of astrocyte communication will provide the infrastructure to develop therapeutic interventions that target multiple cell types in the injured brain.

Revisiting and revising the ‘‘Classical View’’ of astrocytes in CNS trauma As discussed in the introduction, TBI research to date has focused mainly on neuroprotection and given little consideration to the role of glial cells in injury. This has significant implications for the understanding of brain pathology after TBI since the population of glial cells in brain is actually much larger than neurons. Furthermore, the importance of the complex communication and interaction between astrocytes and neurons described above are becoming more apparent in normal brain function and increasingly appear to play a critical role in such perturbed brain functions as seizures and ischemia injury (Vernadakis, 1996; Vesce et al., 1999). Increasingly, research studies are indicating that astroglia cells are adversely affected by trauma and ischemia and that damage to astrocytes also affects the fate of neurons in the traumatically injured brain. Early impairment of astrocyte function after TBI may

compromise critical neuronal–glia interactions and thus, may play a significant role in outcome after injury. Perturbations in astrocyte function can also have indirect effects on normal brain function. For example, alterations in glial glutamate transporter number or function could indirectly alter glutamate-mediated excitability. The recent ﬁndings that D-serine, which is exclusively released from astrocytes in the synapse, preferentially acts as the NMDA receptor co-agonist (Boehning and Snyder, 2003) would suggest that any injuryinduced alteration in astrocyte function could adversely impact normal glutamatergic signaling. The majority of TBI laboratory studies that have investigated astrocyte response to injury have focused on the phenomena of reactive astrocytes characterized by increased GFAP immunoreactivity, hypertrophy, and hyperplasia. The CNS responds within days to injury by producing reactive astrogliosis and glial scarring (McGraw et al., 2001). Although not completely understood, glial scarring is thought to be an attempt by the CNS to restore homeostasis by isolating the damaged region (Fitch et al., 1999). However, the glial scar may interfere with any subsequent neural repair or axonal regeneration (Ridet et al., 1997). Two days after midline ﬂuid percussion TBI in the rat, increased GFAP staining and thickened glia consistent with reactive astrocytes were observed in the CA3 hippocampus (D’Ambrosio et al., 1999) even though midline ﬂuid percussion TBI is generally not associated with CA3 pyramidal cell loss (Lyeth et al., 1990). Reactive astrogliosis and the formation of a glial scar have been viewed as both a promoter of and an impediment to CNS regeneration (Reier et al., 1983; Muller et al., 1995; Ridet et al., 1997). This chapter focuses on events that involve TBI-induced astrocyte damage or death that likely occur prior to the reactive astrocytic responses. Therapeutic interventions targeted at these early mechanisms of astrocyte injury may rescue both astrocytes and neurons and as a consequence may reduce glial scarring. Role of glutamate in astrocyte pathology in CNS trauma During normal neuronal activity extracellular glutamate in the synaptic cleft must be rapidly

65

cleared to optimize the signal-to-noise ratio as well as to prevent neuronal damage from excitotoxicity. Glutamate is removed from the synapse by glutamate transporters located in the plasma membrane of neurons and adjacent astrocytes (Kanner and Schuldiner, 1987). The major glutamate removal mechanism is through the astrocyte-speciﬁc sodium-dependent glutamate transporters, GLT-1 and GLAST (Danbolt, 1994, 2001; Takahashi et al., 1997). GLT-1 is located primarily in the rat forebrain while GLAST is found primarily in the cerebellum (Tanaka et al., 1997; Danbolt, 2001). Glutamate taken up by astrocytes is converted to glutamine by glutamine synthetase, an energy-dependent process requiring one molecule of ATP for each molecule of glutamate converted to glutamine (Daikhin and Yudkoff, 2000). Glutamine then diffuses out of the astrocyte and is taken up by the presynaptic neuron for recycling back into glutamate via the enzyme glutaminase. Glutamate excitotoxicity is a significant determinant of TBI pathophysiology. Numerous laboratory studies (Faden et al., 1989; Katayama et al., 1990; Nilsson et al., 1990; Zhong et al., 2006) and several clinical studies (Zauner et al., 1996; Bullock et al., 1998; Koura et al., 1998) have documented an excessive glutamate release following TBI. The elevated extracellular concentration of glutamate results in an excitotoxicity from excessive activation of ion-channel linked (Hayes et al., 1988; McIntosh et al., 1989, 1998; Hicks et al., 1994) and G-protein linked glutamate receptors (Hayes et al., 1988; Hicks et al., 1994; Mukhin et al., 1996, 1997; Faden et al., 1997; Lyeth et al., 2001; Zwienenberg et al., 2001). Experimental cerebral ischemia is also associated with a large, rapid increase in extracellular glutamate (Choi and Rothman, 1990).

Energy demands and acidosis in CNS trauma The uptake of glutamate rapidly increases energy demands on astrocytes through several mechanisms. Mechanical damage to astrocytes elicits elevated [Na+]i in large part from cotransport with glutamate (Floyd et al., 2005). The sodium cotransported into astrocytes along with glutamate

triggers activation of the Na+-K+ ATPase, which is fueled by the ATP produced by glycolysis (Silver and Erecinska, 1997). Further energy demand is created by the conversion of glutamate to glutamine by glutamine synthetase. Energy demands of the brain are met through several pathways including the oxidative phosphorylation, glycolytic, and glycogen pathways. Recent studies indicate that astrocytic glycolysis plays a major role in supplying rapid energy demands of astrocytes (Schurr et al., 1999; Paemeleire and Leybaert, 2000). Glycolysis also provides an important source for neuronal energy substrates in the form of lactate for subsequent conversion to pyruvate for use in the TCA and oxidative phosphorylation pathways (Schurr et al., 1999; Paemeleire and Leybaert, 2000). The majority of excess lactate measured following TBI is most likely of astrocytic origin and contributes to lactic acidosis (Kawamata et al., 1992, 1995; Lyeth et al., 1996). The lactate released as a by-product of glycolysis in astrocytes is taken up by neurons and used as an aerobic energy substrate. In vitro slice experiments show that adjacent neurons use lactate generated in astrocytes. The elevated levels of lactate in slices exposed to high glutamate and ample glucose are significantly increased when lactate transport into neurons is inhibited (Schurr et al., 1999). Furthermore, neuronal functional viability is also lost when lactate transport into neurons is inhibited (Schurr et al., 1999). Elevated tissue levels of lactate associated with hypoxia or ischemia have often been considered an indicator of anaerobic metabolism. However, many studies now suggest that brain tissue also can produce lactate aerobically under certain physiological conditions when stimulated (Raichle, 1991; Wilsch et al., 1994; Hyder et al., 1996; Hu and Wilson, 1997). Taken together, these results indicate that mechanical injury to astrocytes rapidly elevates [Na+]i, followed by a rapid energy demand, lactate production, and intracellular acidosis. The role of lactacidosis and a fall in intracellular pH in brain injury and ischemia is well documented (Siesjo, 1988; Kraig and Chesler, 1990; Staub et al., 1990; Nedergaard et al., 1991a, b). Electron microscopy studies show that acidosis promotes cytotoxic edema that is localized predominantly in astrocytic endfeet (Garcia et al.,

66

1977; Jenkins et al., 1982). Acidosis-induced glial swelling is attributed to several membrane transporters including Na+/H+ antiporter (Grinstein and Rothstein, 1986), Cl/HCO3 exchanger (Kimelberg et al., 1990), and the Na+/HCO3 cotransporter (Newman, 1991). Activation of these pH-stabilizing mechanisms during TBI-induced acidosis would elicit increases in glial intracellular Na+ and Cl along with osmotically obligated water leading to astrocyte swelling (Kempski et al., 1988; Jakubovicz and Klip, 1989). Recent studies have demonstrated that amiloride analogues that inhibit the NHE-1 significantly reduce hippocampal extracellular glutamate levels following transient cerebral ischemia in rat (Phillis et al., 1998) and reduce CA1 hippocampal neuronal cell death assessed at 6 days after transient cerebral ischemia in the gerbil (Phillis et al., 1999). Phillis et al. (1999) attributed their results to the actions of amiloride in reducing astrocyte swelling and thereby increasing extracellular space. In a study by Plesnila et al. (1999), astrocyte cultures exposed to lactic acid exhibited cell volume increases of 21% when pHi dropped to 6.8–6.2, levels observed in TBI. Swelling was completely blocked with the NHE-1 inhibitor, amiloride. Ringel et al. (2000) found that inhibition of anion transports reduced lactacidosis-induced swelling in astrocyte cultures. Activation of the pH-regulated membrane transporters during periods of intracellular acidosis contribute to further elevations in [Na+]i and may result in reversal of the sodium/calcium exchanger as discussed below.

Acute astrocyte loss following CNS insult Astrocyte swelling is a prominent feature of both cerebral ischemia (Jenkins et al., 1982) and TBI (Castejon, 1998) and is considered to be an ‘‘exaggerated extension of normal astrocyte function’’ (Kimelberg, 1992). Astrocyte swelling is likely a prominent component of raised ICP after both ischemia and TBI and resulting in part from alterations in ion-transport mechanisms that leads to osmotically obligated water entry (Kimelberg et al., 1990). Cortical biopsies of human TBI complicated with subdural hematoma have shown

severe astrocytic edema, dissociation of the perivascular astrocyte endfeet from the capillary basement membrane, and disruption of inter-astrocyte gap junctions (Castejon, 1998). Electron microscopy of brain tissue from human cerebral contusion patients has revealed evidence of astrocytic swelling as early as 3 h after injury, persisting for as long as 3 days (Bullock et al., 1991). It is important to consider that human tissue is essentially never available for analysis during the ﬁrst hours after insult. Thus, there is a distinct possibility that human astrocyte swelling may occur very rapidly after injury as in animal models of ischemia and TBI. In addition to cell swelling and raised intracranial pressure, the alterations in astrocyte ionic homeostasis can likely initiate cell damage and death. There has been a dogma that astrocytes are highly resistant to hypoxic/ischemic insults compared to neurons (Lukaszevicz et al., 2002). This dogma has been challenged by early suggestions and by later experiments demonstrating that astrocytes are indeed vulnerable to ischemic conditions. Over 20 years ago, Plum (Plum, 1983) suggested that the degree of infarction from focal ischemia might depend upon the viability of astrocytes. This proved to be an insightful prediction and subsequent studies indicate that this suggestion applies to TBI as well as cerebral ischemia. A number of in vitro studies have demonstrated that astrocytes are indeed vulnerable to ischemia. Astrocytes exposed to hypoxia alone can require up to 24 h exposure to produce widespread death (Yu et al., 1989; Sochocka et al., 1994) even in the absence of glucose (Kelleher et al., 1993). However, combining hypoxia with acidosis, a more realistic model, accelerates astrocyte death to just a few hours (Swanson et al., 1997). Recent in vitro studies have examined this issue utilizing a novel experimental condition in which cultured astrocytes were exposed to hypoxic, acidic, ionshifted Ringers (HAIR) solution that mimics the environmental characteristics of ischemia and TBI (Bondarenko and Chesler, 2001a). Their microenvironment included glucose and thus mimicked incomplete ischemia as well as severe TBI. They found that as little as 15–20 min exposure to their HAIR solution resulted in death of 60% of the astrocytes (Bondarenko and Chesler, 2001a). The

67

same group subsequently reported that the NHE-1 inhibitor amiloride protected astrocytes exposed to the HAIR protocol. Furthermore, blocking the Na+/Ca2+ exchanger with benzamil or KBR7943, which preferentially inhibits reverse Na+/ Ca2+ exchange, protected astrocytes from the HAIR protocol (Bondarenko and Chesler, 2001b; Bondarenko et al., 2005). These studies provide evidence that astrocytic intracellular acidosis activates NHE-1 leading to Na+ entry and the reversal of the Na+/Ca2+ exchanger, triggering astrocyte cell loss after ischemia. Recent in vivo studies provide further compelling evidence of early astrocyte loss following cerebral ischemia. Liu et al. (1999) reported a 50% decrease in GFAP mRNA and protein in the ischemic core within 5 h after permanent middle cerebral artery occlusion, and almost complete absence of GFAP mRNA by 12 h. The loss of GFAP mRNA preceded the loss of the mRNA for GLUT3, a neuron-speciﬁc glucose transporter. Immunohistochemical experiments using other astrocyte markers including S100beta, GSTYb1, and vimentin produced similar results. Interestingly, mRNA began to increase in the peri-infarct area by 3 h and peaked at 48 h after ischemia, suggestive of reactive astrocyte proliferation in the ischemic penumbra. Using a rat subdural hematoma-induced ischemia model, Jiang et al. (2000) reported complete loss of GFAP immunostaining at 4 h after the insult in the ischemic core. Protoplasmic astrocytes appear to be more vulnerable to ischemia than ﬁbrous astrocytes. For example, following permanent middle cerebral artery occlusion in mice, Lukaszevicz et al. (2002) found that protoplasmic astrocytes (type 1, accounting for most of the astrocytes in gray matter) lost GFAP immunolabeling within minutes in the ischemic core. In contrast, ﬁbrous astrocytes (type 2, found predominantly in white matter) displayed a transient hypertrophy with no conspicuous cell death (Lukaszevicz et al., 2002). These studies provide compelling evidence that ischemia is associated with an early loss of astrocytes. Glutamate excitotoxicity can lead to both necrotic cell death and apoptotic programed cell death following application of glutamate agonists (Du et al., 1997; Nicotera et al., 1997) ischemia (Namura et al., 1998; Graham and Chen, 2001)

and TBI (Rink et al., 1995; Newcomb et al., 1999). Most of these studies have focused on apoptotic cell death of neurons. Astrocytes are also vulnerable to apoptotic cell death that may contribute to the pathogenesis of many acute and chronic neurodegenerative disorders (Takuma et al., 2004). Hypoxia/ischemia induces mitochondrial depolarization in astrocytes (Smith et al., 2003), that triggers caspase-dependent cell death pathways and by poly(ADP-ribose) polymerase-1 cell death pathway (Giffard and Swanson, 2005). Astrocyte apoptosis can be detected by caspase activation by a variety of in vitro insults including staurosporin (Keane et al., 1997), cycloheximide (Tsuchida et al., 2002), and ischemic conditions (Yu et al., 2001; Gabryel et al., 2002). Activation of apoptotic cascades in astrocytes has also been reported after in vivo TBI. For example, accumulation of active caspase-3 was observed for 5 days after controlled cortical impact TBI in rats in both neurons and astrocytes but, interestingly with a higher proportion of activated caspase-3 in astrocytes than in neurons (Johnson et al., 2005). TBI and ischemia share a number of pathophysiological similarities including excessive glutamate release and selective neuronal cell loss. The role of astrocytes in uptake of glutamate may have particular importance to the traumatically injured brain where large ﬂuxes of extracellular glutamate (Faden et al., 1989; Katayama et al., 1990) likely contribute to acute excitotoxic processes (Olney and Ho, 1970; Hayes et al., 1988). Specifically, early impairment of astrocyte function after TBI may compromise maintenance of homeostasis in the extracellular microenvironment and disrupt critical neuronal–glia interactions. Zhao et al. (2003) demonstrated selective loss of GFAP and glutamine synthetase immunoreactivity in rat brains as early as 1 h after lateral ﬂuid percussion TBI. The loss of immunoreactivity was limited to regions directly adjacent to areas of selective neuronal vulnerability, the hippocampus CA3. The loss of immunoreactivity for astrocytes markers preceded ﬂuoro-jade detection of neuronal degeneration by several hours suggesting that loss of supporting astrocytes may contribute to subsequent neuronal cell loss.

68

Unraveling the mechanisms of acute astrocyte damage with in vitro mechanical injury It is clear from the studies described above that in vivo experimental TBI or ischemia produces rapid and profound damage to astrocytes. Moreover, several in vitro models of ischemia also show acute astrocyte pathology. Additionally, several of the astrocytic intracellular events occurring acutely after TBI have been characterized in a series of in vitro studies using a mechanical injury model developed by Ellis et al. (1995). In this model, cells are grown on a membrane that is subjected to a rapid mechanical deformation that approximates the tensile strain or stretch that is a key component in the acceleration/deceleration type of closed human TBI (Meaney et al., 1995) and in vivo brain injury (Margulies et al., 1990). This model has been validated by demonstrating that in vitro strain injury produces many of the post-traumatic responses observed with in vivo injury models including intracellular lesions to the mitochondria, Golgi, and cytoskeletal elements in astrocytes and neurons (Dietrich et al., 1994; Ellis et al., 1995; McKinney et al., 1996), increases in astrocyte and neuronal permeability (Ellis et al., 1995; Weber et al., 1999; Willoughby et al., 2004), activates phospholipases (Lamb et al., 1997; Floyd et al., 2001), and induces free radical and isoprostane formation (McKinney et al., 1996; Lamb et al., 1997; Hoffman et al., 2000) as well as release of the injury markers S-100b and neuron-speciﬁc enolase (NSE; Slemmer et al., 2002, 2004). Strain-injury of cultured neurons also alters the Mg2+ block of the NMDA receptor (Zhang et al., 1996) and produces a novel stretch-induced delayed neuronal depolarization, which may be related to transient neuronal dysfunction observed in vivo (Hamm et al., 1994; Tavalin et al., 1995). Strain-injury in neurons also increases activation of the AMPA receptor by decreasing AMPA receptor desensitization (Goforth et al., 1999). Thus, this well characterized mechanical injury model recapitulates many aspects of in vivo TBI and is a useful tool to examine injury-induced changes in acute astrocyte pathology. A critical element in astrocytic calcium signaling is regulation of intracellular free Ca2+.

Mechanical strain injury to astrocytes produces a rapid rise in [Ca2+]i that returns to near-basal levels within 15 min. However, application of either glutamate or the metabotropic glutamate group I (mGluRI) agonist trans-1-aminocyclopentyl-1-3dicarboxylic acid (tACPD) after injury results in a significantly blunted rise in [Ca2+]i for up to 24 h post-injury, suggesting that the calcium signaling machinery is disturbed by injury (Rzigalinski et al., 1998). Alterations in intracellular calcium dynamics after mechanical strain injury may be related to alterations in inositol trisphosphate (IP3) signaling as strain injury to astrocytes causes a significant increase in IP3 at 5, 15, and 30 min and at 24 and 48 h post-injury (Floyd et al., 2001). Interestingly, the injury-induced increases in IP3 at 15 and 30 min post-injury correspond to time points when [Ca2+]i had returned to near normal levels; however, high IP3 should produce increased [Ca2+]i if Ca2+ signaling were functioning normally. Instead the opposite was observed — elevated IP3 at time points when intracellular calcium has returned to basal levels which indicates that IP3 may be uncoupled from its target, the intracellular Ca2+ store. Additionally, antagonism of mGluRIs and inhibition of PLC attenuated injury-induced uncoupling of IP3-mediated Ca2+ signaling, reduced astrocyte damage (Floyd et al., 2001, 2004), and reduced injury-induced depletion of intracellular calcium stores (Chen et al., 2004). Thus, mechanical strain injury to astrocytes causes acute elevations in intracellular calcium and disruption of IP3-mediated intracellular calcium signaling. Mechanical strain injury to astrocytes also significantly increases intracellular sodium (Floyd et al., 2005). Mild and moderate strain injury produced a rapid rise in intracellular sodium that returned to near-basal levels by 20–25 min postinjury while severe injury produced increased intracellular sodium for at least 50 min (the duration of the experiment). These elevations in intracellular sodium were similar to those induced by exogenous glutamate application and were reduced by pre-injury application of the sodium-dependent glutamate uptake inhibitor TBOA, indicating that sodium which accompanies glutamate uptake contributes to injury-induced increases in intracellular sodium. Comparison of the time course of

69

injury-induced elevations in intracellular sodium and calcium indicated that there is significant overlap in the ﬁrst 30 min post-injury, which raises the possibility that sodium/calcium exchange may be involved in the acute astrocyte pathology after injury. The magnitude and direction of calcium ﬂux of the sodium/calcium exchanger is dependent, in part, on the Na+ electrochemical gradient, and under normal conditions the sodium/calcium exchanger operates in the ‘‘forward’’ or calciumefﬂux mode to extrude calcium. Importantly, manipulations in uninjured astrocytes that raise intracellular sodium have been shown to increase intracellular calcium by ‘‘reversal’’ of the sodium/ calcium exchanger or calcium-inﬂux operation (Goldman et al., 1994). Thus, it was predicted that the elevated intracellular sodium observed after mechanical strain injury may be sufﬁcient to reverse the sodium/calcium exchanger and cause an inﬂux of calcium. It was found that pre-injury blockade of the calcium-inﬂux (reversed) mode with the compound KB-R7943 significantly reduced intracellular calcium in the moderate and high levels of injury (but not at the low levels of injury), the magnitudes of injury with the largest and more sustained elevations in intracellular sodium (Floyd et al., 2005). Additionally, this severity of injury-dependent reversal of the sodium/ calcium exchanger was predicted by mathematical modeling of the reversal potential of the sodium/ calcium exchanger using experimental values of intercellular sodium and calcium concentrations after mechanical strain injury. Taken together, these data suggest that increased intracellular sodium can cause a reversal of the sodium/calcium exchanger to cause inﬂux of calcium into the cells, and that this only occurs with moderate or severe injury. Mechanical strain injury to astrocytes causes release of ATP (Ahmed et al., 2000; Neary et al., 2005). ATP release following mechanical injury not only has consequences for cellular energetics (Bambrick et al., 2004), but also may alter intercellular calcium signaling via P2 purinergic receptors. Cortical astrocytes express both the ionotropic P2X receptors and the G-protein coupled P2Y receptors (Lenz et al., 2000; Weisman et al., 2005); however, P2Y expression is much

more prevalent in astrocytes than P2X (Illes and Ribeiro, 2004). Nucleotides released from injured cells can activate either the P2X or P2Y and cause an increase in intracellular calcium (Illes and Ribeiro, 2004) as well as activation of mitogenactivated protein kinase (MAPK) pathways (Gendron et al., 2003). Importantly, activation of P2X receptors on astrocytes is linked to neurodegenerative events and astrogliosis (for review, see Neary et al., 1996); yet, activation of P2Y receptors on astrocytes is protective. For example, activation of P2Y2 increases Bcl-2 family genes (Chorna et al., 2004), and activates extracellular signal-regulated protein kinase (ERK) which has been linked to the protective activation of CREB (Neary et al., 1996; Chorna et al., 2004). In an eloquent series of studies, Neary and colleagues have detailed the effects of mechanical strain injury to astrocytes on P2 receptor activation and found that injury induces a rapid release of ATP which, via P2 receptors, activates ERK (Neary et al., 2003) and Akt (Neary et al., 2005); increases expression and release of a protein that induces synapse formation, thrombospondin-1 (Tran and Neary, 2006); and also phosphorylates GSK3b at the Ser-9 location, thereby inhibiting activity (Neary and Kang, 2006). Thus, release of ATP from injured astrocytes can act not only as a stimulus for gliosis, but also as a promoter of cell survival and synaptic plasticity. Further work is needed in this exciting area to better elucidate the complex relationship between P2 activation, gliosis, and cell survival. Mechanical strain injury to astrocytes disrupts many of the critical features of intra- and intercellular calcium signaling in astrocytes. The known changes in astrocyte signaling after mechanical injury are summarized schematically in Fig. 1. Strain injury to astrocytes causes a rapid increase in intracellular calcium and sodium that is dependent on the magnitude of injury (Floyd et al., 2005). There are several mechanisms by which calcium is elevated following mechanical injury. Excessive glutamate following injury agonizes mGluRs to cause IP3-mediated calcium release (Floyd et al., 2004). Theoretically, elevated glutamate following injury could also activate the AMPA ionotropic glutamate channel as has been shown in neurons (Goforth et al., 2004). Although the effects of

70 ATP P2YR Ca2+ Propidium Iodide

? AMPA

MITOCHONDRIA

PLC

(reversed)

IP3 Ca2+

GL

T1

NC X

Na+

Hemichannel? ATP

mGluR

Adjacent Astrocyte

Na+ TSP-2 Akt

Ca2+ Store

ERK

Ca2+ Gap Junction

BCL-2 NUCLEUS

Fig. 1. Pathological events known to occur in astrocytes after mechanical strain injury. Mechanical strain injury to astrocytes results in a rapid rise in intracellular calcium and sodium. Sources of injury-induced increases in intracellular calcium include activation of the mGluRs, release of calcium from IP3-mediated intracellular calcium stores, reversal of the sodium/calcium exchanger, and possibly AMPA receptor activation. Increases in intracellular sodium after mechanical injury are largely due to sodium-dependent uptake of excessive glutamate. Injured astrocytes also release ATP. ATP could activate P2YRs and further increase intracellular calcium but also could initiate cell survival pathways such as Bcl-2. Astrocytes are coupled by gap junctions, and small molecules such as IP3 or Ca2+ could travel between cells after injury. Propidium iodide, a marker for dead or damaged cells, could enter the astrocyte through disruptions in the cell membrane and intercalate into the DNA. (See text for further details.) Abbreviations: protein kinase B/Akt (Akt): (RS)-2-amino-3-(3-hydroxy-5-methylisoxazol-4-yl)propionate (AMPA), extracellular signal-regulated protein kinase (ERK), glutamate transporter (GLT-1), inositol trisphosphate (IP3), metabotropic glutamate receptor (mGluR), sodium/calcium exchanger (NCX), purinergic 2Y receptor (P2YR), phospholipase C (PLC).

mechanical injury on the subunit composition in astrocytes has not been examined, yet, if the GluR2 subunit is present, injury induced activation would likely contribute to increase intracellular Na+. However, if the GluR2 subunit were absent, the channel would also be permeable to Ca2+ (Frandsen and Schousboe, 2003). Activation of either conﬁguration of the AMPA receptor would likely contribute to the pathophysiology of

astrocyte injury, as both increased intracellular calcium and sodium are intimately involved. Excessive glutamate from mechanical injury also affects the astrocyte by activation of sodiumdependent glutamate uptake that significantly elevates intracellular sodium. At moderate and severe levels of injury, the elevation in intracellular sodium is sufﬁcient to cause reversal of the sodium/calcium exchanger and produce calcium

Fig. 2. Effect of mechanical injury on functional coupling in astrocytes. Conﬂuent primary astrocyte cultures we mechanically injured and then functional coupling was analyzed using ﬂuorescence recovery after photobleach (FRAP). Panel (A) shows representative micrographs of uninjured (left column) or moderately injured (right column) astrocytes before, immediately after, and 15 s after photobleach. Regions of interest (ROIs) were drawn around four cells per ﬁeld. The ROIs outlined in red and green were bleached but the ROIs in yellow and blue were not. A representative trace of a mildly injured astrocyte at 24 h post-injury is shown in panel (B). Panel (C) shows quantiﬁcation of the maximal recovery at various post-injury times. Mild injury increased and moderate injury decreased functional coupling in astrocytes (n ¼ 3 separate experiments).

71

72

inﬂux, thereby further contributing to elevated intracellular calcium (Floyd et al., 2005). Mechanical strain injury causes release of ATP from damaged mitochondria (Ahmed et al., 2000; Neary et al., 2005), some of which is released into the extracellular space. Although the mechanism of ATP release after injury is still unknown, candidates for release include hemichannels, ATP transporters, or vesicular exocytosis (Schwiebert, 2001). Extracellular ATP following injury can activate purinergic receptors including the P2YRs. Activation of the P2YRs can increase intracellular calcium, and also can have beneﬁcial effects such as activation of cytoprotective pathways and increased expression of Bcl-2 (Neary et al., 2003, 2005). Propidium iodide, a marker of cell damage and death, could enter the cell via perturbations in the cell membrane. As astrocytes are extensively coupled via gap junctions, small molecules such as Ca2+ and IP3 could travel between cells to potentially propagate an injury signal, as will be discussed below. Work in uninjured astrocytes shows that gap junction communication is an important element in intercellular calcium signaling in astrocytes which facilitates intercellular transfer of ions (i.e., Ca2+ or Na+) or second messengers (i.e., IP3). Gap junctions are connections of intercellular channels formed by hemichannels from adjoining cell membranes, and the adult brain astrocytes are so abundantly interconnected that the network is regarded as a syncytium. Gap junctions are formed by the joining of 2 hemichannels, each of which comprises six protein subunits termed connexins. The connexin43 protein (Cx43) is highly expressed in astrocyte gap junctions while other connexin proteins comprise neuronal gap junctions (Yamamoto et al., 1990a, b; Giaume and McCarthy, 1996). Gap junctions are not simply passive channels between cells, but functionally respond to alterations in the cellular milieu that alters gap junction assembly and permeability, mainly via phosphorylation (Musil et al., 1990; Nagy and Li, 2000; Rouach et al., 2002; VanSlyke and Musil, 2005). Most studies of the effect of injury on gap junction coupling have used models of ischemic brain injury with divergent ﬁndings. Increases in intracellular calcium, decreases in pH,

and production of oxygen-free radicals can inhibit gap junction coupling (Bolanos and Medina, 1996; Martinez and Saez, 2000); yet, other evidence suggests that gap junctions remain open during ischemia (Cotrina et al., 1998a). The role of gap junction communication in cell death following an ischemic insult is also unclear. Some groups report that inhibition of astrocyte gap junctions during ischemic injury decreases the area of infarct and is neuroprotective (Lin et al., 1998; Frantseva et al., 2002), suggesting that astrocyte gap junctions amplify damage and spread cell death signals to otherwise viable cells. An alternative hypothesis suggests that gap junctions enable astrocytes to use the syncytium to dissipate otherwise toxic concentrations of glutamate or ions (Blanc et al., 1998; Siushansian et al., 2001). Part of the controversy over the role of gap junction communication in ischemic injury may be related to heterogeneity in gap junction coupling as well as regional or cellular differences in intracellular regulation of gap junctions (Zvalova et al., 2004). For example, gap junction hemichannels that open to the extracellular space have been characterized (Contreras et al., 2002) and may be involved in calcium signaling or ATP release (Stout et al., 2002; Bennett et al., 2003; Dermietzel et al., 2003). The effect of traumatic brain injury on astrocyte gap junction communication has only recently been investigated. Ohsumi et al. (2006) have shown that CX43 immunohistochemistry was increased in the hippocampus 24 and 72 h after lateral ﬂuid percussion injury. We evaluated the effect of mechanical strain injury on functional connectivity of gap junctions in astrocytes. In this experiment, primary astrocyte cultures were grown on Flexs plates and subjected to either a mild or moderate rapid mechanical strain injury as previously described (Floyd et al., 2005). Gap junction coupling was evaluated at 30 min, 4 h, or 24 h after mechanical strain injury using ﬂuorescence recovery after photobleach (FRAP). Astrocytes were bath loaded with 0.05 mg/ml 5-(and 6-) carboxyﬂuorescein diacetate-AM (CFDA) which is a cell membrane and gap junction permeable dye that becomes membrane impermeable after de-esteriﬁcation. Thus, when cells are visualized, the CFDA is trapped within the cells but can

73

transfer between coupled cells via gap junctions. After the ﬂuorescent moieties of the dye are bleached by intense laser excitation in a cell, any subsequent increase in ﬂuorescent intensity, or recovery, is attributed to transfer of dye between gap junctions and represents the extent of gap junction coupling. In this experiment, cells were visualized on an upright Zeiss Axioscop 2 Laser scanning confocal microscope (LSM510) using a Zeiss Achroplan 20 dipping objective. The dye was excited at 400 nm with emitted ﬂuorescence above 520 nm captured by a photo-multiplier tube. Two cells near the middle of the ﬁeld were selected for bleach and regions of interest (ROIs) were drawn. Two additional ROIs were drawn for other cells in the ﬁeld that were not subjected to photobleach. Three ﬁeld scans were obtained prior to bleach. Selected cells were bleached with 100 iterations at 100% laser power. To assess recovery after bleach, ﬁeld scans were acquired every 2 s for the next 2.5 min (75 scans total). Digital images obtained during experiments were saved on a PC for off-line analysis. Mean intensity values for each ROI were divided by mean intensity values of reference (nonbleached) cells to correct for photobleach of the entire ﬁeld. Figure 2(A) shows representative micrographs of uninjured astrocytes (left column) and astrocyte subjected to a moderate mechanical strain injury (6.5 mm deformation) 30 min prior to FRAP analysis. The ROIs outlined in red and green were bleached but the ROIs outlined in yellow and blue were not. A representative trace of the percentage of ﬂuorescence recovery over time for an astrocyte mildly injured 24 h prior to FRAP is shown in Fig. 2(B). Recovery of ﬂuorescence is rapid and begins to plateau by 400 s after bleach. Quantiﬁcation of the maximal recovery at various post-injury times is shown in Fig. 2(C). Uninjured astrocytes were moderately coupled and recovery after photobleach was between 30 and 40%. Neurons are reportedly not extensively coupled and were used as a procedural control with less than 5% ﬂuorescence recovery detected in adjacent neurons in co-cultures. In mildly injured astrocytes, a significant increase in recovery after photobleach was seen at 30 min, and 4 and 24 h after mechanical strain injury. However, with moderate strain injury, recovery after photobleach

was decreased at these same time-points. Recovery of photobleach was assessed after severe strain injury, but many severely injured cells leaked dye into the extracellular compartment, rendering the severe injury data inconclusive (data not shown). In summary, we found that mild injury increases functional coupling and moderate injury decreases coupling. Additional experiments will be conducted to better understand the mechanisms of these changes in gap junction coupling after mechanical strain injury and their role in the pathophysiology of traumatic brain injury.

Conclusion With some notable exceptions such as the study of gliosis and astrocytic swelling, most TBI research has generally focused on the pathophysiology of neurons. However, many recent developments in the ﬁeld of glial biology demonstrate the active nature of neuronal-astrocyte signaling indicating the vital importance of astrocytes in normal brain function. Furthermore, recent studies in Neurotrauma demonstrate the vulnerability of astrocytes to CNS insults. Therefore, deﬁning the time course of astrocyte damage after TBI and uncovering the mechanisms mediating that damage are critical to a more complete understanding of TBI pathophysiology. Collectively, these ﬁndings are leading to a novel departure from traditional approaches to TBI pathophysiology that have generally concentrated on injury to neurons.

Acknowledgments Support provided by UCD Health Systems Research Award (CLF), NIH NS29995 (BGL), NIH NS45136 (BGL).

References Aguado, F., Espinosa-Parrilla, J.F., Carmona, M.A. and Soriano, E. (2002) Neuronal activity regulates correlated network properties of spontaneous calcium transients in astrocytes in situ. J. Neurosci., 22: 9430–9444.

74 Ahmed, S.M., Rzigalinski, B.A., Willoughby, K.A., Sitterding, H.A. and Ellis, E.F. (2000) Stretch-induced injury alters mitochondrial membrane potential and cellular ATP in cultured astrocytes and neurons. J. Neurochem., 74: 1951–1960. Anlauf, E. and Derouiche, A. (2005) Astrocytic exocytosis vesicles and glutamate: a high-resolution immunoﬂuorescence study. Glia, 49: 96–106. Araque, A., Sanzgiri, R.P., Parpura, V. and Haydon, P.G. (1999) Astrocyte-induced modulation of synaptic transmission. Can. J. Physiol. Pharmacol., 77: 699–706. Bambrick, L., Kristian, T. and Fiskum, G. (2004) Astrocyte mitochondrial mechanisms of ischemic brain injury and neuroprotection. Neurochem. Res., 29: 601–608. Bennett, M.V., Contreras, J.E., Bukauskas, F.F. and Saez, J.C. (2003) New roles for astrocytes: gap junction hemichannels have something to communicate. Trends Neurosci., 26: 610–617. Bezzi, P., Carmignoto, G., Pasti, L., Vesce, S., Rossi, D., Rizzini, B.L., Pozzan, T. and Volterra, A. (1998) Prostaglandins stimulate calcium-dependent glutamate release in astrocytes. Nature, 391: 281–285. Bezzi, P. and Volterra, A. (2001) A neuron-glia signalling network in the active brain. Curr. Opin. Neurobiol., 11: 387–394. Blanc, E.M., Bruce-Keller, A.J. and Mattson, M.P. (1998) Astrocytic gap junctional communication decreases neuronal vulnerability to oxidative stress-induced disruption of Ca2+ homeostasis and cell death. J. Neurochem., 70: 958–970. Boehning, D. and Snyder, S.H. (2003) Novel neural modulators. Annu. Rev. Neurosci., 26: 105–131. Bolanos, J.P. and Medina, J.M. (1996) Induction of nitric oxide synthase inhibits gap junction permeability in cultured rat astrocytes. J. Neurochem., 66: 2091–2099. Bondarenko, A. and Chesler, M. (2001a) Calcium dependence of rapid astrocyte death induced by transient hypoxia, acidosis, and extracellular ion shifts. Glia, 34: 143–149. Bondarenko, A. and Chesler, M. (2001b) Rapid astrocyte death induced by transient hypoxia, acidosis, and extracellular ion shifts. Glia, 34: 134–142. Bondarenko, A., Svichar, N. and Chesler, M. (2005) Role of Na+-H+ and Na+-Ca2+ exchange in hypoxia-related acute astrocyte death. Glia, 49: 143–152. Bullock, R., Maxwell, W.L., Graham, D.I., Teasdale, G.M. and Adams, J.H. (1991) Glial swelling following human cerebral contusion: an ultrastructural study. J. Neurol. Neurosurg. Psychiatry, 54: 427–434. Bullock, R., Zauner, A., Woodward, J.J., Myseros, J., Choi, S.C., Ward, J.D., Marmarou, A. and Young, H.F. (1998) Factors affecting excitatory amino acid release following severe human head injury. J. Neurosurg., 89: 507–518. Castejon, O.J. (1998) Morphological astrocytic changes in complicated human brain trauma: a light and electron microscopic study. Brain Inj., 12: 409–427. Chao, T.I., Rickmann, M. and Wolff, J.R. (2002) The synapseastrocyte boundry: an anatomical basis for an inegrative role of glia in synaptic transmission. In: Volterra A., Magistretti P. and Haydon P.G. (Eds.), The Tripartite Synapse: Glia in

Synaptic Transmission. Oxford University Press, Oxford, pp. 3–23. Chen, T., Willoughby, K.A. and Ellis, E.F. (2004) Group I metabotropic receptor antagonism blocks depletion of calcium stores and reduces potentiated capacitative calcium entry in strain-injured neurons and astrocytes. J. Neurotrauma, 21: 271–281. Choi, D.W. and Rothman, S.M. (1990) The role of glutamate neurotoxicity in hypoxic-ischemic neuronal death. Annu. Rev. Neurosci., 13: 171–182. Chorna, N.E., Santiago-Perez, L.I., Erb, L., Seye, C.I., Neary, J.T., Sun, G.Y., Weisman, G.A. and Gonzalez, F.A. (2004) P2Y receptors activate neuroprotective mechanisms in astrocytic cells. J. Neurochem., 91: 119–132. Contreras, J.E., Sanchez, H.A., Eugenin, E.A., Speidel, D., Theis, M., Willecke, K., Bukauskas, F.F., Bennett, M.V. and Saez, J.C. (2002) Metabolic inhibition induces opening of unapposed connexin 43 gap junction hemichannels and reduces gap junctional communication in cortical astrocytes in culture. Proc. Natl. Acad. Sci. U.S.A., 99: 495–500. Cornell-Bell, A.H. and Finkbeiner, S.M. (1991) Ca2+ waves in astrocytes. Cell Calcium, 12: 185–204. Cornell-Bell, A.H., Finkbeiner, S.M., Cooper, M.S. and Smith, S.J. (1990) Glutamate induces calcium waves in cultured astrocytes: long-range glial signaling. Science, 247: 470–473. Cotrina, M.L., Kang, J., Lin, J.H., Bueno, E., Hansen, T.W., He, L., Liu, Y. and Nedergaard, M. (1998a) Astrocytic gap junctions remain open during ischemic conditions. J. Neurosci., 18: 2520–2537. Cotrina, M.L., Lin, J.H., Alves-Rodrigues, A., Liu, S., Li, J., Azmi-Ghadimi, H., Kang, J., Naus, C.C. and Nedergaard, M. (1998b) Connexins regulate calcium signaling by controlling ATP release. Proc. Natl. Acad. Sci. U.S.A., 95: 15735–15740. Cotrina, M.L., Lin, J.H., Lopez-Garcia, J.C., Naus, C.C. and Nedergaard, M. (2000) ATP-mediated glia signaling. J. Neurosci., 20: 2835–2844. Crippa, D., Schenk, U., Francolini, M., Rosa, P., Verderio, C., Zonta, M., Pozzan, T., Matteoli, M. and Carmignoto, G. (2006) Synaptobrevin2-expressing vesicles in rat astrocytes: insights into molecular characterization, dynamics and exocytosis. J. Physiol., 570: 567–582. Daikhin, Y. and Yudkoff, M. (2000) Compartmentation of brain glutamate metabolism in neurons and glia. J. Nutr., 130: 1026S–1031S. D’Ambrosio, R., Maris, D.O., Grady, M.S., Winn, H.R. and Janigro, D. (1999) Impaired K(+) homeostasis and altered electrophysiological properties of post-traumatic hippocampal glia. J. Neurosci., 19: 8152–8162. Danbolt, N.C. (1994) The high afﬁnity uptake system for excitatory amino acids in the brain. Prog. Neurobiol., 44: 377–396. Danbolt, N.C. (2001) Glutamate uptake. Prog. Neurobiol., 65: 1–105. Dani, J.W., Chernjavsky, A. and Smith, S.J. (1992) Neuronal activity triggers calcium waves in hippocampal astrocyte networks. Neuron, 8: 429–440.

75 D’Ascenzo, M., Vairano, M., Andreassi, C., Navarra, P., Azzena, G.B. and Grassi, C. (2004) Electrophysiological and molecular evidence of L-(Cav1), N-(Cav2.2), and R-(Cav2.3) type Ca2+ channels in rat cortical astrocytes. Glia, 45: 354–363. Dermietzel, R., Meier, C., Bukauskas, F. and Spray, D.C. (2003) Following tracks of hemichannels. Cell Commun. Adhes., 10: 335–340. Dietrich, W.D., Alonso, O., Busto, R., Globus, M.Y. and Ginsberg, M.D. (1994) Post-traumatic brain hypothermia reduces histopathological damage following concussive brain injury in the rat. Acta Neuropathol., 87: 250–258. Du, Y., Bales, K.R., Dodel, R.C., Hamilton-Byrd, E., Horn, J.W., Czilli, D.L., Simmons, L.K., Ni, B. and Paul, S.M. (1997) Activation of a caspase 3-related cysteine protease is required for glutamate-mediated apoptosis of cultured cerebellar granule neurons. Proc. Natl. Acad. Sci. U.S.A., 94: 11657–11662. Ellis, E.F., McKinney, J.S., Willoughby, K.A., Liang, S. and Povlishock, J.T. (1995) A new model for rapid stretchinduced injury of cells in culture: characterization of the model using astrocytes. J. Neurotrauma, 12: 325–339. Faden, A.I., Demediuk, P., Panter, S.S. and Vink, R. (1989) The role of excitatory amino acids and NMDA receptors in traumatic brain injury. Science, 244: 798–800. Faden, A.I., Ivanova, S.A., Yakovlev, A.G. and Mukhin, A.G. (1997) Neuroprotective effects of group III mGluR in traumatic neuronal injury. J. Neurotrauma, 14: 885–895. Fitch, M.T., Doller, C., Combs, C.K., Landreth, G.E. and Silver, J. (1999) Cellular and molecular mechanisms of glial scarring and progressive cavitation: in vivo and in vitro analysis of inﬂammation-induced secondary injury after CNS trauma. J. Neurosci., 19: 8182–8198. Floyd, C.L., Gorin, F.A. and Lyeth, B.G. (2005) Mechanical strain injury increases intracellular sodium and reverses Na(+)/ Ca(2+) exchange in cortical astrocytes. Glia, 51: 35–46. Floyd, C.L., Rzigalinski, B.A., Sitterding, H.A., Willoughby, K.A. and Ellis, E.F. (2004) Antagonism of group I metabotropic glutamate receptors and PLC attenuates increases in inositol trisphosphate and reduces reactive gliosis in straininjured astrocytes. J. Neurotrauma, 21: 205–216. Floyd, C.L., Rzigalinski, B.A., Weber, J.T., Sitterding, H.A., Willoughby, K.A. and Ellis, E.F. (2001) Traumatic injury of cultured astrocytes alters inositol (1,4,5)- trisphosphate-mediated signaling. Glia, 33: 12–23. Frandsen, A. and Schousboe, A. (2003) AMPA receptormediated neurotoxicity: role of Ca2+ and desensitization. Neurochem. Res., 28: 1495–1499. Frantseva, M.V., Kokarovtseva, L. and Perez Velazquez, J.L. (2002) Ischemia-induced brain damage depends on speciﬁc gap-junctional coupling. J. Cereb. Blood Flow Metab., 22: 453–462. Gabryel, B., Adamczyk, J., Huzarska, M., Pudelko, A. and Trzeciak, H.I. (2002) Aniracetam attenuates apoptosis of astrocytes subjected to simulated ischemia in vitro. Neurotoxicology, 23: 385–395. Garcia, J.H., Kalimo, H., Kamijyo, Y. and Trump, B.F. (1977) Cellular events during partial cerebral ischemia. I. Electron

microscopy of feline cerebral cortex after middle-cerebralartery occlusion. Virchows Arch. B. Cell Pathol., 25: 191–206. Giaume, C. and McCarthy, K.D. (1996) Control of gap-junctional communication in astrocytic networks. Trends Neurosci., 19: 319–325. Giffard, R.G. and Swanson, R.A. (2005) Ischemia-induced programmed cell death in astrocytes. Glia, 50: 299–306. Glaum, S.R., Holzwarth, J.A. and Miller, R.J. (1990) Glutamate receptors activate Ca2+ mobilization and Ca2+ inﬂux into astrocytes. Proc. Natl. Acad. Sci. U.S.A., 87: 3454–3458. Goforth, P.B., Ellis, E.F. and Satin, L.S. (1999) Enhancement of AMPA-mediated current after traumatic injury in cortical neurons. J. Neurosci., 19: 7367–7374. Goforth, P.B., Ellis, E.F. and Satin, L.S. (2004) Mechanical injury modulates AMPA receptor kinetics via an NMDA receptor-dependent pathway. J. Neurotrauma, 21: 719–732. Goldman, W.F., Yarowsky, P.J., Juhaszova, M., Krueger, B.K. and Blaustein, M.P. (1994) Sodium/calcium exchange in rat cortical astrocytes. J. Neurosci., 14: 5834–5843. Graham, S.H. and Chen, J. (2001) Programmed cell death in cerebral ischemia. J. Cereb. Blood Flow Metab., 21: 99–109. Grinstein, S. and Rothstein, A. (1986) Mechanisms of regulation of the Na+/H+ exchanger. J. Membr. Biol., 90: 1–12. 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. Hamm, R.J., Pike, B.R., O’Dell, D.M. and Lyeth, B.G. (1994) Traumatic brain injury enhances the amnesic effect of an NMDA antagonist in rats. J. Neurosurg., 81(2): 267–271. 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. Hassinger, T.D., Guthrie, P.B., Atkinson, P.B., Bennett, M.V. and Kater, S.B. (1996) An extracellular signaling component in propagation of astrocytic calcium waves. Proc. Natl. Acad. Sci. U.S.A., 93: 13268–13273. Haydon, P.G. (2001) Glia: listening and talking to the synapse. Nat. Rev. Neurosci., 2: 185–193. Hayes, R.L., Jenkins, L.W., Lyeth, B.G., Balster, R.L., Robinson, S.E., Clifton, G.L., Stubbins, J.F. and Young, H.F. (1988) Pretreatment with phencyclidine, an N-methyl-Daspartate antagonist, attenuates long-term behavioral deficits in the rat produced by traumatic brain injury. J. Neurotrauma, 5: 259–574. Hicks, R.R., Smith, D.H., Gennarelli, T.A. and McIntosh, T.K. (1994) Kynurenate is neuroprotective following experimental brain injury in the rat. Brain Res., 655: 91–96. Hoffman, S.W., Rzigalinski, B.A., Willoughby, K.A. and Ellis, E.F. (2000) Astrocytes generate isoprostanes in response to trauma or oxygen radicals. J. Neurotrauma, 17(5): 415–420. Holzwarth, J.A., Gibbons, S.J., Brorson, J.R., Philipson, L.H. and Miller, R.J. (1994) Glutamate receptor agonists stimulate diverse calcium responses in different types of cultured rat cortical glial cells. J. Neurosci., 14: 1879–1891. Hu, Y. and Wilson, G.S. (1997) A temporary local energy pool coupled to neuronal activity: ﬂuctuations of extracellular

76 lactate levels in rat brain monitored with rapid-response enzyme-based sensor. J. Neurochem., 69: 1484–1490. Hyder, F., Chase, J.R., Behar, K.L., Mason, G.F., Siddeek, M., Rothman, D.L. and Shulman, R.G. (1996) Increased tricarboxylic acid cycle ﬂux in rat brain during forepaw stimulation detected with 1H[13C]NMR. Proc. Natl. Acad. Sci. U.S.A., 93: 7612–7617. Illes, P. and Ribeiro, J.A. (2004) Molecular physiology of P2 receptors in the central nervous system. Eur. J. Pharmacol., 483: 5–17. Innocenti, B., Parpura, V. and Haydon, P.G. (2000) Imaging extracellular waves of glutamate during calcium signaling in cultured astrocytes. J. Neurosci., 20: 1800–1808. Jakubovicz, D.E. and Klip, A. (1989) Lactic acid-induced swelling in C6 glial cells via Na+/H+ exchange. Brain Res., 485: 215–224. Jenkins, L.W., Becker, D.P. and Coburn, T.H. (1982) A quantitative analysis of glial swelling and ischemic neuronal injury following complete cerebral ischemia. In: Go K.G. and Baethmann A. (Eds.), Recent Progress in the Study and Therapy of Brain Edema. Plenum Press, New York, pp. 523–537. Jiang, X.B., Ohno, K., Qian, L., Tominaga, B., Kuroiwa, T., Nariai, T. and Hirakawa, K. (2000) Changes in local cerebral blood ﬂow, glucose utilization, and mitochondrial function following traumatic brain injury in rats. Neurol. Med. Chir., 40: 16–28. Johnson, E.A., Svetlov, S.I., Wang, K.K., Hayes, R.L. and Pineda, J.A. (2005) Cell-speciﬁc DNA fragmentation may be attenuated by a survivin-dependent mechanism after traumatic brain injury in rats. Exp. Brain Res., 167: 17–26. Kang, J., Jiang, L., Goldman, S.A. and Nedergaard, M. (1998) Astrocyte-mediated potentiation of inhibitory synaptic transmission. Nat. Neurosci., 1: 683–692. Kanner, B.I. and Schuldiner, S. (1987) Mechanism of transport and storage of neurotransmitters. CRC Crit. Rev. Biochem., 22: 1–38. Katayama, Y., Becker, D.P., Tamura, T. and Hovda, D.A. (1990) Massive increases in extracellular potassium and the indiscriminate release of glutamate following concussive brain injury. J. Neurosurg., 73: 889–900. Kawamata, T., Katayama, Y., Hovda, D.A., Yoshino, A. and Becker, D.P. (1992) Administration of excitatory amino acid antagonists via microdialysis attenuates the increase in glucose utilization seen following concussive brain injury. J. Cereb. Blood Flow Metab., 12: 12–24. Kawamata, T., Katayama, Y., Hovda, D.A., Yoshino, A. and Becker, D.P. (1995) Lactate accumulation following concussive brain injury: the role of ion ﬂuxes induced by excitatory amino acids. Brain Res., 674: 196–204. Keane, R.W., Srinivasan, A., Foster, L.M., Testa, M.P., Ord, T., Nonner, D., Wang, H.G., Reed, J.C., Bredesen, D.E. and Kayalar, C. (1997) Activation of CPP32 during apoptosis of neurons and astrocytes. J. Neurosci. Res., 48: 168–180. Kelleher, J.A., Chan, P.H., Chan, T.Y. and Gregory, G.A. (1993) Modiﬁcation of hypoxia-induced injury in cultured rat astrocytes by high levels of glucose. Stroke, 24: 855–863.

Kempski, O., Staub, F., Jansen, M., Schodel, F. and Baethmann, A. (1988) Glial swelling during extracellular acidosis in vitro. Stroke, 19: 385–392. Kimelberg, H.K. (1992) Astrocytic edema in CNS trauma. J. Neurotrama, 9(Suppl. 1): S71–S81. Kimelberg, H.K., Goderie, S.K., Higman, S., Pang, S. and Waniewski, R.A. (1990) Swelling-induced release of glutamate, aspartate, and taurine from astrocyte cultures. J. Neurosci., 10: 1583–1591. Kondou, H., Inagaki, N. and Fukui, H. (1994) Formation of inositol phosphates mediated by M3 muscarinic receptors in type-1 and type-2 astrocytes from neonatal rat cerebral cortex. Neurosci. Lett., 180: 131–134. Koura, S.S., Doppenberg, E.M., Marmarou, A., Choi, S., Young, H.F. and Bullock, R. (1998) Relationship between excitatory amino acid release and outcome after severe human head injury. Acta Neurochir. Suppl., 71: 244–246. Kraig, R.P. and Chesler, M. (1990) Astrocytic acidosis in hyperglycemic and complete ischemia. J. Cereb. Blood Flow Metab., 10: 104–114. Kreft, M., Stenovec, M., Rupnik, M., Grilc, S., Krzan, M., Potokar, M., Pangrsic, T., Haydon, P.G. and Zorec, R. (2004) Properties of Ca(2+)-dependent exocytosis in cultured astrocytes. Glia, 46: 437–445. Lamb, R.G., Harper, C.C., McKinney, J.S., Rzigalinski, B.A. and Ellis, E.F. (1997) Alterations in phosphatidylcholine metabolism of stretch-injured cultured rat astrocytes. J. Neurochem., 68(5): 1904–1910. Lenz, G., Gottfried, C., Luo, Z., Avruch, J., Rodnight, R., Nie, W.J., Kang, Y. and Neary, J.T. (2000) P(2Y) purinoceptor subtypes recruit different mek activators in astrocytes. Br. J. Pharmacol., 129: 927–936. Lin, J.H., Weigel, H., Cotrina, M.L., Liu, S., Bueno, E., Hansen, A.J., Hansen, T.W., Goldman, S. and Nedergaard, M. (1998) Gap-junction-mediated propagation and ampliﬁcation of cell injury. Nat. Neurosci., 1: 494–500. Liu, D., Smith, C.L., Barone, F.C., Ellison, J.A., Lysko, P.G., Li, K. and Simpson, I.A. (1999) Astrocytic demise precedes delayed neuronal death in focal ischemic rat brain. Brain Res. Mol. Brain Res., 68: 29–41. Liu, Q.S., Xu, Q., Arcuino, G., Kang, J. and Nedergaard, M. (2004a) Astrocyte-mediated activation of neuronal kainate receptors. Proc. Natl. Acad. Sci. U.S.A., 101: 3172–3177. Liu, Q.S., Xu, Q., Kang, J. and Nedergaard, M. (2004b) Astrocyte activation of presynaptic metabotropic glutamate receptors modulates hippocampal inhibitory synaptic transmission. Neuron Glia Biol., 1: 307–316. Lukaszevicz, A.C., Sampaio, N., Guegan, C., Benchoua, A., Couriaud, C., Chevalier, E., Sola, B., Lacombe, P. and Onteniente, B. (2002) High sensitivity of protoplasmic cortical astroglia to focal ischemia. J. Cereb. Blood Flow Metab., 22: 289–298. Lyeth, B.G., Gong, Q.Z., Dhillon, H.S. and Prasad, M.R. (1996) Effects of muscarinic receptor antagonism on the phosphatidylinositol bisphosphate signal transduction pathway after experimental brain injury. Brain Res., 742: 63–70.

77 Lyeth, B.G., Gong, Q.Z., Shields, S., Muizelaar, J.P. and Berman, R.F. (2001) Group I metabotropic glutamate antagonist reduces acute neuronal degeneration and behavioral deficits after traumatic brain injury in rats. Exp. Neurol., 169: 191–199. Lyeth, B.G., Jenkins, L.W., Hamm, R.J., Dixon, C.E., Phillips, L.L., Clifton, G.L., Young, H.F. and Hayes, R.L. (1990) Prolonged memory impairment in the absence of hippocampal cell death following traumatic brain injury in the rat. Brain Res., 526: 249–258. MacVicar, B.A., Hochman, D., Delay, M.J. and Weiss, S. (1991) Modulation of intracellular Ca++ in cultured astrocytes by inﬂux through voltage-activated Ca++ channels. Glia, 4: 448–455. Margulies, S.S., Thibault, L.E. and Gennarelli, T.A. (1990) Physical model simulations of brain injury in the primate. J. Biomech., 23: 823–836. Martinez, A.D. and Saez, J.C. (2000) Regulation of astrocyte gap junctions by hypoxia-reoxygenation. Brain Res. Brain Res. Rev., 32: 250–258. McGraw, J., Hiebert, G.W. and Steeves, J.D. (2001) Modulating astrogliosis after neurotrauma. J. Neurosci. Res., 63: 109–115. McIntosh, T.K., Juhler, M. and Wieloch, T. (1998) Novel pharmacologic strategies in the treatment of experimental traumatic brain injury: 1998. J. Neurotrauma, 15: 731–769. McIntosh, T.K., Vink, R., Noble, L., Yamakami, I., Fernyak, S., Soares, H. and Faden, A.I. (1989) Traumatic brain injury in the rat: characterization of a lateral ﬂuid-percussion model. Neuroscience, 28: 233–244. McKinney, J.S., Willoughby, K.A., Liang, S. and Ellis, E.F. (1996) Stretch-induced injury of cultured neuronal, glial, and endothelial cells: effect of polyethylene glycol-conjugated superoxide dismutase. Stroke, 27: 934–940. Meaney, D.F., Smith, D.H., Shreiber, D.I., Bain, A.C., Miller, R.T., Ross, D.T. and Gennarelli, T.A. (1995) Biomechanical analysis of experimental diffuse axonal injury. J. Neurotrauma, 12: 689–694. Mothet, J.P., Pollegioni, L., Ouanounou, G., Martineau, M., Fossier, P. and Baux, G. (2005) Glutamate receptor activation triggers a calcium-dependent and SNARE protein-dependent release of the gliotransmitter D-serine. Proc. Natl. Acad. Sci. U.S.A., 102: 5606–5611. Mukhin, A., Fan, L. and Faden, A.I. (1996) Activation of metabotropic glutamate receptor subtype mGluR1 contributes to post-traumatic neuronal injury. J. Neurosci., 16: 6012–6020. Mukhin, A.G., Ivanova, S.A. and Faden, A.I. (1997) mGluR modulation of post-traumatic neuronal death: role of NMDA receptors. Neuroreport, 8: 2561–2566. Muller, H.W., Junghans, U. and Kappler, J. (1995) Astroglial neurotrophic and neurite-promoting factors. Pharmacol. Ther., 65: 1–18. Musil, L.S., Cunningham, B.A., Edelman, G.M. and Goodenough, D.A. (1990) Differential phosphorylation of the gap junction protein connexin43 in junctional communication-

competent and -deﬁcient cell lines. J. Cell Biol., 111: 2077–2088. Nagy, J.I. and Li, W.E. (2000) A brain slice model for in vitro analyses of astrocytic gap junction and connexin43 regulation: actions of ischemia, glutamate and elevated potassium. Eur. J. Neurosci., 12: 4567–4572. Namura, S., Zhu, J., Fink, K., Endres, M., Srinivasan, A., Tomaselli, K.J., Yuan, J. and Moskowitz, M.A. (1998) Activation and cleavage of caspase-3 in apoptosis induced by experimental cerebral ischemia. J. Neurosci., 18: 3659–3668. Neary, J.T. and Kang, Y. (2006) P2 purinergic receptors signal to glycogen synthase kinase-3beta in astrocytes. J. Neurosci. Res., 84: 515–524. Neary, J.T., Kang, Y., Tran, M. and Feld, J. (2005) Traumatic injury activates protein kinase B/Akt in cultured astrocytes: role of extracellular ATP and P2 purinergic receptors. J. Neurotrauma, 22: 491–500. Neary, J.T., Kang, Y., Willoughby, K.A. and Ellis, E.F. (2003) Activation of extracellular signal-regulated kinase by stretchinduced injury in astrocytes involves extracellular ATP and P2 purinergic receptors. J. Neurosci., 23: 2348–2356. Neary, J.T., Rathbone, M.P., Cattabeni, F., Abbracchio, M.P. and Burnstock, G. (1996) Trophic actions of extracellular nucleotides and nucleosides on glial and neuronal cells. Trends Neurosci., 19: 13–18. Neary, J.T. and Zhu, Q. (1994) Signaling by ATP receptors in astrocytes. Neuroreport, 5: 1617–1620. Nedergaard, M., Goldman, S.A., Desai, S. and Pulsinelli, W.A. (1991a) Acid-induced death in neurons and glia. J. Neurosci., 11: 2489–2497. Nedergaard, M., Kraig, R.P., Tanabe, J. and Pulsinelli, W.A. (1991b) Dynamics of interstitial and intracellular pH in evolving brain infarct. Am. J. Physiol., 260: R581–R588. Nett, W.J., Oloff, S.H. and McCarthy, K.D. (2002) Hippocampal astrocytes in situ exhibit calcium oscillations that occur independent of neuronal activity. J. Neurophysiol., 87: 528–537. Newcomb, J.K., Zhao, X., Pike, B.R. and Hayes, R.L. (1999) Temporal proﬁle of apoptotic-like changes in neurons and astrocytes following controlled cortical impact injury in the rat. Exp. Neurol., 158: 76–88. Newman, E.A. (1991) Sodium-bicarbonate cotransport in retinal Muller (glial) cells of the salamander. J. Neurosci., 11: 3972–3983. Newman, E.A. (2003) Glial cell inhibition of neurons by release of ATP. J. Neurosci., 23: 1659–1666. Newman, E.A. and Zahs, K.R. (1997) Calcium waves in retinal glial cells. Science, 275: 844–847. Newman, E.A. and Zahs, K.R. (1998) Modulation of neuronal activity by glial cells in the retina. J. Neurosci., 18: 4022–4028. Nicotera, P., Ankarcrona, M., Bonfoco, E., Orrenius, S. and Lipton, S.A. (1997) Neuronal necrosis and apoptosis: two distinct events induced by exposure to glutamate or oxidative stress. Adv. Neurol., 72: 95–101.

78 Nilsson, P., Hillered, L., Ponten, U. and Ungerstedt, U. (1990) Changes in cortical extracellular levels of energy-related metabolites and amino acids following concussive brain injury in rats. J. Cereb. Blood Flow Metab., 10: 631–637. Nimmerjahn, A., Kirchhoff, F., Kerr, J.N. and Helmchen, F. (2004) Sulforhodamine 101 as a speciﬁc marker of astroglia in the neocortex in vivo. Nat. Methods, 1: 31–37. Ohsumi, A., Nawashiro, H., Otani, N., Ooigawa, H., Toyooka, T., Yano, A., Nomura, N. and Shima, K. (2006) Alteration of gap junction proteins (connexins) following lateral ﬂuid percussion injury in rats. Acta Neurochir. Suppl., 96: 148–150. Olney, J.W. and Ho, O.L. (1970) Brain damage in infant mice following oral intake of glutamate, aspartate or cysteine. Nature, 227: 609–611. Paemeleire, K. and Leybaert, L. (2000) Ionic changes accompanying astrocytic intercellular calcium waves triggered by mechanical cell damaging stimulation. Brain Res., 857: 235–245. Panatier, A., Theodosis, D.T., Mothet, J.P., Touquet, B., Pollegioni, L., Poulain, D.A. and Oliet, S.H. (2006) Glia-derived D-serine controls NMDA receptor activity and synaptic memory. Cell, 125: 775–784. Parpura, V., Basarsky, T.A., Liu, F., Jeftinija, K., Jeftinija, S. and Haydon, P.G. (1994) Glutamate-mediated astrocyteneuron signalling. Nature, 369: 744–747. Parpura, V. and Haydon, P.G. (2000) Physiological astrocytic calcium levels stimulate glutamate release to modulate adjacent neurons. Proc. Natl. Acad. Sci. U.S.A., 97: 8629–8634. Parpura, V., Liu, F., Brethorst, S., Jeftinija, K., Jeftinija, S. and Haydon, P.G. (1995) Alpha-latrotoxin stimulates glutamate release from cortical astrocytes in cell culture. FEBS Lett., 360: 266–270. Parri, H.R. and Crunelli, V. (2003) The role of Ca2+ in the generation of spontaneous astrocytic Ca2+ oscillations. Neuroscience, 120: 979–992. 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. Peakman, M.C. and Hill, S.J. (1995) Adenosine A1 receptormediated changes in basal and histamine-stimulated levels of intracellular calcium in primary rat astrocytes. Br. J. Pharmacol., 115: 801–810. Phillis, J.W., Estevez, A.Y., Guyot, L.L. and O’Regan, M.H. (1999) 5-(N-Ethyl-N-isopropyl)-amiloride, an Na(+)-H(+) exchange inhibitor, protects gerbil hippocampal neurons from ischemic injury. Brain Res., 839: 199–202. Phillis, J.W., O’Regan, M.H. and Song, D. (1998) 5-(N-EthylN-isopropyl)-amiloride inhibits amino acid release from the ischemic rat cerebral cortex: role of Na+-H+ exchange. Brain Res., 812: 297–300. Plesnila, N., Haberstok, J., Peters, J., Kolbl, I., Baethmann, A. and Staub, F. (1999) Effect of lactacidosis on cell volume and intracellular pH of astrocytes. J. Neurotrauma, 16: 831–841. Plum, F. (1983) What causes infarction in ischemic brain?: the Robert Wartenberg lecture. Neurology, 33: 222–233.

Porter, J.T. and McCarthy, K.D. (1996) Hippocampal astrocytes in situ respond to glutamate released from synaptic terminals. J. Neurosci., 16: 5073–5081. Raichle, M.E. (1991) The metabolic requirements of functional activity in the human brain: a positron emission tomography study. Adv. Exp. Med. Biol., 291: 1–4. Reier, P.J., Stensaas, L.J. and Guth, L. (1983) The astrocytic scar as an impediment to regeneration in the central nervous system. In: Kao C.C., Bunge R.P. and Reier P.J. (Eds.), Spinal Cord Reconstruction. Raven Press, New York, pp. 163–198. Ridet, J.L., Malhotra, S.K., Privat, A. and Gage, F.H. (1997) Reactive astrocytes: cellular and molecular cues to biological function. Trends Neurosci., 20: 570–577. Ringel, F., Chang, R.C., Staub, F., Baethmann, A. and Plesnila, N. (2000) Contribution of anion transporters to the acidosis-induced swelling and intracellular acidiﬁcation of glial cells. J. Neurochem., 75: 125–132. Rink, A., Fung, K.M., Trojanowski, J.Q., Lee, V.M., Neugebauer, E. and McIntosh, T.K. (1995) Evidence of apoptotic cell death after experimental traumatic brain injury in the rat. Am. J. Pathol., 147: 1575–1583. Rouach, N., Avignone, E., Meme, W., Koulakoff, A., Venance, L., Blomstrand, F. and Giaume, C. (2002) Gap junctions and connexin expression in the normal and pathological central nervous system. Biol. Cell, 94: 457–475. Rzigalinski, B.A., Weber, J.T., Willoughby, K.A. and Ellis, E.F. (1998) Intracellular free calcium dynamics in stretchinjured astrocytes. J. Neurochem., 70: 2377–2385. Schurr, A., Miller, J.J., Payne, R.S. and Rigor, B.M. (1999) An increase in lactate output by brain tissue serves to meet the energy needs of glutamate-activated neurons. J. Neurosci., 19: 34–39. Schwiebert, E.M. (2001) ATP release mechanisms, ATP receptors and purinergic signalling along the nephron. Clin. Exp. Pharmacol. Physiol., 28: 340–350. Shao, Y. and McCarthy, K.D. (1995) Receptor-mediated calcium signals in astroglia: multiple receptors, common stores and all-or-nothing responses. Cell Calcium, 17: 187–196. Siesjo, B.K. (1988) Acidosis and ischemic brain damage. Neurochem. Pathol., 9: 31–88. Silver, I.A. and Erecinska, M. (1997) Energetic demands of the Na+/K+ ATPase in mammalian astrocytes. Glia, 21: 35–45. Siushansian, R., Bechberger, J.F., Cechetto, D.F., Hachinski, V.C. and Naus, C.C. (2001) Connexin43 null mutation increases infarct size after stroke. J. Comp. Neurol., 440: 387–394. Slemmer, J.E., Matser, E.J., De Zeeuw, C.I. and Weber, J.T. (2002) Repeated mild injury causes cumulative damage to hippocampal cells. Brain, 125: 2699–2709. Slemmer, J.E., Weber, J.T. and De Zeeuw, C.I. (2004) Cell death, glial protein alterations and elevated S-100 beta release in cerebellar cell cultures following mechanically induced trauma. Neurobiol. Dis., 15: 563–572. Smith, I.F., Plant, L.D., Boyle, J.P., Skinner, R.A., Pearson, H.A. and Peers, C. (2003) Chronic hypoxia potentiates

79 capacitative Ca2+ entry in type-I cortical astrocytes. J. Neurochem., 85: 1109–1116. Sochocka, E., Juurlink, B.H., Code, W.E., Hertz, V., Peng, L. and Hertz, L. (1994) Cell death in primary cultures of mouse neurons and astrocytes during exposure to and ‘recovery’ from hypoxia, substrate deprivation and simulated ischemia. Brain Res., 638: 21–28. Staub, F., Baethmann, A., Peters, J., Weigt, H. and Kempski, O. (1990) Effects of lactacidosis on glial cell volume and viability. J. Cereb. Blood Flow Metab., 10: 866–876. Stout, C.E., Costantin, J.L., Naus, C.C. and Charles, A.C. (2002) Intercellular calcium signaling in astrocytes via ATP release through connexin hemichannels. J. Biol. Chem., 277: 10482–10488. Swanson, R.A., Farrell, K. and Stein, B.A. (1997) Astrocyte energetics, function, and death under conditions of incomplete ischemia: a mechanism of glial death in the penumbra. Glia, 21: 142–153. Takahashi, M., Billups, B., Rossi, D., Sarantis, M., Hamann, M. and Attwell, D. (1997) The role of glutamate transporters in glutamate homeostasis in the brain. J. Exp. Biol., 200: 401–409. Takuma, K., Baba, A. and Matsuda, T. (2004) Astrocyte apoptosis: implications for neuroprotection. Prog. Neurobiol., 72: 111–127. Tanaka, J., Ichikawa, R., Watanabe, M., Tanaka, K. and Inoue, Y. (1997) Extra-junctional localization of glutamate transporter EAAT4 at excitatory Purkinje cell synapses. Neuroreport, 8: 2461–2464. Tavalin, S.J., Ellis, E.F. and Satin, L.S. (1995) Mechanical perturbation of cultured cortical neurons reveals a stretchinduced delayed depolarization. J. Neurophysiol., 74(6): 2767–2773. Tran, M.D. and Neary, J.T. (2006) Purinergic signaling induces thrombospondin-1 expression in astrocytes. Proc. Natl. Acad. Sci. U.S.A., 103: 9321–9326. Tsuchida, T., Kato, T., Yamada, A. and Kawamoto, K. (2002) Cycloheximide induces apoptosis of astrocytes. Pathol. Int., 52: 181–185. VanSlyke, J.K. and Musil, L.S. (2005) Cytosolic stress reduces degradation of connexin43 internalized from the cell surface and enhances gap junction formation and function. Mol. Biol. Cell, 16: 5247–5257. Verkhratsky, A., Orkand, R.K. and Kettenmann, H. (1998) Glial calcium: homeostasis and signaling function. Physiol. Rev., 78: 99–141. Vernadakis, A. (1996) Glia-neuron intercommunications and synaptic plasticity. Prog. Neurobiol., 49: 185–214. Vesce, S., Bezzi, P. and Volterra, A. (1999) The active role of astrocytes in synaptic transmission. Cell Mol. Life Sci., 56: 991–1000. Volterra, A. and Meldolesi, J. (2005) Astrocytes, from brain glue to communication elements: the revolution continues. Nat. Rev. Neurosci., 6: 626–640. Weber, J.T., Rzigalinski, B.A., Willoughby, K.A., Moore, S.F. and Ellis, E.F. (1999) Alterations in calcium-mediated signal

transduction after traumatic injury of cortical neurons. Cell Calcium, 26: 289–299. Weisman, G.A., Wang, M., Kong, Q., Chorna, N.E., Neary, J.T., Sun, G.Y., Gonzalez, F.A., Seye, C.I. and Erb, L. (2005) Molecular determinants of P2Y2 nucleotide receptor function: implications for proliferative and inﬂammatory pathways in astrocytes. Mol. Neurobiol., 31: 169–183. Willoughby, K.A., Kleindienst, A., Muller, C., Chen, T., Muir, J.K. and Ellis, E.F. (2004) S100B protein is released by in vitro trauma and reduces delayed neuronal injury. J. Neurochem., 91: 1284–1291. Wilsch, V.W., Pidoplichko, V.I., Opitz, T., Shinozaki, H. and Reymann, K.G. (1994) Metabotropic glutamate receptor agonist DCG-IV as NMDA receptor agonist in immature rat hippocampal neurons. Eur. J. Pharmacol., 262: 287–291. Yamamoto, T., Hertzberg, E.L. and Nagy, J.I. (1990a) Epitopes of gap junctional proteins localized to neuronal subsurface cisterns. Brain Res., 527: 135–139. Yamamoto, T., Ochalski, A., Hertzberg, E.L. and Nagy, J.I. (1990b) On the organization of astrocytic gap junctions in rat brain as suggested by LM and EM immunohistochemistry of connexin43 expression. J. Comp. Neurol., 302: 853–883. Yu, A.C., Gregory, G.A. and Chan, P.H. (1989) Hypoxia-induced dysfunctions and injury of astrocytes in primary cell cultures. J. Cereb. Blood Flow Metab., 9: 20–28. Yu, A.C., Wong, H.K., Yung, H.W. and Lau, L.T. (2001) Ischemia-induced apoptosis in primary cultures of astrocytes. Glia, 35: 121–130. Zauner, A., Bullock, R., Kuta, A.J., Woodward, J. and Young, H.F. (1996) Glutamate release and cerebral blood ﬂow after severe human head injury. Acta Neurochir. Suppl., 67: 40–44. Zhang, L., Rzigalinski, B.A., Ellis, E.F. and Satin, L.S. (1996) Reduction of voltage-dependent Mg2+ blockade of NMDA current in mechanically injured neurons. Science, 274(5294): 1921–1923. Zhao, X., Ahram, A., Berman, R.F., Muizelaar, J.P. and Lyeth, B.G. (2003) Early loss of astrocytes after experimental traumatic brain injury. Glia, 44: 140–152. Zhong, C., Zhao, X., Van, K.C., Bzdega, T., Smyth, A., Zhou, J., Kozikowski, A.P., Jiang, J., O’Connor, W.T., Berman, R.F., Neale, J.H. and Lyeth, B.G. (2006) NAAG peptidase inhibitor increases dialysate NAAG and reduces glutamate, aspartate and GABA levels in the dorsal hippocampus following ﬂuid percussion injury in the rat. J. Neurochem., 97: 1015–1025. Zvalova, D., Cordier, J., Mesnil, M., Junier, M.P. and Chneiweiss, H. (2004) p38/SAPK2 controls gap junction closure in astrocytes. Glia, 46: 323–333. Zwienenberg, M., Gong, Q.Z., Berman, R.F., Muizelaar, J.P. and Lyeth, B.G. (2001) The effect of groups II and III metabotropic glutamate receptor activation on neuronal injury in a rodent model of traumatic brain injury. Neurosurgery, 48: 1119–1127.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 6

Rescuing neurons and glia: is inhibition of apoptosis useful? E. Kovesdi1, E. Czeiter1, A. Tamas2, D. Reglodi2, D. Szellar1, J. Pal3, P. Bukovics1,3, T. Doczi1 and A. Buki1, 1

Department of Neurosurgery, University Medical School, Pe´cs University, Pe´cs, Hungary, Re´t u. 2. H-7624, Hungary Department of Anatomy, University Medical School, Pe´cs University, Pe´cs, Hungary, Szigeti u. 12. H-7624, Hungary 3 Neurobiology Research Group of the Hungarian Academy of Sciences, University Medical School, Pe´cs University, Pe´cs, Hungary, Re´t u. 2. H-7624, Hungary 2

Abstract: Traumatic brain injury (TBI) represents a leading cause of death in western countries. Despite all research efforts we still lack any pharmacological agent that could effectively be utilized in the clinical treatment of TBI. Detailed unraveling of the pathobiological processes initiated by/operant in TBI is a prerequisite to the development of rational therapeutic interventions. In this review we provide a summary of those therapeutic interventions purported to inhibit the cell death (CD) cascades ignited in TBI. On noxious stimuli three major forms of CD, apoptosis, autophagia and necrosis may occur. Apoptosis can be induced either via the mitochondrial (intrinsic) or the receptor mediated (extrinsic) pathway; endoplasmic reticular stress is the third trigger of caspase-mediated apoptotic processes. Although, theoretically pancaspase inhibition could be an efﬁcient tool to limit apoptosis and thereby the extent of TBI, potential cross-talk between various avenues of CD suggests that more upstream events, particularly the preservation of the cellular energy homeostasis (cyclosporine-A, poly ADP ribose polymerase (PARP) inhibition, hypothermia treatment) may represent more efﬁcient therapeutic targets hopefully also translated to the clinical care of the severely head injured. Keywords: cell death; caspases; calpain; traumatic brain injury; traumatically induced axonal injury; necrosis; apoptosis per year for TBI is $60 billion. The TBI-related death rate of the population under 35 years of age is 3.5 times that of cancer and heart disease together (Lewin, 1991; Thurman et al., 1999). Despite of the social and economical significance of TBI we still lack any efﬁcient speciﬁc pharmacological approach to interfere with the damaging consequence of injury (Narayan et al., 2002). The therapeutic approaches utilized to date are basically not different from those of the late seventies to early eighties; a noteworthy reduction in mortality was only achieved via the implementation of scientiﬁc evidence based therapeutic guidelines.

Introduction Traumatic brain injury (TBI) puts an extreme burden on societies worldwide. More than ﬁve million Americans have a TBI-related long-term or lifelong need for help for daily activities. Longterm consequences include functional changes such as cognitive deficit, epilepsy, hypopytuitarism and increased risk for Alzheimer’s disease and Parkinson’s disease. The cumulative societal cost Corresponding author. Tel.: +36-30-411-3349; Fax: +36-72-

535931; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61006-6

81

82

While such reorganization of the system and everyday practice of TBI care led to marked reduction in mortality and morbidity, the clinical care for the head injured has reached its limits (Vukic et al., 1999; Ghajar, 2000). Thereby, now basic scientists and/or pharmaceutical research should provide further ammunition for clinicians to ﬁght the silent epidemic represented by TBI. The importance of translational studies in TBI is further highlighted by data from the ﬁeld of health care management, indicating that effective treatment of TBI is considered one of the most cost-efﬁcient medical interventions (Saltman and Findling, 1997). Although detailed analysis of the pathobiological processes activated in TBI are beyond the scope of this work, the authors will provide a short theoretical background to each section addressing various approaches to rescue neurons and glia in TBI, while also referring to other chapters of this work for further relevant information.

Cell death routes On noxious inﬂuences or agents three major forms of cell death (CD) may occur: autophagic, necrotic and apoptotic. Autophagy is characterized by lysosomal sequestration and degradation of cellular components. This mode of cell death has its potential role in physiological processes as well as in neurodegenerative disorders (for review, see Assuncao and Linden, 2004). While primary brain injury leads to focal and diffuse alterations in brain parenchyma, which are predominantly necrotic by nature, secondary brain damage practically triggered and initiated at the moment of TBI is also associated with apoptotic damage. In fact, as time elapses post-injury, apoptotic processes become more dominant and in late phases post-injury autophagy may also participate particularly in axonal demise (Lockshin and Zakeri, 2004; Ringger et al., 2004; Gallyas et al., 2006). Necrosis is characterized with swelling and rupture of the cell (or its axon); membrane damage and activation of inﬂammatory reactions is a consistent feature of necrotic cell death. Such necrotic

processes are also associated with energy depletion/negative energetic state of the cell (Cai et al., 1998; Arden and Betenbaugh, 2004). Recent observations proved that both in focal and diffuse forms of experimental TBI in injured soma as well as in axons, various proteolytic processes are triggered and are responsible for necrotic demise (Bartus, 1997; Yamashima, 2004; Farkas et al., 2006). These studies demonstrated that mechanic or enzymatic alterations of membrane integrity in conjunction with altered function of ionic pumps and channels should lead to intracellular calcium accumulation which in turn results in the activation of neutral proteases, primarily the cysteineaspartate protease calpain (Bartus, 1997; Stys and Jiang, 2002). Besides direct proteolysis associated with calpain activation, indirect proteolytic consequences such as calpain-mediated lysosomal rupture and cathepsin activation (Yamashima, 2004) also contribute to cellular demise. Although the association and feedback connection between permeability changes, activation of proteolytic processes, cellular demise and functional alterations following TBI is not entirely developed, preliminary data suggest that inhibition of necrotic enzymes may facilitate functional recovery (Saatman et al., 1996). Regardless of the initiating factors and the route leading to altered intracellular ionic homeostasis, increased level of intracellular Ca++ represents a permissive environment for the activation of another form of CD that is apoptosis (Cai et al., 1998). Apoptosis, frequently referred to as programmed cell death plays a pivotal role in embryogenesis and is frequently initiated by a non-lethal stimulus leading to cell death via the activation of proteolytic cascades. Characteristic features of apoptosis include DNA laddering (detectable by electrophoresis), DNA fragmentation (TUNEL/ terminal deoxynucleotidyl-transferase-mediated deoxyuridine triphosphate-biotin nick-end labeling), sub-2N DNA (FACS), rounding/blebbing of the cell, fragmentation of nuclei with marginalization of chromatin, exteriorization of phosphatidylserine (annexin V binding) (for review, see Arden and Betenbaugh, 2004; Lockshin and Zakeri, 2004). As apoptosis requires a relatively intact energetic state of the cell, most probably it will not be

83

localized to the core of tissue laceration (focal damage) rather it will involve remote/diffuse (‘‘penumbra’’) area of the central nervous system (CNS). In the last decade multiple lines of evidences pointed to the activation of apoptotic processes in various forms of TBI (Rink et al., 1995; Conti et al., 1998; Buki et al., 2000; Ginis et al., 2000). Studies in more severe, focal forms of TBI have proven cytochrome c (cyto c) release and caspase-3 activation in the cerebral cortex 6–24 h post-injury (Sullivan et al., 2002). Caspase-12 activation was also demonstrated in cortex, hippocampus and cortical astrocytes in this model of TBI implicating endoplasmic reticulum-associated apoptotic processes (vide infra) in the pathogenesis of cell loss (Larner et al., 2004) These authors also demonstrated capsase-7 activation in the same model peaking at day ﬁve post-injury (Larner et al., 2005).

Knoblach et al. (2002), in the model of lateral ﬂuid percussion TBI in accordance with other, previous reports (Yakovlev et al., 1997; Beer et al., 2000; Clark et al., 2000), demonstrated caspase-3 and -9 activation in the cerebral cortex and the hippocampus 1–12 h post-injury. Although these workers were not able to demonstrate considerable activation of caspase-8, in other models including cortical impact TBI caspase-8 (representing the induction of the extrinsic route of apoptosis) was also found activated in both neurons and glia (Beer et al., 2001).

Apoptotic pathways Apoptosis could be initiated via three major pathways (Fig. 1), and in most of these pathways

Fig. 1. The three major routes of apoptosis. Note that different caspases participate at initiation and the execution of the cell death processes. Also note the crucial role of calcium and the mitochondrion (for further details, see text).

84

intracellular Ca++, mitochondria and the caspase enzyme family play a crucial role. The extrinsic receptor-mediated route utilizes speciﬁc cell surface receptors belonging to the tumor necrosis factor (TNF) super family where binding of the ‘‘death signal’’ (such as Fas by Fas ligand) triggers the activation of the caspase enzyme cascade (Chinnaiyan et al., 1995). Caspases are cysteineaspartate proteases which are present in the cell in the inactive proenzyme (zymogen) form. Speciﬁcally, ligand–receptor interaction in the extrinsic route will activate pro-caspase-8, which, in turn activates pro-caspase-3. Similarly to this pathway, the ﬁnal executor of the second, intrinsic pathway is also caspase-3, however, this route is mediated by mitochondrial alterations (Cai et al., 1998; Kruman and Mattson, 1999; Adams and Cory, 2002; Kroemer, 2003). Mitochondria play a key role in apoptosis. Several studies demonstrated both in vitro and in vivo that TBI leads to a perturbation of the energy homeostasis (Giza and Hovda, 2001). Partially due to excessive ﬁring of neurons and/or ionic imbalance due to altered pump/channel function as well as mechanoporation in at least a subpopulation of cell bodies and axons, a clear metabolic dysfunction occurs, leading to a shift toward anaerobic glycolysis, lactate-production and the accumulation of reactive oxygen species (ROS) (Pettus et al., 1994; Stys and Jiang, 2002; Singleton and Povlishock, 2004; Farkas et al., 2006). N-methyl-Daspartate (NMDA) receptor activation in neurons, failure of the Na-K-ATPase in axons and mechanoporation in some soma and axons in conjunction with altered energy homeostasis leads to the accumulation of intracellular (intraaxonal) Ca++ (Siesjo et al., 1999; Stys and Jiang, 2002). Mitochondria accumulate the excess amount of Ca++ leading to swelling and dissipation of their membrane potential, initiating the permealization of their outer membrane for solutes under 1500 Da, widely explained by the opening of the mitochondrial permeability transition pore (MPTP) (Zoratti and Szabo, 1995; Susin et al., 1998; Kruman and Mattson, 1999). This purported pore is consisting of the adenine nucleotide translocase (ANT), voltage dependent anion channel (VDAC), cyclophilin D, peripheral

benzodiazepine receptor, hexokinase, creatine kinase and Bcl-2 proteins (Susin et al., 1999) (some of these Bcl-2 proteins harbor anti-apoptotic potential, while others possess pro-apoptotic activity (Reed, 1998; Scorrano and Korsmeyer, 2003)). Opening of the MPTP leads to the release of proapoptotic substances — cyto c, second mitochondrial activator of caspases (SMAC)/direct inhibitor of apoptosis protein (IAP) binding protein with low pI (DIABLO) and apoptosis inducing factor (AIF). When released, cyto c together with the apoptosis activating factor 1 (Apaf-1), dATP and cytosolic pro-caspase-9 forms the ‘‘apoptosome’’ resulting in the generation of active caspase-9 which, in turn triggers the transition of pro-caspase-3 to active caspase-3 (Cai et al., 1998; Adams and Cory, 2002). The third, most recently described route of apoptosis is initiated by endoplasmatic reticulum (ER)-stress leading to caspase-12 activation. This latter enzyme can act either as an executor caspase or via direct or indirect activation of caspase-3 (Oyadomari et al., 2002; Zong et al., 2003).

Crossroads of cell death To develop rationally targeted therapeutic strategies beyond understanding these basic mechanisms of cell death, it is also mandatory to unravel those potential routes via the pathways initiated by noxious stimuli may communicate or interact. Such connections or ‘‘detours’’ exist within the apoptotic machinery itself as well as between various major pathways of cell death that is necrosis, apoptosis and autophagy (Denecker et al., 2001; Lockshin and Zakeri, 2004). In addition to their role in the maintenance of the ionic and energy homeostasis and triggering the intrinsic pathway of apoptosis, mitochondria also serve as a crossroad or common pathway for all routes of apoptotic cell death (Cai and Jones, 1998; Saikumar et al., 1998) (Fig. 1). To this end, besides pro-apoptotic members of the Bcl-2 family the mitochondrial membrane also contains other pro-apoptotic proteins such as Bax and Bak. These are responsible for the cross-talk between mitochondria (that is the intrinsic

85

pathway of apoptosis) and the other two avenues of apoptotic cell death; the extrinsic route is converging to these pro-apoptotic proteins via the caspase-8 Bid–Bax/Bak interaction while the endoplasmatic reticulum stress also interacts with these proteins as well as via sending direct Ca++ signals to the mitochondria (Jurgensmeier et al., 1998; Schendel et al., 1998; Ferri and Kroemer, 2001). Not only the major intracellular ‘‘death-avenues’’ but also the redundancy of the effector caspases provide considerable versatility for cell death; studies with knock-out mice as well as caspase-inhibitor studies have proven that an injured cell has the potential to select between various effector caspases and such decision is most probably inﬂuenced by various, physiologically active inhibitors of apoptosis (Assuncao and Linden, 2004). Alternative, protease-mediated pathways may also contribute to cell death. Ubiquitination of some caspases and their subsequent destruction by the proteosome may participate in the regulation of apoptosis (Lockshin and Zakeri, 2002; Ditzel et al., 2003; Varshavsky, 2003). Recent observations have proven that disconnection of the cell from its neighboring structures specifically from the extracellular matrix should contribute to the initiation of cell death mechanisms. This process, primarily mediated by matrix metalloproteases is frequently described with the Greek word ‘‘anoikis’’ (homelessness) (Lockshin and Zakeri, 2002; Abe et al., 2004). The above-detailed complexity of the cell death machinery provides multiple choices for a cell to react to a noxious stimulus. Apoptotic cell death requires energy and phagocytes that scavenge the dying cells. In cell cultures, in vitro in the absence of phagocytes the ﬁnal mechanisms of cell death are always necrotic, despite of the initiating agent and the route the cell entered to die (Arden and Betenbaugh, 2004; Assuncao and Linden, 2004). Limited sources of ATP can easily halt the apoptotic processes and the cell may fail to display apoptotic morphology, however, this does not mean an arrest of cell death rather the initiation of autophagic or necrotic processes (Cai et al., 1998; Reed, 1998; Saikumar et al., 1998).

When we consider, how to inhibit apoptosis, we also have to answer, whether inhibition of apoptosis is useful. Theoretically, in a philosophical approach, we may conclude that a lethal or even non-lethal stimulus may ignite any of the above-detailed pathways. Due to the cross-talk between these avenues and the internal diversity of the individual pathways it is quite possible that pharmacological intervention to halt a death process may just shift the cell death mechanism to another pathway of cellular demise. Inhibition of apoptotic cell death can lead to autophagy instead of resulting in cell survival (Lang-Rollin et al., 2003). Similarly, although the application of caspase inhibitors is a popular way to interfere with apoptosis, more upstream targets should be advocated for pharmacological intervention. Speciﬁcally, as caspase-9 and -3 represent the postmitochondrial phase of cell death their activation requires the release of cyto c and other pro-apoptotic substances. Such excessive mitochondrial damage also points to a considerable disruption of the electron transport chain leading to mass generation of free radicals as well as to altered energy homeostasis both facilitating the activation of necrotic processes in the cell. In traumatic axonal injury (TAI) the activation of caspases leads to irreversible digestion of the membrane skeleton and other cytoskeletal structures indicative of imminent, irreversible axonal demise (Wang et al., 1998; Buki et al., 2000) (Fig. 2). The relevance of such co-activation of the ‘‘necrotic’’ and ‘‘apoptotic’’ machinery was also demonstrated in an elegant study by Liu et al. (2006) in hippocampal lysates from rats which have undergone controlled cortical impact injury, where degradome-studies identiﬁed various common targets of calpain-2 and caspase-3 including beta-II-spectrin, synaptotagmin-1 and striatin. Similar ‘‘collaboration’’ between calpain and caspase-3 was also demonstrated by Warren et al. (2005) in metamphetamine induced experimental neurotoxicity.

Inhibition of apoptosis Several studies with caspase-inhibitors corroborate — at least from some aspects — the above-described

86

Fig. 2. Images of damaged axonal proﬁles from the medial lemniscus of rat exposed to impact acceleration brain injury (60 min post-injury). Cytochrome c release from damaged mitochondria (immunoﬂuorescent label on A) and caspase-3 activation (immunoﬂuorescent label on B) are co-localised in the same damaged axons (digital overlay on C) pointing to the contribution of calcium-induced mitochondrial damage and caspase activation to the pathogenesis of traumatically induced diffuse brain injury (axonal injury).

critical attitude to the usefulness of inhibiting the effector-phase of apoptosis alone. Abrahamson et al. (2006) demonstrated that a pan-caspase inhibitor (BAF) significantly reduced the formation of caspase-cleaved amyloid precursor protein fragments also leading to a reduced volume of hippocampal cell loss, however, in their experiment performed in mice applying the cortical impact model of injury significant necrotic areas where still noted in the cortex and the hippocampus. In another work, the pan-caspase inhibitor FK-011 was not able to achieve reduction in cortical lesion size measured 7 days post-injury (Sullivan et al., 2002). Although Knoblach et al.

(2004) were able to prove significant improvement of motor and neurological function using the pan-caspase inhibitor z-VAD-fmk, they also stress that this beneﬁcial effect was purportedly also attributable to the inhibition of calpain that is co-inhibition of necrosis in the dose they applied. In their review Lockshin and Zakeri (2004) also stress that co-inhibition of lysosomal proteases may lead to overestimation of the efﬁcacy of caspase-inhibition. Keeping these warnings in mind one should conclude that inhibition of apoptosis alone might be ‘‘too little, too late’’. Thus, the most promising way to interfere with the progression of the pathobiological processes evoked by/operant in TBI must either be the application of therapeutic agents with multiple targets like the above mentioned pan-caspase inhibitors or targeting premitochondrial/mitochondrial events in the course of cellular demise. As far as multitargetic therapeutic approaches are considered, sex steroids should be included to our list; several studies suggested that steroids could have beneﬁcial effects both on neurons and glia. In ischemic and TBI sex steroids have been proven to inhibit necrotic as well as apoptotic cell death (Roof and Hall, 2000). Although the exact mechanism of their action still requires further investigation, sex steroids are thought to inhibit lipid peroxydation and the formation of ROS while also exerting anti-apoptotic properties facilitating the expression of anti-apoptotic proteins including bcl-2 and Bax (Soustiel et al., 2005). Similarly, estrogen treatment has a positive effect on oligodendroglia function (Curry and Heim, 1966) and astrocytic scar formation (Garcia-Estrada et al., 1999). Testosterone can inﬂuence the expression of the astrocytic water channel protein aquaporin-4 indicating that sex steroids may also exert their action on the formation of brain edema — a major player in secondary brain injury (Gu et al., 2003). In the last two decades several studies suggested that poly(ADP-ribose)polymerase-1 (PARP-1, EC 2.4.2.30) inhibition might be beneﬁcial in the inhibition of CNS injury of various origin (Burkle, 2001; Besson et al., 2003, 2005). This enzyme was originally associated with DNA-repair function: stress induced strand DNA breaks activate PARP

87

that will transfer ADP-ribose units to nuclear proteins from NAD. This energy consuming process was recognized to lead to NAD depletion and loss of ATP leading to a collapse in cellular energy homeostasis and cell death (Burkle, 2001). While long-term inhibition of PARP could theoretically lead to mutagenesis and carcinogenesis, in the acute phase such intervention would be beneﬁcial in terms of restoring/preserving the cellular energy pool. As it has recently been stressed by Komjati et al. (2005) PARP inhibition is not only interfering with ROS-generated necrosis. Beyond representing a potential pathway in initiating and/or modulating necrosis via AIF, PARP has a substantial role in the induction of apoptosis. Via the NF-kappaBpathway PARP is also capable of inﬂuencing the expression of inﬂammatory cytokines and mediators (Komjati et al., 2005). In the last few years several PARP inhibitors have been proven to inhibit TBI. In the lateral ﬂuid percussion head injury model LaPlaca et al. (2001) have demonstrated that the PARP inhibitor GPI 1650 significantly reduced the lesion volume while not inﬂuencing the number of TUNEL positive cells. Other experiments proved that PARP inhibitors were able to improve the neurological score of injured animals without reducing the extent of the cellular lesion (Besson et al., 2005). These data in conjunction with those results indicating that PARP inhibitors reduce the extent of TAI assessed by beta amyloid precursor protein immunoreactivity implicate the inhibition of TAI in the neuroprotective effects of PARP inhibitors. Fluid percussion TBI is not the only model where PARP inhibitors proved their neuroprotective effect: 3-aminobenzanide significantly reduced the lesion size in cold injury (Hortobagyi et al., 2003) and proved beneﬁcial functional effects (PJ34) as well as improved effectiveness of neural stem cell transportation (Lacza et al., 2003). In the last decades, several studies implicated the cytosolic neutral cysteine protease calpain (EC 3.4.22.17) in the pathogenesis of both diffuse and focal TBI (Bartus, 1997; Kampﬂ et al., 1997; Buki et al., 1999b). This Ca++-induced enzyme is capable of reversibly cleaving spectrin and other constituents of the cytoskeleton (Wang et al.,

1998). While classic thought appreciated calpain as an agent contributing to necrotic processes several lines of evidences demonstrated that calpain can participate in the induction of the mitochondrial (intrinsic) pathway of apoptosis. This communication between the classic, Ca++-induced necrotic processes and the apoptotic enzyme system is partially executed that way that calcium accumulation leads to axolemmal permeability alteration through proteolytic modiﬁcation of the subaxolemmal network (cortical cytoskeleton) which in turn at least in some neurons and in axons contributes to further membrane-permeability changes (Buki et al., 1999b). As it has been described in the above passages, excessive inﬂux of Ca++ induces mitochondrial damage via MPTPopening. In addition to this pathway, calpain itself is capable — as it has been demonstrated in liver — to open the MPTP via its direct proteolytic modiﬁcation (Aguilar et al., 1996). Once MPTP is open and the intrinsic route, thus caspase-3 is activated, it can cleave calpastatin, the major inhibitor of calpain, contributing to further uncontrolled activation of calpain. Taking into consideration such interactions and the temporal/ causative relationship between calpain and caspase activation a logical way to inﬂuence the release of apoptotic enzymes should be the inhibition of calcium-mediated proteolytic alterations that is the inhibition of calpain. Recent observations indicated that systemic administration of calpain inhibitors reduced the overall degree of the cerebral ischemia (Bartus, 1997). Further, Saatman et al. (1996) demonstrated that the use of calpain inhibitors resulted in improved behavioral outcome in a contusional model of TBI. Intriguingly enough, in an extension of this work Saatman et al. failed to detect any significant change in the size of contusion, concluding that the beneﬁcial therapeutic effects of AK-295 could be explained by inhibiting TAI in remote areas (Saatman et al., 2000). The correctness of this assumption was further supported by the ﬁnding that in a rodent model of TAI the cell permeable peptidyl-aldehyde calpain inhibitor MDL-28170 (carbobenzylzoxy-Val-Phe-H) proved to prevent axonal injury in the brainstem ﬁber tracts (Buki et al., 2003). Most recent observations

88

indicate that this agent is also potent in terms of reducing the length of axonal segments displaying traumatically induced axolemmal permeability changes frequently associated with the activation of the Ca++-induced protease calpain coupled in some axons to the initial, strain-induced axolemmal mechanoporation (Bu¨ki et al., unpublished) (vide supra). Although physical and chemical properties of MDL-28170 do not favor its application in the clinical care, the family of such cellpermeable calpain-inhibitors that readily cross the blood brain barrier and exert their beneﬁcial effects in a relatively wide therapeutic window (Bartus, 1997; Markgraf et al., 1998) should be considered candidates for further pharmacological modiﬁcations and studies. Calpain inhibitors are of particular interest as recent reports demonstrated that besides decreasing the formation of cleavage products of this cysteine protease primarily participating in necrotic processes, they are capable of inhibiting the formation of caspase-cleaved byproducts, thereby apoptosis too (Kawamura et al., 2005). In the line of therapeutic approaches aiming to inhibit the mitochondrial/premitochondrial phase of apoptosis one of the most promising interventions is the use of the inhibitors of mPTP opening. Among these interventions the immunophilin ligand cyclosporine A (CsA) has been used in a set of experiments targeting TAI in the rodent impact acceleration model of TBI. First, immuno-electron microscopic studies by Okonkwo and Povlishock (1999) demonstrated that CsA significantly attenuated mitochondrial swelling and rupture, translating into reduced numbers of damaged, disconnected axonal segments displaying immonureactivity for the transport protein beta amyloid precursor protein. Subsequent studies also proved that both pre- and post-injury administration of CsA significantly reduced the extent of TAI detected by immunohistochemical markers of calpain-mediated spectrin proteolysis, axonal disconnection and cytoskeletal alterations — neuroﬁlament compaction (NFC). Dose response curves for the beneﬁcial effects of CsA in the experimental model of TBI have also been established (Okonkwo et al., 2003). More recent studies also demonstrated that CsA limits/improves

some aspects of functional outcome following TBI (Riess et al., 2001). Preliminary data regarding the use of CsA in head injury in man indicate that this compound is safe and potentially beneﬁcial (Alves et al., 2003). Another immunophilin ligand FK506, a substance that has no known effect on MPT and exerts its action via inhibition of calcineurin activity (Liu et al., 1991; Kuroda et al., 1999) has also displayed beneﬁcial effects on TAI in a pre-injury administration paradigm (Singleton et al., 2001) and reduced the complications associated with rapid re-warming following hypothermic intervention for TAI (Suehiro et al., 2001). Inhibition of the Ca++-induced phosphatase calcineurin is supposed to be beneﬁcial in axonal injury by preventing dephosphorylization of the neuroﬁlament side arms that should decrease the repelling forces between the neuroﬁlaments thereby permitting NFC leading to impairment of axoplasmatic transport (Povlishock et al., 1997; Okonkwo et al., 1998). Such dephosphorylated neuroﬁlaments would be more susceptible for proteolytic degradation by calpain and caspase (Pant, 1988). It is of note that calcineurine might also participate in the signaling pathways regulating proliferative responses of astrocytes and its inhibition by CsA as well as FK-506 diminishes astrocytic reactions (Pyrzynska et al., 2001). Calcineurin also contributes to the control of synaptic activity by inactivating dynamin I and synapsin I, proteins participating in neurotransmitter release (Cousin and Robinson, 2001) and by dephosporylation-mediated induction of nitric oxide synthase leading to increased nitric oxide level and neurotransmitter release (Dawson et al., 1993). Other important aspects of TBI might also be inﬂuenced by immunophyllin ligands. Specifically, CsA inhibited ischemia-induced edema in vitro, suggesting that such a beneﬁcial effect should also be worth investigating in secondary brain injury where hypoperfusion-hypoxya-induced edema and tissue damage is an issue, too (MacGregor et al., 2003). CsA also proved neuroprotective in the hands of Gabbita et al. (2005) in the controlled cortical impact model where it inhibited the TBI-induced

89

increase of cleaved tau (cytoskeletal protein) level in the ipsilateral hippocampus and its accumulation in the CSF of injured rodents. Van Den et al. (2004) have demonstrated another beneﬁcial effect of CsA in TBI proving that 30 min post-injury administration of CsA reduced APP mRNA at 2 and 6 h post-injury and likewise reduced APP accumulation in injured cell bodies. Nevertheless, they did not prove axonal preservation in terms of APP-immunoreactivity in their experiments. While these results might reﬂect the complexity in the pathogenesis of TBI, it is of note that several studies have proved that APP immunoreactivity alone roughly underestimates the extent of TBI and, in some axon-populations other mechanisms than impaired axoplasmatic transport and axonal swelling should be held accountable for axonal demise (Singleton and Povlishock, 2004; Farkas et al., 2006). While, without doubt, a continuous development of various anti-apoptotic or, even more promising, anti-apoptotic, anti-necrotic drugs should be a major goal of pharmaceutical research, neuroprotective substances physiologically produced in the brain itself should also be taken into account. To this end, a bulk of recently collected evidences point to the neuroprotective effects of pituitary adenylate cyclase activating polypeptide (PACAP). This member of the vasoactive intestinal peptide (VIP)/secretin/glucagon peptide family (Miyata et al., 1989) was discovered as a hypothalamic peptide on its potential of increasing adenylate cyclase activity in the pituitary gland and proved to exert various effects in the central and peripheral nervous systems including neurotrophic and neuroprotective ones (Arimura, 1998; Vaudry et al., 2000). In the extradural static weight compression model of spinal cord injury in rat, post-injury PACAP treatment significantly reduced the number of apoptotic cells assessed by TUNEL in the spinal cord (Katahira et al., 2003). Further, PACAP has been shown to eliminate the increase of TNF after spinal cord transection (Kim et al., 2000) and in experimental TBI postinjury induction of PACAP-mRNA and its upregulation paralleled the decrease in the number of apoptotic cells (Skoglosa et al., 1999). These

studies indicate that PACAP is a promising therapeutic agent in traumatic brain and spinal cord injuries, which is further supported by our observations. Unfortunately, the exact mechanism underlying the in vivo neuroprotective effect of PACAP has not been clariﬁed so far. In vitro studies assign both anti-apoptotic and anti-inﬂammatory actions to PACAP. In cerebellar granule cells, PACAP significantly inhibited the activation of caspase-3 (Vaudry et al., 2004). PACAP was proven to be a potent inactivator of induced microglial release of pro-inﬂammatory cytokines and nitric oxide (Kong et al., 1999; Kim et al., 2000; Delgado et al., 2003). It is also believed to inﬂuence mitochondrial integrity as PACAP inhibited the mitochondrial Ca++ uptake induced inactivation of aconitase, a key mitochondrial enzyme inﬂuencing the viability of neurons (Tabuchi et al., 2003). In the last few years our laboratories extensively investigated the purported neuroprotective role of PACAP in a diffuse model of TBI. Intriguingly enough, we were not able to reproduce neuroprotective effects in terms of preserving axonal integrity assessed by APP-immunoreactivity (vide supra) with the administration paradigm (125 mg pre-injury, iv) that had worked in a middle cerebral artery occlusion model of stroke (Reglodi et al., 2002). Nevertheless, we were able to establish the dose-response curve for intracerebroventricular (icv) administration of PACAP, proving that 100 mg PACAP significantly reduced the density of damaged, immunoreactive axons in the corticospinal tract (Farkas et al., 2004). Our most recent observations proved that a considerable therapeutic window exists for post-injury PACAP treatment demonstrating significant axonal protection in an icv administration paradigm 2 h following impact acceleration TBI (Tamas et al., 2006). A new avenue of pharmacotherapy targeting apoptosis in TBI is represented by the inhibition of cell cycle proteins. Di Giovanni et al. (2005) demonstrated that the cell cycle inhibitor ﬂavopiridol was not only capable of inhibiting etoposideinduced, capase-mediated cell death in cultured primary cortical neurons but also inhibited the proliferation of astrocytes and microglia and decreased the extension of TBI-induced tissue lesion

90

and improved functional recovery in the lateral ﬂuid percussion rodent model. As proteolytic processes are key players in TBIevoked apoptosis, the list of potential therapeutic measures would not be complete without mentioning the neuroprotective capacity of therapeutic hypothermia. This intervention is purported to act via general inhibition or slowing of the proteolytic processes which could provide an extended therapeutic window for further, speciﬁc therapeutic approaches while also enabling the neurons to keep their mitochondria up and running and their local energy homeostasis intact (Buki et al., 1999a). Although the use of controlled hypothermia in the treatment of human TBI remains controversial, on the basis of multiple lines of evidence this therapeutic modality might be worth further investigations and probably further clinical studies planned on the basis of past experience could resolve the controversy concerning the role of hypothermia in the treatment of the severely head injured (Clifton, 2004). Specifically, on the basis of our current knowledge on its action, hypothermia should be introduced relatively earlier in the care of TBI paying special attention to gradual rewarming in conjunction with introduction of other treatment strategies such as FK-506 that has already proved its potential to inhibit diffuse brain injury associated with post-hypothermic rewarming (Suehiro et al., 2001).

Concluding remarks There are several controversies in the development of novel pharmacotherpeutic agents aimed at the inhibition of pathobiological processes evoked by TBI. While everybody in the ﬁeld considers the treatment of TBI an extremely cost efﬁcient medical procedure, we still lack any novel, efﬁcient therapeutic agent. Despite of the promising results achieved with various pre-clinical studies all candidates failed at the clinical phase and so far none of them proved useful for the everyday practice. While detailed analysis of such failure is well beyond the scope of this review, the authors feel appropriate to note that more detailed understanding and appreciation of cell death mechanisms and

experimental models applied should be a prerequisite for successful development of neuroprotective agents. As far as study design is considered we should appreciate that the rather complex and versatile nature of TBI occurring in human is hard to copy in any experimental model existing currently in our hands. In the overwhelming majority of cases human TBI is a mixture of focal and diffuse injuries complicated with secondary brain injury primarily caused by hypoxia and hypoperfusion. The most frequently used models like controlled cortical impact or moderate/severe forms of ﬂuid percussion brain injury produce mass tissue destruction where strain and shearing-force distribution leads to excessive inﬂux and release of Ca++, and excitatory amino acids, activation and generation of ROS. Although such models may mimic the situation in the most severe forms of human TBI, they ignite such a bulk of various pathways that could hardly be controlled with a drug exerting its effect on a single target. While diffuse brain injury models seem to provide a better environment to study more subtle forms of TBI and theoretically provide a more convenient ﬁeld to investigate the effect of pharmaceutical agents, recent observations shed light on the extremely complex and heterogenous nature of both diffuse axonal and neuronal injury. Speciﬁcally, in different ﬁber populations considerably different markers of axonal damage may appear representing heterogeneity in the pathobiology as well as in therapeutic efﬁcacy (Stone et al., 2004). Similarly, acceleration–deceleration injury evokes various forms of neuronal-cellular damage indicating a likewise complex ﬁeld to target therapeutically (Singleton and Povlishock, 2004; Farkas et al., 2006). In light of the above passages and the diversity and interactivity of cell death pathways, a fatalistic approach would be to conclude that when a cell is committed to die due to an extrinsic or intrinsic death signal it will do so regardless of whatever therapeutic measures have been taken; inhibition of apoptosis will ignite autophagic or necrotic cell death, ultimately leading to cellular demise. Nevertheless, the authors still feel appropriate to address this very issue from a different point of view with the believe that combined therapeutic

91

approaches or ‘‘polyhparmacia’’ might successfully target intermingled death pathways with resulting improved morphological as well as functional outcome. To this end one can envision a scenario where ultra-early (pre-hospital) application of therapeutic measures aimed to slow down proteolytic processes (such as hypothermia treatment) and preserving energy homeostasis (PARP-inhibition, CsA-treatment) is followed or joined by caspase and calpain inhibitors, or PACAP treatment. As far as the title of the present article is considered, to date it is hard to predict, whether inhibition of apoptosis is useful; one can only conclude that it is most probably beneﬁcial in conjunction with other measures taken to inhibit various different pathways of traumatically induced neural demise. Abbreviations AIF CD CsA ER IAP Icv MPTP NMDA PARP-1 TAI TBI TNF

apoptosis inducing factor cell death cyclosporine A endoplasmatic reticulum inhibitor of apoptosis protein intracerebroventricular (icv) mitochondrial permeability transition pore N-methyl-D-aspartate poly(ADP-ribose)polymerase 1 traumatic axonal injury traumatic brain injury tumor necrosis factor

Aknowledgments Supported by the Hungarian Science Found (OTKA T048724) and the Bekesi Scholarship of the Hungarian Academy of Sciences. The authors thank for the invaluable advices and help of Professor John T. Povlishock, Department of Anatomy and Neurobiology, Medical College of Virginia, Virginia Commonwealth University, Richmond, VA 23298-0709.

References Abe, Y., Yamamoto, T., Sugiyama, Y., Watanabe, T., Saito, N., Kayama, H. and Kumagai, T. (2004) ‘‘Anoikis’’ of oligodendrocytes induced by Wallerian degeneration: ultrastructural observations. J. Neurotrauma, 21: 119–124. Abrahamson, E.E., Ikonomovic, M.D., Ciallella, J.R., Hope, C.E., Paljug, W.R., Isanski, B.A., Flood, D.G., Clark, R.S. and DeKosky, S.T. (2006) Caspase inhibition therapy abolishes brain trauma-induced increases in Abeta peptide: implications for clinical outcome. Exp. Neurol., 197: 437–450. Adams, J.M. and Cory, S. (2002) Apoptosomes: engines for caspase activation. Curr. Opin. Cell Biol., 14: 715–720. Aguilar, H.I., Botla, R., Arora, A.S., Bronk, S.F. and Gores, G.J. (1996) Induction of the mitochondrial permeability transition by protease activity in rats: a mechanism of hepatocyte necrosis. Gastroenterology, 110: 558–566. Alves, O., Tolias, C.M., Lewis, C., Hayes, R.L., Choi, S.C., Gilman, C., Enriquez, P., Povlishock, J.T. and Bullock, M.R. (2003) Brain neurochemical alterations in patients with severe brain injury treated with cyclosporin-a. A placebo-controlled study. Preliminary report. J. Neurotrauma, 20: 1125. Arden, N. and Betenbaugh, M.J. (2004) Life and death in mammalian cell culture: strategies for apoptosis inhibition. Trends Biotechnol., 22: 174–180. Arimura, A. (1998) Perspectives on pituitary adenylate cyclase activating polypeptide (PACAP) in the neuroendocrine, endocrine, and nervous systems. Jpn. J. Physiol., 48: 301–331. Assuncao, G.C. and Linden, R. (2004) Programmed cell deaths. Apoptosis and alternative deathstyles. Eur. J. Biochem., 271: 1638–1650. Bartus, R. (1997) The calpain hypothesis of neurodegeneration: evidence for a common cytotoxic pathway. Neuroscientist, 3: 314–327. Beer, R., Franz, G., Krajewski, S., Pike, B.R., Hayes, R.L., Reed, J.C., Wang, K.K., Klimmer, C., Schmutzhard, E., Poewe, W. and Kampﬂ, A. (2001) Temporal and spatial proﬁle of caspase 8 expression and proteolysis after experimental traumatic brain injury. J. Neurochem., 78: 862–873. Beer, R., Franz, G., Srinivasan, A., Hayes, R.L., Pike, B.R., Newcomb, J.K., Zhao, X., Schmutzhard, E., Poewe, W. and Kampﬂ, A. (2000) Temporal proﬁle and cell subtype distribution of activated caspase-3 following experimental traumatic brain injury. J. Neurochem., 75: 1264–1273. Besson, V.C., Croci, N., Boulu, R.G., Plotkine, M. and Marchand-Verrecchia, C. (2003) Deleterious poly(ADP-ribose)polymerase-1 pathway activation in traumatic brain injury in rat. Brain Res., 989: 58–66. Besson, V.C., Zsengeller, Z., Plotkine, M., Szabo, C. and Marchand-Verrecchia, C. (2005) Beneﬁcial effects of PJ34 and INO-1001, two novel water-soluble poly(ADP-ribose) polymerase inhibitors, on the consequences of traumatic brain injury in rat. Brain Res., 1041: 149–156. Buki, A., Farkas, O., Doczi, T. and Povlishock, J.T. (2003) Preinjury administration of the calpain inhibitor MDL-28170 attenuates traumatically induced axonal injury. J. Neurotrauma, 20: 261–268.

92 Buki, A., Koizumi, H. and Povlishock, J.T. (1999a) Moderate posttraumatic hypothermia decreases early calpain-mediated proteolysis and concomitant cytoskeletal compromise in traumatic axonal injury. Exp. Neurol., 159: 319–328. Buki, A., Okonkwo, D.O., Wang, K.K. and Povlishock, J.T. (2000) Cytochrome c release and caspase activation in traumatic axonal injury. J. Neurosci., 20: 2825–2834. Buki, A., Siman, R., Trojanowski, J.Q. and Povlishock, J.T. (1999b) The role of calpain-mediated spectrin proteolysis in traumatically induced axonal injury. J. Neuropathol. Exp. Neurol., 58: 365–375. Burkle, A. (2001) Physiology and pathophysiology of poly (ADP-ribosyl)ation. Bioessays, 23: 795–806. Cai, J. and Jones, D.P. (1998) Superoxide in apoptosis. Mitochondrial generation triggered by cytochrome c loss. J. Biol. Chem., 273: 11401–11404. Cai, J., Yang, J. and Jones, D.P. (1998) Mitochondrial control of apoptosis: the role of cytochrome c. Biochim. Biophys. Acta, 1366: 139–149. Chinnaiyan, A.M., O’Rourke, K., Tewari, M. and Dixit, V.M. (1995) FADD, a novel death domain-containing protein, interacts with the death domain of Fas and initiates apoptosis. Cell, 81: 505–512. Clark, R.S., Kochanek, P.M., Watkins, S.C., Chen, M., Dixon, C.E., Seidberg, N.A., Melick, J., Loeffert, J.E., Nathaniel, P.D., Jin, K.L. and Graham, S.H. (2000) Caspase-3 mediated neuronal death after traumatic brain injury in rats. J. Neurochem., 74: 740–753. Clifton, G.L. (2004) Is keeping cool still hot? An update on hypothermia in brain injury. Curr. Opin. Crit. Care, 10: 116–119. Conti, A.C., Raghupathi, R., Trojanowski, J.Q. and McIntosh, T.K. (1998) Experimental brain injury induces regionally distinct apoptosis during the acute and delayed post-traumatic period. J. Neurosci., 18: 5663–5672. Cousin, M.A. and Robinson, P.J. (2001) The dephosphins: dephosphorylation by calcineurin triggers synaptic vesicle endocytosis. Trends Neurosci., 24: 659–665. Curry III, J.J. and Heim, L.M. (1966) Brain myelination after neonatal administration of oestradiol. Nature, 209: 915–916. Dawson, T.M., Steiner, J.P., Dawson, V.L., Dinerman, J.L., Uhl, G.R. and Snyder, S.H. (1993) Immunosuppressant FK506 enhances phosphorylation of nitric oxide synthase and protects against glutamate neurotoxicity. Proc. Natl. Acad. Sci. U.S.A., 90: 9808–9812. Delgado, M., Leceta, J. and Ganea, D. (2003) Vasoactive intestinal peptide and pituitary adenylate cyclase-activating polypeptide inhibit the production of inﬂammatory mediators by activated microglia. J. Leukoc. Biol., 73: 155–164. Denecker, G., Vercammen, D., Declercq, W. and Vandenabeele, P. (2001) Apoptotic and necrotic cell death induced by death domain receptors. Cell Mol. Life Sci., 58: 356–370. Di Giovanni, S., Movsesyan, V., Ahmed, F., Cernak, I., Schinelli, S., Stoica, B. and Faden, A.I. (2005) Cell cycle inhibition provides neuroprotection and reduces glial proliferation and scar formation after traumatic brain injury. Proc. Natl. Acad. Sci. U.S.A., 102: 8333–8338.

Ditzel, M., Wilson, R., Tenev, T., Zachariou, A., Paul, A., Deas, E. and Meier, P. (2003) Degradation of DIAP1 by the N-end rule pathway is essential for regulating apoptosis. Nat. Cell Biol., 5: 467–473. Farkas, O., Lifshitz, J. and Povlishock, J.T. (2006) Mechanoporation induced by diffuse traumatic brain injury: an irreversible or reversible response to injury? J. Neurosci., 26: 3130–3140. Farkas, O., Tamas, A., Zsombok, A., Reglodi, D., Pal, J., Buki, A., Lengvari, I., Povlishock, J.T. and Doczi, T. (2004) Effects of pituitary adenylate cyclase activating polypeptide in a rat model of traumatic brain injury. Regul. Pept., 123: 69–75. Ferri, K.F. and Kroemer, G. (2001) Organelle-speciﬁc initiation of cell death pathways. Nat. Cell Biol., 3: E255–E263. Gabbita, S.P., Scheff, S.W., Menard, R.M., Roberts, K., Fugaccia, I. and Zemlan, F.P. (2005) Cleaved-tau: a biomarker of neuronal damage after traumatic brain injury. J. Neurotrauma, 22: 83–94. Gallyas, F., Pal, J., Farkas, O. and Doczi, T. (2006) The fate of axons subjected to traumatic ultrastructural (neuroﬁlament) compaction: an electron-microscopic study. Acta Neuropathol. (Berl.), 111: 229–237. Garcia-Estrada, J., Luquin, S., Fernandez, A.M. and GarciaSegura, L.M. (1999) Dehydroepiandrosterone, pregnenolone and sex steroids down-regulate reactive astroglia in the male rat brain after a penetrating brain injury. Int. J. Dev. Neurosci., 17: 145–151. Ghajar, J. (2000) Traumatic brain injury. Lancet, 356: 923–929. Ginis, I., Hallenbeck, J.M., Liu, J., Spatz, M., Jaiswal, R. and Shohami, E. (2000) Tumor necrosis factor and reactive oxygen species cooperative cytotoxicity is mediated via inhibition of NF-kappaB. Mol. Med., 6: 1028–1041. Giza, C.C. and Hovda, D.A. (2001) The neurometabolic cascade of concussion. J. Athl. Train, 36: 228–235. Gu, F., Hata, R., Toku, K., Yang, L., Ma, Y.J., Maeda, N., Sakanaka, M. and Tanaka, J. (2003) Testosterone upregulates aquaporin-4 expression in cultured astrocytes. J. Neurosci. Res., 72: 709–715. Hortobagyi, T., Gorlach, C., Benyo, Z., Lacza, Z., Hortobagyi, S., Wahl, M. and Harkany, T. (2003) Inhibition of neuronal nitric oxide synthase-mediated activation of poly(ADP-ribose) polymerase in traumatic brain injury: neuroprotection by 3-aminobenzamide. Neuroscience, 121: 983–990. Jurgensmeier, J.M., Xie, Z., Deveraux, Q., Ellerby, L., Bredesen, D. and Reed, J.C. (1998) Bax directly induces release of cytochrome c from isolated mitochondria. Proc. Natl. Acad. Sci. U.S.A., 95: 4997–5002. Kampﬂ, A., Posmantur, R.M., Zhao, X., Schmutzhard, E., Clifton, G.L. and Hayes, R.L. (1997) Mechanisms of calpain proteolysis following traumatic brain injury: implications for pathology and therapy: implications for pathology and therapy: a review and update. J. Neurotrauma, 14: 121–134. Katahira, M., Yone, K. and Arishima, Y. (2003) The neuroprotective effects of PACAP on spinal cor injury (SCI) in rats. Regul. Pept., 15: 49. Kawamura, M., Nakajima, W., Ishida, A., Ohmura, A., Miura, S. and Takada, G. (2005) Calpain inhibitor MDL 28170

93 protects hypoxic-ischemic brain injury in neonatal rats by inhibition of both apoptosis and necrosis. Brain Res., 1037: 59–69. Kim, W.K., Kan, Y., Ganea, D., Hart, R.P., Gozes, I. and Jonakait, G.M. (2000) Vasoactive intestinal peptide and pituitary adenylyl cyclase-activating polypeptide inhibit tumor necrosis factor-alpha production in injured spinal cord and in activated microglia via a cAMP-dependent pathway. J. Neurosci., 20: 3622–3630. Knoblach, S.M., Alroy, D.A., Nikolaeva, M., Cernak, I., Stoica, B.A. and Faden, A.I. (2004) Caspase inhibitor z-DEVDfmk attenuates calpain and necrotic cell death in vitro and after traumatic brain injury. J. Cereb. Blood Flow Metab., 24: 1119–1132. Knoblach, S.M., Nikolaeva, M., Huang, X., Fan, L., Krajewski, S., Reed, J.C. and Faden, A.I. (2002) Multiple caspases are activated after traumatic brain injury: evidence for involvement in functional outcome. J. Neurotrauma, 19: 1155–1170. Komjati, K., Besson, V.C. and Szabo, C. (2005) Poly (ADPribose) polymerase inhibitors as potential therapeutic agents in stroke and neurotrauma. Curr. Drug Targets CNS Neurol. Disord., 4: 179–194. Kong, L.Y., Maderdrut, J.L., Jeohn, G.H. and Hong, J.S. (1999) Reduction of lipopolysaccharide-induced neurotoxicity in mixed cortical neuron/glia cultures by femtomolar concentrations of pituitary adenylate cyclase-activating polypeptide. Neuroscience, 91: 493–500. Kroemer, G. (2003) Mitochondrial control of apoptosis: an introduction. Biochem. Biophys. Res. Commun., 304: 433–435. Kruman, I.I. and Mattson, M.P. (1999) Pivotal role of mitochondrial Ca++ uptake in neural cell apoptosis and necrosis. J. Neurochem., 72: 529–540. Kuroda, S., Janelidze, S. and Siesjo, B.K. (1999) The immunosuppressants cyclosporin A and FK506 equally ameliorate brain damage due to 30-min middle cerebral artery occlusion in hyperglycemic rats. Brain Res., 835: 148–153. Lacza, Z., Horvath, E.M., Komjati, K., Hortobagyi, T., Szabo, C. and Busija, D.W. (2003) PARP inhibition improves the effectiveness of neural stem cell transplantation in experimental brain trauma. Int. J. Mol. Med., 12: 153–159. Lang-Rollin, I.C., Rideout, H.J., Noticewala, M. and Stefanis, L. (2003) Mechanisms of caspase-independent neuronal death: energy depletion and free radical generation. J. Neurosci., 23: 11015–11025. LaPlaca, M.C., Zhang, J., Raghupathi, R., Li, J.H., Smith, F., Bareyre, F.M., Snyder, S.H., Graham, D.I. and McIntosh, T.K. (2001) Pharmacologic inhibition of poly(ADP-ribose) polymerase is neuroprotective following traumatic brain injury in rats. J. Neurotrauma, 18: 369–376. Larner, S.F., Hayes, R.L., McKinsey, D.M., Pike, B.R. and Wang, K.K. (2004) Increased expression and processing of caspase-12 after traumatic brain injury in rats. J. Neurochem., 88: 78–90. Larner, S.F., McKinsey, D.M., Hayes, R.L. and Wang, K.K. (2005) Caspase 7: increased expression and activation after traumatic brain injury in rats. J. Neurochem., 94: 97–108.

Lewin, I.C.F. (1991) Update of Max, W., Mackenzie, E.J., Rice, D.P. Head injuries: costs and consequences. J. Head Trauma Rehabil., 67: 76–91. Liu, J., Farmer Jr., J.D., Lane, W.S., Friedman, J., Weissman, I. and Schreiber, S.L. (1991) Calcineurin is a common target of cyclophilin-cyclosporin A and FKBP-FK506 complexes. Cell, 66: 807–815. Liu, M.C., Akle, V., Zheng, W., Dave, J.R., Tortella, F.C., Hayes, R.L. and Wang, K.K. (2006) Comparing calpain- and caspase-3-mediated degradation patterns in traumatic brain injury by differential proteome analysis. Biochem. J., 394: 715–725. Lockshin, R.A. and Zakeri, Z. (2002) Caspase-independent cell deaths. Curr. Opin. Cell Biol., 14: 727–733. Lockshin, R.A. and Zakeri, Z. (2004) Apoptosis, autophagy, and more. Int. J. Biochem. Cell Biol., 36: 2405–2419. MacGregor, D.G., Avshalumov, M.V. and Rice, M.E. (2003) Brain edema induced by in vitro ischemia: causal factors and neuroprotection. J. Neurochem., 85: 1402–1411. Markgraf, C.G., Velayo, N.L., Johnson, M.P., McCarty, D.R., Medhi, S., Koehl, J.R., Chmielewski, P.A. and Linnik, M.D. (1998) Six-hour window of opportunity for calpain inhibition in focal cerebral ischemia in rats. Stroke, 29: 152–158. Miyata, A., Arimura, A., Dahl, R.R., Minamino, N., Uehara, A., Jiang, L., Culler, M.D. and Coy, D.H. (1989) Isolation of a novel 38 residue-hypothalamic polypeptide which stimulates adenylate cyclase in pituitary cells. Biochem. Biophys. Res. Commun., 164: 567–574. Narayan, R.K., Michel, M.E., Ansell, B., Baethmann, A., Biegon, A., Bracken, M.B., Bullock, M.R., Choi, S.C., Clifton, G.L., Contant, C.F., Coplin, W.M., Dietrich, W.D., Ghajar, J., Grady, S.M., Grossman, R.G., Hall, E.D., Heetderks, W., Hovda, D.A., Jallo, J., Katz, R.L., Knoller, N., Kochanek, P.M., Maas, A.I., Majde, J., Marion, D.W., Marmarou, A., Marshall, L.F., McIntosh, T.K., Miller, E., Mohberg, N., Muizelaar, J.P., Pitts, L.H., Quinn, P., Riesenfeld, G., Robertson, C.S., Strauss, K.I., Teasdale, G., Temkin, N., Tuma, R., Wade, C., Walker, M.D., Weinrich, M., Whyte, J., Wilberger, J., Young, A.B. and Yurkewicz, L. (2002) Clinical trials in head injury. J. Neurotrauma, 19: 503–557. Okonkwo, D.O., Melon, D.E., Pellicane, A.J., Mutlu, L.K., Rubin, D.G., Stone, J.R. and Helm, G.A. (2003) Doseresponse of cyclosporin A in attenuating traumatic axonal injury in rat. Neuroreport, 14: 463–466. Okonkwo, D.O., Pettus, E.H., Moroi, J. and Povlishock, J.T. (1998) Alteration of the neuroﬁlament sidearm and its relation to neuroﬁlament compaction occurring with traumatic axonal injury. Brain Res., 784: 1–6. Okonkwo, D.O. and Povlishock, J.T. (1999) An intrathecal bolus of cyclosporin A before injury preserves mitochondrial integrity and attenuates axonal disruption in traumatic brain injury. J. Cereb. Blood Flow Metab., 19: 443–451. Oyadomari, S., Araki, E. and Mori, M. (2002) Endoplasmic reticulum stress-mediated apoptosis in pancreatic beta-cells. Apoptosis, 7: 335–345.

94 Pant, H.C. (1988) Dephosphorylation of neuroﬁlament proteins enhances their susceptibility to degradation by calpain. Biochem. J., 256: 665–668. Pettus, E.H., Christman, C.W., Giebel, M.L. and Povlishock, J.T. (1994) Traumatically induced altered membrane permeability: its relationship to traumatically induced reactive axonal change. J. Neurotrauma, 11: 507–522. Povlishock, J.T., Marmarou, A., McIntosh, T., Trojanowski, J.Q. and Moroi, J. (1997) Impact acceleration injury in the rat: evidence for focal axolemmal change and related neuroﬁlament sidearm alteration. J. Neuropathol. Exp. Neurol., 56: 347–359. Pyrzynska, B., Lis, A., Mosieniak, G. and Kaminska, B. (2001) Cyclosporin A-sensitive signaling pathway involving calcineurin regulates survival of reactive astrocytes. Neurochem. Int., 38: 409–415. Reed, J.C. (1998) Bcl-2 family proteins. Oncogene, 17: 3225–3236. Reglodi, D., Tamas, A., Somogyvari-Vigh, A., Szanto, Z., Kertes, E., Lenard, L., Arimura, A. and Lengvari, I. (2002) Effects of pretreatment with PACAP on the infarct size and functional outcome in rat permanent focal cerebral ischemia. Peptides, 23: 2227–2234. Riess, P., Bareyre, F.M., Saatman, K.E., Cheney, J.A., Lifshitz, J., Raghupathi, R., Grady, M.S., Neugebauer, E. and McIntosh, T.K. (2001) Effects of chronic, post-injury Cyclosporin A administration on motor and sensorimotor function following severe, experimental traumatic brain injury. Restor. Neurol. Neurosci., 18: 1–8. Ringger, N.C., O’Steen, B.E., Brabham, J.G., Silver, X., Pineda, J., Wang, K.K., Hayes, R.L. and Papa, L. (2004) A novel marker for traumatic brain injury: CSF alphaIIspectrin breakdown product levels. J. Neurotrauma, 21: 1443–1456. Rink, A., Fung, K.M., Trojanowski, J.Q., Lee, V.M., Neugebauer, E. and McIntosh, T.K. (1995) Evidence of apoptotic cell death after experimental traumatic brain injury in the rat. Am. J. Pathol., 147: 1575–1583. Roof, R.L. and Hall, E.D. (2000) Gender differences in acute CNS trauma and stroke: neuroprotective effects of estrogen and progesterone. J. Neurotrauma, 17: 367–388. Saatman, K.E., Murai, H., Bartus, R.T., Smith, D.H., Hayward, N.J., Perri, B.R. and McIntosh, T.K. (1996) Calpain inhibitor AK295 attenuates motor and cognitive deficits following experimental brain injury in the rat. Proc. Natl. Acad. Sci. U.S.A., 93: 3428–3433. Saatman, K.E., Zhang, C., Bartus, R.T. and McIntosh, T.K. (2000) Behavioral efﬁcacy of posttraumatic calpain inhibition is not accompanied by reduced spectrin proteolysis, cortical lesion, or apoptosis. J. Cereb. Blood Flow Metab., 20: 66–73. Saikumar, P., Dong, Z., Weinberg, J.M. and Venkatachalam, M.A. (1998) Mechanisms of cell death in hypoxia/reoxygenation injury. Oncogene, 17: 3341–3349. Saltman, R.B. and Findling, R.L. (1997) European Health Care Reform. Analysis of current strategies. In: WHO Regional Publications European Series 72 Copenhagen: WHO.

Schendel, S.L., Montal, M. and Reed, J.C. (1998) Bcl-2 family proteins as ion-channels. Cell Death Differ., 5: 372–380. Scorrano, L. and Korsmeyer, S.J. (2003) Mechanisms of cytochrome c release by proapoptotic BCL-2 family members. Biochem. Biophys. Res. Commun., 304: 437–444. Siesjo, B.K., Hu, B. and Kristian, T. (1999) Is the cell death pathway triggered by the mitochondrion or the endoplasmic reticulum? J. Cereb. Blood Flow Metab., 19: 19–26. Singleton, R.H. and Povlishock, J.T. (2004) Identiﬁcation and characterization of heterogeneous neuronal injury and death in regions of diffuse brain injury: evidence for multiple independent injury phenotypes. J. Neurosci., 24: 3543–3553. Singleton, R.H., Stone, J.R., Okonkwo, D.O., Pellicane, A.J. and Povlishock, J.T. (2001) The immunophilin ligand FK506 attenuates axonal injury in an impact acceleration model of traumatic brain injury. J. Neurotrauma, 18: 607–614. Skoglosa, Y., Lewen, A., Takei, N., Hillered, L. and Lindholm, D. (1999) Regulation of pituitary adenylate cyclase activating polypeptide and its receptor type 1 after traumatic brain injury: comparison with brain-derived neurotrophic factor and the induction of neuronal cell death. Neuroscience, 90: 235–247. Soustiel, J.F., Palzur, E., Nevo, O., Thaler, I. and Vlodavsky, E. (2005) Neuroprotective anti-apoptosis effect of estrogens in traumatic brain injury. J. Neurotrauma, 22: 345–352. Stone, J.R., Okonkwo, D.O., Dialo, A.O., Rubin, D.G., Mutlu, L.K., Povlishock, J.T. and Helm, G.A. (2004) Impaired axonal transport and altered axolemmal permeability occur in distinct populations of damaged axons following traumatic brain injury. Exp. Neurol., 190: 59–69. Stys, P.K. and Jiang, Q. (2002) Calpain-dependent neuroﬁlament breakdown in anoxic and ischemic rat central axons. Neurosci. Lett., 328: 150–154. Suehiro, E., Singleton, R.H., Stone, J.R. and Povlishock, J.T. (2001) The immunophilin ligand FK506 attenuates the axonal damage associated with rapid rewarming following posttraumatic hypothermia. Exp. Neurol., 172: 199–210. Sullivan, P.G., Keller, J.N., Bussen, W.L. and Scheff, S.W. (2002) Cytochrome c release and caspase activation after traumatic brain injury. Brain Res., 949: 88–96. Susin, S.A., Lorenzo, H.K., Zamzami, N., Marzo, I., Snow, B.E., Brothers, G.M., Mangion, J., Jacotot, E., Costantini, P., Loefﬂer, M., Larochette, N., Goodlett, D.R., Aebersold, R., Siderovski, D.P., Penninger, J.M. and Kroemer, G. (1999) Molecular characterization of mitochondrial apoptosis-inducing factor. Nature, 397: 441–446. Susin, S.A., Zamzami, N. and Kroemer, G. (1998) Mitochondria as regulators of apoptosis: doubt no more. Biochim. Biophys. Acta, 1366: 151–165. Tabuchi, A., Funaji, K., Nakatsubo, J., Fukuchi, M., Tsuchiya, T. and Tsuda, M. (2003) Inactivation of aconitase during the apoptosis of mouse cerebellar granule neurons induced by a deprivation of membrane depolarization. J. Neurosci. Res., 71: 504–515. Tamas, A., Zsombok, A., Farkas, O., Reglodi, D., Pal, J., Buki, A., Lengvari, I., Povlishock, J.T. and Doczi, T. (2006) Postinjury administration of pituitary adenylate cyclase activating

95 polypeptide (PACAP) attenuates traumatically induced axonal injury in rats. J. Neurotrauma, 23: 686–695. Thurman, D.J., Alverson, C., Dunn, K.A., Guerrero, J. and Sniezek, J.E. (1999) Traumatic brain injury in the United States: A public health perspective. J. Head Trauma Rehabil., 14: 602–615. Van Den, H.C., Donkin, J.J., Finnie, J.W., Blumbergs, P.C., Kuchel, T., Koszyca, B., Manavis, J., Jones, N.R., Reilly, P.L. and Vink, R. (2004) Downregulation of amyloid precursor protein (APP) expression following post-traumatic cyclosporin-A administration. J. Neurotrauma, 21: 1562–1572. Varshavsky, A. (2003) The N-end rule and regulation of apoptosis. Nat. Cell Biol., 5: 373–376. Vaudry, D., Cottet-Rousselle, C., Basille, M., Falluel-Morel, A., Fournier, A., Vaudry, H. and Gonzalez, B.J. (2004) Pituitary adenylate cyclase-activating polypeptide inhibits caspase-3 activity but does not protect cerebellar granule neurons against beta-amyloid (25–35)-induced apoptosis. Regul. Pept., 123: 43–49. Vaudry, D., Gonzalez, B.J., Basille, M., Yon, L., Fournier, A. and Vaudry, H. (2000) Pituitary adenylate cyclase-activating polypeptide and its receptors: from structure to functions. Pharmacol. Rev., 52: 269–324. Vukic, M., Negovetic, L., Kovac, D., Ghajar, J., Glavic, Z. and Gopcevic, A. (1999) The effect of implementation of guidelines for the management of severe head injury on patient

treatment and outcome. Acta Neurochir. (Wien.), 141: 1203–1208. Wang, K.K., Posmantur, R., Nath, R., McGinnis, K., Whitton, M., Talanian, R.V., Glantz, S.B. and Morrow, J.S. (1998) Simultaneous degradation of alphaII- and betaII-spectrin by caspase 3 (CPP32) in apoptotic cells. J. Biol. Chem., 273: 22490–22497. Warren, M.W., Kobeissy, F.H., Liu, M.C., Hayes, R.L., Gold, M.S. and Wang, K.K. (2005) Concurrent calpain and caspase-3 mediated proteolysis of alpha II-spectrin and tau in rat brain after methamphetamine exposure: a similar proﬁle to traumatic brain injury. Life Sci., 78: 301–309. Yakovlev, A.G., Knoblach, S.M., Fan, L., Fox, G.B., Goodnight, R. and Faden, A.I. (1997) Activation of CPP32-like caspases contributes to neuronal apoptosis and neurological dysfunction after traumatic brain injury. J. Neurosci., 17: 7415–7424. Yamashima, T. (2004) Ca2+-dependent proteases in ischemic neuronal death: a conserved ‘calpain-cathepsin cascade’ from nematodes to primates. Cell Calcium, 36: 285–293. Zong, W.X., Li, C., Hatzivassiliou, G., Lindsten, T., Yu, Q.C., Yuan, J. and Thompson, C.B. (2003) Bax and Bak can localize to the endoplasmic reticulum to initiate apoptosis. J. Cell Biol., 162: 59–69. Zoratti, M. and Szabo, I. (1995) The mitochondrial permeability transition. Biochim. Biophys. Acta, 1241: 139–176.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 7

Substance P in traumatic brain injury James J. Donkin1, Renee J. Turner1, Islam Hassan1 and Robert Vink1,2, 1 Discipline of Pathology, University of Adelaide, Adelaide, South Australia Centre for Neurological Diseases, The Hanson Institute, Adelaide, South Australia

2

Abstract: Recent evidence has suggested that neuropeptides, and in particular substance P (SP), may play a critical role in the development of morphological injury and functional deﬁcits following acute insults to the brain. Few studies, however, have examined the role of SP, and more generally, neurogenic inﬂammation, in the pathophysiology of traumatic brain injury and stroke. Those studies that have been reported suggest that SP is released following injury to the CNS and facilitates the increased permeability of the blood brain barrier, the development of vasogenic edema and the subsequent cell death and functional deﬁcits that are associated with these events. Inhibition of the SP activity, either through inhibition of the neuropeptide release or the use of SP receptor antagonists, have consistently resulted in profound decreases in edema formation and marked improvements in functional outcome. The current review summarizes the role of SP in acute brain injury, focussing on its properties as a neurotransmitter and the potential for SP to adversely affect outcome. Keywords: neurotrauma; traumatic brain injury; edema; neuropeptides such as helmets, airbags and seatbelts, little can be done to prevent primary injury, and such injury may be regarded as irreversible. In contrast, secondary injury is made up of the delayed biochemical and physiological factors that are initiated by the primary event, and these secondary injury factors are thought to account for much of the morbidity following brain injury (McIntosh et al., 1996). This secondary injury cascade evolves over minutes to days and even months after the initial event, and as such, there are opportunities for interventional pharmacology to prevent further injury and improve outcome. As a result, research has focused on the identiﬁcation of secondary injury factors and the development of novel therapies that attenuate, or even prevent, their action. A number of secondary injury factors have been identiﬁed to date including blood brain barrier (BBB) opening, edema formation, release of neurotransmitters such as excitatory amino acids, ion changes, oxidative stress and bioenergetic failure,

Introduction Traumatic brain injury is the leading cause of death and disability in people under the age of 40 years (Fleminger and Ponsford, 2005) with incidence rates estimated at 150–250 cases per 100,000 populations per year (Leon-Carrion et al., 2005). The cost for rehabilitation and care of such individuals to the community runs into billion of dollars annually. Despite the enormity of this public health problem, no effective treatment currently exists. It is now accepted that brain injury results in the development of neurologic deﬁcits through two main mechanisms. Firstly, the primary event includes the mechanical processes such as shearing, laceration and stretching of nerve ﬁbres that occurs at the time of the injury (Graham et al., 1992, 1996). Besides the use of preventive measures Corresponding author. Tel.: +61-8–8222-3092; Fax: +61-8-

8222-3093; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61007-8

97

98

amongst others. At the cellular level, the initial effect of mechanical impact is to increase the selective permeability of the cell membrane and this occurs to varying degrees depending on the severity of injury. This effect, known as mechanoporation (Gennarelli and Graham, 1998), allows for the increased movement of ions into and out of cells along their natural concentration gradients. Thus, calcium (Ca2+), sodium (Na+) and chloride (Cl) ions enter cells whilst potassium (K+) and magnesium (Mg2+) ions are lost from the cells. From this point, the pathological changes might be considered to differentiate into two subroutines according to whether these alterations in ion concentration cause effects due to their chemical properties (the enzymatic subroutine) or due to their physical properties (the osmotic subroutine). The enzymatic subroutine revolves around the inﬂux of calcium ions, which activates several cellular enzyme cascades. These enzyme cascades mediate cellular dysfunction, including activation of calpains, axonal injury, accumulation of free radical species, increased production of nitric oxide and induction of proinﬂammatory gene expression, which can potentially culminate in cell death (Obrenovitch and Urenjak, 1997; Xiong et al., 1997; Vespa et al., 1998). Among these different mechanisms of delayed cell damage in TBI, inﬂammation is the predominant mechanism in the case of contusions (Graham et al., 2002). The inﬂammatory reaction consists of various components that evolve at their own speciﬁc rate and according to their own speciﬁc pattern as the age of the lesion increases (Oehmichen and Raff, 1980; Oehmichen et al., 1986; Cervos-Navarro and Lafuente, 1991). For example, in terms of inﬂammatory cell inﬁltration, several microscopic studies of human injury have demonstrated a distinct time course (Holmin et al., 1998; Hausmann et al., 1999; Engel et al., 2000). In lesions aged up to 24 h, the cellular component of inﬂammation was represented by margination of neutrophils (also referred to polymorphonuclear leukocytes or PMNLs) in the vessels, whereas at 3–5 days of survival, the inﬂammatory cell reaction consisted of tissue inﬁltration of not only neutrophils, but also monocyte/macrophages and CD4- and CD8positive T-lymphocytes, as well as an activation of

resident microglia. Changes in inﬂammatory cells are paralleled by proliferation of astrocytes (Hausmann and Betz, 2000), proliferation of capillaries, swelling of their endothelium and by the formation of perivascular edema (Bullock et al., 1991; Vaz et al., 1997). The changes often culminate in a gliotic scar studded with hemosiderinladen macrophages. The osmotic subroutine occurs because the net inﬂux of ions is much greater than the net efﬂux of ions. Consequently, water is osmotically obligated to follow the passage of ions into cells. This leads to cellular swelling, referred to as cytotoxic edema. Glia also swell due to the fact that they function in the uptake of the K+ accumulating in the extracellular ﬂuid (Reilly, 2001). This glial swelling may further compromise cerebral perfusion by compressing the small blood vessels running amidst the glial cells. Alternatively, water may be obligated to follow an osmotic gradient generated by the passage of proteins and ions from the vasculature to the brain interstitium. This edema is known as vasogenic edema and is associated with an increased permeability of the BBB, best observed in the ﬁrst 5 h after the TBI (O’Connor et al., 2003). The microvasculature in the injury zone is affected such that capillaries exhibit increased permeability and arterioles lose their capacity to regulate blood ﬂow (Dietrich et al., 1994). Although the exact mechanisms of BBB disruption are unknown, it is hypothesised that inﬂammatory mediators play a role, possibly through receptor-mediated actions. Among these inﬂammatory mediators, neuropeptides such as substance P (SP), released from perivascular axons, are prime candidates. It is clear that the development of edema is common to both the enzymatic and osmotic subroutines of injury following TBI, and its adverse consequences on outcome through effects on intracranial pressure (ICP) have been well described (Marmarou et al., 2000). Current protocols for the management of raised ICP include pharmacological regimens such as administration of hyperosmotic agents and barbiturates, or induction of hyperventilation and hypothermia, as well as surgical procedures such as drainage of cerebrospinal ﬂuid (CSF) and decompressive craniotomy (Graham et al., 2002). Unfortunately, in terms of

99

improving patient survival rates and functional outcome, these interventions have essentially been inadequate, largely because they do not address the fundamental issue of what speciﬁc mechanisms are associated with edema development after TBI. Recent studies have suggested that neuropeptides, and in particular SP, may play a critical role in edema formation, not only in terms of vasogenic edema associated with increased BBB permeability, but also in the later cytotoxic phase of edema development (Nimmo et al., 2004). Its involvement in the pathophysiology of TBI therefore seems to straddle both the enzymatic and osmotic subroutines of injury.

Substance P SP was ﬁrst identiﬁed in the early part of 1930 (Von Euler and Gaddum, 1931), initially as a crude extract isolated from equine brain and gut. The letter P derives from the ‘powder’ they extracted that contained the active substance. It was found to have potent hypotensive and smooth muscle contractile properties (Von Euler and Gaddum, 1931), and was identiﬁed in high concentrations in the dorsal root of the spinal cord, leading to the proposal that it was a neuronal sensory transmitter associated with pain transmission (Lembeck, 1953). Today, it is accepted that SP is released from both central and peripheral endings of primary afferent neurons and functions as a neurotransmitter (Otsuka and Yoshioka, 1993). Susan Leeman and colleagues (Chang et al., 1971) identiﬁed SP as an undecapeptide in the early 1970s, and were the ﬁrst to synthesise the compound (Tregear et al., 1971) and set up radioimmunoassays (Powell et al., 1973). Such advances allowed the effects of SP to be tested in physiological models (Takahashi et al., 1974; Henry, 1976), using antibodies to monitor SP by radioimmunoassay and immunohistochemistry (Hokfelt et al., 1975; Nilsson et al., 1975; Takahashi and Otsuka, 1975; Cuello and Kanazawa, 1978; Ljungdahl et al., 1978; Costa et al., 1980; Schultzberg et al., 1980) and demonstrating controlled neuronal release (Otsuka and Konishi, 1976; Olgart et al., 1977). It’s role as a neurotransmitter was subsequently

widely accepted (Otsuka and Takahashi, 1977; Nicoll et al., 1980; Pernow, 1983; Otsuka and Yoshioka, 1993). In terms of structure, the fact that certain neuropeptides share speciﬁc amino acid sequences allows this vast collection of molecules to be sorted into families, such as the tachykinin family which includes SP, calcitonin gene-related peptide (CGRP) and neurokinin A (NKA). The tachykinin peptide family certainly represents one of the largest peptide families described in animals (Severini et al., 2002). Synthesis The formation of the peptide bonds of neuropeptides necessitates that they be synthesised on ribosomes, structures present exclusively in the cell body. It is common for the mRNA encoding neuropeptides to initially be translated into a larger protein precursor. Two genes exist, the preprotachykinin A (PPTA) gene and PPTB gene. The PPTA gene can express four different forms of mRNA through alternative splicing, two of which (the b and g forms) encode synthesis of both SP and NKA. Expression of SP and its mRNA is widely abundant in both CNS and PNS (Harrison and Geppetti, 2001). aPPTA expression is more abundant in the brain, whilst bPPT-A and gPPTA mRNAs predominate in peripheral tissues (Kotani et al., 1986). The b and g forms of PPTA mRNA also encode synthesis of neuropeptide K (NPK) and neuropeptide g (NPg), which are elongated forms of NKA. However their function has not been fully elucidated. The PPTB gene gives rise to neurokinin B (NKB) (Hokfelt et al., 2001). Localisation SP immunoreactivity has been demonstrated in the rhinencephalon, telencephalon, basal ganglia, hippocampus, amygdala, septal areas, diencephalon, hypothalamus, mesencephalon, metencephalon, pons, myelencephalon and spinal cord (Shults et al., 1984). Nerve ﬁbres containing SPlike immunoreactivity are common in most autonomic ganglia (Helke et al., 1982; Helke and Phillips, 1988; Bergner et al., 2000), and SP

100

immunoreactivity has also been described in trigeminal and dorsal root ganglia (Lee et al., 1985; Gibbins et al., 1987) and intrinsic neurons of the gut (Sternini et al., 1995). In the autonomic ganglia SP is thought to play a modulatory role, the best characterised response being observed in guinea pig inferior mesenteric ganglion. In these neurons SP mimics a slow depolarisation, which can be evoked by repetitive afferent nerve stimulation (Dun and Minota, 1981). Peripheral inﬂammation also leads to an increase in SP immunoreactivity within the superﬁcial spinal cord (Marlier et al., 1991) and increased SP release (Schaible et al., 1990). Activation or damage to neurons leads to changes in neuropeptide biosynthesis that results from induction of neuropeptide gene expression (Hokfelt et al., 1994). Speciﬁcally, the expression of PPT mRNA (Noguchi et al., 1988) and SP (NK1) receptor mRNA (McCarson, 1999) are upregulated in the periphery during noxious stimulation or neurogenic inﬂammation (Harrison and Geppetti, 2001). SP is released from its precursor by the actions of proteases called convertases. Cleavage points for the convertases on the PPT gene are doublets of cationic residues (Harrison and Geppetti, 2001). After release, the only mechanisms to terminate the action of neuropeptides are diffusion away from the receptor site or degradation by extracellular peptidases. The slow nature of these processes accounts for the prolonged effects of neuropeptides (Kandel and Squire, 2000).

Metabolism Enzymes involved in metabolising SP include neutral endopeptidase (NEP) (Matsas et al., 1984), SP-degrading enzyme (Probert and Hanley, 1987), angiotensin-converting enzyme (ACE) (Skidgel and Erdos, 1987), dipeptidyl aminopeptide IV (Heymann and Mentlein, 1978), post-proline endopeptidase (Blumberg et al., 1980), cathepsin-D (Azaryan and Galoyan, 1988) and cathepsin-E (Kageyama, 1993). While all of these enzymes are known to cleave SP in vitro, their individual cellular localisation suggests that NEP and/or ACE are most likely to be involved in the

cleavage of SP in vivo (Nadel, 1991). NEP has been demonstrated to metabolise SP in the brain (Hooper and Turner, 1987), spinal cord (Sakurada et al., 1990) and peripheral tissues (Di Maria et al., 1998), while ACE has been reported to degrade SP in plasma (Wang et al., 1991), CSF and substantia nigra. ACE has also been shown to contribute to the degradation of fragments released from NEP. Both NEP and ACE catalyse the hydrolysis bonds of SP, leaving the peptide lacking the carboxyl terminal regions required to bind to the tachykinin receptors (Skidgel and Erdos, 1987). Receptors The biological actions of SP are mediated by tachykinin (neurokinin: NK) receptors (Harrison and Geppetti, 2001), rhodopsin-like membrane structures consisting of seven hydrophobic transmembrane domains, connected by extracellular and intracellular loops and coupled to G-proteins (Nakanishi, 1991; Gerard et al., 1993; Maggi and Schwartz, 1997). There are three types of mammalian tachykinin receptors that have been cloned: NK1, NK2 and NK3 exhibiting preferences for SP, NKA and NKB, respectively (Regoli et al., 1994). Endogenous tachykinins are not highly selective for any given receptor, and may act on all three receptors with varying afﬁnities under certain conditions such as receptor availability or high peptide concentrations. For this reason SP activates not only NK1 receptors, but also NK2 and NK3 receptors in a number of tissues (Regoli et al., 1994). Functions SP has been implicated in memory and reinforcement processes. The amino terminus (NH2) of SP has been found to be involved in the memory promoting effects of SP while the carboxy terminus (COOH) is involved in reinforcing properties. An NK1 antagonist (WIN 51708) was found to block these actions, indicating that the behavioural effects of SP are mediated by NK1 receptor (Hasenohrl et al., 2000). SP is also expressed widely in areas of the brain involved in fear producing pathways,

101

including the amygdala, septum, hippocampus, hypothalamus and periaqueductal gray (Rupniak and Kramer, 1999), and accordingly is released in response to aversive stimuli. As expected, injection of SP into regions such as the periaqueductal gray modulates defensive reactions, while administration of an NK1 antagonist inhibits long-lasting audible vocalisation, which is a sign of anxiety and fear (Severini et al., 2002). Intracerebroventricular injection of SP results in many diverse effects in rodents including increased blood pressure and heart rate, increased hindlimb rearing behaviour, scratching, skin biting and grooming. In bronchial smooth muscle, activation of sensory neurons leads to vasodilation and increased vascular permeability, effects that are abolished by pretreatment with capsaicin, an agent that causes neuropeptide depletion, or by treatment with a SP antagonist. Histamine antagonists also blocked nearly all of the stimulatory effects of SP. Therefore, SP containing neurons play a major role in the local regulation of airway resistance and interstitial ﬂuid transfer in various pathological conditions (Lundberg et al., 1983). Trigeminovascular system Cerebral blood vessels are innervated by a combination of sympathetic, parasympathetic and trigeminal somatic nerve ﬁbres all of which are important in cerebrovascular regulation (Atalay et al., 2002). The trigeminal component of this innervation is commonly referred to as the trigeminovascular system. This system has been shown to transmit pain sensation from the dura mater and cranial vessels (Huber, 1899; Penﬁeld, 1932, 1934, 1940; Feindel et al., 1960). The perivascular endings of trigeminovascular ﬁbres contain several neurotransmitters including SP (Edvinsson et al., 1989), CGRP (Uddman et al., 1985; McCulloch et al., 1986), NKA (Edvinsson et al., 1988), nitric oxide and amylin (Edvinsson et al., 2001). Neurogenic inﬂammation Bayliss (1901) initially described vasodilatation of lower limb vessels following stimulation of the

dorsal root ganglia. The concept of neurogenic inﬂammation has since evolved to encompass vasodilatation, plasma extravasation and neuronal hypersensitivity caused by the release of neuropeptides, including SP and CGRP, from sensory neurons (Black, 2002). The effects of sensory neuropeptides are particularly prominent at the level of the vasculature where they cause vasodilation of arterioles, plasma protein extravasation in post-capillary venules and leukocyte adhesion to endothelial cells of venules (Geppetti et al., 1995). Additional tissue-speciﬁc responses produced by neurogenic inﬂammation include smooth muscle relaxation/contraction in the urinary bladder, ureter and iris, inotropic and chronotropic effect on the heart and bronchoconstriction in the airways amongst others (Geppetti et al., 1995). Peptide-containing primary sensory neurons are characterised by their unique sensitivity to capsaicin, the pungent ingredient found in capsicum (Szallasi and Blumberg, 1999). The recent cloning of the channel operated by capsaicin, the transient receptor potential vanilloid receptor-1 (TRPVR-1) (Caterina et al., 1997), has clariﬁed the molecular basis of the selective action of capsaicin on sensory neurons. This seven transmembrane domain protein is a non-selective cation channel, whose endogenous stimulants include heat (>431C) and protons. Subsets of primary sensory neurons are stimulated selectively by capsaicin, causing the release of sensory neuropeptides and promoting neurogenic inﬂammation. At higher concentrations capsaicin kills neurons, blocking the genesis of subsequent neurogenic inﬂammatory responses (Szallasi and Blumberg, 1999). These neurons are deﬁned as ‘‘capsacin-sensitive’’ due to the speciﬁc excitatory/desensitisation effect of capsaicin (Szolcsanyi and Mozsik, 1984).

Edema Of all the secondary injury factors involved in the development of neuronal dysfunction, edema formation is thought to be central to outcome following injury (Lobato et al., 1988; Sarabia et al., 1988). This is particularly the case in younger victims of TBI where formation of edema within the

102

brain has been found to be responsible for 50% of all death and disability (Feickert et al., 1999). The mechanisms associated with cerebral edema formation remain largely unknown, however, investigation of peripheral tissues (Woie et al., 1993) has demonstrated an association between neuropeptides, development of increased vascular permeability and subsequent edema formation. Termed neurogenic inﬂammation, the process involves the stimulation of the slow velocity C-ﬁbres (nociceptors), initiating the release of neuropeptides (Woie et al., 1993) including SP, CGRP and NKA, which produce vasodilation, edema and tissue swelling. While SP has been recognised as being primarily associated with increased vascular permeability, SP is stored and co-released from sensory nerve endings with CGRP, which is a most potent endogenous vasodilator, and displays potent edema producing activity in the presence of SP (Severini et al., 2002). When left unchecked cerebral edema results in an increase in ICP that may lead to a decrease in tissue perfusion, localised hypoxia and ischemia, and in severe cases brain herniation and death. It is therefore of paramount importance that edema genesis is inhibited following TBI. Whilst a number of studies have investigated the role of classical inﬂammation in edema formation following TBI (Stahel et al., 1998; Lenzlinger et al., 2001; Stein and Hoffman, 2003; Besson et al., 2005), to date only one study has examined the role of neurogenic inﬂammation post-trauma. Nimmo et al. (2004) demonstrated that capsaicin administered prior to TBI signiﬁcantly attenuated BBB opening, edema formation and the development of both motor and cognitive deﬁcits. The authors concluded that a capsaicin-induced depletion of neuropeptides clearly prevented the development of neurogenic inﬂammation, and inhibited development of vasogenic edema (Fig. 1). Although, this study demonstrated a role for neurogenic inﬂammation in TBI, the identiﬁcation of which neuropeptide is primarily involved in the formation of increased BBB permeability and edema formation was not established. Previous studies of peripheral edema have shown that SP is the neuropeptide most closely associated with capsaicin sensitivity (Saria, 1984;

Fig. 1. Alterations in brain water content at 5 h following diffuse traumatic brain injury in rats. Vehicle treated controls showed signiﬁcantly more (po0.001) edema than either sham or capsaicin pre-treated animals.

Yonehara et al., 1987; De and Ghosh, 1990; Laird et al., 2000). It is also well known that SP is the neuropeptide responsible for increased vascular permeability, whereas CGRP is primarily associated with vasodilation. Finally, Kramer et al. (2003) have demonstrated in studies of cardiac ischaemia that SP release is increased with magnesium depletion; declines in magnesium concentration have been widely described following TBI (Vink et al., 1987, 1988). Therefore it is feasible to propose that drugs acting on SP receptors, and in particular NK1 receptor (Hokfelt et al., 2001), may be beneﬁcial in disease treatment. Studies of neurogenic inﬂammation following acute brain injury have provided evidence supporting such a possibility.

Acute CNS injury Virtually all blood vessels of the body are surrounded by sensory nerve ﬁbres that contain both CGRP and SP. Cerebral arteries, in particular, appear to receive a dense supply of these neurones, and it is therefore consistent that these neurones have a role as mediators of the inﬂammatory process following injury. Studies of migraine (Ferrari, 1998) have indeed demonstrated that neuropeptides are a therapeutic target to reduce vascular permeability. However, only a limited number of studies have demonstrated that

103

neurogenic inﬂammation may play a role in acute injury. Signiﬁcant amounts of SP release has been detected in the nervous system following both peripheral nerve injury (Malcangio et al., 2000), traumatic spinal cord injury (Sharma et al., 1990) and more recently, in vitro studies of endothelium stimulation (Annunziata et al., 2002). In our own TBI studies, perivascular SP immunoreactivity has been shown to increase after TBI (Fig. 2), irrespective of the injury model (focal versus diffuse) or severity of injury. In ischaemia, SP immunoreactivity has been shown to increase in GABAergic interneurons around regions of infarction, and transiently expressed in cerebrovenular endothelium (Stumm et al., 2001). Our own studies have shown increased SP immunoreactivity following ischaemic brain injury in the rat, which was associated with a signiﬁcant increase in edema formation (Turner et al., 2006). With respect to the glial response after acute injury, receptor-binding sites for SP have been shown to increase on glia after neuronal injury (Mantyh et al., 1989). Because SP is known to regulate inﬂammatory and immune responses in peripheral tissue, it therefore may regulate the glial response to injury. Subsequent studies have conﬁrmed that SP receptors are expressed on astrocytes after injury and may therefore be linked to their transformation to reactive astrocytes (Lin, 1995). This increase was not observed in undamaged areas. Several studies using NK1 receptor antagonists also support a role for SP in neurogenic inﬂammation following ischaemic injury, although few have been applied to studies of the CNS. For

example, in a rodent cerebral stroke model, the NK1 receptor antagonist (SR140333) reduced infarct volume after focal ischaemia, implying that that SP might play a role in exacerbating ischaemic damage (Yu et al., 1997). However, despite these positive ﬁndings there has been no further work published in this area. Interestingly, serum levels of SP have been measured in humans with both transient ischaemic attack and complete stroke, and these have been found to be signiﬁcantly elevated compared with controls (Bruno et al., 2003). In other organs, post-ischaemic blockade of tachykinin receptors have been shown to inhibit vascular permeability, neutrophil recruitment, intestinal haemorrhage and neutropaenia following ischaemia and reperfusion of the superior mesenteric artery in the rat (Souza et al., 2002). Similarly, SP antagonists reduced post-ischaemic myocardial injury in rats with dietary Mg deﬁciency (Kramer et al., 1997), and the authors suggest that SP may play an early critical role in inﬂammatory/pro-oxidant responses following ischaemia. This ﬁnding is of particular interest to TBI as traumatic injury produces sustained decline in intracellular magnesium, implying that SP may signiﬁcantly contribute to the detrimental effects of magnesium deﬁciency (Heath and Vink, 1996).

NK1 receptor antagonists A number of groups have hypothesised that tachykinin receptor antagonists may have several therapeutic applications (Watling, 1992; Lowe

Fig. 2. SP immunoreactivity at 5 h following diffuse traumatic brain injury in the rat in (A) sham, uninjured rat cortex and (B) injured rat cortex. Note the qualitative increase in perivascular SP immunoreactivity (bar ¼ 100 mm).

104

et al., 1994; Rupniak et al., 2000). The notion of antagonising SP was ﬁrst raised by Leban et al. (1979) when examining the effects of SP agonists in the guinea pig ileum. Subsequently, Folkers et al. (1981) discussed the chemical design of SP antagonists, before Engberg et al. (1981) developed the ﬁrst synthetic peptide antagonist (D-Pro, D-Trp)-SP for use in the CNS, speciﬁcally the block of locus coeruleus (LC) neurones. However, the afﬁnity and metabolic instability of peptides limited their usefulness for in vivo studies. It was the development of the ﬁrst non-peptide SP antagonist by Snider et al. (1991), who showed that CP-96345 was a potent, competitive and highly selective antagonist of the NK1 receptor, that initiated a new wave of interest in these compounds. This antagonist was further reﬁned to create the highly potent n-acetyl-L-tryptophan benzyl esters (MacLeod et al., 1993, 1995), and it was the 3,5-bis (triﬂuoromethyl) benzyl ester (L-732138) that led to the inhibition of SP-induced inositol phosphate accumulation in Chinese hamster ovarian cells expressing the human NK1 receptor. L-732138 was then used as a starting point to identify highafﬁnity SP receptor antagonists with improved in vivo activity. Rupniak and Kramer (1999) ﬁrst described the efﬁcacy of SP antagonists in the treatment of experimental depression and emesis, with the NK1 receptor antagonists being able to decrease anxiety (Santarelli et al., 2001) and depression with fewer side effects than other drugs of choice for the treatment of depression (Severini et al., 2002). Ranga and Krishnan (2002) subsequently published the ﬁrst use of the SP antagonist MK-0869 in the treatment of clinical depression and anxiety. NK1 receptor antagonists have also been tested in dental pain, osteoarthritis, neuropathic pain and migraine however no analgesic effects have been reported in studies to date (Hokfelt et al., 2001). An alternative to inhibiting SP binding is to decrease neuropeptide release and depleting neuropeptide stores using the vanilloid agonist capsaicin (Cadieux et al., 1986; Kashiba et al., 1997). Nimmo et al. (2004) were able to demonstrate that neuropeptide-depletion prior to induction of TBI attenuated the development of BBB dysfunction and vasogenic edema formation associated with

neurogenic inﬂammation, as well as improve motor and cognitive function after TBI. In guinea pig skin, administration of the NK1 receptor antagonist RP 67580 was able to inhibit SP-induced edema formation and white blood cell accumulation (Campos and Calixto, 2000). In vitro studies of endothelial injury have supported a potential role for SP antagonists in the treatment of BBB dysfunction by demonstrating that such antagonists neutralised increased BBB permeability, upregulation of MHC-II molecules, reduced expression of ICAM-1 and prevented associated cell morphological changes (Annunziata et al., 2002). Conclusion While a role for neurogenic inﬂammation in vascular permeability and edema formation has been described in peripheral tissues for a number of years, few studies have examined the potential for neurogenic inﬂammation to inﬂuence BBB permeability and edema formation after traumatic brain injury. Those studies that have investigated a role for neuropeptides in acute brain injury have demonstrated that inhibition of release attenuates BBB permeability and edema formation after injury, and results in an associated improvement in functional outcome. Immunohistochemistry studies have demonstrated that increased SP levels are observed perivascularly, conﬁrming a potential role in vasogenic edema formation. Given the apparent lack of side effects from this class of compound, and the potential to improve posttraumatic anxiety and depression, inhibition of the SP pathway using NK1 receptor antagonists is expected to provide a novel approach to the management of edema formation and improvement of functional outcome after TBI. References Annunziata, P., Cioni, C., Santonini, R. and Paccagnini, E. (2002) Substance P antagonist blocks leakage and reduces activation of cytokine-stimulated rat brain endothelium. J. Neuroimmunol., 131: 41–49. Atalay, B., Bolay, H., Dalkara, T., Soylemezoglu, F., Oge, K. and Ozcan, O.E. (2002) Transcorneal stimulation of trigeminal nerve afferents to increase cerebral blood ﬂow in rats with

105 cerebral vasospasm: a noninvasive method to activate the trigeminovascular reﬂex. J. Neurosurg., 97: 1179–1183. Azaryan, A.V. and Galoyan, A.A. (1988) Substrate speciﬁcity of cerebral cathepsin D and high-Mr aspartic endopeptidase. J. Neurosci. Res., 19: 268–271. Bayliss, W.M. (1901) On the origin from the spinal cord of the vaso-dilator of the hindlimb, and on the nature of these ﬁbers. J. Physiol. (Lond.), 32: 1025–1043. Bergner, A.J., Murphy, S.M. and Anderson, C.R. (2000) After axotomy, substance P and vasoactive intestinal peptide expression occurs in pilomotor neurons in the rat superior cervical ganglion. Neuroscience, 96: 611–618. Besson, V.C., Chen, X.R., Plotkine, M. and MarchandVerrecchia, C. (2005) Fenoﬁbrate, a peroxisome proliferator-activated receptor alpha agonist, exerts neuroprotective effects in traumatic brain injury. Neurosci. Lett., 388: 7–12. Black, P.H. (2002) Stress and the inﬂammatory response: a review of neurogenic inﬂammation. Brain Behav. Immun., 16: 622–653. Blumberg, S., Teichberg, V.I., Charli, J.L., Hersh, L.B. and Mckelvy, J.F. (1980) Cleavage of substance P to an N-terminal tetrapeptide and a C-terminal heptapeptide by a post-proline cleaving enzyme from bovine brain. Brain Res., 192: 477–486. Bruno, G., Tega, F., Bruno, A., Graf, U., Corelli, F., Molfetta, R. and Barucco, M. (2003) The role of substance P in cerebral ischemia. Int. J. Immunopathol. Pharmacol., 16: 67–72. Bullock, R., Maxwell, W.L., Graham, D.I., Teasdale, G.M. and Adams, J.H. (1991) Glial swelling following human cerebral contusion: an ultrastructural study. J. Neurol. Neurosurg. Psychiatry, 54: 427–434. Cadieux, A., Springall, D.R., Mulderry, P.K., Rodrigo, J., Ghatei, M.A., Terenghi, G., Bloom, S.R. and Polak, J.M. (1986) Occurrence, distribution and ontogeny of CGRP immunoreactivity in the rat lower respiratory tract: effect of capsaicin treatment and surgical denervations. Neuroscience, 19: 605–627. Campos, M.M. and Calixto, J.B. (2000) Neurokinin mediation of edema and inﬂammation. Neuropeptides, 34: 314–322. Caterina, M.J., Schumacher, M.A., Tominaga, M., Rosen, T.A., Levine, J.D. and Julius, D. (1997) The capsaicin receptor: a heat-activated ion channel in the pain pathway. Nature, 389: 816–824. Cervos-Navarro, J. and Lafuente, J.V. (1991) Traumatic brain injuries: structural changes. J. Neurol. Sci., 103(Suppl): S3–S14. Chang, M.M., Leeman, S.E. and Niall, H.D. (1971) Aminoacid sequence of substance P. Nat. New Biol., 232: 86–87. Costa, M., Cuello, A.C., Furness, J.B. and Franco, R. (1980) Distribution of enteric neurons showing immunoreactivity for substance P in the guinea-pig ileum. Neuroscience, 5: 323–331. Cuello, A.C. and Kanazawa, I. (1978) The distribution of substance P immunoreactive ﬁbers in the rat central nervous system. J. Comp. Neurol., 178: 129–156. De, A.K. and Ghosh, J.J. (1990) Inﬂammatory responses induced by substance P in rat paw. Indian J. Exp. Biol., 28: 946–948.

Di Maria, G.U., Belloﬁore, S. and Geppetti, P. (1998) Regulation of airway neurogenic inﬂammation by neutral endopeptidase. Eur. Respir. J., 12: 1454–1462. Dietrich, W.D., Alonso, O. and Halley, M. (1994) Early microvascular and neuronal consequences of traumatic brain injury: a light and electron microscopic study in rats. J. Neurotrauma, 11: 289–301. Dun, N.J. and Minota, S. (1981) Effects of substance P on neurones of the inferior mesenteric ganglia of the guinea-pig. J. Physiol., 321: 259–271. Edvinsson, L., Brodin, E., Jansen, I. and Uddman, R. (1988) Neurokinin A in cerebral vessels: characterization, localization and effects in vitro. Regul. Pept., 20: 181–197. Edvinsson, L., Ekman, R. and Thulin, T. (1989) Reduced levels of calcitonin gene-related peptide (CGRP) but not substance P during and after treatment of severe hypertension in man. J. Hum. Hypertens, 3: 267–270. Edvinsson, L., Elsas, T., Suzuki, N., Shimizu, T. and Lee, T.J. (2001) Origin and co-localization of nitric oxide synthase, CGRP, PACAP, and VIP in the cerebral circulation of the rat. Microsc. Res. Tech., 53: 221–228. Engberg, G., Svensson, T.H., Rosell, S. and Folkders, K. (1981) A synthetic peptide as an antagonist of substance P. Nature, 293: 222–223. Engel, S., Schluesener, H., Mittelbronn, M., Seid, K., Adjodah, D., Wehner, H.D. and Meyermann, R. (2000) Dynamics of microglial activation after human traumatic brain injury are revealed by delayed expression of macrophage-related proteins MRP8 and MRP14. Acta Neuropathol. (Berl.), 100: 313–322. Feickert, H.J., Drommer, S. and Heyer, R. (1999) Severe head injury in children: impact of risk factors on outcome. J. Trauma, 47: 33–38. Feindel, W., Penﬁeld, W. and Mc, N.F. (1960) The tentorial nerves and localization of intracranial pain in man. Neurology, 10: 555–563. Ferrari, M.D. (1998) Migraine. Lancet, 351: 1043–1051. Fleminger, S. and Ponsford, J. (2005) Long term outcome after traumatic brain injury. BMJ, 331: 1419–1420. Folkers, K., Horig, J., Rosell, S. and Bjorkroth, U. (1981) Chemical design of antagonists of substance P. Acta Physiol. Scand., 111: 505–506. Gennarelli, T.A. and Graham, D.I. (1998) Neuropathology of the head injuries. Semin. Clin. Neuropsychiatry, 3: 160–175. Geppetti, P., Bertrand, C., Ricciardolo, F.L. and Nadel, J.A. (1995) New aspects on the role of kinins in neurogenic inﬂammation. Can. J. Physiol. Pharmacol., 73: 843–847. Gerard, N.P., Bao, L., Xiao-Ping, H. and Gerard, C. (1993) Molecular aspects of the tachykinin receptors. Regul. Pept., 43: 21–35. Gibbins, I.L., Furness, J.B. and Costa, M. (1987) Pathwayspeciﬁc patterns of the co-existence of substance P, calcitonin gene-related peptide, cholecystokinin and dynorphin in neurons of the dorsal root ganglia of the guinea-pig. Cell Tissue Res., 248: 417–437. Graham, D.I., Clark, J.C., Adams, J.H. and Gennarelli, T.A. (1992) Diffuse axonal injury caused by assault. J. Clin. Pathol., 45: 840–841.

106 Graham, D.I., Gennarelli, T.A. and McIntosh, T.K. (2002) Greenﬁeld’s neuropathology. Arnold, New York. Graham, D.I., Gentleman, S.M., Nicoll, J.A., Royston, M.C., Mckenzie, J.E., Roberts, G.W. and Grifﬁn, W.S. (1996) Altered beta-APP metabolism after head injury and its relationship to the aetiology of Alzheimer’s disease. Acta Neurochir. Suppl., 66: 96–102. Harrison, S. and Geppetti, P. (2001) Substance P. Int. J. Biochem. Cell Biol., 33: 555–576. Hasenohrl, R.U., Souza-Silva, M.A., Nikolaus, S., Tomaz, C., Brandao, M.L., Schwarting, R.K. and Huston, J.P. (2000) Substance P and its role in neural mechanisms governing learning, anxiety and functional recovery. Neuropeptides, 34: 272–280. Hausmann, R. and Betz, P. (2000) The time course of the vascular response to human brain injury: an immunohistochemical study. Int. J. Legal Med., 113: 288–292. Hausmann, R., Kaiser, A., Lang, C., Bohnert, M. and Betz, P. (1999) A quantitative immunohistochemical study on the time-dependent course of acute inﬂammatory cellular response to human brain injury. Int. J. Legal Med., 112: 227–232. Heath, D.L. and Vink, R. (1996) Traumatic brain axonal injury produces sustained decline in intracellular free magnesium concentration. Brain Res., 738: 150–153. Helke, C.J., Neil, J.J., Massari, V.J. and Loewy, A.D. (1982) Substance P neurons project from the ventral medulla to the intermediolateral cell column and ventral horn in the rat. Brain Res., 243: 147–152. Helke, C.J. and Phillips, E.T. (1988) Substance P antagonistinduced spinal cord vasoconstriction: effects of thyrotropinreleasing hormone and substance P agonists. Peptides, 9: 1307–1315. Henry, J.L. (1976) Effects of substance P on functionally identiﬁed units in cat spinal cord. Brain Res., 114: 439–451. Heymann, E. and Mentlein, R. (1978) Liver dipeptidyl aminopeptidase IV hydrolyzes substance P. FEBS Lett., 91: 360–364. Hokfelt, T., Kellerth, J.O., Nilsson, G. and Pernow, B. (1975) Experimental immunohistochemical studies on the localization and distribution of substance P in cat primary sensory neurons. Brain Res., 100: 235–252. Hokfelt, T., Pernow, B. and Wahren, J. (2001) Substance P: a pioneer amongst neuropeptides. J. Intern. Med., 249: 27–40. Hokfelt, T., Zhang, X. and Wiesenfeld-Hallin, Z. (1994) Messenger plasticity in primary sensory neurons following axotomy and its functional implications. Trends Neurosci., 17: 22–30. Holmin, S., Soderlund, J., Biberfeld, P. and Mathiesen, T. (1998) Intracerebral inﬂammation after human brain contusion. Neurosurgery, 42: 291–298; discussion 298–299. Hooper, N.M. and Turner, A.J. (1987) Isolation of two differentially glycosylated forms of peptidyl-dipeptidase A (angiotensin converting enzyme) from pig brain: a re-evaluation of their role in neuropeptide metabolism. Biochem. J., 241: 625–633. Huber, G.C. (1899) Observations on the innervation of intracranial vessels. J. Comp. Neurol., 9: 1–34.

Kageyama, T. (1993) Rabbit procathepsin E and cathepsin E. Nucleotide sequence of cDNA, hydrolytic speciﬁcity for biologically active peptides and gene expression during development. Eur. J. Biochem., 216: 717–728. Kandel, E.R. and Squire, L.R. (2000) Neuroscience: breaking down scientiﬁc barriers to the study of brain and mind. Science, 290: 1113–1120. Kashiba, H., Ueda, Y. and Senba, E. (1997) Systemic capsaicin in the adult rat differentially affects gene expression for neuropeptides and neurotrophin receptors in primary sensory neurons. Neuroscience, 76: 299–312. Kotani, Y., Hirota, Y., Sugiyama, K., Joh, S., Shibutani, T., Matsuura, H. and Inoki, R. (1986) Effects of noxious stimuli and anesthetic agents on substance P content in rat central nervous system. Jpn. J. Pharmacol., 40: 143–147. Kramer, J.H., Mak, I.T., Phillips, T.M. and Weglicki, W.B. (2003) Dietary magnesium intake inﬂuences circulating proinﬂammatory neuropeptide levels and loss of myocardial tolerance to postischemic stress. Exp. Biol. Med. (Maywood), 228: 665–673. Kramer, S.D., Jakits-Deiser, C. and Wunderli-Allenspach, H. (1997) Free fatty acids cause pH-dependent changes in drug–lipid membrane interactions around physiological pH. Pharm. Res., 14: 827–832. Laird, J.M., Olivar, T., Roza, C., De Felipe, C., Hunt, S.P. and Cervero, F. (2000) Deﬁcits in visceral pain and hyperalgesia of mice with a disruption of the tachykinin NK1 receptor gene. Neuroscience, 98: 345–352. Leban, J., Rackur, G., Yamaguchi, I., Folkers, K., Bjorkroth, U., Rosell, S., Yanaihara, N. and Yanaihara, C. (1979) Synthesis of substance P analogs and agonistic and antagonistic activities. Acta Chem. Scand. B, 33: 664–668. Lee, Y., Kawai, Y., Shiosaka, S., Takami, K., Kiyama, H., Hillyard, C.J., Girgis, S., Macintyre, I., Emson, P.C. and Tohyama, M. (1985) Coexistence of calcitonin gene-related peptide and substance P-like peptide in single cells of the trigeminal ganglion of the rat: immunohistochemical analysis. Brain Res., 330: 194–196. Lembeck, F. (1953) Central transmission of afferent impulses. III. Incidence and signiﬁcance of the substance P in the dorsal roots of the spinal cord. Naunyn Schmiedebergs Arch. Exp. Pathol. Pharmakol., 219: 197–213. Lenzlinger, P.M., Morganti-Kossmann, M.C., Laurer, H.L. and McIntosh, T.K. (2001) The duality of the inﬂammatory response to traumatic brain injury. Mol. Neurobiol., 24: 169–181. Leon-Carrion, J., Dominguez-Morales Mdel, R., Barroso, Y.M.J.M. and Murillo-Cabezas, F. (2005) Epidemiology of traumatic brain injury and subarachnoid hemorrhage. Pituitary, 8: 197–202. Lin, R.C. (1995) Reactive astrocytes express substance-P immunoreactivity in the adult forebrain after injury. Neuroreport, 7: 310–312. Ljungdahl, A., Hokfelt, T. and Nilsson, G. (1978) Distribution of substance P-like immunoreactivity in the central nervous system of the rat-I. Cell bodies and nerve terminals. Neuroscience, 3: 861–943.

107 Lobato, R.D., Sarabia, R., Cordobes, F., Rivas, J.J., Adrados, A., Cabrera, A., Gomez, P., Madera, A. and Lamas, E. (1988) Posttraumatic cerebral hemispheric swelling. Analysis of 55 cases studied with computerized tomography. J. Neurosurg., 68: 417–423. Lowe III, J.A., Drozda, S.E., Mclean, S., Bryce, D.K., Crawford, R.T., Snider, R.M., Longo, K.P., Nagahisa, A. and Tsuchiya, M. (1994) Aza-tricyclic substance P antagonists. J. Med. Chem., 37: 2831–2840. Lundberg, J.M., Martling, C.R., Saria, A., Folkers, K. and Rosell, S. (1983) Cigarette smoke-induced airway oedema due to activation of capsaicin-sensitive vagal afferents and substance P release. Neuroscience, 10: 1361–1368. Macleod, A.M., Cascieri, M.A., Merchant, K.J., Sadowski, S., Hardwicke, S., Lewis, R.T., Macintyre, D.E., Metzger, J.M., Fong, T.M., Shepheard, S., et al. (1995) Synthesis and biological evaluation of NK1 antagonists derived from L-tryptophan. J. Med. Chem., 38: 934–941. Macleod, A.M., Merchant, K.J., Cascieri, M.A., Sadowski, S., Ber, E., Swain, C.J. and Baker, R. (1993) N-acyl-L-tryptophan benzyl esters: potent substance P receptor antagonists. J. Med. Chem., 36: 2044–2045. Maggi, C.A. and Schwartz, T.W. (1997) The dual nature of the tachykinin NK1 receptor. Trends Pharmacol. Sci., 18: 351–355. Malcangio, M., Ramer, M.S., Jones, M.G. and Mcmahon, S.B. (2000) Abnormal substance P release from the spinal cord following injury to primary sensory neurons. Eur. J. Neurosci., 12: 397–399. Mantyh, P.W., Johnson, D.J., Boehmer, C.G., Catton, M.D., Vinters, V.H., Maggio, J.E., Too, H.P. and Vigna, S.R. (1989) Substance P receptor binding sites are expressed by glia in vivo after neuronal injury. Proc. Natl. Acad. Sci. U.S.A., 86: 5193–5197. Marlier, L., Poulat, P., Rajaofetra, N. and Privat, A. (1991) Modiﬁcations of serotonin-, substance P- and calcitonin gene-related peptide-like immunoreactivities in the dorsal horn of the spinal cord of arthritic rats: a quantitative immunocytochemical study. Exp. Brain Res., 85: 482–490. Marmarou, A., Fatouros, P.P., Barzo, P., Portella, G., Yoshihara, M., Tsuji, O., Yamamoto, T., Laine, F., Signoretti, S., Ward, J.D., Bullock, M.R. and Young, H.F. (2000) Contribution of edema and cerebral blood volume to traumatic brain swelling in head-injured patients. J. Neurosurg., 93: 183–193. Matsas, R., Kenny, A.J. and Turner, A.J. (1984) The metabolism of neuropeptides. The hydrolysis of peptides, including enkephalins, tachykinins and their analogues, by endopeptidase-24.11. Biochem. J., 223: 433–440. McCarson, K.E. (1999) Central and peripheral expression of neurokinin-1 and neurokinin-3 receptor and substance P-encoding messenger RNAs: peripheral regulation during formalin-induced inﬂammation and lack of neurokinin receptor expression in primary afferent sensory neurons. Neuroscience, 93: 361–370. McCulloch, J., Uddman, R., Kingman, T.A. and Edvinsson, L. (1986) Calcitonin gene-related peptide: functional role in

cerebrovascular regulation. Proc. Natl. Acad. Sci. U.S.A., 83: 5731–5735. McIntosh, T.K., Smith, D.H., Meaney, D.F., Kotapka, M.J., Gennarelli, T.A. and Graham, D.I. (1996) Neuropathological sequelae of traumatic brain injury: relationship to neurochemical and biomechanical mechanisms. Lab. Invest., 74: 315–342. Nadel, J.A. (1991) Neutral endopeptidase modulates neurogenic inﬂammation. Eur. Respir. J., 4: 745–754. Nakanishi, S. (1991) Mammalian tachykinin receptors. Annu. Rev. Neurosci., 14: 123–136. Nicoll, R.A., Schenker, C. and Leeman, S.E. (1980) Substance P as a transmitter candidate. Annu. Rev. Neurosci., 3: 227–268. Nilsson, G., Larsson, L.I., Hakanson, R., Brodin, E., Pernow, B. and Sundler, F. (1975) Localization of substance P-like immunoreactivity in mouse gut. Histochemistry, 43: 97–99. Nimmo, A.J., Cernak, I., Heath, D.L., Hu, X., Bennett, C.J. and Vink, R. (2004) Neurogenic inﬂammation is associated with development of edema and functional deﬁcits following traumatic brain injury in rats. Neuropeptides, 38: 40–47. Noguchi, K., Morita, Y., Kiyama, H., Ono, K. and Tohyama, M. (1988) A noxious stimulus induces the preprotachykininA gene expression in the rat dorsal root ganglion: a quantitative study using in situ hybridization histochemistry. Brain Res., 464: 31–35. O’Connor, C.A., Cernak, I. and Vink, R. (2003) Interaction between anesthesia, gender, and functional outcome task following diffuse traumatic brain injury in rats. J. Neurotrauma, 20: 533–541. Obrenovitch, T.P. and Urenjak, J. (1997) Is high extracellular glutamate the key to excitotoxicity in traumatic brain injury? J. Neurotrauma, 14: 677–698. Oehmichen, M., Eisenmenger, W., Raff, G. and Berghaus, G. (1986) Brain macrophages in human cortical contusions as indicator of survival period. Forensic Sci. Int., 30: 281–301. Oehmichen, M. and Raff, G. (1980) Timing of cortical contusion. Correlation between histomorphologic alterations and post-traumatic interval. Z. Rechtsmed., 84: 79–94. Olgart, L., Gazelius, B., Brodin, E. and Nilsson, G. (1977) Release of substance P-like immunoreactivity from the dental pulp. Acta Physiol. Scand., 101: 510–512. Otsuka, M. and Konishi, S. (1976) Release of substance P-like immunoreactivity from isolated spinal cord of newborn rat. Nature, 264: 83–84. Otsuka, M. and Takahashi, T. (1977) Putative peptide neurotransmitters. Annu. Rev. Pharmacol. Toxicol., 17: 425–439. Otsuka, M. and Yoshioka, K. (1993) Neurotransmitter functions of mammalian tachykinins. Physiol. Rev., 73: 229–308. Penﬁeld, W. (1932) Intracerebral vascular nerves. Arch. Neurol. Psychiatry, 27: 30–44. Penﬁeld, W. (1934) A contribution to the mechanism of intracranial pain. Proc. Assoc. Res. Nerv. Mental Dis., 15: 399–415. Penﬁeld, W. (1940) Dural headache and the innervation of the dura mater. Arch. Neurol. Psychiatry, 44: 43–75. Pernow, B. (1983) Substance P. Pharmacol. Rev., 35: 85–141.

108 Powell, D., Leeman, S., Tregear, G.W., Niall, H.D. and Potts Jr., J.T. (1973) Radioimmunoassay for substance P. Nat. New Biol., 241: 252–254. Probert, L. and Hanley, M.R. (1987) The immunocytochemical localisation of ‘substance-P-degrading enzyme’ within the rat spinal cord. Neurosci. Lett., 78: 132–137. Ranga, K. and Krishnan, R. (2002) Clinical experience with substance P receptor (NK1) antagonists in depression. J. Clin. Psychiatry, 63(Suppl 11): 25–29. Regoli, D., Boudon, A. and Fauchere, J.L. (1994) Receptors and antagonists for substance P and related peptides. Pharmacol. Rev., 46: 551–599. Reilly, P.L. (2001) Brain injury: the pathophysiology of the ﬁrst hours. ‘Talk and die revisited’. J. Clin. Neurosci., 8: 398–403. Rupniak, N.M., Carlson, E.C., Harrison, T., Oates, B., Seward, E., Owen, S., De Felipe, C., Hunt, S. and Wheeldon, A. (2000) Pharmacological blockade or genetic deletion of substance P (NK(1)) receptors attenuates neonatal vocalisation in guinea-pigs and mice. Neuropharmacology, 39: 1413–1421. Rupniak, N.M. and Kramer, M.S. (1999) Discovery of the antidepressant and anti-emetic efﬁcacy of substance P receptor (NK1) antagonists. Trends Pharmacol. Sci., 20: 485–490. Sakurada, T., Hara, A., Matsumura, H., Yamada, H., Sakurada, S. and Kisara, K. (1990) A substance P analogue reduces amino acid contents in the rat spinal cord. Pharmacol. Toxicol., 66: 75–76. Santarelli, L., Gobbi, G., Debs, P.C., Sibille, E.T., Blier, P., Hen, R. and Heath, M.J. (2001) Genetic and pharmacological disruption of neurokinin 1 receptor function decreases anxiety-related behaviors and increases serotonergic function. Proc. Natl. Acad. Sci. U.S.A., 98: 1912–1917. Sarabia, R., Lobato, R.D., Rivas, J.J., Cordobes, F., Rubio, J., Cabrera, A., Gomez, P., Munoz, M.J. and Madera, A. (1988) Cerebral hemisphere swelling in severe head injury patients. Acta Neurochir. Suppl. (Wien.), 42: 40–46. Saria, A. (1984) Substance P in sensory nerve ﬁbres contributes to the development of oedema in the rat hind paw after thermal injury. Br. J. Pharmacol., 82: 217–222. Schaible, H.G., Jarrott, B., Hope, P.J. and Duggan, A.W. (1990) Release of immunoreactive substance P in the spinal cord during development of acute arthritis in the knee joint of the cat: a study with antibody microprobes. Brain Res., 529: 214–223. Schultzberg, M., Hokfelt, T., Olson, L., Alund, M., Nilsson, G., Terenius, L., Elde, R., Goldstein, M. and Said, S. (1980) Substance P, VIP, enkephalin and somatostatin immunoreactive neurons in intestinal tissue transplanted to the anterior eye chamber. J. Auton. Nerv. Syst., 1: 291–303. Severini, C., Improta, G., Falconieri-Erspamer, G., Salvadori, S. and Erspamer, V. (2002) The tachykinin peptide family. Pharmacol. Rev., 54: 285–322. Sharma, H.S., Nyberg, F., Olsson, Y. and Dey, P.K. (1990) Alteration of substance P after trauma to the spinal cord: an experimental study in the rat. Neuroscience, 38: 205–212. Shults, C.W., Quirion, R., Chronwall, B., Chase, T.N. and O’donohue, T.L. (1984) A comparison of the anatomical

distribution of substance P and substance P receptors in the rat central nervous system. Peptides, 5: 1097–1128. Skidgel, R.A. and Erdos, E.G. (1987) Cleavage of peptide bonds by angiotensin I converting enzyme. Agents Actions Suppl., 22: 289–296. Snider, R.M., Longo, K.P., Drozda, S.E., Lowe III, J.A. and Leeman, S.E. (1991) Effect of CP-96,345, a nonpeptide substance P receptor antagonist, on salivation in rats. Proc. Natl. Acad. Sci. U.S.A., 88: 10042–10044. Souza, D.G., Mendonca, V.A., De, A.C.M.S., Poole, S. and Teixeira, M.M. (2002) Role of tachykinin NK receptors on the local and remote injuries following ischaemia and reperfusion of the superior mesenteric artery in the rat. Br. J. Pharmacol., 135: 303–312. Stahel, P.F., Morganti-Kossmann, M.C. and Kossmann, T. (1998) The role of the complement system in traumatic brain injury. Brain Res. Brain Res. Rev., 27: 243–256. Stein, D.G. and Hoffman, S.W. (2003) Concepts of CNS plasticity in the context of brain damage and repair. J. Head Trauma Rehabil., 18: 317–341. Sternini, C., Su, D., Gamp, P.D. and Bunnett, N.W. (1995) Cellular sites of expression of the neurokinin-1 receptor in the rat gastrointestinal tract. J. Comp. Neurol., 358: 531–540. Stumm, R., Culmsee, C., Schafer, M.K., Krieglstein, J. and Weihe, E. (2001) Adaptive plasticity in tachykinin and tachykinin receptor expression after focal cerebral ischemia is differentially linked to gabaergic and glutamatergic cerebrocortical circuits and cerebrovenular endothelium. J. Neurosci., 21: 798–811. Szallasi, A. and Blumberg, P.M. (1999) Vanilloid (capsaicin) receptors and mechanisms. Pharmacol. Rev., 51: 159–212. Szolcsanyi, J. and Mozsik, G. (1984) Effects of capsaicin on the development of gastric mucosal damage by different necrotizing agents and of gastric cytoprotection by PGI2 atropine and cimetidine on rats. Acta Physiol. Hung., 64: 287–291. Takahashi, T., Konishi, S., Powell, D., Leeman, S.E. and Otsuka, M. (1974) Identiﬁcation of the motoneuron-depolarizing peptide in bovine dorsal root as hypothalamic substance P. Brain Res., 73: 59–69. Takahashi, T. and Otsuka, M. (1975) Regional distribution of substance P in the spinal cord and nerve roots of the cat and the effect of dorsal root section. Brain Res., 87: 1–11. Tregear, G.W., Niall, H.D., Potts Jr, J.T., Leeman, S.E. and Chang, M.M. (1971) Synthesis of substance P. Nat. New Biol., 232: 87–89. Turner, R.J., Blumbergs, P.C., Sims, N.R., Helps, S.C., Rodgers, K.M. and Vink, R. (2006) Increased substance P immunoreactivity and edema formation following reversible ischemic stroke. Acta Neurochir. Suppl., 96: 263–266. Uddman, R., Edvinsson, L., Ekman, R., Kingman, T. and McCulloch, J. (1985) Innervation of the feline cerebral vasculature by nerve ﬁbers containing calcitonin gene-related peptide: trigeminal origin and co-existence with substance P. Neurosci. Lett., 62: 131–136. Vaz, R., Sarmento, A., Borges, N., Cruz, C. and Azevedo, I. (1997) Ultrastructural study of brain microvessels in patients

109 with traumatic cerebral contusions. Acta Neurochir. (Wien.), 139: 215–220. Vespa, P., Prins, M., Ronne-Engstrom, E., Caron, M., Shalmon, E., Hovda, D.A., Martin, N.A. and Becker, D.P. (1998) Increase in extracellular glutamate caused by reduced cerebral perfusion pressure and seizures after human traumatic brain injury: a microdialysis study. J. Neurosurg., 89: 971–982. Vink, R., McIntosh, T.K., Demediuk, P. and Faden, A.I. (1987) Decrease in total and free magnesium concentration following traumatic brain injury in rats. Biochem. Biophys. Res. Commun., 149: 594–599. Vink, R., McIntosh, T.K., Demediuk, P., Weiner, M.W. and Faden, A.I. (1988) Decline in intracellular free Mg2+ is associated with irreversible tissue injury after brain trauma. J. Biol. Chem., 263: 757–761. Von Euler, U.S. and Gaddum, J.H. (1931) An unidentiﬁed depressor substance in certain tissue extracts. J. Physiol. (Lond.), 72: 74–87. Wang, L.H., Ahmad, S., Benter, I.F., Chow, A., Mizutani, S. and Ward, P.E. (1991) Differential processing of substance P

and neurokinin A by plasma dipeptidyl(amino)peptidase IV, aminopeptidase M and angiotensin converting enzyme. Peptides, 12: 1357–1364. Watling, K.J. (1992) Nonpeptide antagonists herald new era in tachykinin research. Trends Pharmacol. Sci., 13: 266–269. Woie, K., Koller, M.E., Heyeraas, K.J. and Reed, R.K. (1993) Neurogenic inﬂammation in rat trachea is accompanied by increased negativity of interstitial ﬂuid pressure. Circ. Res., 73: 839–845. Xiong, Y., Peterson, P.L., Muizelaar, J.P. and Lee, C.P. (1997) Amelioration of mitochondrial function by a novel antioxidant U-101033E following traumatic brain injury in rats. J. Neurotrauma, 14: 907–917. Yonehara, N., Shibutani, T. and Inoki, R. (1987) Contribution of substance P to heat-induced edema in rat paw. J. Pharmacol. Exp. Ther., 242: 1071–1076. Yu, Z., Cheng, G., Huang, X., Li, K. and Cao, X. (1997) Neurokinin-1 receptor antagonist SR140333: a novel type of drug to treat cerebral ischemia. Neuroreport, 8: 2117–2119.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 8

Current concepts of cerebral oxygen transport and energy metabolism after severe traumatic brain injury B.H. Verweij1,, G.J. Amelink1 and J.P. Muizelaar2 1

Rudolf Magnus Institute of Neuroscience, Department of Neurosurgery, University Medical Center Utrecht, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands 2 Department of Neurosurgery, University of California at Davis Medical Center, 4860 Y Street, Suite 3740, Sacramento, CA 95817, USA

Abstract: Before energy metabolism can take place, brain cells must be supplied with oxygen and glucose. Only then, in combination with normal mitochondrial function, sufﬁcient energy (adenosine tri-phosphate (ATP)) can be produced. Glucose is virtually the sole fuel for the human brain. The brain lacks fuel stores and requires a continuous supply of glucose and oxygen. Therefore, continuous cerebral blood ﬂow (CBF), cerebral oxygen tension and delivery, and normal mitochondrial function are of vital importance for the maintenance of brain function and tissue viability. This review focuses on three main issues: (1) Cerebral oxygen transport (CBF, and oxygen partial pressure (PO2) and delivery to the brain); (2) Energy metabolism (glycolysis, mitochondrial function: citric acid cycle and oxidative phosphorylation); and (3) The role of the above in the pathophysiology of severe head injury. Basic understanding of these issues in the normal as well as in the traumatized brain is essential in developing new treatment strategies. These issues also play a key role in interpreting data collected from monitoring techniques such as cerebral tissue PO2, jugular bulb oxygen saturation (SjvO2), near infra red spectroscopy (NIRS), microdialysis, intracranial pressure monitoring (ICP), laser Doppler ﬂowmetry, and transcranial Doppler ﬂowmetry — both in the experimental and in the clinical setting. Keywords: traumatic brain injury; metabolism; mitochondrial function; mitochondria; lactate; cerebral blood ﬂow; ischemia

electro-encephalographic slowing and at 20 ml/ 100 g/min loss of consciousness occurs, but may be tolerated without long term functional consequences. Below a CBF of 18 ml/100 g/min, ionic homeostasis becomes jeopardized and neurons convert to anaerobic metabolism (Jones et al., 1981; Siesjo, 1992; Schroder et al., 1996). At a CBF of 10 ml/100 g/min, membrane integrity is lost and irreversible brain damage is inevitable. Tissue infarction is related to CBF but is also time dependent as shown in Fig. 1 (Jones et al., 1981). Arterial blood contains 13 vol% of O2, while jugular venous blood contains 6.7 vol%, for an

Cerebral oxygen transport Cerebral blood flow Under normal conditions, the brain has critical thresholds for cerebral blood ﬂow (CBF) as well as oxygen tension (PO2). If CBF is reduced, neuronal events will result (Astrup, 1982). Normal CBF is 50 ml/100 g brain tissue/min (Sokoloff, 1960). If CBF is reduced to 25 ml/100 g/min there is Corresponding author. Tel.: +31-30-2507059;

Fax: +31-30-2542100; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61008-X

111

112

Fig. 1. Schematic showing the relation between CBF and ischemic duration/tissue infarction. (Adapted with permission from Jones et al., 1981.)

arterio venous difference of oxygen (AVDO2) of (136.7) ¼ 6.3 vol% (ml of O2/100 ml of blood). Knowing how much blood is ﬂowing to the brain (50 ml/100 g brain tissue/min) and how much oxygen the brain extracts from this blood (AVDO2), one can calculate cerebral metabolic rate of oxygen (CMRO2): CMRO2 ¼ CBF AVDO2 which is generally 3.2 ml (of O2/100 g of brain tissue/min) (Gibbs et al., 1942).

Cerebrovascular reactivity There are two circumstances in which the brain governs its own ﬂow. In one, CBF changes proportionally with changes in CMRO2 — metabolic autoregulation. In the other, CBF remains constant despite changes in blood pressure or ICP (together CPP ¼ MABP ICP, where CPP is cerebral perfusion pressure and MABP is mean arterial blood pressure) — pressure autoregulation — and blood viscosity — viscosity autoregulation (McHenry et al., 1974; Muizelaar et al., 1986; Muizelaar, 1989). Metabolic-, pressure-, and viscosity-autoregulation have in common that AVDO2 remains essentially constant (under physiologic conditions), considering CMRO2 ¼ CBF AVDO2.

Another important factor to be considered is carbon dioxide (CO2) reactivity: with hyperventilation (resulting in low blood CO2 levels) cerebral vasoconstriction occurs with an ensuing lower CBF and higher AVDO2, whereas with high blood CO2 levels the reverse occurs (Muizelaar et al., 1991). Autoregulation is fundamentally different from CO2 reactivity: while in both metabolic and pressure autoregulation vessel diameter changes are compensatory responses to maintain a constant AVDO2, in CO2 reactivity the diameter changes are primary and CBF and AVDO2 follow passively. Thus, CO2 reactivity differs from any before-mentioned type of autoregulation in that AVDO2 changes. It appears not to be an adaptive response of the brain to changing circumstances.

Physiology of cerebral blood flow (CBF) and cerebral blood volume (CBV) The factors governing CBF are expressed in the Hagen–Poiseuille equation: CBF ¼ k

CPP d 4 8lv

where k is a constant, CPP cerebral perfusion pressure in turn deﬁned by mean arterial blood pressure (MABP) minus intracranial pressure (ICP), d diameter of the blood vessel, l the length

113

of the blood vessel (which is practically constant), and v blood viscosity. The most powerful factor in this equation is vessel diameter. For instance, the maximum constriction that can be obtained by hyperventilation is 20% from normal baseline (Kontos et al., 1977). This however, leads to a decrease in CBF of 60% from a normal value of 50 ml/100 g/min to 20 ml/100 g/min. Practically all of this diameter regulation takes place in the microcirculation, especially in the arterioles with a diameter of 300–15 m (Kontos et al., 1977, 1987). The most intense changes in diameter and therefore, cerebral blood volume occur in the microcirculation. Although it is unclear how much blood is to be found in this part of the cerebral circulation, it is estimated to be one-third of 60 ml of total blood volume in the brain, i.e., 20 ml under normal conditions (Muizelaar, 1989). With diameter ranging between 80 and 160% of baseline, this translates into volume ranging between 64 and 256% of the baseline 20 ml; 13 ml with maximum vasoconstriction, and 51 ml with maximal vasodilatation. Although many different factors are essential in maintaining adequate CBF, it is suggested that local metabolic factors are of primary importance in the local tissue regulation. Under normal circumstances, in areas of increased brain activity, vasoactive substances are released which alter vascular tone and local perfusion. Increased perfusion then creates a washout effect, which leads to reduction of perfusion. This feedback system allows for the modiﬁcation of CBF for short periods during times of increased metabolic requirements. Several metabolic factors play a role in the autoregulation of CBF under normal conditions, such as CO2/H+ (pH), K+, adenosine, prostaglandins, and nitric oxide (NO), as well as serotonin, histamine, neuropeptide Y, vasoactive intestinal peptide, calcitonin generated peptide, and others.

Brain tissue oxygen tension Under physiological conditions, a linear relationship exists between arterial PO2 and brain PO2 with arterial levels being 90 mm Hg and cerebrovenous levels 35 mm Hg. Because oxygen is consumed in

the tissue, cerebral tissue PO2 is best described as a continuum that can vary from 90 mm Hg very close to capillaries to much less than 34 mm Hg in more distal regions (Zauner et al., 2002). Reductions of partial pressure of arterial oxygen (PaO2) with normal rates of CBF also lead to functional deﬁcits. A reduction of PaO2 to 65 mm Hg induces in humans an impaired ability to perform complex tasks. Short-term memory is impaired at 55 mm Hg. A PaO2 of 30 mm Hg causes loss of consciousness (Siesjo¨, 1978). In animal models, PaO2 reduction to 36 mm Hg cause intracellular acidosis, reductions in phosphocreatine (PCr) and ATP, and increases in intracellular lactate levels (Xiong et al., 1997). Normal human brain has a critical tissue PO2 between 15 and 20 mm Hg, below which infarction (depending on the duration) of tissue may occur (Fleckenstein et al., 1990; Maas et al., 1993; Meixenberger et al., 1993; Kiening et al., 1996; van Santbrink et al., 1996; Wu and Saggau, 1997; Zauner et al., 1997; Van den Brink et al., 2000). Neuronal mitochondria require an intracellular PO2 of at least 1.5 mm Hg to maintain aerobic metabolism (Chance et al., 1973; Siesjo¨ and Siesjo¨, 1996). If cellular PO2 is low, the driving force to deliver oxygen to the mitochondria is dramatically reduced. The minimum tissue PO2 required to provide sufﬁcient intracellular oxygen is unknown. In addition, it has been proposed that the diffusion distance for oxygen from the microvasculature may increase after TBI due to astrocytic swelling, generalized tissue swelling, and tissue damage. In these situations the brain might require higher tissue oxygen tensions to maintain sufﬁcient tissue oxygenation (Zauner et al., 2002).

Oxygen and hemoglobin Flow of oxygen from the alveoli to the mitochondria in the brain is dependent on hemoglobin and PO2. Once oxygen has diffused from the alveoli into pulmonary blood, it is transported by hemoglobin to the cerebral tissue capillaries where it is released for use, by mitochondria. The presence of hemoglobin in the erythrocytes of the blood allows

114

Fig. 2. The oxygen–hemoglobin saturation curve, including the effect of hypothermia and hyperventilation. (Adapted with permission from Guyton, 1986.)

transporting 30–100 times as much oxygen as could be transported simply in the form of dissolved oxygen in blood without hemoglobin. The afﬁnity of oxygen for hemoglobin is best expressed by the oxygen–hemoglobin saturation curve as seen in Fig. 2 (Guyton, 1986). This physiological system is unique in that it favors the binding of oxygen to hemoglobin in the lungs and the release of oxygen in the periphery. In the lungs, the curve favors 100% hemoglobin saturation at or around normal alveolar oxygen tensions, whereas in the periphery the rate of oxygen delivery is proportional to the difference in oxygen partial pressure (PO2) between capillary blood and the tissue cells. Oxygen use, by the mitochondria, is responsible for creating the driving force for oxygen delivery. It is interesting to note the effect of temperature and hyperventilation on tissue oxygenation. A decrease in temperature as well as hyperventilation (alkalosis), shifts P50 (i.e., the oxygen tension at which hemoglobin is 50% saturated) to the left thus reducing tissue oxygenation due to increased Hb-O2 afﬁnity and thus, decreased O2 unloading to tissues.

Cerebral energy metabolism Under normal circumstances, the brain requires large amounts of energy. Although the brain comprises only 2–3% of whole body weight, up to 20% of energy generated in the whole body is used by the brain. Fifty percent of the energy produced by the brain is needed for synaptic activity; 25% is used for restoring ionic gradients across the cell membrane. The remaining energy is spent on biosynthesis such as maintaining membrane integrity and other processes. If the synthesis of ATP is insufﬁcient, homeostatic mechanisms deteriorate, intracellular concentration of calcium increases, and cell death is inevitable. Most of the energy is consumed by the neurons. Although glial cells account for almost half of the brain volume, they have a much lower metabolic rate and account for less than 10% of total cerebral energy consumption (Siesjo, 1984). Under normal conditions, almost all energy in our body is produced by aerobic metabolism

115

(Stryer, 1988). Krebs described three stages in the generation of energy: – Large molecules in food are broken down into smaller units. Proteins are hydrolyzed to amino acids, polysaccharides are hydrolyzed to simple sugars such as glucose, and fats are hydrolyzed to glycerol and fatty acids. No useful energy is generated in this phase. – These numerous small molecules are degraded to a few simple units that play a central role in metabolism. Most of them are converted in the acetyl unit of acetyl co-enzyme A (acetyl CoA). A small amount of ATP is generated at this stage. – Acetyl-CoA brings acetyl units into the citric acid cycle, where they are completely oxidized to CO2. Four pairs of electrons are transferred to NAD+ and FAD for each acetyl group that is oxidized. Then ATP is generated as electrons ﬂow from the reduced forms of these carriers to O2 during oxidative phosphorylation. Thus, most of the energy is generated in the third stage. The metabolic patterns of the brain are strikingly different from other organs in their use of fuel to meet their energy needs (Guyton, 1986). Glucose is virtually the sole fuel for the human brain, except during prolonged starvation. The brain lacks fuel stores and hence requires a continuous supply of glucose, which enters freely at all times. It consumes 120 g daily, which corresponds to an energy input of 420 kcal. The brain accounts for some 60% of utilization of glucose by the whole body in resting state. During starvation, ketone bodies (acetoacetate and 3-hydroxybutyrate) partly replace glucose as fuel for the brain. Acetoacetate is activated by the transfer of CoA from succinyl CoA to give acetoacetyl CoA. Cleavage by thiolase then yields two molecules of acetyl CoA, which enter the citric acid cycle. Fatty acids do not serve as fuel for the brain because they are bound to albumin in plasma and so they do not traverse the blood-brain barrier. In essence, ketone bodies are transportable equivalents of fatty acids.

Glycolysis Figure 3b illustrates the steps in the glycolysis that are carried out in the cytoplasm (Stryer, 1988; Zauner et al., 2002). GLUT-1 transports glucose across the blood-brain barrier. GLUT-1 also mediates uptake into the astrocyte while GLUT-3 does the same for neurons. The expression of these GLUT transporters is up-regulated in experimental models of hypoxia. The latter results in increased import of glucose for energy production. Glycolysis is regulated by the enzyme phosphofructokinase-1. Increased ATP demand will activate this enzyme by increasing cellular cAMP levels and thereby increasing the rate of glycolytic ATP generation. If mitochondria become dysfunctional, even after restoring blood ﬂow, a small amount of ATP can still be formed by glycolysis because this process does not require oxygen. In this process only a few percent of the total energy in the glucose molecule is released. The law of mass action states that as the end products of a chemical reaction build up in the reacting medium the rate of the reaction approaches zero, thus preventing further production of ATP. The two end products in the glycolytic reactions (pyruvate and hydrogen ions) are combined with NAD+ to form NADH and H+. The quantities of these end products increase and react with each other to form lactic acid. This lactic acid can diffuse readily into the extracellular ﬂuids and even into the intracellular ﬂuids of other less active cells. Therefore, lactic acid represents an ‘‘escape’’ into which the glycolytic end products can be directed, thus allowing glycolysis to proceed far longer than would be possible if the pyruvate and hydrogen were not removed from the reacting medium. Glycolysis could proceed for only seconds without this conversion. Instead, it can proceed for several minutes, supplying the body with ‘‘considerable’’ quantities of ATP. In the human at rest, 5–10% of the glucose consumed by the body manifests as a net output of lactate into blood (Siesjo¨, 1978; Guyton, 1986; Sokoloff, 1989; Andersen and Marmarou, 1992). Once pyruvate has been synthesized it can either reversibly be converted to lactate and accumulated,

116

Fig. 3. (a) Schematic showing glycolysis and Krebs cycle. (Adapted with permission from Magistretti et al., 1999; Zauner et al., 2002.) (b) Schematic showing mitochondrial electron transport. (Adapted with permission from Magistretti et al., 1999; Zauner et al., 2002.) (c) Schematic showing the Magistretti model of coupled metabolism. (Adapted with permission from Magistretti et al., 1999; Zauner et al., 2002.)

117

converted to amino acid alanine, or enter the citric acid cycle to produce energy via oxidative phosphorylation. Citric acid cycle The citric acid cycle that takes place in the mitochondria is shown in Fig. 3a (Stryer, 1988; Zauner et al., 2002). Under aerobic conditions, the next step in the aerobic generation of energy from glucose is the oxidative decarboxylation of pyruvate to form acetyl CoA. This activated acetyl unit is then completely oxidized to CO2 by the citric acid cycle, a series of reactions that is also known as the tricarboxylic acid cycle or the Krebs cycle. The citric acid cycle is the ﬁnal common pathway for the oxidation of fuel molecules; it also serves as a source of building blocks for biosynthesis. Oxidative phosphorylation Figure 3b shows the electron transport chain (Stryer, 1988; Zauner et al., 2002). The NADH and FADH2 formed in glycolysis, fatty acid oxidation, and the citric acid cycle are energy-rich molecules because each contains a pair of electrons with a high transfer potential. These electrons are subsequently donated to molecular oxygen, resulting in a large amount of free energy, which can be used to generate ATP. Oxidative phosphorylation is the process in which ATP is formed as electrons are transferred from NADH or FADH2 to O2 by a series of electron carriers. This process acts as a major source of ATP in aerobic organisms. Some salient features of this process are (Stryer, 1988): – Respiratory assemblies that are located in the inner membrane of mitochondria carry out oxidative phosphorylation. The citric acid cycle and the pathway of the fatty acid oxidation, which supply most of the NADH and FADH2, are in the adjacent mitochondrial matrix. – The oxidation of NADH yields 3 ATP, whereas the oxidation of FADH2 yields 2 ATP. Oxidation and phosphorylation are coupled processes.

– Respiratory assemblies contain numerous electron carriers, such as the cytochromes. The step-by-step transfer of electrons from NADH or FADH2 to O2 through these carriers leads to pumping of protons out of the mitochondrial matrix. A proton-motive force is generated consisting of a pH gradient and a transmembrane electric potential. ATP is synthesized when protons ﬂow back to the mitochondrial matrix through an F0F1 ATP synthase complex. Thus oxidation and phosphorylation are coupled by a proton gradient across the inner mitochondrial membrane. Traditionally cerebral energy production has been considered to consist mainly of aerobic metabolism of glucose. Although it has long been assumed that glia and neurons use glucose as their sole energy source, recent information has suggested otherwise; astrocytes may have the ability to transport glucose across the blood-brain barrier via GLUT-1 and anaerobically metabolize it to lactate. Lactate is then released into the extracellular space, where it is taken up by neurons and consumed aerobically to generate energy as seen in Fig. 3c (Vibulsreth et al., 1987; Walz and Mukerji, 1988; Magistretti et al., 1999). With increasing neuronal activity, potassium and glutamate are released into the extracellular space and are taken up by the astrocytes in an energydependent fashion causing increased astrocytic glycolysis. In traumatic brain injury conditions, aerobic metabolism is diminished due to reductions in cellular oxygen, or due to mitochondrial dysfunction, causing increased lactate accumulation. Mitochondria Mitochondria are oval-shaped organelles, typically 2 mm in length and 0.5 mm in diameter. Techniques for isolating mitochondria were devised in the late 1940s. Eugene Kennedy and Albert Lehninger subsequently discovered that mitochondria contain the respiratory assembly, the enzymes of both the citric acid cycle and fatty acid oxidation. Electron microscopic studies by George Palade and Fritjof Sjo¨strand revealed that mitochondria have two membrane systems: an outer membrane and a

118

Fig. 4. Schematic of mitochondrion.

highly folded inner membrane. The inner membrane is folded into a series of internal ridges called cristae. Hence, there are two compartments in mitochondria: the intermembrane space between the outer and inner membranes, and the matrix, which is bounded by the inner membrane (Fig. 4). Oxidative phosphorylation takes place in the inner mitochondrial membrane, in contrast with most of the reactions of the citric acid cycle and fatty acid oxidation, which occur in the matrix. The outer membrane is quite permeable to most small molecules and ions because it contains many copies of porin, a transmembrane protein with a large pore. In contrast, the inner membrane is intrinsically impermeable to nearly all ions and polar molecules. Speciﬁc protein carriers transport molecules such as adenosine di-phosphate (ADP) and long chain fatty acids across the inner mitochondrial membrane.

Pathophysiology Traumatic brain damage injury may be divided into primary and secondary types of injury (Verweij and Muizelaar, 1996). Mechanical forces acting at the moment of injury damage the blood vessels, axons, neurons, and glia of the brain initiating an evolving sequence of secondary changes that result in complex cellular, inﬂammatory, neurochemical, and metabolic alterations. It is not within the scope of this review to describe all the changes occurring in the brain after severe head injury; only those directly related to

cerebral oxygen transport and energy metabolism will be mentioned: Traumatic intracranial hematomas (intraparenchymal, subdural, and epidural) are also common after head injury. Epidural hematomas (in up to 5% of all patients admitted to hospitals for head injury, and 9% of those with severe head injury) are often the result of skull fractures causing rupture of underlying arteries or veins. If arterial in origin they can enlarge very quickly and cause rapid neurological deterioration. If surgical intervention is prompt, and no other brain injury is present, outcome can be favorable. Acute subdural hematoma (ASDH) occurs in up to 25% of all patients with severe head injury (Richards and Hoff, 1974; Hubschmann and Nathanson, 1985). In comatose patients with TBI, ASDH carries the highest mortality rate: 60–90 (Jamieson and Yelland, 1972; Gennarelli et al., 1989; Wilberger et al., 1991). In this group outcome is strikingly unfavorable due to decreased energy metabolism. On one hand, decreased energy metabolism is due to ischemia caused by increased intracranial pressure and therefore decreased perfusion pressure; on the other hand due to the underlying damaged brain being unable to use oxygen because of damaged mitochondria (Verweij et al., 2000b, 2001).

Hypotension and hypoxia The majority of the potential clinical events after neurotrauma have been investigated with respect to their frequency of occurrence and impact on outcome, both for the prehospital and intensive care unit (ICU) periods (Chesnut et al., 1993; Jones et al., 1994). These studies have uniformly identiﬁed hypotension (SABPo90 mm Hg) and hypoxia (PaO2o60 torr) as the most inﬂuential. These parameters, amenable to therapeutic manipulation, seem to be the most signiﬁcant predictors of poor outcome, independent of their etiologies and pre-resuscitation of secondary insults. Admission to the ICU does not eliminate secondary brain injury. Jones et al. (1994) have reported the results of computerized online evaluation of 14 variables in 124 head-injured patients of a variety of grades admitted to the

119

neurosurgical ICU. More than one episode of hypotension occurred in 73% of all patients, with median durations of 29 min (SABPo90 mm Hg), 22 min (SABPo80 mm Hg), and 32 min (SABPo70 mm Hg). In 40% of all cases, more than one episode of hypoxia occurred with an average duration of 12 min (PaO2o60 torr), 19 min (PaO2o52 torr), and 20 min (PaO2o45 torr). It has also been demonstrated that these secondary insults do not only occur in the ICU, but also during patient transport in X-ray and in OR suites (Andrews et al., 1990).

ICP and cerebral circulation According to the Monro–Kellie doctrine, ICP is governed by the interplay between the volumes of brain (including cytotoxic edema), cerebrospinal ﬂuid (including vasogenic or extracellular edema) and the blood within the cerebral blood vessels, all of which is within the conﬁnes of the rigid skull (Monro, 1783; Kellie, 1824). An increase in volume in one of these compartments (or epidural, subdural, or intracerebral hematoma) leads to a rise in ICP, unless this increase is compensated by an equal decrease in one of the two remaining compartments. The natural defense against rising ICP with brain swelling is the displacement of CSF from the skull: hence small compressed ventricles and absence of basal cisterns on computer tomography (CT) scans after severe trauma. Sometimes this process can be palliated by drainage through a ventricular catheter. When one considers the Hagen–Poiseuille equation, there are two practical methods of maintaining CBF during vasoconstriction. The ﬁrst is to increase CPP, which can be done by raising the blood pressure. When pressure autoregulation is intact, this maneuver in and of itself will cause vasoconstriction, and this has occasionally been used to decrease ICP (Muizelaar, 1989). More important, however, is the need to avoid low blood pressure, and hence, the effect of ‘‘perfusion pressure therapy’’ may be due in part to the simple avoidance of arterial hypotension (Rosner and Daughton, 1990; Bouma et al., 1992a; Rosner et al., 1995). The second method to maintain CBF in the face of decreasing

vessel diameter is to decrease blood viscosity. Again, this decrease itself leads to vasoconstriction if viscosity autoregulation is intact, and it has been argued that the viscosity lowering effect mediates a good deal of mannitol’s effect on ICP (besides the osmotic effect) (Muizelaar et al., 1983, 1984, 1986). When viscosity autoregulation is not intact, lowering viscosity with mannitol can maintain CBF despite cerebral vasoconstriction associated with hyperventilation (Cruz et al., 1990). As stated previously, CBV can vary between 13 and 51 ml in the microcirculation. To comprehend what these differences in volume mean to ICP, one must consider the pressure volume index (PVI) (Marmarou et al., 1978). Pressure volume index is deﬁned as the amount of ﬂuid (in ml) needed to add to the intracranial space to make ICP rise tenfold (or withdraw to decrease ICP tenfold): PVI ¼

DV Log½ICPbefore =ICPafter

Normal PVI is 20–25 ml (Shapiro et al., 1980). However PVI has been observed as low as 6 ml after severe head injury, which indicates that in going from normoventilation (PaCO2 ¼ 30 mm Hg) to strong hyperventilation (PaCO2 ¼ 18 mm Hg) resulting in vasoconstriction, ICP could theoretically be decreased tenfold (Bouma et al., 1992a). (That this is not always desirable may be clear from the following example: PaCO2 36 mm Hg, ICP 40 mm Hg, MABP 100 mm Hg, CBF 30 ml/100 g/min; AVDO2 6 vol% - CMRO2 1.8 ml/100 g/min; now with hyperventilation to get ICP below the desired 20 mm Hg: PaCO2 ¼ 22, ICP 20 mm Hg, MABP 100 mg Hg - CPP from 60 to 80 - CBF to 40 ml/100 g/min.) However, because of 20% vasoconstriction and diameter being to the fourth power in the Hagen–Poiseuille equation, CBF drops to 16 ml — well below the threshold for infarction (Jones et al., 1981), and especially so after severe head injury. Moreover, as AVDO2 cannot rise above 10, CMRO2 will drop to 1.6 ml/ 100 g/min. If ICP is uncontrollable barbiturates are sometimes administered, lowering the cerebral metabolic demands. If metabolic autoregulation is still intact this will result in decreased blood volume and therefore reduced ICP.

120

All of these examples are common in clinical practice, and thus it may be obvious that a good monitor is required to guide the management of ICP, blood pressure, ventilatory parameters, and blood viscosity. Although monitoring of CBF and/ or AVDO2 is ideal, there are no practical ways to do this continuously; therefore the authors revert to a derivate of AVDO2, i.e., SjvO2 (Cruz, 1988; Robertson et al., 1989; Gopinath et al., 1994).

Ischemia and CMRO2 Secondary cerebral ischemia is very common after severe head injury and is associated with an unfavorable outcome (Graham and Adams, 1971; Bouma et al., 1991; Siesjo¨ and Siesjo¨, 1996). As discussed earlier, CBF is closely regulated by a number of mechanisms. However, after trauma these mechanisms can fail resulting in ischemia. This is especially important as the brain seems more vulnerable to ischemia after trauma and this vulnerability persists for at least 24 h (Jenkins, 1989). It has also been demonstrated that CBF is low in the hyperacute post-traumatic period (Bouma et al., 1991, 1992b). Increased intracranial pressure (as in cerebral edema and subdural hematoma) and therefore decreased cerebral perfusion pressure appears also to be an important cause of ischemia as well as too vigorous hyperventilation (Muizelaar et al., 1991; Verweij et al., 2001). Salvant and Muizelaar suggested that a parallel reduction in CBF and CMRO2 without an increase in AVDO2 is consistent with diminished metabolism and may be due to mitochondrial dysfunction. Verweij et al. (2000b) demonstrated mitochondrial dysfunction in isolated mitochondria from human tissue. In a normal coupled relationship between AVDO2 and CBF (AVDO2 ¼ CMRO2/CBF), AVDO2 remains unchanged when the CMRO2 changes. If however CMRO2 remains constant, changes in AVDO2 reﬂect uncoupled changes in CBF (Robertson et al., 1989). If CBF decreases following head injury, AVDO2 will increase as the brain compensates by extracting a greater amount of oxygen. A further uncompensated decline in CBF leads to ischemia and a fall in CMRO2.

Lactate/hyperglycolysis In animals and humans it has been repeatedly shown that TBI induces increased brain lactate production, which normalizes gradually after the ﬁrst few days in those who survive, but remains at levels ﬁve to ten times normal in those who succumb. Microdialysis studies have demonstrated that extracellular ﬂuid glucose declines to extremely low levels when lactate is increasing (Goodman et al., 1999; Zauner et al., 2002). If aerobic metabolism fails, anaerobic metabolism remains, resulting in hyperglycolysis and lactate production. A shift to anaerobic metabolism will occur if ischemia is present. If brain oxygen tension levels fall below 20 mm Hg aerobic metabolism ceases to occur (Zauner et al., 1997). However, if mitochondrial dysfunction occurs, aerobic metabolism will also shift toward anaerobic metabolism inducing the same phenomena (Verweij et al., 2000b). In human positron emission tomography (PET) studies, increases in glucose metabolism are demonstrated especially in the zone around contusions and in the hemisphere underlying hematomas (Bergsneider et al., 1997). Such tissues are often adjacent to ‘‘low-density’’ cytotoxic edema areas on CT scan. Human PET studies at later time points (1–4 weeks post-ictus) and mitochondrial analyses in the acute stage have shown uniformly decreased metabolism, both for glucose and oxygen in humans (Verweij et al., 2000b, 2001). Marmarou as well as others have shown through animal experiments that anaerobic cerebral metabolism with generation of lactate appears to occur even in the absence of blood ﬂow limitation (De Salles et al., 1987). It has also been found by measuring brain oxygenation, CO2 generation, pH and temperature in parallel with extracellular ﬂuid lactate, and glucose levels measured by microdialysis that lactate generation is increased in 65% of measurements, even in the presence of adequate CBF and brain oxygen levels (Zauner et al., 1997). Mitochondrial dysfunction will explain lactate production in these circumstances (Verweij et al., 2000b). High levels of ECF lactate might be harmful to the injured brain. Marked cerebral acidosis may exacerbate calcium-mediated damage to

121

intracellular enzyme systems and may also interfere with ion-channel function (Siesjo, 1992). High tissue lactate levels could foster a decline in brain pH, as has been shown in numerous posttraumatic animal and human studies. The restoration of circulation after ischemia is accompanied by normalization of the tissue lactate concentration and the lactate-to-pyruvate ratio. Initially there is an increase in pyruvate concentration as lactate is converted back to pyruvate. Animal studies have demonstrated that if the resuscitation interval is prolonged, tissue lactate remains elevated during reperfusion, suggesting that residual tissue lactic acidosis is a sign of mitochondrial dysfunction. Since mitochondrial dysfunction is reversible, therapeutic intervention might be possible (Verweij et al., 1997; Xiong et al., 1998; Berman et al., 2000). Rapid reversal of acidosis may be unfavorable, as mild acidosis might be beneﬁcial (pH paradox) during recovery from hypoxia. There is little direct evidence to demonstrate that lactate alone, or substantial intracellular acidosis alone, is toxic to normal cerebral tissues. This may be attributable to preserved high-energy phosphate concentrations, which allow potential intracellular buffering and transport of hydrogen ions from the cell. During ischemia, however, acidosis may injure neurons by denaturation of proteins, lead to damage of astrocytes owing to failure of membrane transport systems, and cause promotion of irondependent free radical formation. These events can cause the inhibition of glycolysis by the complete inhibition of the glycolytic phosphofructokinase at a pH of 6.5 or below. Mild acidosis might be protective in vitro and in vivo in models of ischemia/hypoxia, by slowing enzymatic processes and reducing energy consumption and free radical production. Nonetheless, tissue acidosis is consistently associated with worsened ischemic outcome in vivo, which may be augmented by hyperglycemia. Cerebral tissue PO2/SjvO2 Cerebral oxygenation is currently monitored in two ways: (1) measurement of jugular bulb oxygen saturation, which reﬂects the relationship between

oxygen delivery to the brain and the extraction of oxygen by the brain (Cruz, 1988; Cruz et al., 1990; Gopinath et al., 1994); (2) local brain tissue oximetry by a single Clark-type electrode (or multielectrode also measuring PCO2, pH, and temperature), which gives information of local cerebral tissue PO2. The relationship between SjvO2, cerebral tissue PO2, and CBF has been well documented. An increase in the concentration of tissue lactate in the brain indicates a shift from aerobic to anaerobic metabolism in an attempt to maintain ATP production. This shift occurs in case of low cerebral tissue PO2 that has been shown to occur in 25–39% of patients with severe TBI in the ﬁrst 12 h post- injury. Low brain tissue PO2 also closely correlates with low regional CBF. It has also been shown that low brain tissue PO2 is strongly correlated with high levels of dialysate lactate in the brain (Zauner et al., 2002). Increasing the concentration of inspired oxygen to 100% has shown to increase brain tissue PO2. In a study by Menzel et al. (1999), statistically signiﬁcant correlation existed between brain tissue PO2 and CPP. Brain lactate measured by microdialysis remained high during the entire period rising later on. No correlations were found between ICP, CPP, brain tissue PO2, or lactate (brain tissue PO2 increases over the ﬁrst 30 h with an overshoot at 36–48 h). The latter could be explained by mitochondrial dysfunction (Verweij et al., 2000b). Mitochondrial function after severe head injury Mitochondria play a vital role in cell survival and tissue development by virtue of their role in energy metabolism and apoptosis (Nicholls and Budd, 2000; Friberg and Wieloch, 2002). Since neuronal tissue stores and anaerobic glycolysis provides ATP sufﬁcient to maintain cellular function for only 1–2 min, mitochondrial generation of ATP is of vital importance (Siesjo, 1992). Neuronal mitochondria have a high capacity to store calcium ions, thereby protecting neurons against transient elevations in intracellular calcium concentrations during neuronal hyperactivity. This calcium sequestration requires negative mitochondrial matrix potential and therefore full functional mitochondria with access to oxygen and pyruvate.

122

In the past, research efforts in patients with severe head injury have focused on optimizing the delivery of oxygen and glucose to the injured brain in an attempt to maintain the ATP supply and to avoid neuronal damage. However, limiting factors in synthesizing ATP are not only inadequate delivery of oxygen and glucose but also impairment of mitochondrial function (Verweij et al., 1997, 2000b). Severe TBI with or without hypoxia and or ischemia results in a number of biochemical processes such as amino acid efﬂux and oxygen free radical production. This ultimately leads to massive ion shifts with increased calcium in the intracellular compartment (Fineman et al., 1993). It has previously been demonstrated that experimental TBI perturbs calcium homeostasis with an overload of cytosolic calcium and excessive calcium adsorption by the mitochondrial membranes (Sciamanna et al., 1992; Verweij et al., 1997; Xiong et al., 1997). This inhibits mitochondrial function, even when there are sufﬁcient oxygen and substrate present. In rats, mitochondrial dysfunction begins 1 h after TBI and persists for at least 14 days, with the maximum level of dysfunction occurring at 12–72 h (Verweij et al., 1997). The same phenomena occur in patients (Verweij et al., 2000b). A number of critical mechanisms have been elucidated by which mitochondria are involved in cell death. Elevated cytosolic Ca2+ and oxidative stress both contribute to the opening of the mitochondrial permeability transition pore (PTP), which depolarizes the mitochondrion and leads to mitochondrial swelling and subsequent release of cytochrome ‘‘c’’ from the intermembrane space. Cytochrome ‘‘c’’ normally functions as part of the respiratory chain, but when released into the cytosol (as a result of PTP opening) it becomes a critical component of the apoptosis execution machinery, where it activates caspases (cysteine aspartate proteases) and (if ATP is available) causes apoptotic cell death (Young, 1992). Thus, a number of new targets for therapeutic intervention have emerged (Verweij et al., 1997, 2000a; Xiong et al., 1998; Berman et al., 2000). These possible interventions have recently been described in a well-written review by Merenda and Bullock (2006).

Abbreviations ADP AMP ASDH ATP AVDO2 CBF CBV CMRO2 CoA CPP CSF CT EEG FAD GCS GLUT ICP MABP NAD NIRS NO PaCO2 PaO2 PET PO2 PVI SABP SjvO2 TBI

adenosine diphosphate adenosine monophosphate traumatic acute subdural hematoma adenosine triphosphate arteriovenous oxygen content difference cerebral blood ﬂow cerebral blood volume cerebral metabolic rate of oxygen co-enzyme A cerebral perfusion pressure cerebrospinal ﬂuid computerized tomography electro-encephalogram ﬂavin adenine dinucleotide Glasgow coma scale glucose transporter intracranial pressure mean arterial blood pressure nicotinamide adenine dinucleotide near infra red spectroscopy nitric oxide arterial PCO2 tension arterial oxygen partial pressure positron emission tomography oxygen partial pressure pressure volume index systolic arterial blood pressure jugular venous oxygen saturation traumatic brain injury

References Andersen, B.J. and Marmarou, A. (1992) Functional compartmentalization of energy production in neural tissue. Brain Res., 585: 190–195. Andrews, P.J., Piper, I.R., Dearden, N.M., et al. (1990) Secondary insults during intrahospital transport of head injured patients. Lancet, 335(8685): 327–330. Astrup, J. (1982) Energy requiring cell functions in the ischemic brain: their critical supply and possible inhibition in protective therapy. J. Neurosurg., 56: 482–497. Bergsneider, M., Hovda, D.A., Shalmon, E., et al. (1997) Cerebral hyperglycolysis following severe traumatic brain injury in humans: a positron emission tomographic study. J. Neurosurg., 86: 241–251.

123 Berman, R.F., Verweij, B.H. and Muizelaar, J.P. (2000) Neurobehavioral protection by the neuronal calcium channel blocker ziconotide in a model of traumatic diffuse brain injury in rats. J. Neurosurg., 93(5): 821–828. Bouma, G.J., Muizelaar, J.P., Bandoh, K., et al. (1992a) Blood pressure and intracranial pressure-volume dynamics in severe head injury: relationship with cerebral blood ﬂow. J. Neurosurg., 77(1): 15–19. Bouma, G.J., Muizelaar, J.P., Choi, S.C., et al. (1991) Cerebral bloodﬂow and metabolism after severe traumatic brain injury: the elusive role of ischemia. J. Neurosurg., 75(5): 685–693. Bouma, G.J., Muizelaar, J.P., Stringer, W.A., et al. (1992b) Ultra-early evaluation of regional cerebral blood ﬂow in severely head-injured patients using xenon-enhanced computirized tomography. J. Neurosurg., 77: 360–368. van den Brink, W.A., van Santbrink, H., Steyerberg, E.W., et al. (2000) Brain oxygen tension in severe head injury. Neurosurgery, 46(4): 868–876. Chance, B., Oshimo, N., Sugamo, T. and Mayersky, A. (1973) Basic principles of tissue oxygen determination from mitochondrial signals. In: Bruley D.F. (Ed.), Oxygen Transport to Tissue. Plenum Publishing, New York, pp. 277–292. Chesnut, R.M., Marshall, S.B., Piek, J., et al. (1993) Early and late systemic hypotension as a frequent source of cerebral ischemia following severe brain injury in the traumatic coma data bank. Acta Neurochir. Suppl. (Wein.), 59: 121–125. Cruz, J. (1988) Continuous versus serial global hemometabolic monitoring: application in acute brain trauma. Acta Neurochir. (Wien.), 42: 35–39. Cruz, J., Miner, M.E., Allen, S.J., et al. (1990) Continuous monitoring of cerebral oxygenation in acute brain injury: injection of mannitol during hyperventilation. J. Neurosurg., 73(5): 725–730. De Salles, A.A.F., Muizelaar, J.P. and Young, H.F. (1987) Hyperglycemia, cerebrospinal ﬂuid lactic acidosis, and cerebral blood ﬂow in severely head injured patients. Neurosurgery, 21: 45–50. Fineman, I., Hovda, D.A. and Smith, M. (1993) Concussive brain injury is associated with a prolonged accumulation of calcium: a 45Ca autoradiographic study. Brain Res., 624: 94–102. Fleckenstein, W., Maas, A.I., Nollert, G. and de Long, D.A. (1990) Oxygen pressure in cerebrospinal ﬂuid. In: Fleckenstein W. and Maas A.I. (Eds.), Oxygen Transport. Blackwell, Berlin, pp. 368–393. Friberg, H. and Wieloch, T. (2002) Mitochondrial permeability transition in neurodegeneration. Biochimie, 84: 241–250. Gennarelli, T.A., Champion, H.R. and Sacco, W.J. (1989) Mortality of patients with head injury and extracranial injury treated in trauma centers. J. Trauma, 29: 1193–1202. Gibbs, E.L., Lennox, W.G., Nims, L.F., et al. (1942) Arterial and cerebral venous blood: arterial venous differences in man. J. Biol. Chem., 144: 325–332. Goodman, J.C., Valadka, A.B., Gopinath, S.P., Uzura, M. and Robertson, C.S. (1999) Extracellular lactate and glucose alterations in the brain after head injury measured by microdialysis. Crit. Care Med., 27: 1965–1973.

Gopinath, S.P., Robertson, C.S., Contant, C.F., et al. (1994) Jugular venous desaturation and outcome after head injury. J. Neurol. Neurosurg. Psychiatry, 57(6): 717–723. Graham, D.I. and Adams, J.H. (1971) Ischemic brain damage in fatal head injuries. Lancet, 1: 265–266. Guyton, A.C. (1986) Textbook of Medical Physiology (Seventh edition). W.B. Saunders Company, Philadelphia, pp. 481–525. Hubschmann, O.R. and Nathanson, D.C. (1985) The role of calcium and cellular membrane dysfunction in experimental trauma and subarachnoid hemorrhage. J. Neurosurg., 6: 698–703. Jamieson, K.G. and Yelland, J.D.N. (1972) Surgically treated traumatic subdural hematomas. J. Neurosurg., 37: 137–149. Jones, P.A., Andrews, P.J., Midgley, S., et al. (1994) Measuring the burden of secondary insults in head-injured patients during intensive care. J. Neurosurg. Anesthesiol., 6(1): 4–14. Jones, T., Morawetz, R., Crowell, R.M., et al. (1981) Threshold of focal cerebral ischemia in awake monkeys. J. Neurosurg., 54: 773–782. Kellie, G. (1824) On death from cold, and on congestions of the brain: an account of the appearances observed in the dissection of two of three individuals presumed to have perished in the storm of 3rd November, 1821; with some reﬂections on the pathology of the brain. TRANS Med. Chir. Soc., Edinburgh, 1: 84–169. Kiening, K.L., Unterberg, A.W., Bardt, T.F., Schneider, G.H. and Lanksch, W.R. (1996) Monitoring of cerebral oxygenation in patients with severe head injuries. J. Neurosurg., 85: 751–757. Kontos, H.A., Raper, A.J. and Patterson, J.L. (1977) Analysis of vasoactivity of local pH, PCO2, and bicarbonate on pial vessels. Stroke, 8(3): 358–360. Kontos, H.A., Wei, E.P., Navari, R.M., et al. (1987) Responses of cerebral arteries and arterioles to acute hypotension and hypertension. Am. J. Physiol., 234(4): H371–H383. Maas, A.I., Fleckenstein, W., de Jong, D.A. and van Santbrink, H. (1993) Monitoring cerebral oxygenation. Acta Neurochir. Suppl. (Wien.), 59: 50–57. Magistretti, P.J., Pellerin, L., Rothman, D.L. and Shulman, R.G. (1999) Energy on demand. Science, 283: 496–497. Marmarou, A., Shulman, K. and Rosende, R. (1978) A nonlinear analysis of the cerebrospinal ﬂuid system and intracranial pressure dynamics. J. Neurosurg., 48: 332–344. McHenry Jr., L.C., West, J.W., Coopers, E.S., et al. (1974) Cerebral autoregulation in man. Stroke, 5(6): 695–706. Meixenberger, J., Dings, J., Kuhnigk, H. and Roosen, K. (1993) Studies of tissue PO2 in normal and pathophysiological human brain cortex. Acta Neurochir. Suppl. (Wien.), 59: 58–63. Menzel, M., Doppenberg, E.M.R., Zauner, A., et al. (1999) Increased inspired oxygen concentration as a factor in improved brain tissue oxygenation and tissue lactate levels after severe human head injury. J. Neurosurg., 91: 1–10. Merenda, A. and Bullock, R. (2006) Clinical treatments for mitochondrial dysfunction after brain injury. Curr. Opin. Crit. Care, 12: 90–96. Monro, A. (1783) Observations on the structure and functions of the nervous system. Creech and Johnson, Edinburgh, p. 5.

124 Muizelaar, J.P. (1989) Cerebral blood ﬂow, cerebral blood volume, and cerebral metabolism after severe head injury. In: Becker D.P. and Gudeman S.K. (Eds.), Textbook of Head Injury. W.B. Saunders Company, Philadelphia, PA, pp. 221–240. Muizelaar, J.P., Lutz, H.A. and Becker, D.P. (1984) Effect of mannitol on ICP and CBF and correlation with pressure autoregulation in severely head-injured patients. J. Neurosurg., 61(4): 700–706. Muizelaar, J.P., Marmarou, A., Ward, J.D., et al. (1991) Adverse effects of prolonged hyperventilation in patients with severe head injury: a randomized clinical trial. J. Neurosurg., 75: 731–739. Muizelaar, J.P., Wei, E.P., Kontos, H.A., et al. (1983) Mannitol causes compensatory cerebral vasoconstriction and vasodilatation in response to blood viscosity changes. J. Neurosurg., 59(5): 822–828. Muizelaar, J.P., Wei, E.P., Kontos, H.A., et al. (1986) Cerebral blood ﬂow is regulated by changes in blood pressure and in blood viscosity alike. Stroke, 17(1): 44–48. Nicholls, D.G. and Budd, S.L. (2000) Mitochondrial and neuronal survival. Physiol. Rev., 80: 315–360. Richards, T. and Hoff, J. (1974) Factors affecting survival from acute subdural hematoma. Surgery, 75: 253–258. Robertson, C.S., Narayan, R.K., Gokaslan, Z.L., et al. (1989) Cerebral arteriovenous oxygen difference as an estimate of cerebral blood ﬂow in comatose patients. J. Neurosurg., 70(2): 222–230. Rosner, M.J. and Daughton, S. (1990) Cerebral perfusion pressure management in head injury. J. Trauma, 30(8): 933–940. Rosner, M.J., Rosner, S.D. and Johnson, A.H. (1995) Cerebral perfusion pressure: management protocol and clinical results. J. Neurosurg., 83(6): 949–962. Schroder, M., Muizelaar, J.P., Kuta, A., et al. (1996) Thresholds for cerebral ischemia after severe head injury: relationship with late CT ﬁndings and outcome. J. Neurotrauma, 13: 17–23. van Santbrink, H., Maas, A.I. and Avezaat, C.J. (1996) Continuous monitoring of partial pressure of brain tissue oxygen in patients with severe head injury. Neurosurgery, 38(1): 21–31. Sciamanna, M.A., Zinkel, J., Fabi, A.Y., et al. (1992) Ischemic injury to rat forebrain mitochondria and cellular calcium homeostasis. Biochim. Biophys. Acta, 1134: 223–232. Shapiro, K., Marmarou, A. and Shulman, K. (1980) Characterization of clinical CSF dynamics and neural axis compliance using the pressure-volume index: I. The normal pressure volume index. Ann. Neurol., 7(6): 508–514. Siesjo¨, B.K. (1978) Brain Energy Metabolism. Wiley, Chichester. Siesjo¨, B.K. (1984) Cerebral circulation and metabolism. J. Neurosurg., 60(5): 883–908. Siesjo¨, B.K. (1992) Pathofysiology and treatment of focal cerebral ischemia. Part I: pathofysiology. J. Neurosurg., 77: 169–184. Siesjo¨, B.K. and Siesjo¨, P. (1996) Mechanisms of secondary brain injury. Eur. J. Anaesthesiol., 13: 247–268. Sokoloff, L. (1960) Metabolism of the central nervous system in vivo. In: Field J., Magoun H.W. and Hall V.E. (Eds.),

Handbook of Physiology. Section 11: Neurophysiology, Vol. 3. American Physiological Society, Washington DC, pp. 1843–1864. Sokoloff, L. (1989) Circulation and Energy Metabolism of the Brain. Raven Press, New York, pp. 541–564. Stryer, L. (1988) Biochemistry (Third edition). W.H. Freeman and Company, New York, pp. 313–544. Verweij, B.H. and Muizelaar, J.P. (1996) Avoiding secondary brain injury after severe head trauma: monitoring and management. J. Craniomaxillofac. Trauma, 2(3): 8–17. Verweij, B.H., Muizelaar, J.P. and Vinas, F.C. (2001) Hyperacute measurement of intracranial pressure, cerebral perfusion pressure, jugular venous oxygen saturation, and laser Doppler ﬂowmetry, before and during removal of traumatic acute subdural hematoma. J. Neurosurg., 95(4): 569–572. Verweij, B.H., Muizelaar, J.P., Vinas, F.C., Peterson, P.L., Xiong, Y. and Lee, C.P. (1997) Mitochondrial dysfunction after experimental and human brain injury and its possible reversal with a selective N-type calcium channel antagonist (SNX-111). Neurol. Res., 19(3): 334–339. Verweij, B.H., Muizelaar, J.P., Vinas, F.C., Peterson, P.L., Xiong, Y. and Lee, C.P. (2000a) Improvement in mitochondrial dysfunction as a new surrogate efﬁciency measure for preclinical trials: dose-response and time-window proﬁles for administration of the calcium channel blocker Ziconotide in experimental brain injury. J. Neurosurg., 93(5): 829–834. Verweij, B.H., Muizelaar, J.P., Vinas, F.C., Peterson, P.L., Xiong, Y. and Lee, C.P. (2000b) Impaired cerebral mitochondrial function after traumatic brain injury in humans. J. Neurosurg., 93(5): 815–820. Vibulsreth, S., Hefti, F., Ginsberg, M.D., Dietrich, W.D. and Busto, R. (1987) Astrocytes protect cultured neurons from degeneration induced by anoxia. Brain Res., 422: 303–311. Walz, W. and Mukerji, S. (1988) Lactate production and release in cultured astrocytes. Neurosci. Lett., 86: 296–300. Wilberger Jr., J.E., Harris, M. and Diamond, D.L. (1991) Acute subdural hematoma: morbidity, mortality, and operative timing. J. Neurosurg., 74: 212–218. Wu, L.G. and Saggau, P. (1997) Presynaptic inhibition of elicited neurotransmitter release. Trends Neurosci., 20: 204–212. Xiong, Y., Gu, Q. and Peterson, P.L. (1997) Mitochondrial dysfunction and calcium perturbation induced by traumatic brain injury. J. Neurotrauma, 14: 23–34. Xiong, Y., Peterson, P.L., Verweij, B.H., et al. (1998) Mitochondrial dysfunction after experimental traumatic brain injury: combined efﬁcacy of SNX-111 and U-101033E. J. Neurotrauma, 15(7): 531–544. Young, W. (1992) Role of calcium in central nervous systems injuries. J. Neurotrauma, 9(Suppl 1): S9–S25. Zauner, A., Daugherty, W.P., Bullock, M.R., et al. (2002) Brain oxygenation and energy metabolism: Part 1-Biological function and pathophysiology. Neurosurgery, 51: 289–302. Zauner, A., Doppenberg, E.M., Woodward, J.J., Choi, S.C., Young, H.F. and Bullock, R. (1997) Continuous monitoring of cerebral substrate delivery and clearance. Neurosurgery, 41: 1082–1091.

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 9

Progressive damage after brain and spinal cord injury: pathomechanisms and treatment strategies Helen M. Bramlett and W. Dalton Dietrich Department of Neurological Surgery, Neurotrauma Research Center, The Miami Project to Cure Paralysis, University of Miami Miller School of Medicine, 1095 NW 14th Terrace, Miami, FL 33136, USA

Abstract: The pathophysiology of brain and spinal cord injury (SCI) is complex and involves multiple injury mechanisms that are spatially and temporally speciﬁc. It is now appreciated that many of these injury mechanisms remain active days to weeks after a primary insult. Long-term survival studies in clinically relevant experimental studies have documented the structural changes that continue at the level of the insult as well as in remote brain structures. After traumatic brain injury (TBI), progressive atrophy of both gray and white matter structures continues up to 1 year post-trauma. Progressive changes may therefore underlie some of the long-term functional deficits observed in this patient population. After SCI, similar features of progressive injury are observed including delayed cell death of neurons and oligodendrocytes, axonal demyelination of intact ﬁber tracts and retrograde tract degeneration. SCI also leads to supraspinal changes in cell survival and remote brain circuitry. The progressive changes in multiple structures after brain and SCI are important because of their potential consequences on chronic or developing neurological deficits associated with these insults. In addition, the better understanding of these injury cascades may one day allow new treatments to be developed that can inhibit these responses to injury and hopefully promote recovery. This chapter summarizes some of the recent data regarding progressive damage after CNS trauma and mechanisms underlying these changes. Keywords: traumatic brain injury; spinal cord injury; progressive damage; pathophysiology; treatment important from the prospective of clarifying mechanisms underlying cell death, but more importantly provide new targets for therapeutic intervention. Indeed, the observation that processes which potentially affect long-term outcome may be active days or even months after injury provides new targets to improve outcome after CNS injury. This new way of assessing and treating acute injury is also important as we think about how acute insults such as ischemic or traumatic injuries may enhance the vulnerability of the aging brain to later occurring neurodegenerative diseases. The main objective of this chapter is to summarize recent date emphasizing the progressive nature of lesion pathology after brain and

Introduction Recent studies from both experimental and clinical investigations have emphasized the progressive nature of central nervous system (CNS) injury. In contrast to the initial concept that the majority of damage occurs at the time of the primary ischemic or traumatic insult, new evidence emphasizes that acute injury can initiate a variety of pathophysiological cascades that lead to secondary injury mechanisms associated with subacute as well as progressive injury. These ﬁndings are Corresponding author. Tel.: +1-305-243-8926; Fax: +1-305-

243-3914; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61009-1

125

126

spinal cord injury (SCI) and highlight potential therapeutic strategies that may be relevant to these devastating insults.

Traumatic brain injury Traumatic brain injury (TBI) is a leading cause of morbidity and mortality in the United States (Langlois et al., 2000). In 2000, half of a million new cases of moderate and severe TBI were reported. Traumatic insults commonly occur in young adults as a consequence of trafﬁc and sporting accidents. More recently, there has been an increase in the number of traumatic brain insults in the elderly due to falls and other traumatic insults. In developing countries, a significant rise in vehicular accidents has also been observed. Thus, the magnitude of the problem internationally merits increased concern regarding the prevention as well as treatment of TBI. The pathophysiology of TBI is complicated and involves both primary and secondary insults (Graham et al., 2000a; Bramlett and Dietrich, 2004). Primary insults due to impact injury result in the rupture of membranes that lead to acute damage of neuronal, glial and vascular components. These membrane disturbances also cause metabolic stress that initiates a cascade of cellular and molecular mechanisms that can initiate both reparative as well as destructive processes. Acute consequences of TBI include alterations in the blood-brain barrier (BBB) as well as complex changes in local cerebral blood ﬂow (LCBF) and metabolism (Cortez et al., 1989; Hovda et al., 1991; Dietrich et al., 1994). Trauma-induced glial swelling as a consequence of both vasogenic and cytotoxic edema is also a common early consequence to TBI (Bullock et al., 1991; Kimbelberg and Norenberg, 1994). Early metabolic and hemodynamic events on their own can lead to necrotic cell death in speciﬁc vulnerable populations of cells. Severe reductions in LCBF are observed in some TBI patients and may indicate ischemic events (Bramlett and Dietrich, 2004). In human TBI, speciﬁc gray and white matter structures including the corpus callosum and hippocampus are frequently damaged after moderate and severe

TBI (Povlishock and Christman, 1995; Graham et al., 2000b; Leclereq et al., 2001). Diffuse axonal injury underlies a major mechanism for the morbidity and mortality associated with TBI (Adams et al., 1989, 1991). Various animal models have been developed that mimic some of the structural and behavioral consequences of human TBI (Gennarelli et al., 1982; Cernak, 2005). Although no one model of TBI exactly represents the human condition of brain trauma, these models are important in the understanding of critical pathomechanisms responsible for traumatic events and the eventual testing of novel therapeutic interventions. One model, ﬂuid-percussion (F-P) brain injury has been used commonly in these preclinical TBI studies (Dixon et al., 1987). For example, after moderate parasagital F-P brain injury, the extravasation of the protein tracer horseradish peroxidase (HRP) is observed in speciﬁc brain regions associated with acute vascular damage due to shearing forces produced by the insult (Dietrich et al., 1994). In other studies, enduring changes in axolemmal permeability has been reported in white matter tracts, not necessarily associated with overt damage (Povlishock and Pettus, 1996). Thus, abnormal permeability of a variety of cellular membranes are rendered leaky to tracers after TBI (Pettus and Povlishock, 1996). More recently, evidence for apoptotic cell death has been demonstrated in models of TBI as well as SCI (Crowe et al., 1997; Beattie et al., 2002; Lu et al., 2003). McIntosh and colleagues (Rink et al., 1995) ﬁrst provided ultrastructural evidence indicating that neurons may die by apoptotic mechanisms after F-P brain injury. In that study, neurons demonstrated the classical appearance of apoptotic bodies within the cell nucleus. More recently, molecular and biochemical approaches have been used to shed light on the various receptor and intracellular cell signaling mechanisms responsible for apoptotic cell death (Keane et al., 2001a, b). The importance of this line of research to the present discussion is that emerging evidence supports the concept that apoptotic pathways may be activated for long periods after CNS injury and that mechanisms similar to programmed cell death may be responsible for some of the

127

progressive changes seen in the brain late after trauma (Williams et al., 2001). In addition to apoptosis, inﬂammatory and immune processes also appear to play an important role in pathophysiology of TBI (Shohami et al., 1994; Schwab et al., 2001; Morganti-Kosmann et al., 2002). Following brain injury, indicators of the activation of pro-inﬂammatory processes including increased expression of pro-inﬂammatory cytokines such as IL-1b, TNFa and IL-6 have been described (Holmin et al., 1997; MorgantiKossman et al., 1997; Kinoshita et al., 2002). Also, speciﬁc receptors that interact with these cytokines are expressed after injury and lead to the stimulation of inﬂammatory signaling cascades responsible for the expression of inﬂammatory genes (Lotocki et al., 2004). Thus, experimental studies that have targeted inﬂammatory processes by applying antibodies or blockers to inﬂammation have in some cases lead to improved outcome in the subacute post-traumatic setting. For example, in several studies, treatment with the interleukin-1 receptor antagonist has been shown to lead to improved histopathological and behavioral outcome after TBI and cerebral ischemia (Toulmond and Rothwell, 1995; Rothwell and Luheshi, 2000). In reference to the present discussion, recent evidence for inﬂammatory processes remaining activated up to 1 year after experimental TBI (Nonaka et al., 1999) and even longer after clinical TBI (Gentleman et al., 2004) emphasizes the potential role of inﬂammation in progressive damage after trauma.

Clinical evidence for progressive injury In patients following TBI, magnetic resonance imaging (MRI) approaches have identiﬁed evidence for progressive atrophy of speciﬁc brain regions at chronic periods after trauma (Cullum and Bigler, 1986; Anderson and Bigler, 1995; van der Naalt et al., 1999). In these clinical studies, evidence for the enlargement of ventricle structures and atrophy of speciﬁc gray and white matter structures has been demonstrated. In one study, enlargement of the lateral ventricle was identiﬁed as a relatively late occurrence after TBI. This delayed response

occurred without any evidence of a systemic secondary insult and therefore had no obvious underlying mechanism. In experimental studies of TBI, the majority of investigations have evaluated traumatic outcome in the acute or subacute post-traumatic period. Thus, until recently, little information was available to determine the more chronic consequences of experimental TBI. However, in 1997, the chronic histopathological consequences of moderate F-P brain injury were ﬁrst assessed at 2 months after trauma. Bramlett et al. (1997a) reported a significant enlargement of the lateral ventricle and with associated atrophy of cerebral cortical areas within the traumatized hemisphere. This study was the ﬁrst to provide histopathological evidence for structural changes continuing to occur weeks to months after TBI. In subsequent studies, other investigators using similar or different injury models have complemented and extended these initial ﬁndings by assessing even longer periods of posttraumatic injury (Smith et al., 1997; Pierce et al., 1998; Dixon et al., 1999; Bramlett and Dietrich, 2002). For example, Smith et al. (1997) reported progressive injury in various forebrain areas 1 year after moderate lateral F-P injury. In a model of controlled cortical impact (CCI) injury, Dixon et al. (1999) also reported progressive damage and long lasting behavioral deficits in that trauma model of focal damage. Thus, in various laboratories using different injury models, evidence of progressive damage has been demonstrated. Most recently, Bramlett and Dietrich (2002) assessed alterations in white matter tracts 1 year after moderate F-P brain injury (Fig. 1). In that study, significant atrophy of various white matter structures within the traumatized hemisphere provided direct evidence of progressive white matter pathology in widespread forebrain circuits. These newer data emphasize the important fact that white matter as well as gray matter structures may be highly vulnerable to progressive damage after relatively moderate degrees of trauma. Thus, therapies may have to be developed that target both gray and white matter pathology after TBI (Medana and Esiri, 2003). The importance of white matter vulnerability will again be emphasized in the SCI discussion.

128

Fig. 1. Double-stained H&E and Luxol-fast blue sections 1 year after TBI or sham procedure. (A) TBI animal showing gross atrophy with marked expansion of the ipsilateral lateral ventricle. (B) Sham operated animal appearing unremarkable. (C) Higher magniﬁcation of external capsule thinning (arrows) after TBI. Reprinted from Bramlett and Dietrich (2002) with kind permission of Springer Science and Business Media.

Pathomechanisms underlying progressive injury As previously discussed, both apoptotic and inﬂammatory cascades are felt to underlie some of the acute and subacute pathophysiological mechanisms that are responsible for cellular dysfunction and death. Recent data also indicate that these mechanisms may be active weeks and months after trauma and also participate in the progressive nature of TBI. For example, prolonged apoptotic cell death has been demonstrated in several CNS injury models (Emery et al., 1998; Beattie et al., 2002). In these studies neuronal, microglial and oligodendrocyte apoptosis has been reported days to weeks after injury. Delayed neuronal apoptosis may lead to the continued removal of axon projections that could ultimately lead to patterns of deafferentation syndromes in brain regions remote from the primary site of damage. A good example of this type of progressive and remote damage is the delayed thalamic pathology observed after parasagittal or lateral F-P brain injury (Bramlett et al., 1997a, b). Apoptotic death of oligodendrocytes could also have devastating effects on both

structure and function of an axon and/or circuit. Oligodendrocyte death would lead to demyelination of projecting axons that would be expected to result in axonal dysfunction and possible progressive damage (Waxman, 1989). Indeed, electrophysiological studies have demonstrated the adverse effects of demyelination on action potentials and axonal survival (Waxman et al., 1994). Thus, progressive apoptotic cell death can have a variety of adverse consequences on the structure and function of both gray and white matter structures. In the area of inﬂammation, recent data has also emphasized the progressive nature of that response to TBI. Several studies have provided evidence for long-term inﬂammatory responses to TBI. In brain sections stained for macrophage activation for example, evidence for inﬂammatory events is sometimes also seen many weeks after a traumatic brain insult. Gentleman et al. (2004) reported microglial activation, weeks after TBI. Similarly, inﬂammatory cells have been observed chronically in an animal model of TBI (Rodriguez-Paez et al., 2005). In this study,

129

Fig. 2. Light level micrographs of toluidine blue stained thick plastic sections of control (A) and traumatized tissue, specifically the lateral posterior thalamic nuclei, (100 ) at 3 days (B), 15 days (C), 3 months (D), 9 months (E) and 12 months (F) after TBI. (A) Control tissue shows myelinated ﬁgures oriented perpendicular and parallel to the plain of the section. Normal appearing neuronal cell bodies (N), astrocytes (black arrow) and blood vessels (V) are also apparent. (B) At 3 days, axonal abnormalities including changes in axoplasmic density, unraveling of the myelin sheath (black pointer) and irregular swollen myelinated proﬁles (arrowhead) were observed. There is also a proliferation of microglial cells (open arrow) associated with neuronal (N) and astrocytic (black arrow) swelling. Normal appearing oligodendrocytes (double arrows) are observed. (C) At 15 days, there is an apparent increase in overall tissue vacuolation, which appears to be associated with axonal (arrowhead) and non-axonal (black arrow) proﬁles. An increase in microglial cells (open arrows) as well as unraveled myelin ﬁgures (black pointer) are observed. (D) At 3 months, parenchymal vacuolation is still apparent with increased numbers of microglial cells (open arrows), swollen axonal ﬁgures (arrowhead) and irregular myelin proﬁles (black pointer). (E) At 9 months, a dramatic number of vacuolated proﬁles are observed related to inﬂammatory cells scattered among tissue debris. An amorphous crystallized material (black star) was observed inside the vacuolated proﬁles. There appears to be a progressive increase in the size of the vacuolated proﬁles as the post-injury time increases. (F) At 12 months, the vacuolization (large black arrows) continues to progress in association with a further decrease in numbers of myelinated axons without the presence of the amorphous material described above. Reprinted from Rodriguez-Paez et al. (2005) with kind permission of Springer Science and Business Media.

130

macrophage/microglia inﬁltration and swollen axons were observed as late as 6 months using both light and electron microscopic analysis within several vulnerable structures (Fig. 2). Additionally, a temporal decline in the number of myelinated axons within the cerebral cortex (Fig. 3) and thalamus were reported which may be due to the prolonged inﬂammatory response observed in this study. Nonaka et al. (1999) also provided evidence for inﬂammatory processes being active up to 1 year after TBI. In that study, NF-kappaB was seen in both cortical and sub-cortical brain regions undergoing progressive atrophy. Thus, the potential for macrophage/microglia released toxic substances leading to tissue damage is a real possibility even weeks after TBI. To support this assumption, recent biomarker studies that have measured levels of pro-inﬂammatory mediators in the cerebral spinal ﬂuid (CSF) and plasma of TBI patients have reported elevated levels, days

after injury (Rancan et al., 2004). In the future, surrogate biochemical markers of tissue damage may be developed to predict and treat progressive injury mechanisms. Abnormal protein aggregation has been implicated in the pathogenesis of a number of neurological diseases including Alzheimer’s and Parkinson’s disease (Chaudhuri and Paul, 2006). Recent evidence indicates that abnormal protein aggregation also occurs in models of cerebral ischemia and TBI (Blumbergs et al., 1994; Graham et al., 1995; Lewen et al., 1995; Bramlett et al., 1997b; Hamberger et al., 2003). In models of TBI, immunocytochemical localization of beta APP and other proteins has been identiﬁed in gray as well as white matter tracts (Sherriff et al., 1994; Bramlett et al., 1997b). Whether evidence for prolonged periods of protein aggregation can be correlated with progressive tissue damage merits continued study.

Fig. 3. Estimation of myelinated axons in the cerebral cortex of animals following TBI or Sham surgery at 3 days, 15 days, 1 month, 3 months, 6 months, 9 months and 12 months. A significant difference was found for both group (po0.001) and time (po0.001) between TBI and Sham animals. TBI results in a decrease in the number of myelinated axons compared to Sham animals at all time points analyzed. In addition, there was a temporal decrease in the number of myelinated axons in the Sham animals as well. The decrease in the axonal numbers of the Sham animals may be due to a normal aging process. Reprinted from Rodriguez-Paez et al. (2005) with kind permission of Springer Science and Business Media.

131

Spinal cord injury Each year in the United States, approximately 11,000 new spinal cord injuries are recorded. Currently there are over 250,000 individuals living with chronic SCI and its devastating consequences in the United States. Similar to what has been described with TBI, evidence for lesion progression has also been demonstrated in both experimental and clinical conditions of SCI (Wallace et al., 1987; Bunge et al., 1993; Crowe et al., 1997; Bruce et al., 2000; Hill et al., 2001; Guest et al., 2005; Totoiu and Keirstead, 2005). Because SCI frequently occurs in young people, it is equally important that we understand the mechanisms underlying chronic progressive damage and develop therapeutic interventions to retard cell death, axonal degeneration and demyelination in SCI patients. The pathophysiology of acute SCI is multifactorial and like TBI, includes both primary and secondary injury mechanisms (Nashmi and Fehlings, 2001; Keane et al., 2006). Primary injury mechanisms include acute SC compression, impaction, laceration, shear damage and missile injury (Norenberg et al., 2004). These acute injury mechanisms initiate a complex cascade of secondary injury mechanisms that may remain activated for months or years after injury. These initial traumatic events lead to vascular damage and hemorrhage, alterations in spinal cord blood ﬂow, vascular thrombosis, vasospasm and loss of autoregulation. As a consequence to cellular membrane damage, metabolic abnormalities and electrolytic shifts in ions occur (LoPachin and Lehning, 1997; Li et al., 2000). In addition, neurotransmitters are released into the extracellular space leading to the abnormal activation of various receptors and intracellular signaling processes (Keane et al., 2006). Other classical injury cascades that are activated after SCI include free radical formation, lipid peroxidation and edema formation. All of these acute processes lead to the acute destruction of gray and white matter structures (Blight, 1985; Schwab and Bartholdi, 1996; Rosenberg and Wrathall, 1997). As in brain injury, inﬂammatory cascades are also activated after SCI (Blight, 1992; Crowe et al., 1997; Popovich et al., 2002; Keane et al., 2006). In

addition, calpain activation has also been emphasized as an acute injury mechanism (Ray et al., 2003). Many of these injury cascades have been targets for therapeutic interventions (Blight and Zimber, 2001). More recently, evidence for apoptotic cell death has been reported in a number of SCI models (Springer et al., 1999; Ozawa et al., 2002; Keane et al., 2006). A complex integrated apoptotic pathway involving both extrinsic and intrinsic apoptotic cascades has been reported. Also, the study of pro- and anti-apoptotic molecules, which can control apoptotic cell death, is an exciting area of current investigation. Recent work has concentrated on mechanisms that allow for the communication between the external environment with intracellular processes involved in cell survival and death. After trauma for example, speciﬁc death receptors have been shown to accumulate within speciﬁc areas of the plasma membrane called lipid rafts that contain high concentrations of cholesterol and sphingolipid (Lotocki et al., 2004). Thus, after SCI and brain trauma, TNF1 receptors accumulate in lipid rafts and assist in the formation of cytoplasmic platforms that allow scaffolding proteins to accumulate; leading to the increased interactions of proteins that lead to the activation of intracellular cascades associated with cell survival and death (Keane et al., 2006). Thus, research is currently being undertaken to understand this relatively acute injury response to neuronal vulnerability and survival. Because a large amount of cell death occurs in the acute post-traumatic period, evidence for these pathomechanisms has been commonly reported during these periods. However, since cell death during delayed post-traumatic periods is spread out and more difﬁcult to identify, other approaches including immunocytochemistry must be used to investigate later occurring injury mechanisms. In addition, experimental and clinical investigations have also demonstrated that apoptotic cell death, demyelination, remyelination and axonal degeneration may occur weeks to months after experimental and clinical SCI (Bunge et al., 1961; Blight, 1993; Quencer and Bunge, 1996; Schwab and Bartholdi, 1996; Crowe et al., 1997; Emery

132

et al., 1998; Bruce et al., 2000; Hill et al., 2001; Guest et al., 2005; Totoiu and Keirstead, 2005). For example, in a study by Crowe et al. (1997), apoptotic cells were identiﬁed from 6 h to 3 weeks after experimental SCI. Apoptotic cell death was specifically shown to be present in the white matter tracts where apoptotic cells were shown to be positive for cellular markers of oligodendrocytes. This observation is important because it indicates oligodendrocyte cell death with resulting demyelination could be an active mechanism in the progressive injury cascades associated with human SCI. Following this particular observation, various laboratories have also reported apoptotic cell death in the spinal cord including neurons, oligodendrocytes and inﬂammatory cells (Emery et al., 1998; Beattie et al., 2002). In addition to clarifying mechanisms of injury, these results are also important because they provide new potential targets for therapeutic interventions directed toward the acute as well as chronic post-traumatic period (Springer et al., 1999; Ozawa et al., 2002; Demjen et al., 2004). Evidence for apoptotic cell death and Schwannosis has also been reported in specimens obtained from patients that survived long periods after SCI. In a study by Emery et al. (1998), apoptotic cells were identiﬁed using the apoptotic marker caspase-3. In that study, oligodendrocytes stained positively for this indicator of apoptotic injury in speciﬁc white matter tracts. Again, this clinical observation is important because of the role of the oligodendrocyte in axonal myelination. We know from various electrophysiological studies that if an axon is demyelinated by a toxic or traumatic insult, that axon’s ability to propagate axon potentials is severely affected (Blight and Young, 1989; Waxman, 1989). Thus, strategies that would target degenerative mechanisms or axonal conduction blockage could potentially improve outcome in SCI subjects (Blight and Young, 1989). Hill et al. (2001) have reported on extensive dieback of the corticospinal tract chronically after SCI (Fig. 4). However, there was a regenerative response of this tract along with the reticulospinal tract evidenced by an extension of collaterals into the lesion matrix. Recently, Totoiu and Keirstead (2005) have assessed chronic progressive

demyelination in an animal model of SCI. Data from that study indicated that chronic progressive demyelination after thoracic injury in rats does occur for days after injury (Fig. 5). Interesting, a process of secondary demyelination was reported around 120–450 days post injury. These studies underscore the importance of targeting demyelination in the development of therapeutic interventions and again emphasize the progressive nature of neurodegeneration after SCI. In human tissue, evidence for chronic demyelination has also been reported (Bunge et al., 1993; Norenberg et al., 2004; Guest et al., 2005). In the study by Guest et al. (2005), evidence for axonal demyelination, even a decade after human traumatic SCI was presented. Although the response was very heterogeneous among the SCI specimens evaluated, the potential for demyelination in speciﬁc spinal cord circuits in which axon proﬁles remained intact was observed. In this regard, evidence for spontaneous remyelination of intact CNS axons by invading Schwann cells has also been observed (Guest et al., 2005; Totoiu and Keirstead, 2005). This observation is important because other studies have reported similar cellular responses to injury that indicate some endogenous reparative mechanisms including remyelination are activated after SCI (Hagg and Oudega, 2006). In a clinical study by Bruce et al. (2000), evidence for Schwannosis was seen in 32 out of 65 cases that survived after 24 years. The incidence of Schwannosis rose to 82% in SCI patients who survived more than 4 months. Associated with Schwannosis was intense chondroitin sulfate proteoglycan (CSPG) staining. This observation is important because the CSPGs are thought to be inhibitory molecules that reduce axonal regeneration after CNS injury (Snow et al., 1990). Thus the continued clariﬁcation of both the positive and negative consequences of Schwann cell responses to acute and chronic SCI remains an important area of investigation. The use of human spinal cords donated for research is also allowing other injury mechanisms to be evaluated in human tissues (Norenberg et al., 2004). At The Miami Project to Cure Paralysis, approximately 115 human spinal cords have been collected from individuals surviving from 24 h to

133

Fig. 4. Appearance of impact site after a 12.5 mm contusion injury 1 dpi (A–C), 3 dpi (A, D, E), 8 dpi (A, F, G), 21 dpi (A, H, I) and 14 weeks (A, J, K, L) after injury. (A) The progression of cavitation. At 1 day, the impact site is hemorrhagic but no cavitation is present, and RBCs extend up to an additional 5 mm rostral (arrow); between 3, 8 and 21 days the injury site becomes progressively more deﬁned, and it appears as a dark region in low magniﬁcation at 8 dpi and is clearly deﬁned at 21 days. At 21 days and 14 weeks a cellular mass is present within the cavity attached to the spared rim of white matter at several points via trabeculae (arrow). One dpi (B) RBCs and (C) damaged axons at the impact site. Three dpi (D) damaged axons (small arrow) intermixed with blood cell inﬁltrate (large arrow) at the impact site and (E) some macrophages (large arrowhead) have inﬁltrated the rim of spared white matter (small arrows). Eight dpi (F) macrophages are densely packed at the center of the cavity while (G) open spaces (small arrow) are present between groups of macrophages and other nonﬂuorescently labeled cells (large arrowhead) that appear in scattered bands within the cavity. Twenty-one dpi (H) trabeculae, thin tissue bridges with regions of no tissue beside them, are beginning to form (small arrow), and macrophages (large arrow head) are still present within the cavity in association with the trabeculae or as (I) macrophage rafts. Fourteen weeks after injury (J–L) autoﬂuorescing macrophages (large arrowhead) are present within (J) ﬁbrous trabeculae and (K) cellular trabeculae as well as (L) the cellular inﬁltrate rostral to the cavity. Bar in (A), 1 mml; all other bars, 100 mm. (C, E, F), same magniﬁcation. (B, D, G–L), same magniﬁcation. dpi ¼ days post injury. Reprinted from Hill et al. (2001) with permission from Elsevier.

134

Fig. 5. Quantiﬁcation of demyleinated and remyelinated axons at different time points post injury. The numbers of axons at the six cranial and six caudal levels examined for all animals within a group were averaged to generate each point on the graph. The number of demyleinated axons was highest at 1 day post injury and decreased substantially by 7 days post injury. Thereafter, demyelination was a chronic and progressive phenomenon. Both oligodendrocyte and Schwann cell remyelination was present at all subsequent time points post injury. Reprinted from Totoiu and Keirstead (2005) with permission from John Wiley & Sons, Inc.

24 years after injury. Corresponding MRI and immunocytochemical techniques are used to analyze the injured tissue (Bunge et al., 1993; Becerra et al., 1995; Emery et al., 1998; Guest et al., 2005). With this approach, investigators have assessed the acute as well as more chronic immunohistopathological consequences of traumatic human SCI. For example, in early specimens taken 3 days after injury H&E staining shows damage to gray and white matter areas with a well identiﬁed central hemorrhagic lesion (Bunge et al., 1993). If specimens are stained with various immunocytochemical approaches, evidence of gliosis in the form of increased glial ﬁbrillary acidic protein (GFAP) staining is seen surrounding the contusive site. Also, the accumulation of invading inﬂammatory cells such as polymorphonuclear leukocytes (PMNL) and macrophages can be visualized (Fleming et al., 2006). These types of studies conducted on human tissues are important because they verify some of the acute and more chronic histopathological changes that are reported in preclinical animal studies. This information is currently of importance to the development of new treatments targeting inﬂammatory processes,

which is an important secondary injury mechanism after SCI (Gris et al., 2004). Evidence for retrograde degeneration of cortical spinal tracts (CST) has also been shown in human specimens. Following damage to thoracic cortical spinal tracts for example, evidence for retrograde degeneration can be seen in remote spinal tracts (Bunge et al., 1993; Norenberg et al., 2004). In addition to retrograde degeneration, evidence for Wallerian degeneration can also be obtained using repetitive MR imaging procedures in SCI patients (Quencer and Bunge, 1996). These studies emphasize the widespread changes that occur as a consequence to local SCI. With the development and improvement of new imaging techniques, acute damage as well as the progressive nature of human SCI will be more easily investigated. In acute injury settings, better evidence for local cord compression will provide important information to help guide surgical approaches to target ischemic events. In chronic postoperative SCI cases, both cord compression and the formation and progression of cavities can be visualized in patients experiencing neurological symptoms including chronic pain. With time, some lesions expand to form cysts

135

that can have a dramatic effect on progression of neurological symptoms. These imaging techniques will continue to be improved and allow specialized treatments to be developed for an individual patient undergoing progressive injuries.

Supraspinal alterations after SCI In addition to changes occurring at the level of the injured spinal cord, it is also clear that SCI leads to alterations in supraspinal areas of the neuroaxis (Jain et al., 1997; Raineteau et al., 2001; Hains et al., 2003; Hubscher and Johnson, 2006; Kim et al., 2006). In this regard, there is a rich literature on experimental SCI lesioning studies in rodents and non-human primates showing evidence for degeneration of neurons in the cerebral cortex after injury. Some data indicate that cell death occurs through apoptotic mechanisms whereas more recently, additional evidence indicates that cortical cell bodies as a consequence of SCI may only undergo severe atrophy but not actually die. In these cases, the local addition of neurotrophic factors to a particular brain region reverses some of these morphological changes. In addition to cellular changes, evidence for circuit reorganization and plasticity resulting in deactivation and reactivation of certain brain areas has been demonstrated after clinical and experimental SCI (Jain et al., 1997; Raineteau et al., 2001). Chronic SCI has also been shown to induce changes in the response of thalamic neurons to physiological activation (Hubscher and Johnson, 2006). Thus, SCI leads to alterations in brain circuits that are responsible for assessing normal and abnormal sensation. In this regard, some researchers believe that this circuit plasticity in addition to other events including chronic inﬂammation may account for the late development of chronic neurogenic pain in patients with SCI. A clearer understanding of local circuit changes in response to SCI may provide important information regarding how to treat these patients with neurogenic pain. As a consequence of chronic SCI, the cerebral cortex may also undergo plastic changes in terms of shifts in cortical map architecture (Topka et al., 1991; Jain et al., 1997; Bruehlmeier et al., 1998;

McDonald et al., 2002; Sabbah et al., 2002; Lotze et al., 2006). As speciﬁc inputs to the cerebral cortex are removed after SCI, other cortical areas may take over function. Thus, electrophysiological and metabolic studies have shown evidence for cortical plasticity after SCI (Hoffman and Field-Fote, 2007; Kim et al., 2006). These observations not only emphasize the potential for plasticity occurring in remote brain regions after SCI, but also may be important as we think about repairing the nervous system in the chronically injured spinal cord. If brain circuits and/or cortical maps were significantly altered due to chronic SCI, would such a consequence affect the ability to repair the nervous system and return function? Only after successful regenerative approaches are developed in the chronically injured state can these questions be answered.

Therapeutic interventions targeting progressive injury Based on the complexity of brain and SCI, it is clear that the injury mechanisms are multifactorial and may require a combinational therapeutic approach. In the area of brain and SCI, various neuroprotective agents have been evaluated (Narayan et al., 2002). Neuroprotective agents such as methylprednisolone, GM1 gangliosides, lazeroids, calcium and sodium channel blockers, growth factors as well as blockers of excitototoxic process have been reported to be effective in some animal models. More recently, the use of anti-inﬂammatory strategies, calpain antagonists, antiapoptotic strategies as well as agents targeting cAMP have also been investigated with various results (Blight and Zimber, 2001). Mild hypothermia that targets multiple injury cascades has been tested with various degrees of success after both SCI and TBI (Hayashi et al., 2004; Guest and Dietrich, 2005). Although hypothermia was used in the 60 s to target both brain and SCI, profound hypothermia (271C) commonly produced severe effects on cardiac function and increased infection rates in patients. However, in the mid-80 s, the importance of mild to moderate hypothermic treatment was demonstrated in

136

ischemic and traumatic animal models. Busto et al. (1987) ﬁrst reported that a reduction of just 2 or 3 degrees in the temperature of the brain provided dramatic protection of CA1 neurons within the post-ischemic hippocampus. Subsequent studies demonstrated that mild hypothermia was also protective when initiated at various times after cerebral ischemia or TBI. Recently, mild hypothermia has been shown to have a dramatic effect on SCI. In a study by Yu et al. (2000), the beneﬁcial effects of systemic hypothermia on locomotor outcome and histopathological damage were reported after contusion SCI in rats. In that study, mild hypothermia (331C) initiated 30 min after SCI significantly improved open locomotor function and significantly reduced contusion volume weeks after injury. Because of these exciting preclinical ﬁndings, multiple clinical trials have been initiated in patients with various neurological insults including cardiac arrest, stroke, TBI and SCI. Recent multicenter trials in cardiac arrest patients have shown a dramatic beneﬁt with mild hypothermia in that patient population (Bernard et al., 2002; Hozler, 2002). Also, individual and multicenter trials have reported that mild hypothermia shows promise in improving outcome in severe TBI patients (Jiang et al., 2006). Clinical investigations are currently determining the effects of mild hypothermia involving SCI (Guest and Dietrich, 2005). In this regard, mild hypothermia may be used during elective surgeries or in the early post-injury periods. The beneﬁcial effects of mild hypothermia involve multiple injury pathomechanisms. Various studies have shown that modifying brain or SC temperatures significantly affects injury-induced excitatory neurotransmitter levels, free fatty acid formation, BBB breakdown and the formation of edema. Mild hypothermia after SCI has been shown to significantly reduce the acute inﬂammatory response to injury (Chatzipanteli et al., 2000). Because the early inﬂammatory response to injury is thought to represent an important secondary injury mechanism, the ability of mild hypothermia to limit the accumulation of inﬂammatory cells after SCI represents an important therapeutic target for treatment. Free radical formation can come from multiple sources including prostaglandin and nitric

oxide (NO) synthesis. The inducible form of nitric oxide synthase (iNOS) produces free radicals that can be destructive to tissue survival. Post-traumatic hypothermia following TBI also reduces iNOS activity and may improve outcome by affecting free radical generation as well (Chatzipanteli et al., 1999). Mild hypothermia has been shown to be neuroprotective and promote functional outcome after TBI (Hayashi et al., 2004). Bramlett et al. (1995) ﬁrst reported that post-traumatic hypothermia improved cognitive function in rats. These studies are important because cognitive deficits are some of the most severe consequences of mild, moderate and severe TBI. In regards to potential treatments for progressive injury, post-traumatic hypothermia has also been shown to reduce the progressive nature of F-P brain injury in rats (Bramlett et al., 1997a). In that study, a period of 3 h of moderate hypothermia significantly protected against progressive damage and enlargement of the lateral ventricle at 2 months after TBI. Thus, in addition to the acute beneﬁts of therapeutic hypothermia, this therapy may also limit the amount of progressive damage after trauma. Additional studies will be required to determine what duration of the hypothermic therapy is most beneﬁcial in providing long-term beneﬁts in terms of functional outcome. Finally, recent data indicate that different types of cellular transplantation strategies can promote repair and improve recovery in animal models of brain and SCI (Schouten et al., 2004; Schwab et al., 2006; Thuret et al., 2006). In some cases, cell transplantation strategies have been initiated to replace dysfunctional or dead neurons injured by the insult. Thus, an exciting ﬁeld of research is currently directed toward stem cell biology and the potential for neural stem cells to repopulate the injured nervous system and improve function. Cell survival, migration, control of cellular differentiation and the integration of new cells into existing circuits are some of the challenges currently being addressed with these approaches. Alternatively, the transplantation of speciﬁc populations of restricted progenitor cells that differentiate into myelin-forming cells is another important research direction (Cao et al., 2005; Keirstead et al., 2005;

137

Karimi-Abdolrezaee et al., 2006). In recent studies, grafted cells have been shown to form morphologically normal-appearing myelin sheaths around damaged axons. Importantly, experimental treatments targeting neuronal or glia damage after CNS injury leads to improvements in functional recovery. Because progressive damage after brain and SCI can affect a variety of different cell types, these cellular therapies may be helpful in both neuroprotective as well as reparative strategies targeting the long-term consequences of CNS injury. Continued research into the progressive nature of brain and SCI should provide new targets for treatment and better outcome in patients that sustain these devastating injuries.

Acknowledgments This work was supported in part by NIH grants NS30291, NS38665 and DAMD17-02-1-0190. The authors also thank the members of the Bramlett/ Dietrich laboratory for their important contributions to this research ﬁeld.

References Adams, J.H., Doyle, D., Ford, I., Gennarelli, T.A., Graham, D.I. and McLellan, D.R. (1989) Diffuse axonal injury in head injury: definition, diagnosis and grading. Histopathology, 15: 49–59. Adams, J.H., Graham, D.I., Gennarelli, T.A. and Maxwell, W.L. (1991) Diffuse axonal injury in non-missile head injury. J. Neurol. Neurosurg. Psychiatry, 54: 481–483. Anderson, C.V. and Bigler, E.D. (1995) Ventricular dilation, cortical atrophy, and neuropsychological outcome following traumatic brain injury. J. Neuropsychiatry Clin. Neurosci., 7: 42–48. Beattie, M.S., Hermann, G.E., Rogers, R.C. and Bresnahan, J.C. (2002) Cell death in models of spinal cord injury. In: McKerracher L., Doucet G. and Rossignol S. (Eds.), Progress in Brain Research, Vol. 137. Elsevier, Amsterdam, pp. 37–47. Becerra, J.L., Puckett, W.R., Hiester, E.D., Quencer, R.M., Marcillo, A.E., Post, M.J. and Bunge, R.P. (1995) MRpathologic comparisons of wallerian degeneration in spinal cord injury. AJNR Am. J. Neuroradiol., 16: 125–133. Bernard, S.A., Gray, T.W., Buist, M.D., Jones, B.M., Silvester, W., Gutteridge, G. and Smith, K. (2002) Treatment of comatose survivors of out-of-hospital cardiac arrest with induced hypothermia. N. Engl. J. Med., 346: 557–563.

Blight, A.R. (1985) Delayed demyelination and macrophage invasion: a candidate for secondary cell damage in spinal cord injury. Cent. Nerv. Syst. Trauma, 2: 299–315. Blight, A.R. (1992) Macrophages and inﬂammatory damage in spinal cord injury. J. Neurotrauma, 9(Suppl 1): S83–S91. Blight, A.R. (1993) Remyelination, revascularization, and recovery of function in experimental spinal cord injury. Adv. Neurol., 59: 91–104. Blight, A.R. and Young, W. (1989) Central axons in injured cat spinal cord recover electrophysiological function following remyelination by Schwann cells. J. Neurol. Sci., 91: 15–34. Blight, A.R. and Zimber, M.P. (2001) Acute spinal cord injury: pharmacotherapy and drug development perspectives. Curr. Opin. Investig. Drugs, 2: 801–808. Blumbergs, P.C., Scott, G., Manavis, J., Wainwright, H., Simpson, D.A. and McLean, A.J. (1994) Staining of amyloid precursor protein to study axonal damage in mild head injury. Lancet, 344: 1055–1056. Bramlett, H.M. and Dietrich, W.D. (2002) Quantitative structural changes in white and gray matter 1 year following traumatic brain injury in rats. Acta Neuropathol. (Berl.), 103: 607–614. Bramlett, H.M. and Dietrich, W.D. (2004) Pathophysiology of cerebral ischemia and brain trauma: similarities and differences. J. Cereb. Blood Flow Metab., 24: 133–150. Bramlett, H.M., Dietrich, W.D., Green, E.J. and Busto, R. (1997a) Chronic histopathological consequences of ﬂuidpercussion brain injury in rats: effects of post-traumatic hypothermia. Acta Neuropathol. (Berl.), 93: 190–199. Bramlett, H.M., Green, E.J., Dietrich, W.D., Busto, R., Globus, M.Y-T. and Ginsberg, M.D. (1995) Post-traumatic brain hypothermia provides protection from sensorimotor and cognitive behavioral deficits. J. Neurotrauma, 12: 289–298. Bramlett, H.M., Kraydieh, S., Green, E.J. and Dietrich, W.D. (1997b) Temporal and regional patterns of axonal damage following traumatic brain injury: a beta-amyloid precursor protein immunocytochemical study in rats. J. Neuropathol. Exp. Neurol., 56: 1132–1141. Bruce, J.H., Norenberg, M.D., Kraydieh, S., Puckett, W., Marcillo, A. and Dietrich, W.D. (2000) Schwannosis: role of gliosis and proteoglycan in human spinal cord injury. J. Neurotrauma, 17: 781–788. Bruehlmeier, M., Dietz, V., Leenders, K.L., Roelcke, U., Missimer, J. and Curt, A. (1998) How does the human brain deal with a spinal cord injury? Eur. J. Neurosci., 10: 3918–3922. Bullock, R., Maxwell, W.L., Graham, D.I., Teasdale, G.M. and Adams, J.H. (1991) Glial swelling following human cerebral contusion: an ultrastructural study. J. Neurol. Neurosurg. Psychiatry, 54: 427–434. Bunge, M.B., Bunge, R.P. and Ris, H. (1961) Ultrastructural study of remyelination in an experimental lesion in adult cat spinal cord. J. Biophys. Biochem. Cytol., 10: 67–94. Bunge, R.P., Puckett, W.R., Becerra, J.L., Marcillo, A. and Quencer, R.M. (1993) Observations on the pathology of human spinal cord injury. A review and classiﬁcation of 22 new cases with details from a case of chronic cord compression with extensive focal demyelination. Adv. Neurol., 59: 75–89.

138 Busto, R., Dietrich, W.D., Globus, M.Y.-T., Valdes, I., Scheinberg, P. and Ginsberg, M.D. (1987) Small differences in intraischemic brain temperature critically determine the extent of ischemic neuronal injury. J. Cereb. Blood Flow Metab., 7: 729–738. Cao, Q., Xu, X.M., DeVries, W.H., Enzmann, G.U., Ping, P., Tsoulfas, P., Wood, P.M., Bunge, M.B. and Whittemore, S.R. (2005) Functional recovery in traumatic spinal cord injury after transplantation of multineurotrophin-expressing glial-restricted precursor cells. J. Neurosci., 25: 6947–6957. Cernak, I. (2005) Animal models of head trauma. NeuroRx., 2: 410–422. Chatzipanteli, K., Wada, K., Busto, R. and Dietrich, W.D. (1999) Effects of moderate hypothermia on constitutive and inducible nitric oxide synthase activities after traumatic brain injury in the rat. J. Neurochem., 72: 2047–2052. Chatzipanteli, K., Yanagawa, Y., Marcillo, A.E., Kraydieh, S., Yezierski, R.P. and Dietrich, W.D. (2000) Posttraumatic hypothermia reduces polymorphonuclear leukocyte accumulation following spinal cord injury in rats. J. Neurotrauma, 17: 321–332. Chaudhuri, T.K. and Paul, S. (2006) Protein-misfolding diseases and chaperone-based therapeutic approaches. FEBS J., 273: 1331–1349. Cortez, S.C., McIntosh, T.K. and Noble, L.J. (1989) Experimental ﬂuid percussion brain injury: vascular disruption and neuronal and glial alternations. Brain Res., 482: 271–282. Crowe, M.J., Bresnahan, J.C., Shuman, S.L., Masters, J.N. and Beattie, M.S. (1997) Apoptosis and delayed degeneration after spinal cord injury in rats and monkeys. Nat. Med., 3: 73–76. Cullum, C.M. and Bigler, E.D. (1986) Ventricle size, cortical atrophy and the relationship with neuropsychological status in closed head injury: a quantitative analysis. J. Clin. Exp. Neuropsychol., 8: 437–452. Demjen, D., Klussmann, S., Kleber, S., Zulani, C., Stieltjes, B., Metzger, C., Hirt, U.A., Walczak, H., Falk, W., Essig, M., Edler, L., Krammer, P.H. and Martin-Villalba, A. (2004) Neutralization of CD95 ligand promotes regeneration and functional recovery after spinal cord injury. Nat. Med., 10: 389–395. Dietrich, W.D., Alonso, O. and Halley, M. (1994) Early microvascular and neuronal consequences of traumatic brain injury: a light and electron microscopic study in rats. J. Neurotrauma, 11: 289–301. Dixon, C.E., Kochanek, P.M., Yan, H.Q., Schiding, J.K., Grifﬁth, R.G., Baum, E., Marion, D.W. and DeKosky, S.T. (1999) One-year study of spatial memory performance, brain morphology, and cholinergic markers after moderate controlled cortical impact in rats. J. Neurotrauma, 16: 109–122. Dixon, C.E., Lyeth, B.G., Povlioshock, J.T., Findling, R.L., Hamm, R.J., Marmarou, A., Young, H.F. and Hayes, R.L. (1987) A ﬂuid percussion injury model of experimental brain injury in the rat. J. Neurosurg., 67: 110–119. Emery, E., Aldana, P., Bunge, M.B., Puckett, W., Srinivasan, A., Keane, R.W., Bethea, J. and Levi, A.D. (1998) Apoptosis after traumatic human spinal cord injury. J. Neurosurg., 89: 911–920.

Fleming, J.C., Norenberg, M.D., Ramsay, D.A., Dekaban, G.A., Marcillo, A.E., Saenz, A.D., Pasquale-Styles, M., Dietrich, W.D. and Weaver, L.C. (2006) The cellular inﬂammatory response in human spinal cords after injury. Brain, 129: 3259–3269. Gennarelli, T.A., Thibault, L.E., Adams, J.H., Graham, D.I., Thompson, C.J. and Marcincin, R.P. (1982) Diffuse axonal injury and traumatic coma in the primate. Ann. Neurol., 12: 564–574. Gentleman, S.M., Leclercq, P.D., Moyes, L., Graham, D.I., Smith, C., Grifﬁn, W.S. and Nicoll, J.A. (2004) Long-term intracerebral inﬂammatory response after traumatic brain injury. Forensic Sci. Int., 146: 97–104. Graham, D.I., Gentleman, S.M., Lynch, A. and Roberts, G.W. (1995) Distribution of beta-amyloid protein in the brain following severe head injury. Neuropathol. Appl. Neurobiol., 21: 27–34. Graham, D.I., McIntosh, T.K., Maxwell, W.L. and Nicoll, J.A. (2000a) Recent advances in neurotrauma. J. Neuropathol. Exp. Neurol., 59: 641–651. Graham, D.I., Raghupathi, R., Saatman, K.E., Meaney, D. and McIntosh, T.K. (2000b) Tissue tears in the white matter after lateral ﬂuid percussion brain injury in the rat: relevance to human brain injury. Acta Neuropathol. (Berl.), 99: 117–124. Gris, D., March, D.R., Oatway, M.A., Chen, Y., Hamilton, E.F., Dekaban, G.A. and Weaver, L.C. (2004) Transient blockade of the CD11d/CD18 integrin reduces secondary damage after spinal cord injury, improving sensory, autonomic, and motor function. J. Neurosci., 24: 4043–4051. Guest, J.D. and Dietrich, W.D. (2005) Spinal cord ischemia and trauma. In: Tisherman S.A. and Sterz F. (Eds.), Therapeutic Hypothermia. Springer, New York, pp. 101–118. Guest, J.D., Hiester, E.D. and Bunge, R.P. (2005) Demyelination and Schwann cell responses adjacent to injury epicenter cavities following chronic human spinal cord injury. Exp. Neurol., 192: 384–393. Hagg, T. and Oudega, M. (2006) Degenerative and spontaneous regenerative processes after spinal cord injury. J. Neurotrauma, 23: 264–280. Hains, B.C., Willis, W.D. and Hulsebosch, C.E. (2003) Temporal plasticity of dorsal horn somatosensory neurons after acute and chronic spinal cord hemisection in rat. Brain Res., 970: 238–241. Hamberger, A., Huang, Y.L., Zhu, H., Bao, F., Ding, M., Blennow, K., Olsson, A., Hansson, H.A., Viano, D. and Haglid, K.G. (2003) Redistribution of neuroﬁlaments and accumulation of beta-amyloid protein after brain injury by rotational acceleration of the head. J. Neurotrauma, 20: 169–178. Hayashi, N., Bullock, R., Dietrich, W.D. and Maekawa, T. (2004) Hypothermia for Acute Brain Damage. Springer, New York. Hill, C.E., Beattie, M.S. and Bresnahan, J.C. (2001) Degeneration and sprouting of identiﬁed descending supraspinal axons after contusive spinal cord injury in the rat. Exp. Neurol., 171: 153–169.

139 Hoffman, L.R. and Field-Fote, E.C. (2007) Cortical reorganization following bimanual training and somatosensory stimulation in cervical spinal cord injury: a case report. Phys. Ther., 87: 208–223. Holmin, S., Schalling, M., Hojeberg, B., Nordqvist, A.C., Skeftruna, A.K. and Mathiesen, T. (1997) Delayed cytokine expression in rat brain following experimental contusion. J. Neurosurg., 86: 493–504. Hovda, D.A., Yoshino, A., Kawamata, T., Katayama, Y. and Becker, D.P. (1991) Diffuse prolonged depression of cerebral oxidative metabolism following concussive brain injury in the rat: a cytochrome oxidase histochemistry study. Brain Res., 567: 1–10. Hozler M., for the Hypothermia after Cardiac Arrest Study Group. (2002) Mild therapeutic hypothermia to improve the neurological outcome after cardiac arrest. N. Engl. J. Med., 346: 549–556. Hubscher, C.H. and Johnson, R.D. (2006) Chronic spinal cord injury induced changes in the responses of thalamic neurons. Exp. Neurol., 197: 177–188. Jain, N., Catania, K.C. and Kaas, J.H. (1997) Deactivation and reactivation of somatosensory cortex after dorsal spinal cord injury. Nature, 386: 495–498. Jiang, J.Y., Xu, W., Li, W.P., Gao, G.Y., Bao, Y.H., Liang, Y.M. and Luo, Q.Z. (2006) Effect of long-term mild hypothermia or short-term mild hypothermia on outcome of patients with severe traumatic brain injury. J. Cerebral Blood Flow Metab., 26: 771–776. Karimi-Abdolrezaee, S., Eftekharpour, E., Wang, J., Morshead, C.M. and Fehlings, M.G. (2006) Delayed transplantation of adult neural precursor cells promotes remyelination and functional neurological recovery after spinal cord injury. J. Neurosci., 26: 3377–3389. Keane, R.W., Davis, A.R. and Dietrich, W.D. (2006) Inﬂammatory and apoptotic signaling after spinal cord injury. J. Neurotrauma, 23: 335–344. Keane, R.W., Kraydieh, S., Lotocki, G., Alonso, O.F., Aldana, P. and Dietrich, W.D. (2001a) Apoptotic and antiapoptotic mechanisms after traumatic brain injury. J. Cereb. Blood Flow Metab., 21: 1189–1198. Keane, R.W., Kraydieh, S., Lotocki, G., Bethea, J.R., Krajewski, S., Reed, J.C. and Dietrich, W.D. (2001b) Apoptotic and anti-apoptotic mechanisms following spinal cord injury. J. Neuropathol. Exp. Neurol., 60: 422–429. Keirstead, H.S., Nistor, G., Bernal, G., Totoiu, M.O., Cloutier, F., Sharp, K. and Steward, O. (2005) Human embryonic stem cell-derived oligodendrocyte progenitor cell transplants remyelinate and restore locomotion after spinal cord injury. J. Neurosci., 25: 4694–4705. Kim, B.G., Dai, H.N., McAtee, M., Vicini, S. and Bregman, B.S. (2006) Remodeling of synaptic structures in the motor cortex following spinal cord injury. Exp Neurol., 198: 401–415. Kimbelberg, H.K. and Norenberg, M.D. (1994) Astroglial response to CNS trauma. In: Salzman S.K. and Faden A.I. (Eds.), The Neurobiology of Central Nervous System Trauma. Oxford University Press, New York, pp. 193–208.

Kinoshita, K., Chatzipanteli, K., Vitarbo, E., Truettner, J.S., Alonso, O.F. and Dietrich, W.D. (2002) Interleukin-1beta messenger ribonucleic acid and protein levels after ﬂuidpercussion brain injury in rats: importance of injury severity and brain temperature. Neurosurgery, 51: 195–203. Langlois, J.A., Rutland-Brown, W. and Thomas, K.E. (2000) Traumatic Brain Injury in the United States: Emergency Department Visits, Hospitalizations, and Deaths. Centers for Disease Control and Prevention, National Center for Injury Prevention and Control, Atlanta, GA. Leclercq, P.D., McKenzie, J.E., Graham, D.I. and Gentleman, S.M. (2001) Axonal injury is accentuated in the caudal corpus callosum of head-injured patients. J. Neurotrauma, 18: 1–9. Lewen, A., Li, G.L., Nilsson, P., Olsson, Y. and Hillered, L. (1995) Traumatic brain injury in rat produces changes of beta-amyloid precursor protein immunoreactivity. Neuroreport, 6: 357–360. Li, S., Jiang, Q. and Stys, P.K. (2000) Important role of reverse Na(+)-Ca(2+) exchange in spinal cord white matter injury at physiological temperature. J. Neurophysiol., 84: 1116–1119. LoPachin, R.M. and Lehning, E.J. (1997) Mechanism of calcium entry during axon injury and degeneration. Toxicol. Appl. Pharmacol., 143: 233–244. Lotocki, G., Alonso, O.F., Dietrich, W.D. and Keane, R.W. (2004) TNF receptor 1 and its signaling intermediates are recruited to lipid rafts in the traumatized brain. J. Neurosci., 24: 11010–11016. Lotze, M., Laubis-Herrmann, U. and Topka, H. (2006) Combination of TMS and fMRI reveals a speciﬁc pattern of reorganization in M1 in patients after complete spinal cord injury. Restor. Neurol. Neurosci., 24: 97–107. Lu, J., Moochhala, S., Shirhan, M., Ng, K.C., Tan, M.H., Teo, A.L. and Ling, E.A. (2003) Nitric oxide induces macrophage apoptosis following traumatic brain injury in rats. Neurosci. Lett., 339: 147–150. McDonald, J.W., Becker, D., Sadowsky, C.L., Jane, Sr., J.A., Conturo, T.E., Schultz, L.M. (2002) Late recovery following spinal cord injury. Case report and review of the literature. J. Neurosurg., 97(Suppl 2): 252–265 Medana, I.M. and Esiri, M.M. (2003) Axonal damage: a key predictor of outcome in human CNS diseases. Brain, 126: 515–530. Morganti-Kossman, M.C., Lenzlinger, P.M., Hans, V., Stahel, P., Csuka, E., Ammann, E., Stocker, R., Trentz, O. and Kossmann, T. (1997) Production of cytokines following brain injury: beneﬁcial and deleterious for the damaged tissue. Mol. Psychiatry, 2: 133–136. Morganti-Kossmann, M.C., Rancan, M., Stahel, P.F. and Kossmann, T. (2002) Inﬂammatory response in acute traumatic brain injury: a double-edged sword. Curr. Opin. Crit. Care, 8: 101–105. van der Naalt, J., Hew, J.M., van Zomeren, A.H., Sluiter, W.J. and Minderhoud, J.M. (1999) Computed tomography and magnetic resonance imaging in mild to moderate head injury: early and late imaging related to outcome. Ann. Neurol., 46: 70–78.

140 Narayan, R.K., Michel, M.E., Ansell, B., Baethmann, A., Biegon, A., Bracken, M.B., Bullock, M.R., Choi, S.C., Clifton, G.L., Contant, C.F., Coplin, W.M., Dietrich, W.D., Ghajar, J., Grady, S.M., Grossman, R.G., Hall, E.D., Heetderks, W., Hovda, D.A., Jallo, J., Katz, R.L., Knoller, N., Kochanek, P.M., Maas, A.I., Majde, J., Marion, D.W., Marmarou, A., Marshall, L.F., McIntosh, T.K., Miller, E., Mohberg, N., Muizelaar, J.P., Pitts, L.H., Quinn, P., Riesenfeld, G., Robertson, C.S., Strauss, K.I., Teasdale, G., Temkin, N., Tuma, R., Wade, C., Walker, M.D., Weinrich, M., Whyte, J., Wilberger, J., Young, A.B. and Yurkewics, L. (2002) Clinical trials in head injury. J. Neurotrauma, 19: 503–557. Nashmi, R. and Fehlings, M.G. (2001) Mechanisms of axonal dysfunction after spinal cord injury: with an emphasis on the role of voltage-gated potassium channels. Brain Res. Brain Res. Rev., 38: 65–191. Nonaka, M., Chen, X.H., Pierce, J.E., Leoni, M.J., McIntosh, T.K., Wolf, J.A. and Smith, D.H. (1999) Prolonged activation of NF-kappaB following traumatic brain injury in rats. J. Neurotrauma, 16: 1023–1034. Norenberg, M.D., Smith, J. and Marcillo, A. (2004) The pathology of human spinal cord injury: deﬁning the problems. J. Neurotrauma, 21: 429–440. Ozawa, H., Keane, R.W., Marcillo, A.E., Diaz, P.H. and Dietrich, W.D. (2002) Therapeutic strategies targeting caspase inhibition following spinal cord injury in rats. Exp. Neurol., 177: 306–313. Pettus, E.H. and Povlishock, J.T. (1996) Characterization of a distinct set of intra-axonal ultrastructural changes associated with traumatically induced alteration in axolemmal permeability. Brain Res., 722: 1–11. Pierce, J.E., Smith, D.H., Trojanowski, J.Q. and McIntosh, T.K. (1998) Enduring cognitive, neurobehavioral and histopathological changes persist for up to one year following severe experimental brain injury in rats. Neuroscience, 87: 359–369. Popovich, P.G., Guan, Z., McGaughy, V., Fisher, L., Hickey, W.F. and Basso, D.M. (2002) The neuropathological and behavioral consequences of intraspinal microglial/macrophage activation. J. Neuropathol. Exp. Neurol., 61: 623–633. Povlishock, J.T. and Christman, C.W. (1995) The pathobiology of traumatically induced axonal injury in animals and humans: a review of current thoughts. J. Neurotrauma, 12: 555–564. Povlishock, J.T. and Pettus, E.H. (1996) Traumatically induced axonal damage: evidence for enduring changes in axolemmal permeability with associated cytoskeletal change. Acta Neurochir. Suppl., 66: 81–86. Quencer, R.M. and Bunge, R.P. (1996) The injured spinal cord: imaging, histopathologic clinical correlates, and basic science approaches to enhancing neural function after spinal cord injury. Spine, 21: 2064–2066. Raineteau, O. and Schwab, M.E. (2001) Plasticity of motor systems after incomplete spinal cord injury. Nat. Rev. Neurosci., 2: 263–273.

Raineteau, O., Fouad, K., Noth, P., Thallmair, M. and Schwab, M.E. (2001) Functional switch between motor tracts in the presence of the mAb IN-1 in the adult rat. Proc. Natl. Acad. Sci. USA, 98: 6929–6934. Rancan, M., Bye, N., Otto, V.I., Trentz, O., Kossmann, T., Frentzel, S. and Morganti-Kossmann, M.C. (2004) The chemokine fractalkine in patients with severe traumatic brain injury and a mouse model of closed head injury. J. Cereb. Blood Flow Metab., 24: 1110–1118. Ray, S.K., Hogan, E.L. and Banik, N.L. (2003) Calpain in the pathophysiology of spinal cord injury: neuroprotection with calpain inhibitors. Brain Res. Rev., 42: 169–185. Rink, A., Fung, K.M., Trojanowski, J.Q., Lee, V.M., Neugebauer, E. and McIntosh, T.K. (1995) Evidence of apoptotis cell death after experimental traumatic brain injury. Am. J. Pathol., 147: 1575–1583. Rodriguez-Paez, A.C., Brunschwig, J.P. and Bramlett, H.M. (2005) Light and electron microscopic assessment of progressive atrophy following moderate traumatic brain injury in the rat. Acta Neuropathol., 109: 603–616. Rosenberg, L.J. and Wrathall, J.R. (1997) Quantitative analysis of acute axonal pathology in experimental spinal cord contusion. J. Neurotrauma, 14: 823–838. Rothwell, N.J. and Luheshi, G.N. (2000) Interleukin 1 in the brain: biology, pathology and therapeutic target. TINS, 23: 618–625. Sabbah, P., De, S.S., Leveque, C., Gay, S., Pfefer, F., Nioche, C., Sarrazin, J.L., Barouti, H., Tadie, M. and Cordoliani, Y.S. (2002) Sensorimotor cortical activity in patients with complete spinal cord injury: a functional magnetic resonance imaging study. J. Neurotrauma, 19: 53–60. Schouten, J.W., Fulp, C.T., Royo, N.C., Saatman, K.E., Watson, D.J., Snyder, E.Y., Trojanowski, J.Q., Prockop, D.J., Maas, A.I. and McIntosh, T.K. (2004) A review and rationale for the use of cellular transplantation as a therapeutic strategy for traumatic brain injury. J. Neurotrauma, 21: 1501–1538. Schwab, M.E. and Bartholdi, D. (1996) Degeneration and regeneration of axons in the lesioned spinal cord. Physiol. Rev., 76: 319–370. Schwab, J.M., Brechtel, K., Mueller, C.A., Failli, V., Kaps, H.P., Tuli, S.K. and Schluesener, H.J. (2006) Experimental strategies to promote spinal cord regeneration — an integrative perspective. Prog. Neurobiol., 78: 91–116. Schwab, J.M., Seid, K. and Schluesener, H.J. (2001) Traumatic brain injury induces prolonged accumulation of cyclooxygenase-1 expressing microglia/brain macrophages in rats. J. Neurotrauma, 18: 881–890. Sherriff, F.E., Bridges, L.R. and Sivaloganathan, S. (1994) Early detection of axonal injury after human head trauma using immunocytochemistry for beta-amyloid precursor protein. Acta Neuropathol. (Berl.), 87: 55–62. Shohami, E., Novikov, M., Bass, R., Yamin, A. and Gallily, R. (1994) Closed head injury triggers early production of TNF alpha and IL-6 by brain tissue. J. Cereb. Blood Flow Metab., 14: 615–619.

141 Smith, D.H., Chen, X.H., Pierce, J.E., Wolf, J.A., Trojanowski, J.Q., Graham, D.I. and McIntosh, T.K. (1997) Progressive atrophy and neuron death for one year following brain trauma in the rat. J. Neurotrauma, 14: 715–727. Snow, D.M., Lemmon, V., Carrino, D.A., Caplan, A.I. and Silver, J. (1990) Sulfated proteoglycans in astroglial barriers inhibit neurite outgrowth in vitro. Exp. Neurol., 109: 111–130. Springer, J.E., Azbill, R.D. and Knapp, P.E. (1999) Activation of the caspase-3 apoptotic cascade in traumatic spinal cord injury. Nat. Med., 5: 943–946. Thuret, S., Moon, L.D. and Gage, F.H. (2006) Therapeutic interventions after spinal cord injury. Nat. Rev. Neurosci., 7: 628–643. Topka, H., Cohen, L.G., Cole, R.A. and Hallet, M. (1991) Reorganization of corticospinal pathways following spinal cord injury. Neurology, 41: 1276–1283. Totoiu, M.O. and Keirstead, H.S. (2005) Spinal cord injury is accompanied by chronic progressive demyelination. J. Comp. Neurol., 486: 373–383. Toulmond, S. and Rothwell, N.J. (1995) Interleukin-1 receptor antagonist inhibits neuronal damage caused by ﬂuid percussion injury in the rat. Brain Res., 671: 261–266.

Wallace, M.C., Tator, C.H. and Lewis, A.J. (1987) Chronic regenerative changes in the spinal cord after cord compression injury in rats. Surg. Neurol., 27: 209–219. Waxman, S.G. (1989) Demyelination in spinal cord injury. J. Neurol. Sci., 91: 1–14. Waxman, S.G., Utzschneider, D.A. and Kocsis, J.D. (1994) Enhancement of action potential conduction following demyelination: experimental approaches to restoration of function in multiple sclerosis and spinal cord injury. Prog. Brain Res., 100: 233–243. Williams, S., Raghupathi, R., MacKinnon, M.A., McIntosh, T.K., Saatman, K.E. and Graham, D.I. (2001) In situ DNA fragmentation occurs in white matter up to 12 months after head injury in man. Acta Neuropathol. (Berl.), 102: 581–590. Yu, C.G., Jimenez, O., Marcillo, A.E., Weider, B., Bangerter, K., Dietrich, W.D., Castro, S. and Yezierski, R.P. (2000) Beneﬁcial effects of modest systemic hypothermia on locomotor function and histopathological damage following contusion-induced spinal cord injury in rats. J. Neurosurg., 93(Suppl 1): 85–93.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 10

Injury-induced alterations in CNS electrophysiology $

Akiva S. Cohen1,, Bryan J. Pﬁster2, Elizabeth Schwarzbach3, M. Sean Grady4, Paulette B. Goforth5 and Leslie S. Satin5 1

Department of Pediatrics, University of Pennsylvania, School of Medicine and Division of Neurology, Children’s Hospital of Philadelphia, Philadelphia, PA, USA 2 Department of Biomedical Engineering, New Jersey Institute of Technology, Newark, NJ, USA 3 Department of Pharmacology, University of Pennsylvania, School of Medicine, Philadelphia, PA, USA 4 Department of Neurosurgery, University of Pennsylvania, School of Medicine, Philadelphia, PA, USA 5 Department of Pharmacology and Toxicology, VCU School of Medicine, Virginia Commonwealth University, Richmond, VA, USA

Abstract: Mild to moderate cases of traumatic brain injury (TBI) are very common, but are not always associated with the overt pathophysiogical changes seen following severe trauma. While neuronal death has been considered to be a major factor, the pervasive memory, cognitive and motor function deﬁcits suffered by many mild TBI patients do not always correlate with cell loss. Therefore, we assert that functional impairment may result from alterations in surviving neurons. Current research has begun to explore CNS synaptic circuits after traumatic injury. Here we review signiﬁcant ﬁndings made using in vivo and in vitro models of TBI that provide mechanistic insight into injury-induced alterations in synaptic electrophysiology. In the hippocampus, research now suggests that TBI regionally alters the delicate balance between excitatory and inhibitory neurotransmission in surviving neurons, disrupting the normal functioning of synaptic circuits. In another approach, a simpliﬁed model of neuronal stretch injury in vitro, has been used to directly explore how injury impacts the physiology and cell biology of neurons in the absence of alterations in blood ﬂow, blood brain barrier integrity, or oxygenation associated with in vivo models of brain injury. This chapter discusses how these two models alter excitatory and inhibitory synaptic transmission at the receptor, cellular and circuit levels and how these alterations contribute to cognitive impairment and a reduction in seizure threshold associated with human concussive brain injury. Keywords: TBI; electrophysiology; hippocampus; cortical neurons; excitation; inhibition; stretch injury; synapse upon healthcare systems worldwide. Many who survive the initial traumatic incident endure continuing problems with memory, cognition and motor function that are debilitating to their daily lives. While TBI has been extensively studied from both a clinical and a basic science standpoint, many of these studies are descriptive rather than mechanistic, often focusing on the behavioral

Introduction Traumatic brain injury (TBI) is a major public health problem, which has had a signiﬁcant impact $

Supported by NIH-NINDS RO1 NS 45975 (ASC), Commonwealth Neurotrauma Initiative Fund CNI 04-094 (LSS).

Corresponding author. Tel.: +1-215-590-1472;

E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61010-8

143

144

consequences and neuronal cell loss associated with severe TBI in human patients or animal models. However, the extensive and widespread primary neuronal destruction seen in severe cases of TBI is generally irreversible and thus, in all likelihood, therapeutically intractable. Mild and moderate TBI, on the other hand, is much more commonly encountered than severe cases. From the authors’ standpoint, the pervasive deﬁcits associated with mild injury do not necessarily require the frank loss of neuronal or glial cells in the CNS, but result from the disruption of the synaptic function of surviving neurons, negatively impacting the brain function of TBI patients. It is our contention that understanding how the delicate balance of excitation and inhibition in CNS synaptic circuits is disturbed after TBI will help explain and ultimately treat disturbances of cognition and memory in mildly and moderately injured patients. In contrast to the large number of histological, behavioral or biochemical studies that have appeared in the literature to date, there have been relatively few studies of CNS electrophysiology after trauma. The aim of the present review, is to emphasize and discuss the signiﬁcant progress made recently using in vivo and in vitro models of TBI. Furthermore, to understand the mechanisms of brain injury at the level of synaptic electrophysiology, we highlight the need for more mechanistically designed studies of synaptic electrophysiology in the future. The chapter is divided into two major sections. The ﬁrst part is concerned with how TBI results in speciﬁc changes in neuronal function that disrupt the function of synaptic circuits in the brain, affecting the balance between excitatory and inhibitory neurotransmission. Here we focus on studies of hippocampal slices from rodents subjected to lateral ﬂuid percussion injury (LFPI) in vivo, a widely used experimental model of TBI (Dixon et al., 1987; McIntosh et al., 1989; Thompson et al., 2005). In the second part, we discuss studies obtained using a simpliﬁed model of TBI, where cortical pyramidal neurons, cocultured with glia, are subjected to a deﬁned degree of tensile stretch in vitro. Use of this model has allowed perturbations to be delivered in the absence of changes in blood ﬂow, blood brain barrier integrity or oxygenation, and facilitate

understanding how injury impacts the physiology and cell biology of neurons and neural circuits. We believe that comparing and contrasting results obtained using these two very different models is productive and leads to a deeper understanding of the mechanisms evoked by physical trauma. In the last section, we summarize what has been learned to date, and suggest some directions for future research.

In vivo models of TBI Fluid percussion injury LFPI is a reliable and reproducible experimental model of concussive TBI implemented in rodents. LFPI is widely used since it reproduces many pathophysiological features of mild to moderate human TBI including blood brain barrier breakdown, neuronal cell loss, gliosis, perturbation of ionic homeostasis and long-term alterations in cognitive and motor function (Dixon et al., 1987; McIntosh et al., 1989; Thompson et al., 2005). In this model, a pendulum strikes a piston sending a brief (15 ms) percussion pulse of saline into the extradural space of the closed cranial cavity causing a brief displacement and deformation of the brain mimicking concussive injury. The severity of injury is set by adjusting the height of the pendulum. LFPI results in both focal as well as diffuse brain injury consisting of a cortical contusion at the impact site and cell death and axonal damage in the hippocampus, thalamus and cortex (Dixon et al., 1987; McIntosh et al., 1989; Carbonell et al., 1998; Thompson et al., 2005). LFPI has been extensively characterized and widely used in rats as a model, but has recently been adapted for use in mice (Carbonell et al., 1998; Witgen et al., 2005). Importantly, the use of mice to study TBI will allow investigators to exploit transgenic and knockout mouse technology to directly test the role of speciﬁc genes or gene mutations on TBI, as well as lowering the costs associated with the use of rats. In this review chapter, we will include data obtained using ﬂuid percussion to injure mice to demonstrate that these results are consistent with the rat LFPI model.

145

Experimental LFPI and the hippocampus In humans, TBI results in heterogeneous changes in brain function, including cognitive impairments in learning and memory. Damage to the hippocampus associated with signiﬁcant impairment of learning and memory was ﬁrst recognized in studies of bilateral damage to the hippocampus in humans (Rempel-Clower et al., 1996; Deweer et al., 2001). Subsequently, a large body of knowledge has accumulated indicating that, while not exclusively responsible for memory function, the hippocampus plays a central role in memory consolidation and retrieval (Cave and Squire, 1991; Miller et al., 1998; Bohbot et al., 2000). Moreover, the hippocampus is frequently damaged in TBI patients, thus likely contributing to observed cognitive deﬁcits including impaired learning and memory (Kotapka et al., 1992, 1993, 1994; Bigler et al., 1997). The dependence of memory acquisition on the hippocampus is similar in humans and rats, which both show selective loss of recently acquired memory and the failure to imprint new memories after hippocampal damage (He et al., 2001; Liu and Bilkey, 2001). In rodents, LFPI produces a consistent injury to the hippocampus including neuronal loss within all hippocampal subregions as well as physiologic, ionic and neurochemical changes in the dentate hilus and areas CA1 and CA3. This damage has been correlated to visuo-spatial memory deﬁcits observed in the Morris water maze assay (Smith et al., 1991; Gorman et al., 1993) and to hippocampaldependent cognitive impairment demonstrated by changes in the conditioned fear response (Hogg et al., 1998a, b; Witgen et al., 2005). Accordingly, LFPI in the rat and mouse is currently the most widely accepted experimental model to mimic hippocampal injury.

Functional electrophysiological changes in the hippocampus after TBI Normal hippocampal function is directly determined by a delicate balance between neuronal excitation and inhibition. Perturbation of this state of equilibrium can have catastrophic consequences,

including impaired cognitive function and the development of seizures (Cave and Squire, 1991; Annegers et al., 1996; Asikainen et al., 1999). Within the hippocampus there are three interconnected subregions referred to collectively as the ‘Trisynaptic Circuit’. This circuit consists of the dentate gyrus (DG), area CA3, and area CA1, each having distinct physiological roles (see Fig. 1). The DG ‘‘ﬁlters’’ out aberrant or excessive input to the hippocampus from propagating further along the hippocampal circuit. Area CA3 ‘‘ampliﬁes’’ the activity that the DG has allowed to enter, while area CA1 ‘‘transduces’’ processed hippocampal information outﬂow to the cerebral cortex. It is believed that the DG acts as a ﬁlter because it is itself extraordinarily resistant to the generation of synchronized bursting that is characteristic of epileptic seizures (Heinemann et al., 1992; Lothman et al., 1992). This dampening of synaptic input may thus be critical in protecting the fragile neurons within the hippocampus from excessive activation and seizure formation. The ﬁltering or gatekeeper behavior of the DG is due to three main mechanisms: (1) the intrinsic membrane properties of dentate granule cells (DGCs) tend to resist excessive excitation (Fricke and Prince, 1984), (2) at the circuit level, the dearth of interconnections between DGCs impedes synchronization since neurons ﬁring inappropriately cannot easily transmit this overexcitation to their neighbors (Claiborne et al., 1986) and (3) the efﬁcacy of DG-mediated ﬁltering is further enhanced by robust surround inhibition of the DG (Sloviter, 1994). In area CA3, pyramidal neurons are highly interconnected to each other via recurrent axon collaterals (Dudek et al., 1986; Wong et al., 1986), providing the anatomical substrate for efﬁcacious synchronization and signal ampliﬁcation. Thus, area CA3 is highly susceptible to recurrent excitation and is consequently prone to seizure generation and injury-related pathology (i.e., if the DG does not ﬁlter excessive input). This activity is in turn transmitted via CA3 axons (Schaffer collaterals) to area CA1, which excite both pyramidal neurons as well as inhibitory interneurons, thereby regulating both inhibition and excitation in area CA1.

146

Fig. 1. Illustration of the hippocampal ‘Trisynaptic circuit’. This circuit consists of three interconnected subregions: the dentate gyrus (DG), area CA3 and area CA1. The DG is thought to ﬁlter aberrant or excessive input from the entorhinal cortex via the perforant pathway. Filtered activity is relayed to area CA3 via the mossy ﬁbers where it is ampliﬁed and sent onward to area CA1 via the Schaffer collaterals where the information is then transduced out to the cerebral cortex.

The role of GABAergic inhibition in the hippocampus Pathological alterations in the effectiveness of GABAA-mediated inhibition are implicated in many neurological disease processes, including TBI (Reeves et al., 1995; Toth et al., 1997; Santhakumar et al., 2001). GABA is the principal inhibitory neurotransmitter in the mammalian brain and not only controls the degree of neuronal activation but, importantly, is also critical in

initiating, synchronizing and terminating both normal and pathological neuronal network activity in the central nervous system. A majority of inhibitory synapses are localized on neuronal somata and axon initial segments, since synaptic input to these regions are highly efﬁcacious in regulating neuronal output. Synaptic localization, coupled with the heterogeneous patterns of interneuronal innervation of excitatory (principal) neurons (Freund and Buzsaki, 1996), endows individual GABAAergic interneurons with the

147

capability of synchronizing the activity of hundreds to thousands of principal neurons within the hippocampus (Cobb et al., 1995). The synchronization of neocortical structures is known to subserve important functions, such as the processing of sensory information (Lytton and Sejnowski, 1991; Singer and Gray, 1995) and memory consolidation and formation (Joliot et al., 1994). Therefore, oscillations which are synchronized by GABAAergic inhibition can be viewed as being critical for normal hippocampal function. GABAA receptors are pentameric structures composed of several related subunit families. At present, eight different GABAA receptor subunit families have been cloned (with many families having multiple members), including a (1–6), b (1–4), g (1–3), d, r(1–3), e, p and y (Barnard et al., 1998). The subunit stoichiometry is considered to be a primary determinant of the pharmacology of the receptor. Although subunit composition is modulable under different developmental and pathological conditions, to date no injury-induced alterations in GABAA receptor subunit composition have been reported. Surprisingly, there have been few studies exploring functional changes in hippocampal circuitry following TBI. Instead, most studies to date have attempted to correlate regional hippocampal cell loss with behavioral changes induced by TBI, despite the gap between these two divergent endpoints. While hippocampal cell loss is often hypothesized to alter the balance between neuronal excitation and inhibition which is necessary for proper hippocampal function, iterations in function do not always correlate with the degree of observed cell death (Lyeth et al., 1990). Recent studies have supported the alternative hypothesis that alterations in the behavior of surviving neurons may be a key. Speciﬁcally, a few experimental studies using LFPI indicate that TBI affects excitatory and inhibitory synaptic transmission, giving rise to dysfunctional hippocampal circuits in rodents despite the absence of signiﬁcant cell death (Reeves et al., 1995; Toth et al., 1997; D’Ambrosio et al., 1998; Witgen et al., 2005). However, the precise cellular mechanism(s) responsible for the TBI-induced pathological alterations in hippocampal function currently remain unknown.

Injury-induced functional alterations in area CA1 Area CA1: injury-induced alterations in hippocampal area CA1 excitability To evaluate hippocampal function after injury, several groups have speciﬁcally investigated changes in area CA1 excitability. Electrophysiological studies on hippocampal slices from injured rodents using LFPI generally show a decrease in excitability in hippocampal area CA1. D’Ambrosio et al. (1998) initially found, 2 days post-LFPI, that increased simulation intensity is required to overcome the threshold of postsynaptic population spikes (PS) as well as ﬁeld extracellular postsynaptic potentials (fEPSP). In addition, the potentials exhibit smaller amplitudes and decreased linear slopes after injury. These results suggest a decrease in area CA1 excitability, which was further corroborated and expanded upon using mice 1 week after LFPI (Witgen et al., 2005). Input/output (I/O) curves, relating the stimulus intensity as a function of the linear slope of the fEPSP are shifted downward after injury compared with curves obtained from naı¨ ve/sham mice, Fig. 2 (Witgen et al., 2005). In addition, the threshold for evoking PS in slices from injured animals is correspondingly increased. The observed shifts in the I/O curve and the PS threshold denote a decrease in excitability, and are indicative of diminished synaptic communication between pre and postsynaptic neurons. While these studies in hippocampal slices establish reduced excitability in area CA1 after LFPI, earlier in vivo studies reported increased excitability during the acute period of 2 h–2 days following injury. Stimulation thresholds for the generation of PS in vivo decreases within 2–3 h after LFPI (Miyazaki et al., 1992), and the decrease persists for 2 days, returning to control levels by 7 days after LFPI (Reeves et al., 1995, 1997b). Although lowered thresholds suggest increased excitability, the amplitudes of the PS and fEPSPs are reduced compared with controls. The discrepant nature of these results could be explained by differences in injury severity in the respective studies, the use of in vivo vs. in vitro slice preparations, differences in the recording techniques used, or the difference in post-injury time points chosen. An additional

148

Fig. 2. Extracellular input/output curves for area CA1 evoked by afferent Schaffer collateral stimulation. LFPI in mice results in smaller fEPSP slopes compared with those from sham animals. Reprinted from Neuroscience, Witgen et al. (2005), with permission from Elsevier.

report recently suggested increased excitability of area CA1 at 1 week after LFPI (Akasu et al., 2002).

Area CA1: Glutamate evoked currents Alterations in excitatory synaptic transmission are one potential mechanism underlying the injuryinduced shifts in area CA1 excitability seen following TBI. Supporting the hypothesis that excitatory neurotransmission is decreased after injury, a variety of immunocytochemical studies of the hippocampus indicate that the number of N-methyl-D-aspartate (NMDA) receptors decreases within 12 –24 h post-injury, with no associated changes in non-NMDA receptors such as AMPA or kainate receptors (Miller et al., 1990; Sihver et al., 2001; Osteen et al., 2004). The NMDA receptor is important for the induction of excitatory synaptic plasticity (see below). Therefore, to examine whether injury causes NMDA receptor dysfunction, extracellular recording techniques were used to measure evoked

NMDA-mediated depolarizing potentials (AMPAand GABAergic activity were pharmacologically blocked) (Cohen and Abraham, 1996) in brain slices from sham vs. LFPI mice, Fig. 3 (Schwarzbach et al., 2006). These data show that in mild to moderately injured mice, the amplitude of NMDA potentials (Figs. 3B, C) is signiﬁcantly smaller than in slices from sham animals (Figs. 3A, C). To evaluate NMDA receptor alterations in individual hippocampal neurons, whole-cell patch voltageclamp recordings were performed in visually identiﬁed pyramidal neurons in hippocampal slices (Schwarzbach et al., 2006). Focal application of glutamate results in signiﬁcantly smaller NMDA isolated excitatory current (sodium, AMPA and GABA currents blocked) in slices from LFPI (Fig. 3E) vs. sham (Fig. 3D) mice. Furthermore, by subtracting the NMDA-mediated currents from the total glutamate-evoked currents, the AMPAergic component could be determined. This demonstrates there is a signiﬁcant reduction in AMPA as well as NMDA currents in LFPI slices compared with sham controls (Fig. 3G).

149

A

B

C

*

FPI

SHAM

D

E

BMI TTX glycine

CNQX

APV

BMI TTX glycine

G

F

0

APV

80

40

20

CNQX

* FPI SHAM

40

0

* FPI SHAM

Fig. 3. Injury diminishes isolated NMDA receptor potentials and EPSCs following injury. (A) Sham; (B) FPI, trace of fEPSP, isolated NMDA potential, recording after addition of APV to ensure the waveform is NMDA dependent. Scale bar: 0.5 mV, 10 ms. (C) Histogram showing the quantiﬁcation of the reduction in the peak amplitude of the NMDA potential recorded in slices from FPI mice compared with sham animals (n ¼ 11 and 7 for FPI and sham, respectively, po0.05, denoted by *). (D) Sham; (E) FPI, glutamatergic currents were recorded by focal application of glutamate (100 mM, 50 ms, 40 psi), the NMDA component was isolated with the addition of CNQX to the bath, and at the end of each experiment, APV was perfused to ensure the current is NMDAdependent. Scale bar: 15 pA, 100 ms. (F) The histogram shows FPI causes a signiﬁcant reduction in the peak amplitude of the NMDA current. (G) The histogram is created by subtracting the isolated NMDA-mediated current from the entire glutamate evoked current to determine the AMPA-mediated current. This shows FPI also causes a signiﬁcant reduction in the peak amplitude of the AMPA currents (* denotes signiﬁcance, po0.05).

Area CA1: changes in inhibitory synaptic transmission While we have presented evidence showing decreases in excitatory synaptic transmission after injury, alterations in overall synaptic excitation may also reﬂect changes in inhibitory circuits as these effects may be additive postsynaptically. Speciﬁcally, interneurons innervate CA1 pyramidal neurons to provide GABAA-mediated inhibition control over the initiation, synchronization and

termination of excitatory neuronal activity. Therefore, augmentation of inhibitory activity could also contribute to decreased area CA1 excitability (Fig. 2). Furthermore, because synaptic plasticity, especially long-term potentiation (LTP), requires a precise functional balance of inhibition and excitation, alterations in either excitation or inhibition may interfere with functional information processing. A role for GABAA alteration as a cause for memory impairments is supported by the ﬁndings that GABAA receptor immunolocalization in area

150

CA1 increases after LFPI (Reeves et al., 1997a, b) and that the administration of GABAA receptor antagonists reduces cognitive deﬁcits in rats after FPI (O’Dell and Hamm, 1995). To test whether LFPI leads to augmented GABAA-mediated inhibition, miniature inhibitory postsynaptic currents (mIPSCs) were recorded in slices from sham vs. injured mice 1 week postinjury. Miniature IPSCs are spontaneously occurring events, reﬂecting the spontaneous release of single packets of GABA from individual presynaptic terminals of GABAergic interneurons. Analysis of mIPSCs provides detailed information about the properties of postsynaptic receptors because mIPSCs result from the activation of individual synapses, in contrast to stimulation studies which involve the activation of tens to

hundreds of synapses (Mody et al., 1994). Speciﬁcally, changes in the number of postsynaptic receptors are generally accepted to cause changes in mIPSC amplitude, while shifts in receptor subunit composition and/or GABA release kinetics alter mIPSC kinetics. In addition, a change in the total number of active synapses (i.e., presynaptic release sites) is reﬂected by variations in mIPSC frequency. The median mIPSC amplitude in CA1 neurons recorded from FPI animals is signiﬁcantly greater than in sham animals (Figs. 4A, B) (Witgen et al., 2005). The signiﬁcant increase in mIPSC amplitude is further demonstrated by calculating total mIPSC charge transfer (i.e., the amount of current ﬂowing during an individual mIPSC and calculated by integrating the total area associated with

Fig. 4. Whole-cell voltage clamp recordings from area CA1 pyramidal neurons from FPI animals demonstrate increases in spontaneous miniature (mIPSC) activity 1 week after LFPI. (A) Cumulative frequency amplitude histogram for neurons from sham and LFPI animals. (B) The mIPSCs are larger in CA1 pyramidal neurons from LFPI mice. Histograms of mean median (C) mIPSC amplitude and (D) net charge transfer demonstrate signiﬁcant increases in inhibitory activity of neurons from LFPI mice compared with control animals. (E) Histogram demonstrating that mIPSC frequency is not signiﬁcantly different in sham compared with LFPI slices. Reprinted from Neuroscience, Witgen et al. (2005), with permission from Elsevier.

151

an mIPSC) from the two populations of neurons. Miniature IPSC charge transfer in neurons in slices from FPI animals is 155% of that calculated for the sham population (Fig. 4D). Interestingly, neither the 50% decay time (T50), weighted decay or the frequency of occurrence of mIPSCs recorded in slices from injured animals are signiﬁcantly different from values obtained in slices from sham animals (Fig. 4E) suggesting that concussive brain injury does not alter the number of active presynaptic inhibitory synapses or GABAA receptor subunit composition. The enhanced action potential-independent spontaneous inhibitory activity in area CA1 following FPI would be expected to contribute to the reduced excitability in this region. Qualitatively similar results were seen in stretch injured cortical neurons where the amplitude of GABAA-mediated currents are directly increased following injury (see below; Kao et al., 2004).

Area CA1: excitatory synaptic plasticity Long-term potentiation Persistent and life-long cognitive impairment is a hallmark of brain injury pathology (Barth et al., 1983; Bennett-Levy, 1984; Lyeth et al., 1990; McAllister, 1992; McAllister et al., 1999) and even mild TBI can damage the ﬁne structure of the hippocampus (Kotapka et al., 1991; Lowenstein et al., 1992; Grady et al., 2003), contributing to a reduction in information processing (Mathias et al., 2004) and learning and memory impairments. LTP, a form of activity-dependent plasticity, is currently our best physiological correlate of learning and memory. LTP can be experimentally induced in area CA1 and is characterized by a long-lasting potentiation of fEPSP slope above baseline in response to tetanic stimulation of the Schaffer collaterals. Several rodent LFPI models report a consistent inability to induce and maintain LTP in area CA1 in the ipsilateral hippocampus in vivo (Miyazaki et al., 1992; Reeves et al., 1995, 1997b) and in vitro beginning at 2 h and persisting to 8 weeks post-injury (D’Ambrosio et al., 1998; Sick et al., 1998; Sanders et al., 2000; Schwarzbach et al., 2006) and may contribute to learning and

memory deﬁcits observed following TBI (Lyeth et al., 1990; Smith et al., 1991; Gorman et al., 1993). The complement of LTP is long-term depression (LTD), which is a decrease in net synaptic efﬁcacy seen after low frequency stimulation (Christie et al., 1994; Bear and Abraham, 1996). Unlike LTP, this form of synaptic plasticity can be generated in slices from LFPI mice and rats (D’Ambrosio et al., 1998; Schwarzbach et al., 2006), demonstrating the ability of neurons to express some forms of synaptic plasticity even after LFPI.

Mechanisms of LTP The progression of LTP under physiological conditions consists of pre to postsynaptic induction, postsynaptic expression and subsequent long-term expression and maintenance. Results reported in the early literature were mixed and often discrepant as to the mechanisms involved in injuryinduced loss of LTP. Sanders et al. (2000) suggests that while there are major deﬁcits in the maintenance phase, measurements at 5 min show no impairment in potentiation and that TBI may not affect the mechanisms essential to LTP induction and expression. This is the only study, however, that does not show a depression of the fEPSP immediately after injury. Post-tetanic potentiation (PTP), an initial component of LTP also cannot be observed after LFPI (Sick et al., 1998; Schwarzbach et al., 2006). Since, PTP is believed to be due to residual presynaptic calcium buildup resulting in increased neurotransmitter release (Del Castillo and Katz, 1954; Martin and Pilar, 1964; Barrett and Stevens, 1972; Hirst et al., 1981), this led to the hypothesis that injury is affecting LTP via action on presynaptic mechanisms. To determine if there are alterations in presynaptic transmitter release following TBI, paired pulse studies (two identical stimulus pulse separated by less than 100 ms) were compared in sham vs. injured slices. Paired pulse facilitation (PPF) in area CA1 results in a modest augmentation of the second fEPSP from the excess calcium in the presynaptic terminal, whereas impaired PPF would denote alterations in release probability

152

have occurred. Reeves et al. (2000) demonstrated that the second response is enhanced within 1 h after LFPI, returning to control levels by 7 days after injury. Schwarzbach et al. (2006), however, using the same interpulse interval, did not observe differences in PPF between sham vs. injured 7 days after injury. Methodological or different recording conﬁgurations may underlie this discrepancy. Synaptic plasticity, like proper synaptic function in general, requires a balance between inhibition

and excitation (Liu et al., 2004). As previously reviewed, injury leads to a downward amplitude shift in the area CA1 I/O curve (see Fig. 2), demonstrating decreased excitability due to augmented inhibition (Witgen et al., 2005). Thus, an initial hypothesis postulates that augmented inhibition is the reason why LTP cannot be induced after TBI in area CA1. However, fully blocking inhibition with the GABAA antagonist bicuculline methiodide does not rescue LTP (Fig. 5C) refuting

Fig. 5. Inability to induce area CA1 LTP but not LTD following lateral LFPI in injured mice (open circles) compared with sham (ﬁlled circles). (A) High frequency tetanic stimulation (arrows) results in facilitation of the fEPSP in hippocampal slices from sham but not from injured mice. (B) Low frequency stimulation resulted in LTD of the fEPSP in slices derived from both injured and sham animals. (C) Histogram demonstrating that LTP can be induced in slices from sham and naı¨ ve animals using either the standard HFS or theta burst stimulation (TBS); however, with both protocols, LTP cannot be induced in slices from LFPI animals. However, LTP can be induced and maintained in area CA1 in the hippocampus contralateral to injury. Area CA1 LTP cannot be rescued by varying the stimulation protocol, altering extracellular Ca2+ and Mg2+ concentrations or bath application of BMI. (D) LTD induced by a LFS (line) in slices from FPI animals does not make subsequent LTP induction by HFS (arrows) possible. Responses were followed for an additional 30 min post-tetanic stimulation, no potentiation (even a return to baseline) was observed. Reprinted from Hippocampus, Schwarzbach et al. (2006), with permission from Wiley-Liss, Inc.

153

the hypothesis that overwhelming inhibition alone underlies dysfunctional LTP (Schwarzbach et al., 2006). Data suggesting that LTD can be induced in slices from injured animals show that plasticity is possible in mice that have undergone FPI. D’Ambrosio et al. (1998) suggested that it is not alterations in the induction mechanisms which induce LTP, but rather that injury itself potentiates transmission leading to a saturation of LTP. Therefore, if injury prepotentiates synapses, tetanic presynaptic stimulation will be unable to further potentiate the fEPSP. To test this hypothesis, slices from injured animals were ﬁrst ‘‘depotentiated’’ (using the same low frequency stimulation protocol used to classically induce LTD) and slices were then subsequently tetanized with a high frequency, presynaptic stimulus in an attempt to induce LTP. However, using this protocol did not result in either LTP or, a return to baseline in slices from FPI animals (Fig. 5D). This suggests that the mechanism(s) speciﬁc for the induction of LTP, and not for synaptic plasticity in general, are in fact altered by injury (Schwarzbach et al., 2006). The anatomical substrate for LTP induction and expression is the dendritic spine (Leuner et al., 2003; Segal, 2005). Anatomical studies using lucifer yellow illustrate that following FPI, the mean apical and basal dendritic spine size (mm2) signiﬁcantly increases (Fig. 6; Schwarzbach et al., 2006). This increased spine size could effect the threshold level of calcium required for LTP induction, which may be kept to a minimum due to small spine size (Malenka and Nicoll, 1999; Gazzaley et al., 2002). It is known from many studies that the calcium and calmodulin-dependent protein kinase a-CaMKII, which is abundant postsynaptically, is critical for the induction of LTP (Malenka et al., 1989). Hence, a functional decrease in a-CaMKII or a decrease in its phosphorylating activity after injury could contribute to an inability to induce LTP. Western blot analysis demonstrates a signiﬁcant decrease in the alpha form of CaMKII protein expression in regionally dissected area CA1 tissue from FPI injured animals (Schwarzbach et al., 2006). However, CaMKII activity has been

shown in another study to increase after FPI in rat brain (Atkins et al., 2006). Thus, Western blot analysis of subcellular fractions from the hippocampus or parietal cortex regions reveal a redistribution of phosphorylated a-CaMKII from the cytosolic to membrane fractions 30 min after moderate lateral FPI (Atkins et al., 2006). Presently more studies are therefore needed to determine whether reduced CaMKII function contributes to injury-induced deﬁcits in LTP expression. The data presented above describe a consistent inability to induce area CA1 LTP following LFPI and explore many possible contributing mechanisms. The work has presented many valuable injury-induced alterations in the cascade leading to increased synaptic efﬁcacy, however the inability to induce and maintain LTP does not seem to be due to one ﬁnite cause.

Injury-induced functional alterations in the DG DG: injury-induced alterations in DG excitability The DG is thought to act as a ﬁlter to dampen synaptic input from the perforant pathway (PP) before it is ampliﬁed by area CA3. The lack of interconnections between DGCs (Claiborne et al., 1986) and robust feed-forward and feed-back inhibition (Sloviter, 1994) function to impede synchronization and endow the DG with its ﬁltering ability. Studies on the electrophysiological properties of the DG have consistently shown that LFPI leads to increased DG excitability in rodents, effectively reducing its ﬁltering capacity. While signiﬁcant cell death has been observed in the hilar region, the preferential survival of excitatory neurons is not found (Lowenstein et al., 1992; Toth et al., 1997; Santhakumar et al., 2000). In injured animals, stimulation of the PP evokes repetitive PS ﬁring (Lowenstein et al., 1992; Santhakumar et al., 2000; Golarai et al., 2001), reduced PS threshold and increased spike amplitude (Toth et al., 1997; Santhakumar et al., 2000, 2001; Witgen et al., 2005) compared with non-injured animals. Studies have also shown that the amplitude and slope of fEPSPs substantially increases in injured animals (Coulter et al., 1996; Golarai et al., 2001;

154

Fig. 6. Increased dendritic spine size in slices from LFPI mice. (A) CA1 pyramidal neuron and its spiny dendrites from an injured mouse loaded with Lucifer yellow. (pcl, pyramidal cell layer; sl, stratum lucidum; so, stratum oriens.) Scale bar: 10 mm. (B) Higher magniﬁcation of area highlighted by box in section A showing fourth order basal dendrite segment spines. Scale bar: 1 mm. (C) Higher magniﬁcation of area highlighted by box in section A showing second order apical dendrite segment spines. Scale bar: 1 mm. (D) Neurolucida drawing of the spiny segment shown in B. Scale bar: 1 mm. (E) Comparison of the distribution of dendritic spine size of sham (black) and FPI (gray) apical (E) and basal (F) CA1 pyramidal cells. Reprinted from Hippocampus, Schwarzbach et al. (2006), with permission from Wiley-Liss, Inc.

155

Fig. 7. Extracellular input/output curves for the dentate gyrus evoked by afferent perforant path stimulation. LFPI in mice results in increased fEPSP slopes compared with those from sham animals. Reprinted from Neuroscience, Witgen et al. (2005), with permission from Elsevier.

Witgen et al., 2005). Consistent with rat models, 1 week after LFPI in the mouse, the I/O curve (measuring the normalized slope of the fEPSP vs. the stimulation intensity), is signiﬁcantly shifted upward with respect to amplitude as compared with sham animals (Fig. 7). This injury-induced shift in the I/O curve suggests that net synaptic efﬁcacy increases in the mouse DG after LFPI. Since the discharge of DGCs is under tight inhibitory control by GABAergic interneurons, it is constructive to completely antagonize inhibition, in order to test whether injury-induced alterations in DG excitability are due to changes in glutamatergic synaptic activity. One study presents data supporting injury-induced alterations in the excitatory limb of the DG (Santhakumar et al., 2000). Using slices from injured rats 1 week post-LFPI, afferent PP stimulation causes DG neurons to respond with protracted depolarizations and augments action potential discharges in the absence of GABAAergic inhibition. The late depolarizations exhibited in DGCs are mediated by a polysynaptic

pathway and could be blocked with AMPA receptor antagonist(s) (Santhakumar et al., 2000). The overall conclusion of this study is that excitatory mossy cells become ‘‘irritable’’ and lead to an increased excitatory drive onto both DG excitatory and inhibitory neurons (Santhakumar et al., 2000). More recently studies have supported this hypothesis showing mossy ﬁber sprouting at 2–3 and 14–15 weeks after a weight drop model of injury (Golarai et al., 2001). In another study, mossy ﬁber sprouting led to increased output from the excitatory principal cells of the DG to other granule cells and interneurons augmenting excitation (Santhakumar et al., 2001). This was based on ﬁndings that injury caused an increase in the evoked PS lasting 1 month post-LFPI and sIPSC from granule cells exhibited a higher frequency through 6 months after LFPI. The cause of this increased sIPSC was demonstrated to be due to the glutamatergic excitatory drive to interneurons.

156

DG: changes in inhibitory synaptic transmission While it has been shown that surviving DGCs sustain TBI-induced increases in excitability, alterations in inhibitory function cannot be ruled out. Many studies demonstrate neuronal loss in the hilar region, however, no preferential survival of inhibitory neurons is observed, including somatostatin, parvalbumin and cholecystokinin positive interneurons (Lowenstein et al., 1992; Toth et al., 1997; Santhakumar et al., 2000). While inhibitory neuronal loss could result in a decrease of GABAergic inhibitory efﬁcacy, functional changes in GABA receptors is also a possibility. Indeed, studies show injury-induced alterations in the inhibitory network of the DG. By 1 week post-injury in the rat, there is a decrease in mIPSCs frequency, limiting GABAergic inhibition without altering mIPSC kinetics or amplitude (Toth et al., 1997). Similar results are found in the mouse but the amplitudes of mIPSCs recorded in slices from injured animals 1 week post-LFPI are smaller than sham animals (Figs. 8A–D), while the kinetics of mIPSCs are not altered. The consequence of the decrease in mIPSC amplitude is demonstrated by a signiﬁcant reduction in total mIPSC net charge transfer in neurons from FPI animals, which was 76% of that calculated for neurons from sham slices (Fig. 8E). Furthermore, the mean frequency of occurrence of mIPSCs is signiﬁcantly lower in neurons from FPI animals than that present in DG neurons in sham slices (Fig. 8F). These alterations in GABAAergic function demonstrate that spontaneous inhibitory activity (known to be mediated by basket and axo-axonic interneurons) regulating feed-forward inhibition on dentate granule neurons is signiﬁcantly compromised 1 week post-LFPI. Interestingly, the frequency of spontaneous action potential-mediated inhibitory postsynaptic currents is enhanced up to 5 months after injury. This increase in spontaneous inhibitory postsynaptic current (sIPSC) frequency is due to enhanced excitatory drive (from irritable mossy cells) onto inhibitory interneurons (Santhakumar et al., 2000). Additionally, LFPI also causes a transient (approximately 10 mV) depolarized shift in DG interneurons exhibited immediately after injury

(1–4 h) and lasting up to 4 days post-LFPI (Ross and Soltesz, 2000). This depolarization is mediated by an injury-induced reduction in the activity of the Na+/K+ pump (ATPase), which contributes to maintaining the resting membrane potential. This 10 mV depolarization transiently raises the excitability of the GABAAergic network in the early post-traumatic DG. TBI effects seizure threshold As previously discussed, the DG’s ﬁltering efﬁcacy is crucial to preventing seizure activity. Thus, a dysfunction in this region is a plausible explanation for the post-traumatic epilepsy observed (Annegers et al., 1996). Santhakumar et al. (2001) show that there is a decrease in the threshold to generate sustained (30+ min) of seizure like activity to tetanic stimulation, however no spontaneous seizure like activity is present in area CA1 at 3 months (after recovery period). Furthermore, slices generated from LFPI animals have a lower threshold for developing self-sustained seizure activity 1 week post-injury (Coulter et al., 1996) and persisting up to 3 months (Santhakumar et al., 2001). Coulter et al. (1996) found greater disinhibition in the DG of slices from FPI animals 1 week after injury when compared with both sham-operated and epileptic animals, which could be a possible explanation for the observed decreased seizure threshold. In addition, using the weight drop model of TBI, Golarai et al. (2001) demonstrate that injured animals have a lower seizure threshold to pentylenetetrazoleinduced convulsions 15 weeks post-TBI (Golarai et al., 2001). Cellular mechanisms of abnormal electrophysiology after trauma: the effect of tensile strain on ionotropic (excitatory and inhibitory) receptors of cortical neurons in an in vitro model of TBI As illustrated above, TBI induces regional alterations in excitatory and inhibitory synaptic function in hippocampal slices from rats and mice subjected to LFPI. Elucidation of the molecular sites and biochemical pathways that mediate synaptic

157

Fig. 8. Whole-cell voltage clamp recordings from dentate granule neurons from injured mice demonstrate reductions in spontaneous miniature (mIPSC) activity 1 week after LFPI. The mIPSCs are smaller and less frequent in dentate granule neurons from LFPI animals. (A) Continuous sweeps of mIPSCs from sham (left panel) and LFPI (right panel) dentate granule cells at 60 mV. (B) Cumulative frequency amplitude histogram for neurons in (A) with representative mIPSCs from sham and injured dentate granule neurons. Histograms of mean median (D) mIPSC amplitude, (E) net charge transfer, and (F) mIPSC frequency demonstrate signiﬁcant reductions in inhibitory activity of neurons from LFPI mice compared with sham animals. Reprinted from Neuroscience, Witgen et al. (2005), with permission from Elsevier.

dysfunction in the CNS would provide favorable targets for putative pharmacological therapeutic treatments directed toward improving cognitive function and memory after TBI. Yet, presently, little is known in detail regarding the cellular mechanisms underlying dysfunctional neurotransmission. For example, it is still unresolved whether

mechanical forces directly impact upon neurons during traumatic impact to trigger secondary consequences of injury or affect neuronal function indirectly by the release of excitatory and other CNS neurotransmitters during the so-called ‘depolarization impact’ phase of trauma (Katayama et al., 1990). Investigators have thus begun to examine

158

the cellular processes that result from mechanical trauma in further detail and more directly by using simpliﬁed in vitro cell injury models. In the earliest in vitro models of mechanical trauma, pyramidal neurons were injured by scratching a cultured neuronal cell layer with a glass sylet (Tecoma et al., 1989), transecting cells in a manner similar to tissue damage incurred during penetrating TBI. Alternative in vitro models have been developed to simulate the forces (tensile, compressional or torsional) acting on brain tissue during blunt trauma or acceleration/deceleration injury, in order to more fully characterize the mechanisms by which mechanical deformation (or tensile strain) leads to changes in neurons and glia (Pike, 2001). Models of tensile strain (stretch) typically use pneumatic or mechanical translational devices to deform a stretchable substrate upon which various types of neurons or brain slices are cultured and are thus adherent (Ellis et al., 1995; Morrison et al., 1998, 2006; Arundine et al., 2003). Shear strain or torque, which approximates the inertial loading caused by the rapid acceleration of brain tissue during TBI, has been modeled by forcing ﬂuid over stationary cells using a cell shearing injury device (CSID) (LaPlaca and Thibault, 1998), or a newly developed electro-mechanical cell shearing model, which can be used to examine the effects of torsional stress applied to three dimensional neuronal cultures (LaPlaca et al., 2005). As simpliﬁed systems, in vitro models allow injury to be studied in the absence of ischemia, blood/brain barrier considerations, or other changes, and the models provide increased experimental control of tissue oxygenation, the cell type to be studied, availability of nutrients, temperature and ionic and drug concentrations. For electrophysiological studies, in vitro models enable rapid drug or agonist application, patch clamping and intracellular calcium measurements to be carried out with relative ease. In vitro models are thus well suited for examining the direct effects of a mechanical insult on ion channel or receptor function. Using the in vitro model of stretch injury developed by Ellis et al. (1995) several studies reveal alterations in the function of neurotransmitter receptors, including AMPA, NMDA and GABAA receptors following mechanical strain (Zhang et al.,

1996; Goforth et al., 1999, 2004; Kao et al., 2004). Ellis et al. (1995) developed an in vitro TBI model in which neonatal rat cortical neurons are cocultured with astrocytes in 6-well plates having deformable SILASTIC (Dow Corning) bottoms (FlexCell Plates; Ellis et al., 1995). An injury controller used in this model applies a controlled pulse of compressed nitrogen to individual wells via an airtight gasket (Fig. 9; Ellis et al., 1995). Displacement of the SILASTIC membrane with its adherent cells thereby rapidly stretches the cells within 100 ms, mimicking the deformation forces that contribute to rapid acceleration/deceleration injury (Pike, 2001). It is important to emphasize that the forces produced by this model are likely to be highly relevant and akin to those encountered by the brain in vivo following acceleration/ deceleration induced TBI, as evidenced by biomechanical studies and simulations carried out using computer modeling (Margulies et al., 1990; Meaney et al., 1995). The Ellis model produces a range of injury levels, ranging from mild to moderate to severe. Signiﬁcantly, the model generates an injury in a very reproducible manner. Access to the multiple wells provides good experimental control, and facilitates studies of pharmacological agents to probe the mechanisms of trauma or to evaluate potential therapeutic agents. The model also facilitates studies of cell viability by using propidium iodide staining (McKinney et al., 1996), mitochondrial membrane potential measurement using rhodamine 123 (Ahmed et al., 2000, 2002), cell calcium levels measured with fura-2 (Rzigalinski et al., 1997, 1998a, b; Weber et al., 2001, 2002), electrophysiology (Tavalin et al., 1995, 1997; Zhang et al., 1996; Goforth et al., 1999, 2004; Lea et al., 2002; Kao et al., 2004) and biochemical assays of either the attached cells or the bathing media (Lamb et al., 1997; Rzigalinski et al., 1999; Hoffman et al., 2000). Mild injury due to a 5.7 mm displacement of the Silastic membrane (corresponding to 31% stretch of brain cells) releases glutamate from astrocytes, with peak glutamate occurring within 5 min after injury, and then declining within an hour (Rzigalinski et al., 1998a, b). The model also produces a transient rise in neuronal [Ca2+]i which persists for 410 min and

159

Fig. 9. In vitro TBI model of mechanical injury. An injury controller applies a timed pulse of nitrogen to cultured cells grown on a deformable SILASTIC membrane to mimic stretch deformation in TBI. Reprinted from J. Neurotrauma, Ellis et al. (1995), with permission from Mary Ann Liebert, Inc., Publishers.

is partially attenuated by the NMDA antagonist MK-801 (Ahmed et al., 2002). In addition [Ca2+]i responses of mildly injured neurons to exogenous glutamate are potentiated 15 min–24 h post-injury (Weber et al., 1999). The Ellis model reproduces many features of whole animal TBI including changes in cell viability (McKinney et al., 1996), neuronal and glial [Ca2+]i homeostasis (McKinney et al., 1996; Zhang et al., 1996; Rzigalinski et al., 1998a, b; Weber et al., 1999, 2002), the status of endoplasmic reticulum (ER) Ca stores (Weber et al., 1999; Floyd et al., 2001), free radicals (McKinney et al., 1996), phosphoinositide metabolism (Floyd et al., 2001), mitochondrial function (Ahmed et al., 2000, 2002) and electrophysiology (Tavalin et al., 1995, 1997; Ahmed et al., 2000).

Stretch injury alters cortical glutamate receptor function Excitatory synaptic transmission is mediated by AMPA and NMDA glutamate receptor subtypes. Both NMDA and AMPA receptors are heteromultimers, composed of combinations of NR1 and NR2A-D or GluR1-4 subunits, respectively. The NMDA current of control, uninjured neurons is well known to be characterized by its ‘‘J-shaped’’ voltage dependence, which is due to external voltage-dependent Mg2+ block of NMDA receptors (NMDARs) (Mayer et al., 1984; McBain and Mayer, 1994; Dingledine et al., 1999). However, in mildly injured neurons, voltage-dependent Mg2+ block of NMDARs is greatly reduced (Fig. 10;

160

Fig. 10. Stretch injury reduces Mg2+ block of NMDARs, as evidenced by linearization of the J-shaped NMDA current (I)–voltage (V) curve. I–V currents generated by the application of 200 mM NMDA and 10 mM glycine to an uninjured control and a mildly stretch-injured cortical pyramidal neuron. Reprinted from Science, Zhang et al. (1996), with permission from AAAS.

Zhang et al., 1996). Thus, injury shifts the IC50 of Mg2+ inhibition of NMDA current by 100-fold, and reduces its Bmax by 50% (Zhang et al., 1996). Moreover, raising [Mg2+]o restores some of the block to the channels (Zhang et al., 1996). The loss of Mg2+ sensitivity of NMDARs results in turn in potentiated [Ca2+]i responses to exogenously applied NMDA (Zhang et al., 1996). This enhanced NMDA-elicited [Ca2+]i response was recently duplicated in a different in vitro model of stretch injury (Geddes-Klein et al., 2006), and in other studies by Lea et al. (2002, 2003, see below). Importantly, pretreating neurons with the PKC inhibitor calphostin C, partially prevents the loss of Mg2+ block (Zhang et al., 1996), consistent with an earlier report that PKC modulates the Mg2+ sensitivity of NMDA receptors in nodose neurons (Chen and Huang, 1992). Using the stretch-injury model, Faden and coworkers conﬁrmed that in vitro injury indeed decreases external Mg2+ block of NMDARs and increases NMDA current in cultured neurons (Lea et al., 2002). The effect of stretch injury on NMDA receptors was shown to be altered by the activation of the Group I metabotropic glutamate receptors (mGluRs), mGluR1 and mGluR5 (Lea et al., 2002,

2003). Furthermore, the effect of mGluR agonists and antagonists on injury-induced NMDA potentiation is dependent on the presence or absence of glial cells (Lea et al., 2002, 2003). In pure neuronal cultures, inhibition of either mGluR1 or mGluR5 prevents stretch-induced NMDA receptor potentiation, suggesting a contributing role of Group I mGluRs in NMDA alteration (Lea et al., 2002, 2003). However, while mGluR5 inhibition also attenuates NMDA alterations in neurons grown in mixed cultures with glia, inhibition of mGluR 1 augments injury-induced NMDA potentiation (Lea et al., 2002, 2003). Activation of mGluR1 also differentially modulates the effect of injury on NMDA receptors, depending on the presence or absence of glia (Lea et al., 2003). The above data highlight the complexity of the biochemical and cellular pathways leading to alterations in NMDA receptors after mechanical injury and suggest the possible involvement of not only neuronal but also glial mGluRs in this process. Reduced Mg2+ block of NMDARs has been observed in other pathological conditions, including crush injury of motoneuron axons (Furukawa, 2000), axotomization of spinal cord motoneurons (Abdrachmanova et al., 2002), kainate lesioning of hippocampal neurons (Chen et al., 1999) and peripheral inﬂammation of dorsal horn neurons (Guo and Huang, 2001). The neurohormone somatostatin also reduces the Mg2+ block of hippocampal NMDARs by increasing PKC activity (Pittaluga et al., 2000). Thus, loss of Mg2+ block appears to be a general mechanism of traumatic injury and neuromodulation. Stretch injury via the Ellis in vitro model also modulates AMPARs in cortical neurons. Goforth et al. (1999) found mild injury similar to that used to modify NMDA current also increases whole-cell steady-state AMPAR current density (Goforth et al., 1999; Fig. 11A). Stretch does not appear to modify the sensitivity of AMPARs to AMPA, AMPAR current–voltage properties or the apparent number of AMPA receptors on pyramidal neurons. However, pharmacological reduction of AMPAR desensitization using cyclothiazide or activation of AMPARs with the weakly AMPAR desensitizing agonist kainate, offsets the observed differences between control and injured AMPA

161

Fig. 11. Stretch injury alters AMPAR kinetics and currents. (A) Whole-cell AMPA-elicited currents in control and injured cultured cortical pyramidal neurons. Stretch injury reduces fast desensitization of AMPARs, resulting in larger steady-state currents. (B) AMPAR currents recorded from outside/out patches in response to a 100 ms pulse of 1 mM glutamate+20 mM APV. (C) Mean 20–80% activation time (top) and desensitization rate (bottom) of AMPAR currents recorded from outside/out patches (*po0.05). Reprinted from J. Neurosci., copyright 1999 by the Society for Neuroscience, and J. Neurotrauma, Goforth et al. (2004), with permission from Mary Ann Liebert, Inc., Publishers.

currents, suggesting that injury suppresses AMPAR desensitization (Goforth et al., 1999). Changes in AMPAR kinetics surprisingly even persist in cell free membrane patches excised from injured neurons vs. controls and include slower AMPAR activation as well as desensitization (Fig. 11B). AMPA current alterations persist for at least 24 h post-injury and stretching neurons with the NMDA antagonist APV present prevents the alterations in AMPA currents (Goforth et al., 2004). This suggests that NMDAR activation is again a prerequisite for AMPAR modulation by injury. CaMKII is well known to be an important mediator of synaptic modulation in the CNS (see above in CA1 plasticity section). CaMKII also appears to play an important role in mediating changes in ionotropic receptors associated with TBI in cortical neurons. Thus, pretreatment of cultured cortical neurons with the CaMKII inhibitor KN-93 (N-[2-[[[3-(4-chlorophenyl)-2-propenyl] methylamino]methyl]phenyl]-N-(2-hydroxyethyl)-4 methoxybenzenesulphonamide) in the Ellis TBI model also prevents the alterations in AMPAR

kinetics observed following mild mechanical stretch (Goforth et al., 2004). CaMKII is [Ca2+]i -activated, is known to phosphorylate the GluR1 subunit of the AMPA receptor and is thought to participate in the induction of LTP in area CA1 of the hippocampus (Lisman et al., 1997; Malenka and Nicoll, 1997, 1999; Malinow et al., 2000; Pickard et al., 2001; Fink and Meyer, 2002; Malinow and Malenka, 2002; Song and Huganir, 2002). In support of the involvement of CAMKII in injury, measurements of CaMKII activity in cell lysates from control and stretch-injured neurons show an increase in the percent of autophosphorylated autonomous CaMKII activity at 10 min post-injury (R. Tombes, P. Goforth and L. Satin, unpublished data). This is consistent with a demonstrated increase in CamKII after FPI in vivo, which is accompanied by increased expression of phosphorylated GluR1, an AMPA receptor subunit and CaMKII substrate (Atkins et al., 2006). Although studies indicate that the effect of stretch injury on NMDA and AMPA receptors appear to involve the activation of the protein

162

kinases PKC and CaMKII, respectively (Zhang et al., 1996; Goforth et al., 2004), it is not known whether activation of these kinases after injury leads to direct phosphorylation of ionotropic glutamate receptors or associated proteins. Modulation of AMPARs by phosphorylation reﬂects a complex balance of kinase and phosphatase activity that can affect not only receptor subunits but anchoring or modulatory proteins that interact with AMPARS. Thus, it is important to determine the phosphorylation state of speciﬁc AMPA and NMDA subunits after injury in a highly quantitative manner. It is also possible that injury alters the expression and/or localization of ionotropic glutamate receptors and these changes may or may not be dependent on phosphorylation. Changes in glutamate receptor function such as those described above following stretch injury would be expected to contribute to dysfunctional synaptic transmission in CNS circuits impacted by mechanical trauma. In fact preliminary data demonstrate decreased mEPSC amplitude, and slowed mEPSC rise time and duration in cortical pyramidal neurons observed immediately after mild stretch injury (Satin, unpublished data). Injury-induced reductions in AMPA rise time may contribute to decreased mEPSC amplitudes, while the slowing of the mEPSC time course may be effected by reduced AMPA receptor desensitization as well increased contribution of NMDA receptor activation after mechanical injury (Zhang et al., 1996; Goforth et al., 1999, 2004). Spontaneous glutamatergic EPSCs of cortical pyramidal neurons also exhibit reduced amplitude following mild stretch injury (Satin, unpublished data). These data are in general agreement with studies that show attenuation of glutamate-evoked currents recorded from pyramidal CA1 neurons from mice subjected to FPI (Schwarzbach et al., 2006). The discrepant ﬁndings regarding the relative contribution of NMDA receptors to EPSCs after injury may be due, in part, to the examination of different brain regions (cortex vs. hippocampus), or the post-injury time point at which studies were conducted (immediately vs. 7 days post-injury). To further elucidate the role of direct glutamate receptor alterations in dysfunctional neurotransmission after TBI, it will be important to determine

whether similar alterations in AMPA and NMDA receptors occur in different brain regions (i.e., hippocampal CA1, CA3 and DG vs. cortical neurons) and determine the time course of these alterations.

Stretch injury alters cortical GABAA receptor function As discussed above, many studies have demonstrated injury-induced dysfunctional inhibitory synaptic transmission in the hippocampus (Toth et al., 1997; Witgen et al., 2005). As with glutamate receptors, a study of cortical GABAA receptor function suggests that direct alterations in GABAA receptors likely contributes to inhibitory dysfunction after trauma (Kao et al., 2004). Using the Ellis in vitro model, Kao et al. (2004) demonstrate that mild mechanical injury augments GABA-activated currents, which were blocked by bicuculline, conﬁrming their mediation by GABAA receptors (Fig. 12). Stretch injury does not shift the EC50 or current–voltage relationship of these GABAA currents. However, as reported for AMPA receptors, GABA current potentiation is prevented by injuring neurons in the presence of the NMDA receptor antagonist APV or the CaMKII inhibitor KN-93 (Kao et al., 2004). CaMKII has been shown to phosphorylate GABAA channels (Swope et al., 1999), and inhibiting the expression of the CaMKII alpha subunit is known to decrease GABAA receptor function (Churn et al., 2000). More studies are needed to determine the mechanism mediating enhanced GABA current, which, in theory, could result from either an increase in the number of functional GABAA channels in the neuronal plasma membrane, an increase in the unitary current through open GABAA channels or an increase in the probability of opening of a ﬁxed number of GABAA channels. As with glutamate receptors, direct alterations in GABAA channels would likely contribute to dysfunctional inhibitory synaptic transmission after TBI. In fact, miniature as well as spontaneous IPSCs recorded from stretch-injured cortical neurons exhibit increased amplitude (Goforth and Satin, unpublished data), but no change in mIPSC frequency, such as observed in hippocampal

163

Fig. 12. Stretch injury potentiates GABA-elicited whole-cell current. (A) Patch-clamp recordings of whole-cell currents elicited by the application of 50 mM GABA (ﬁlled bar) for a control neuron (left) and a stretched-injured neuron (right). Neurons were voltageclamped to –60 mV, and the rapid application of GABA-induced inward current. GABA-elicited currents of mildly injured neurons were increased in amplitude compared with uninjured control neurons. (B) Amplitude histograms of GABA current density shown for control neurons (left) and stretch-injured neurons (right). Currents were elicited by 50 mM GABA, and were normalized by neuron capacitance to yield GABA current density (pA/pF). Mean current densities obtained were 20.271.7 pA/pF (mean7SE, n ¼ 69) for control neurons and 41.272.6 pA/pF (n ¼ 82) for injured neurons (po0.001; Student’s t test). Reprinted from J. Neurotrauma, Kao et al. (2004), with permission from Mary Ann Liebert, Inc., Publishers.

CA1 neurons from mice subjected to FPI in vivo (Witgen et al., 2005). The data discussed clearly demonstrate that there are prominent injury-induced alterations in the function of glutamate and GABA receptors

in cortical neurons. Such changes occurring in postsynaptic neurotransmitter receptor function would thus be expected to contribute to dysfunctional synaptic transmission in vivo or in brain slices from TBI rodents. To underscore an

164

enduring theme of this chapter, we stress that the changes observed in synaptic function in vitro following injury were seen using levels of injury unable to cause signiﬁcant derangement of cell attachment, cell morphology or outright cell death. This emphasizes the importance of studying changes in synaptic function after mild and moderate TBI, which as a topic needs far more emphasis by the neurotrauma ﬁeld. Thus, the detailed study of CNS synaptic mechanisms and their exquisite sensitivity to mechanical deformation remains highly signiﬁcant for the ﬁeld and will likely result in new insights of importance to the development of new approaches for treating TBI patients.

blocker TTX (tetrodotoxin) or a bathing solution lacking calcium. However, SIDD is unaffected by the AMPA receptor (AMPAR) antagonist CNQX (6-cyano-7-nitroquinoxaline-2,3-dione) indicating that the requirement for glutamate receptor activation is subtype speciﬁc (Tavalin et al., 1997). Soltesz and colleagues (Ross and Soltesz, 2000) found that the membrane potentials of DGCs in brain slices from TBI rats are also depolarized due to the same ionic mechanism identiﬁed using the Ellis model by Tavalin et al. (Ross and Soltesz, 2000). Functionally, SIDD would be expected to contribute to the disruptions in ionic homeostasis, and changes in neuronal excitability known to occur after TBI.

Stretch injury alters Na+/K+ ATPase activity

Conclusions

In addition to abnormal neurotransmitter receptor function, neurotransmission and neuronal excitability in the CNS may also be modulated by trauma-induced alterations in ion pump and transporter activity. Under normal conditions, the resting membrane potentials of many pyramidal neurons includes an electrogenic component mediated by the basal activity of plasma membrane Na+/K+ ATPases or ‘Na pumps’ (Tavalin et al., 1995, 1997). This occurs because of unequal (i.e., electrogenic) ﬂuxes of Na+ out of the neuron and K+ into the neuron driven by the Na+/K+ ATPase, which is mediated energetically by ATP hydrolysis. In vitro studies of hippocampal neurons in brain slices from TBI rats and cortical neurons subjected to mild stretch result in clear injuryinduced decrease in Na+/K+ ATPase activity, which in turn leads to depolarized neuronal resting membrane potentials recorded in both the in vitro stretch and LFPI injury models (Tavalin et al., 1997; Ross and Soltesz, 2000). In the mildly stretch-injured neurons, resting membrane potential depolarization (magnitude 10–12 mV), developed 15–60 min after injury and was termed ‘‘stretch-induced delayed depolarization’’ (or ‘‘SIDD’’) (Tavalin et al., 1995). SIDD results from a Ca2+-inﬂux dependent loss of electrogenic Na/K ATPase activity and is prevented by pretreatment with the NMDA antagonist APV, the Na+ channel

Normal brain function is critically dependent on the activation and tight regulatory control of highly organized and precise neural circuits, the fundamental unit of which is the synapse. It is well documented that TBI disrupts normal brain function, resulting in behavioral, motor and cognitive deﬁcits such as impaired memory, attention and executive function, as well as slowed processing speed (Salmond and Sahakian, 2005). We postulate that such deﬁcits most likely reﬂect abnormal synaptic transmission after TBI. Synapses are either excitatory or inhibitory, deﬁned by activation of the speciﬁc neurotransmitter released. Overall neuronal activity is shaped by the integration of inhibitory and excitatory synaptic potentials, and slight changes in the balance of excitation and inhibition, or their speed of occurrence can produce profound changes in brain function by affecting synaptic information processing. Synaptic inputs moreover are not static, but rather quite plastic and are capable of being modulated in response to physiological stimulation, resulting in changes in the number or pattern of synaptic connections, presynaptic release of neurotransmitter or the functional properties of neuronal postsynaptic receptors (Malenka and Bear, 2004). Pathological alterations in the strength, timing or number of synaptic inputs would be expected to affect overall integration of synaptic potentials in individual

165

neurons, leading to abnormal cell ﬁring and dysfunctional information processing and transmission of signals within the important neural circuits of the cerebral cortex and hippocampus, as well as other brain regions. In order to understand the mechanisms underlying the cognitive changes that occur after TBI, it is thus critically important to elucidate the effects of injury on each key component of brain function: neuronal circuitry, cell to cell synaptic transmission and intrinsic cellular electrical activity. Despite important advances made in this area of investigation, most of the mechanisms involved in mediating the effects of TBI on the electrophysiology of neurons and synapses in the CNS still remain unknown. These mechanisms and their molecular and genetic underpinnings are therefore a major gap in our knowledge of the effects of TBI on the brain and should therefore be considered the next frontier of brain injury research. We thus believe that they deserve much more attention by investigators working in this ﬁeld and hope that the present review will stimulate new investigations of synaptic function and neuronal electrophysiology following TBI.

References Abdrachmanova, G., Teisinger, J. and Vyklicky Jr., L. (2002) Axotomy-induced changes in the properties of NMDA receptor channels in rat spinal cord motoneurons. J. Physiol., 538(Pt 1): 53–63. Ahmed, S.M., Rzigalinski, B.A., Willoughby, K.A., Sitterding, H.A. and Ellis, E.F. (2000) Stretch-induced injury alters mitochondrial membrane potential and cellular ATP in cultured astrocytes and neurons. J. Neurochem., 74(5): 1951–1960. Ahmed, S.M., Weber, J.T., Liang, S., Willoughby, K.A., Sitterding, H.A., Rzigalinski, B.A. and Ellis, E.F. (2002) NMDA receptor activation contributes to a portion of the decreased mitochondrial membrane potential and elevated intracellular free calcium in strain-injured neurons. J. Neurotrauma, 19(12): 1619–1629. Akasu, T., Muraoka, N. and Hasuo, H. (2002) Hyperexcitability of hippocampal CA1 neurons after ﬂuid percussion injury of the rat cerebral cortex. Neurosci. Lett., 329(3): 305–308. Annegers, J.F., Rocca, W.A. and Hauser, W.A. (1996) Causes of epilepsy: contributions of the Rochester epidemiology project. Mayo Clin. Proc., 71(6): 570–575. Arundine, M., Chopra, G.K., Wrong, A., Lei, S., Aarts, M.M., MacDonald, J.F. and Tymianski, M. (2003) Enhanced vulnerability to NMDA toxicity in sublethal traumatic neuronal injury in vitro. J. Neurotrauma, 20(12): 1377–1395.

Asikainen, I., Kaste, M. and Sarna, S. (1999) Early and late posttraumatic seizures in traumatic brain injury rehabilitation patients: brain injury factors causing late seizures and inﬂuence of seizures on long-term outcome. Epilepsia, 40(5): 584–589. Atkins, C.M., Chen, S., Alonso, O.F., Dietrich, W.D. and Hu, B.R. (2006) Activation of calcium/calmodulin-dependent protein kinases after traumatic brain injury. J. Cereb. Blood Flow Metab., 26(12): 1507–1518. Barnard, E.A., Skolnick, P., Olsen, R.W., Mohler, H., Sieghart, W., Biggio, G., Braestrup, C., Bateson, A.N. and Langer, S.Z. (1998) International Union of Pharmacology. XV. Subtypes of gamma-aminobutyric acid a receptors: classiﬁcation on the basis of subunit structure and receptor function. Pharmacol. Rev., 50(2): 291–313. Barrett, E.F. and Stevens, C.F. (1972) The kinetics of transmitter release at the frog neuromuscular junction. J. Physiol., 227(3): 691–708. Barth, J.T., Macciocchi, S.N., Giordani, B., Rimel, R., Jane, J.A. and Boll, T.J. (1983) Neuropsychological sequelae of minor head injury. Neurosurgery, 13(5): 529–533. Bear, M.F. and Abraham, W.C. (1996) Long-term depression in hippocampus. Annu. Rev. Neurosci., 19: 437–462. Bennett-Levy, J.M. (1984) Long-term effects of severe closed head injury on memory: evidence from a consecutive series of young adults. Acta Neurol. Scand., 70(4): 285–298. Bigler, E.D., Blatter, D.D., Anderson, C.V., Johnson, S.C., Gale, S.D., Hopkins, R.O. and Burnett, B. (1997) Hippocampal volume in normal aging and traumatic brain injury. AJNR Am. J. Neuroradiol., 18(1): 11–23. Bohbot, V.D., Allen, J.J. and Nadel, L. (2000) Memory deﬁcits characterized by patterns of lesions to the hippocampus and parahippocampal cortex. Ann. N.Y. Acad. Sci., 911: 355–368. Carbonell, W.S., Maris, D.O., McCall, T. and Grady, M.S. (1998) Adaptation of the ﬂuid percussion injury model to the mouse. J. Neurotrauma, 15(3): 217–229. Cave, C.B. and Squire, L.R. (1991) Equivalent impairment of spatial and nonspatial memory following damage to the human hippocampus. Hippocampus, 1(3): 329–340. Chen, L. and Huang, L.Y. (1992) Protein kinase C reduces Mg2+ block of NMDA-receptor channels as a mechanism of modulation. Nature, 356(6369): 521–523. Chen, Y., Chad, J.E., Cannon, R.C. and Wheal, H.V. (1999) Reduced Mg2+ blockade of synaptically activated N-methylD-aspartate receptor-channels in CA1 pyramidal neurons in kainic acid-lesioned rat hippocampus. Neuroscience, 88(3): 727–739. Christie, B.R., Kerr, D.S. and Abraham, W.C. (1994) Flip side of synaptic plasticity: long-term depression mechanisms in the hippocampus. Hippocampus, 4(2): 127–135. Churn, S.B., Sombati, S. and Delorenzo, R.J. (2000) Inhibition of CaMKII alpha subunit expression results in decreased GABAA receptor function. 2000 Abstract Viewer and Itinerary Planner. Washington D.C. Society for Neuroscience. Online. Claiborne, B.J., Amaral, D.G. and Cowan, W.M. (1986) A light and electron microscopic analysis of the mossy ﬁbers of the rat dentate gyrus. J. Comp. Neurol., 246(4): 435–458.

166 Cobb, S.R., Buhl, E.H., Halasy, K., Paulsen, O. and Somogyi, P. (1995) Synchronization of neuronal activity in hippocampus by individual GABAergic interneurons. Nature, 378(6552): 75–78. Cohen, A.S. and Abraham, W.C. (1996) Facilitation of longterm potentiation by prior activation of metabotropic glutamate receptors. J. Neurophysiol., 76(2): 953–962. Coulter, D.A., Raﬁq, A., Shumate, M., Gong, Q.Z., DeLorenzo, R.J. and Lyeth, B.G. (1996) Brain injury-induced enhanced limbic epileptogenesis: anatomical and physiological parallels to an animal model of temporal lobe epilepsy. Epilepsy Res., 26(1): 81–91. D’Ambrosio, R., Maris, D.O., Grady, M.S., Winn, H.R. and Janigro, D. (1998) Selective loss of hippocampal long-term potentiation, but not depression, following ﬂuid percussion injury. Brain Res., 786(1–2): 64–79. Del Castillo, J. and Katz, B. (1954) Statistical factors involved in neuromuscular facilitation and depression. J. Physiol., 124(3): 574–585. Deweer, B., Pillon, B., Pochon, J.B. and Dubois, B. (2001) Is the HM story only a ‘‘remote memory’’? Some facts about hippocampus and memory in humans. Behav. Brain Res., 127(1–2): 209–224. Dingledine, R., Borges, K., Bowie, D. and Traynelis, S.F. (1999) The glutamate receptor ion channels. Pharmacol. Rev., 51(1): 7–61. Dixon, C.E., Lyeth, B.G., Povlishock, J.T., Findling, R.L., Hamm, R.J., Marmarou, A., Young, H.F. and Hayes, R.L. (1987) A ﬂuid percussion model of experimental brain injury in the rat. J. Neurosurg., 67(1): 110–119. Dudek, F.E., Snow, R.W. and Taylor, C.P. (1986) Role of electrical interactions in synchronization of epileptiform bursts. Adv. Neurol., 44: 593–617. Ellis, E.F., McKinney, J.S., Willoughby, K.A., Liang, S. and Povlishock, J.T. (1995) A new model for rapid stretchinduced injury of cells in culture: characterization of the model using astrocytes. J. Neurotrauma, 12(3): 325–339. Fink, C.C. and Meyer, T. (2002) Molecular mechanisms of CaMKII activation in neuronal plasticity. Curr. Opin. Neurobiol., 12(3): 293–299. Floyd, C.L., Rzigalinski, B.A., Weber, J.T., Sitterding, H.A., Willoughby, K.A. and Ellis, E.F. (2001) Traumatic injury of cultured astrocytes alters inositol (1,4,5)-trisphosphatemediated signaling. Glia, 33(1): 12–23. Freund, T.F. and Buzsaki, G. (1996) Interneurons of the hippocampus. Hippocampus, 6(4): 347–470. Fricke, R.A. and Prince, D.A. (1984) Electrophysiology of dentate gyrus granule cells. J. Neurophysiol., 51(2): 195–209. Furukawa, Y. (2000) Reduction of voltage-dependent magnesium block of N-methyl-D-aspartate receptor-mediated current by in vivo axonal injury. Neuroscience, 96(2): 385–392. Gazzaley, A., Kay, S. and Benson, D.L. (2002) Dendritic spine plasticity in hippocampus. Neuroscience, 111(4): 853–862. Geddes-Klein, D.M., Schiffman, K.B. and Meaney, D.F. (2006) Mechanisms and consequences of neuronal stretch injury in vitro differ with the model of trauma. J. Neurotrauma, 23(2): 193–204.

Goforth, P.B., Ellis, E.F. and Satin, L.S. (1999) Enhancement of AMPA-mediated current after traumatic injury in cortical neurons. J. Neurosci., 19(17): 7367–7374. Goforth, P.B., Ellis, E.F. and Satin, L.S. (2004) Mechanical injury modulates AMPA receptor kinetics via an NMDA receptor-dependent pathway. J. Neurotrauma, 21(6): 719–732. Golarai, G., Greenwood, A.C., Feeney, D.M. and Connor, J.A. (2001) Physiological and structural evidence for hippocampal involvement in persistent seizure susceptibility after traumatic brain injury. J. Neurosci., 21(21): 8523–8537. Gorman, L.K., Shook, B.L. and Becker, D.P. (1993) Traumatic brain injury produces impairments in long-term and recent memory. Brain Res., 614(1–2): 29–36. Grady, M.S., Charleston, J.S., Maris, D., Witgen, B.M. and Lifshitz, J. (2003) Neuronal and glial cell number in the hippocampus after experimental traumatic brain injury: analysis by stereological estimation. J. Neurotrauma, 20(10): 929–941. Guo, H. and Huang, L.Y. (2001) Alteration in the voltage dependence of NMDA receptor channels in rat dorsal horn neurones following peripheral inﬂammation. J. Physiol., 537(Pt 1): 115–123. He, J., Yamada, K., Zou, L.B. and Nabeshima, T. (2001) Spatial memory deﬁcit and neurodegeneration induced by the direct injection of okadaic acid into the hippocampus in rats. J. Neural. Transm., 108(12): 1435–1443. Heinemann, U., Beck, H., Dreier, J.P., Ficker, E., Stabel, J. and Zhang, C.L. (1992) The dentate gyrus as a regulated gate for the propagation of epileptiform activity. Epilepsy Res. Suppl., 7: 273–280. Hoffman, S.W., Rzigalinski, B.A., Willoughby, K.A. and Ellis, E.F. (2000) Astrocytes generate isoprostanes in response to trauma or oxygen radicals. J. Neurotrauma, 17(5): 415–420. Hogg, S., Perron, C., Barneoud, P., Sanger, D.J. and Moser, P.C. (1998a) Neuroprotective effect of eliprodil: attenuation of a conditioned freezing deﬁcit induced by traumatic injury of the right parietal cortex in the rat. J. Neurotrauma, 15(7): 545–553. Hogg, S., Sanger, D.J. and Moser, P.C. (1998b) Mild traumatic lesion of the right parietal cortex in the rat: characterisation of a conditioned freezing deﬁcit and its reversal by dizocilpine. Behav. Brain Res., 93(1–2): 157–165. Joliot, M., Ribary, U. and Llinas, R. (1994) Human oscillatory brain activity near 40 Hz coexists with cognitive temporal binding. Proc. Natl. Acad. Sci. U.S.A., 91(24): 11748–11751. Kao, C.Q., Goforth, P.B., Ellis, E.F. and Satin, L.S. (2004) Potentiation of GABA(A) currents after mechanical injury of cortical neurons. J. Neurotrauma, 21(3): 259–270. Katayama, Y., Becker, D.P., Tamura, T. and Hovda, D.A. (1990) Massive increases in extracellular potassium and the indiscriminate release of glutamate following concussive brain injury. J. Neurosurg., 73(6): 889–900. Kotapka, M.J., Gennarelli, T.A., Graham, D.I., Adams, J.H., Thibault, L.E., Ross, D.T. and Ford, I. (1991) Selective vulnerability of hippocampal neurons in acceleration-induced experimental head injury. J. Neurotrauma, 8(4): 247–258.

167 Kotapka, M.J., Graham, D.I., Adams, J.H. and Gennarelli, T.A. (1992) Hippocampal pathology in fatal non-missile human head injury. Acta Neuropathol. (Berl.), 83(5): 530–534. Kotapka, M.J., Graham, D.I., Adams, J.H., Doyle, D. and Gennarelli, T.A. (1993) Hippocampal damage in fatal paediatric head injury. Neuropathol. Appl. Neurobiol., 19(2): 128–133. Kotapka, M.J., Graham, D.I., Adams, J.H. and Gennarelli, T.A. (1994) Hippocampal pathology in fatal human head injury without high intracranial pressure. J. Neurotrauma, 11(3): 317–324. Lamb, R.G., Harper, C.C., McKinney, J.S., Rzigalinski, B.A. and Ellis, E.F. (1997) Alterations in phosphatidylcholine metabolism of stretch-injured cultured rat astrocytes. J. Neurochem., 68(5): 1904–1910. LaPlaca, M.C., Cullen, D.K., McLoughlin, J.J. and Cargill II, R.S. (2005) High rate shear strain of three-dimensional neural cell cultures: a new in vitro traumatic brain injury model. J. Biomech., 38(5): 1093–1105. LaPlaca, M.C. and Thibault, L.E. (1998) Dynamic mechanical deformation of neurons triggers an acute calcium response and cell injury involving the N-methyl-D-aspartate glutamate receptor. J. Neurosci. Res., 52(2): 220–229. Lea, P.M., Custer, S.J., Vicini, S. and Faden, A.I. (2002) Neuronal and glial mGluR5 modulation prevents stretchinduced enhancement of NMDA receptor current. Pharmacol. Biochem. Behav., 73(2): 287–298. Lea, P.M.T., Custer, S.J., Stoica, B.A. and Faden, A.I. (2003) Modulation of stretch-induced enhancement of neuronal NMDA receptor current by mGluR1 depends upon presence of glia. J. Neurotrauma, 20(11): 1233–1249. Leuner, B., Falduto, J. and Shors, T.J. (2003) Associative memory formation increases the observation of dendritic spines in the hippocampus. J Neurosci, 23(2): 659–665. Lisman, J., Malenka, R.C., Nicoll, R.A. and Malinow, R. (1997) Learning mechanisms: the case for CaM-KII. Science, 276(5321): 2001–2002. Liu, L., Wong, T.P., Pozza, M.F., Lingenhoehl, K., Wang, Y., Sheng, M., Auberson, Y.P. and Wang, Y.T. (2004) Role of NMDA receptor subtypes in governing the direction of hippocampal synaptic plasticity. Science, 304(5673): 1021–1024. Liu, P. and Bilkey, D.K. (2001) The effect of excitotoxic lesions centered on the hippocampus or perirhinal cortex in object recognition and spatial memory tasks. Behav. Neurosci., 115(1): 94–111. Lothman, E.W., Stringer, J.L. and Bertram, E.H. (1992) The dentate gyrus as a control point for seizures in the hippocampus and beyond. Epilepsy Res. Suppl., 7: 301–313. Lowenstein, D.H., Thomas, M.J., Smith, D.H. and McIntosh, T.K. (1992) Selective vulnerability of dentate hilar neurons following traumatic brain injury: a potential mechanistic link between head trauma and disorders of the hippocampus. J. Neurosci., 12(12): 4846–4853. Lyeth, B.G., Jenkins, L.W., Hamm, R.J., Dixon, C.E., Phillips, L.L., Clifton, G.L., Young, H.F. and Hayes, R.L. (1990) Prolonged memory impairment in the absence of hippocampal

cell death following traumatic brain injury in the rat. Brain Res., 526(2): 249–258. Lytton, W.W. and Sejnowski, T.J. (1991) Simulations of cortical pyramidal neurons synchronized by inhibitory interneurons. J. Neurophysiol., 66(3): 1059–1079. Malenka, R.C. and Bear, M.F. (2004) LTP and LTD: an embarrassment of riches. Neuron, 44(1): 5–21. Malenka, R.C., Kauer, J.A., Perkel, D.J. and Nicoll, R.A. (1989) The impact of postsynaptic calcium on synaptic transmission — its role in long-term potentiation. Trends Neurosci., 12(11): 444–450. Malenka, R.C. and Nicoll, R.A. (1997) Silent synapses speak up. Neuron, 19(3): 473–476. Malenka, R.C. and Nicoll, R.A. (1999) Long-term potentiation — a decade of progress? Science, 285(5435): 1870–1874. Malinow, R., Mainen, Z.F. and Hayashi, Y. (2000) LTP mechanisms: from silence to four-lane trafﬁc. Curr. Opin. Neurobiol., 10(3): 352–357. Malinow, R. and Malenka, R.C. (2002) AMPA receptor trafﬁcking and synaptic plasticity. Annu. Rev. Neurosci., 25: 103–126. Margulies, S.S., Thibault, L.E. and Gennarelli, T.A. (1990) Physical model simulations of brain injury in the primate. J. Biomech., 23(8): 823–836. Martin, A.R. and Pilar, G. (1964) Presynaptic and post-synaptic events during post-tetanic potentiation and facilitation in the avian ciliary ganglion. J. Physiol., 175: 17–30. Mathias, J.L., Bigler, E.D., Jones, N.R., Bowden, S.C., BarrettWoodbridge, M., Brown, G.C. and Taylor, D.J. (2004) Neuropsychological and information processing performance and its relationship to white matter changes following moderate and severe traumatic brain injury: a preliminary study. Appl. Neuropsychol., 11(3): 134–152. Mayer, M.L., Westbrook, G.L. and Guthrie, P.B. (1984) Voltage-dependent block by Mg2+ of NMDA responses in spinal cord neurones. Nature, 309(5965): 261–263. McAllister, T.W. (1992) Neuropsychiatric sequelae of head injuries. Psychiatr. Clin. North Am., 15(2): 395–413. McAllister, T.W., Saykin, A.J., Flashman, L.A., Sparling, M.B., Johnson, S.C., Guerin, S.J., Mamourian, A.C., Weaver, J.B. and Yanofsky, N. (1999) Brain activation during working memory 1 month after mild traumatic brain injury: a functional MRI study. Neurology, 53(6): 1300–1308. McBain, C.J. and Mayer, M.L. (1994) N-methyl-D-aspartic acid receptor structure and function. Physiol. Rev., 74(3): 723–760. McIntosh, T.K., Vink, R., Noble, L., Yamakami, I., Fernyak, S., Soares, H. and Faden, A.L. (1989) Traumatic brain injury in the rat: characterization of a lateral ﬂuid-percussion model. Neuroscience, 28(1): 233–244. McKinney, J.S., Willoughby, K.A., Liang, S. and Ellis, E.F. (1996) Stretch-induced injury of cultured neuronal, glial, and endothelial cells. Effect of polyethylene glycol-conjugated superoxide dismutase. Stroke, 27(5): 934–940. Meaney, D.F., Smith, D.H., Shreiber, D.I., Bain, A.C., Miller, R.T., Ross, D.T. and Gennarelli, T.A. (1995) Biomechanical

168 analysis of experimental diffuse axonal injury. J. Neurotrauma, 12(4): 689–694. Miller, L.A., Lai, R. and Munoz, D.G. (1998) Contributions of the entorhinal cortex, amygdala and hippocampus to human memory. Neuropsychologia, 36(11): 1247–1256. Miller, L.P., Lyeth, B.G., Jenkins, L.W., Oleniak, L., Panchision, D., Hamm, R.J., Phillips, L.L., Dixon, C.E., Clifton, G.L. and Hayes, R.L. (1990) Excitatory amino acid receptor subtype binding following traumatic brain injury. Brain Res., 526(1): 103–107. Miyazaki, S., Katayama, Y., Lyeth, B.G., Jenkins, L.W., DeWitt, D.S., Goldberg, S.J., Newlon, P.G. and Hayes, R.L. (1992) Enduring suppression of hippocampal long-term potentiation following traumatic brain injury in rat. Brain Res., 585(1–2): 335–339. Mody, I., De Koninck, Y., Otis, T.S. and Soltesz, I. (1994) Bridging the cleft at GABA synapses in the brain. Trends Neurosci., 17(12): 517–525. Morrison, III, B., Cater, H.L., Benham, C.D. and Sundstrom, L.E. (2006) An in vitro model of traumatic brain injury utilising two-dimensional stretch of organotypic hippocampal slice cultures. J. Neurosci. Methods, 150(2): 192–201. Morrison, III, B., Meaney, D.F. and McIntosh, T.K. (1998) Mechanical characterization of an in vitro device designed to quantitatively injure living brain tissue. Ann. Biomed. Eng., 26(3): 381–390. O’Dell, D.M. and Hamm, R.J. (1995) Chronic postinjury administration of MDL 26,479 (Suritozole), a negative modulator at the GABAA receptor, and cognitive impairment in rats following traumatic brain injury. J. Neurosurg., 83(5): 878–883. Osteen, C.L., Giza, C.C. and Hovda, D.A. (2004) Injuryinduced alterations in N-methyl-D-aspartate receptor subunit composition contribute to prolonged 45 calcium accumulation following lateral ﬂuid percussion. Neuroscience, 128(2): 305–322. Pickard, L., Noel, J., Duckworth, J.K., Fitzjohn, S.M., Henley, J.M., Collingridge, G.L. and Molnar, E. (2001) Transient synaptic activation of NMDA receptors leads to the insertion of native AMPA receptors at hippocampal neuronal plasma membranes. Neuropharmacology, 41(6): 700–713. Pike, B.R. (2001) In vitro models of traumatic brain injury. In: Miller L.P. and Hayes R.L. (Eds.), Head Trauma: Basic, Preclinical and Clinical Directions. Wiley, New York, pp. 37–64. Pittaluga, A., Bonfanti, A. and Raiteri, M. (2000) Somatostatin potentiates NMDA receptor function via activation of InsP(3) receptors and PKC leading to removal of the Mg(2+) block without depolarization. Br. J. Pharmacol., 130(3): 557–566. Reeves, T.M., Kao, C.Q., Phillips, L.L., Bullock, M.R. and Povlishock, J.T. (2000) Presynaptic excitability changes following traumatic brain injury in the rat. J. Neurosci. Res., 60(3): 370–379. Reeves, T.M., Lyeth, B.G. and Povlishock, J.T. (1995) Longterm potentiation deﬁcits and excitability changes following traumatic brain injury. Exp. Brain Res., 106(2): 248–256.

Reeves, T.M., Lyeth, B.G., Phillips, L.L., Hamm, R.J. and Povlishock, J.T. (1997a) The effects of traumatic brain injury on inhibition in the hippocampus and dentate gyrus. Brain Res., 757(1): 119–132. Reeves, T.M., Zhu, J., Povlishock, J.T. and Phillips, L.L. (1997b) The effect of combined ﬂuid percussion and entorhinal cortical lesions on long-term potentiation. Neuroscience, 77(2): 431–444. Rempel-Clower, N.L., Zola, S.M., Squire, L.R. and Amaral, D.G. (1996) Three cases of enduring memory impairment after bilateral damage limited to the hippocampal formation. J. Neurosci., 16(16): 5233–5255. Ross, S.T. and Soltesz, I. (2000) Selective depolarization of interneurons in the early posttraumatic dentate gyrus: involvement of the Na(+)/K(+)-ATPase. J. Neurophysiol., 83(5): 2916–2930. Rzigalinski, B.A., Liang, S., McKinney, J.S., Willoughby, K.A. and Ellis, E.F. (1997) Effect of Ca2+ on in vitro astrocyte injury. J. Neurochem., 68(1): 289–296. Rzigalinski, B.A., Weber, J.T., Willoughby, K.A. and Ellis, E.F. (1998a) Intracellular free calcium dynamics in stretchinjured astrocytes. J. Neurochem., 70(6): 2377–2385. Rzigalinski, B.A., Weber, J.T., Willoughby, K.A., L.S., Woodward, J.J. and Ellis, E.F. (1998b). Glutamate release and enhanced receptor sensitivity in stretch-injured neurons. J Neurotrauma, 15: 894. Rzigalinski, B.A., Willoughby, K.A., Hoffman, S.W., Falck, J.R. and Ellis, E.F. (1999) Calcium inﬂux factor, further evidence it is 5,6-epoxyeicosatrienoic acid. J. Biol. Chem., 274(1): 175–182. Salmond, C.H. and Sahakian, B.J. (2005) Cognitive outcome in traumatic brain injury survivors. Curr. Opin. Crit. Care, 11: 111–116. Sanders, M.J., Sick, T.J., Perez-Pinzon, M.A., Dietrich, W.D. and Green, E.J. (2000) Chronic failure in the maintenance of long-term potentiation following ﬂuid percussion injury in the rat. Brain Res., 861(1): 69–76. Santhakumar, V., Bender, R., Frotscher, M., Ross, S.T., Hollrigel, G.S., Toth, Z. and Soltesz, I. (2000) Granule cell hyperexcitability in the early post-traumatic rat dentate gyrus: the ‘irritable mossy cell’ hypothesis. J. Physiol., 524(Pt 1): 117–134. Santhakumar, V., Ratzliff, A.D., Jeng, J., Toth, Z. and Soltesz, I. (2001) Long-term hyperexcitability in the hippocampus after experimental head trauma. Ann. Neurol., 50(6): 708–717. Schwarzbach, E., Bonislawski, D.P., Xiong, G. and Cohen, A.S. (2006) Mechanisms underlying the inability to induce area CA1 LTP in the mouse after traumatic brain injury. Hippocampus, 16(6): 541–550. Segal, M. (2005) Dendritic spines and long-term plasticity. Nat. Rev. Neurosci., 6(4): 277–284. Sick, T.J., Perez-Pinzon, M.A. and Feng, Z.Z. (1998) Impaired expression of long-term potentiation in hippocampal slices 4 and 48 h following mild ﬂuid-percussion brain injury in vivo. Brain Res., 785(2): 287–292. Sihver, S., Marklund, N., Hillered, L., Langstrom, B., Watanabe, Y. and Bergstrom, M. (2001) Changes in mACh,

169 NMDA and GABA(A) receptor binding after lateral ﬂuidpercussion injury: in vitro autoradiography of rat brain frozen sections. J. Neurochem., 78(3): 417–423. Singer, W. and Gray, C.M. (1995) Visual feature integration and the temporal correlation hypothesis. Annu. Rev. Neurosci., 18: 555–586. Sloviter, R.S. (1994) The functional organization of the hippocampal dentate gyrus and its relevance to the pathogenesis of temporal lobe epilepsy. Ann. Neurol., 35(6): 640–654. Smith, D.H., Okiyama, K., Thomas, M.J., Claussen, B. and McIntosh, T.K. (1991) Evaluation of memory dysfunction following experimental brain injury using the Morris water maze. J. Neurotrauma, 8(4): 259–269. Song, I. and Huganir, R.L. (2002) Regulation of AMPA receptors during synaptic plasticity. Trends Neurosci., 25(11): 578–588. Swope, S.L., Moss, S.J., Raymond, L.A. and Huganir, R.L. (1999) Regulation of ligand-gated ion channels by protein phosphorylation. Adv. Second Messenger Phosphoprotein Res., 33: 49–78. Tavalin, S.J., Ellis, E.F. and Satin, L.S. (1995) Mechanical perturbation of cultured cortical neurons reveals a stretchinduced delayed depolarization. J. Neurophysiol., 74(6): 2767–2773. Tavalin, S.J., Ellis, E.F. and Satin, L.S. (1997) Inhibition of the electrogenic Na pump underlies delayed depolarization of cortical neurons after mechanical injury or glutamate. J. Neurophysiol., 77(2): 632–638. Tecoma, E.S., Monyer, H., Goldberg, M.P. and Choi, D.W. (1989) Traumatic neuronal injury in vitro is attenuated by NMDA antagonists. Neuron, 2(6): 1541–1545.

Thompson, H.J., Lifshitz, J., Marklund, N., Grady, M.S., Graham, D.I., Hovda, D.A. and McIntosh, T.K. (2005) Lateral ﬂuid percussion brain injury: a 15-year review and evaluation. J. Neurotrauma, 22(1): 42–75. Toth, Z., Hollrigel, G.S., Gorcs, T. and Soltesz, I. (1997) Instantaneous perturbation of dentate interneuronal networks by a pressure wave-transient delivered to the neocortex. J. Neurosci., 17(21): 8106–8117. Weber, J.T., Rzigalinski, B.A., Willoughby, K.A., Moore, S.F. and Ellis, E.F. (1999) Alterations in calcium-mediated signal transduction after traumatic injury of cortical neurons. Cell Calcium, 26(6): 289–299. Weber, J.T., Rzigalinski, B.A. and Ellis, E.F. (2001) Traumatic injury of cortical neurons causes changes in intracellular calcium stores and capacitative calcium inﬂux. J. Biol. Chem., 276(3): 1800–1807. Weber, J.T., Rzigalinski, B.A. and Ellis, E.F. (2002) Calcium responses to caffeine and muscarinic receptor agonists are altered in traumatically injured neurons. J. Neurotrauma, 19(11): 1433–1443. Witgen, B.M., Lifshitz, J., Smith, M.L., Schwarzbach, E., Liang, S.L., Grady, M.S. and Cohen, A.S. (2005) Regional hippocampal alteration associated with cognitive deﬁcit following experimental brain injury: a systems, network and cellular evaluation. Neuroscience, 133(1): 1–15. Wong, R.K., Traub, R.D. and Miles, R. (1986) Cellular basis of neuronal synchrony in epilepsy. Adv. Neurol., 44: 583–592. Zhang, L., Rzigalinski, B.A., Ellis, E.F. and Satin, L.S. (1996) Reduction of voltage-dependent Mg2+ blockade of NMDA current in mechanically injured neurons. Science, 274(5294): 1921–1923.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 11

Traumatic injury of the spinal cord and nitric oxide Jozef Marsˇ ala, Judita Orenda´cˇova´, Nadezˇda Luka´cˇova´ and Ivo Vanicky´ Institute of Neurobiology, Slovak Academy of Sciences, Kosˇice, Slovak Republic

Abstract: In the current report, we summarize our ﬁndings related to the involvement of nitric oxide (NO) in the pathology of spinal cord trauma. We initially studied the distribution of nitric oxide synthase (NOS)immunolabeled and/or nicotinamide adenine dinucleotide phosphate diaphorase (NADPHd; which is highly colocalized with NOS)-stained somata and ﬁbers in the spinal cord of the rabbit. Segmental and laminar distribution of NADPHd-stained neurons in the rabbit revealed a large number of NADPHdstained neurons in the spinal cord falling into six categories, N1–N6, while others could not be classiﬁed. Large numbers of NADPHd-stained neurons were identiﬁed in the superﬁcial dorsal horn and around the central canal. Four morphologically distinct kinds of NADPHd-stained axons 2.5–3.5 mm in diameter were identiﬁed throughout the white matter in the spinal cord. Moreover, a massive occurrence of axonal NADPHd-staining was detected in the juxtagriseal layer of the ventral funiculus along the rostrocaudal axis. The prominent NADPHd-stained ﬁber bundles were identiﬁed in the mediobasal and central portion of the ventral funiculus. The sulcomarginal fasciculus was found in the basal and medial portion of the ventral funiculus in all cervical and thoracic segments. Since the discovery that NO may act as a neuronal transmitter, an increasing interest has focused on its ability to modulate synaptic function. NO passes through cell membranes without speciﬁc release or uptake mechanisms inducing changes in signal-related functions by several means. In particular, the activation of the soluble guanylyl cyclases (sGC), the formation of cyclic guanosine 30 ,50 -monophosphate (cGMP) and the action of cGMP-dependent protein kinases has been identiﬁed as the main signal transduction pathways of NO in the nervous system including spinal cord. It is known that the intracellular level of cGMP is strictly controlled by its rate of synthesis via guanylyl cyclases (GC) and/or by the rate of its degradation via 30 ,50 -cyclic nucleotide phosphodiesterases (PDE). GC can be divided into two main groups, i.e., the membrane-bound or particular guanylyl cyclase (pGC) and the cytosolic or sGC. In the spinal cord, the activation of pGC has only been demonstrated for natriuretic peptides, which stimulate cGMP accumulation in GABA-ergic structures in laminae I–III of the rat cervical spinal cord. These neurons are involved in controlling the action of the locomotor circuit. In view of the abundance of NO-responsive structures in the brain, it is proposed that NO–cGMP signaling will be part of neuronal information processing at many levels. In relation to this, we found that surgically induced Th7 constriction of 24 h duration stimulated both the constitutive NOS activity and cGMP level by 120 and 131%, respectively, in non-compartmentalized white matter of Th8–Th9 segments, located just caudally to the site of injury. NO-mediated cGMP formation was only slightly increased in the dorsal funiculus of Th5–Th9 segments. There are some other sources that may inﬂuence the NO-mediated cGMP formation in spinal cord. A high level of glutamate produced at the site of the lesion and an excessive accumulation of intracellular Ca2+ may stimulate NOS activity and create suitable conditions for NO synthesis and its adverse effect on white matter. An increased interest has focused on the role of NO at the Corresponding author. Tel.: +421 55 6785 062; Fax: +421 55

6785 074; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61011-X

171

172

site of injury and in areas located close to the epicenter of the impact site and, in these connections an upregulation of NOS was noted in neurons and interneurons. However, the upregulation of NOS expression was also seen in interneurons located just rostrally and caudally to the lesion. A quantitative analysis of laminar distribution of multiple cauda equina constriction (MCEC) induced NADPHd-stained neurons revealed a considerable increase in these neurons in laminae VIII–IX 8 h postconstriction, and a highly statistically significant increase of such neurons in laminae VII–X 5 days postconstriction in the lumbosacral segments. Concurrently, the number of NADPHd-stained neurons on laminae I–II in LS segments was greatly reduced. It is concluded that a greater understanding of NO changes after spinal cord trauma is essential for the possibility of targeting this pathway therapeutically. Keywords: spinal cord; trauma; cauda equina constriction; nitric oxide The occurrence of NADPH diaphorase-stained and/ or bNOS immunoreactive neurons in the gray matter of the spinal cord Small, morphologically heterogeneous populations of neurons containing nicotinamide adenine dinucleotide phosphate diaphorase (NADPHd), and/or brain nitric oxide synthase (bNOS), an enzyme known to generate nitric oxide (NO), have been identiﬁed at various sites in the central nervous system (CNS) in a number of mammalian species including the human brain (Kowall et al., 1985, 1987; Kowall and Mueller, 1988; Mizukawa et al., 1989; Bredt et al., 1990; Vincent and Kimura, 1992; Egberongbe et al., 1994; Vincent, 1994). Immunohistochemistry of bNOS revealed that the occurrence of this enzyme is almost completely homotopic with the localization of neurons stained for NADPHd (Luka´cˇova´ et al., 1999). It has long been known that NADPHd-stained neurons are present in the spinal cord, especially in the substantia gelatinosa, lamina X, and the intermediolateral cell column (Thomas and Pearse, 1964; Mizukawa et al., 1989; Blottner and Baumgarten, 1992; Bredt and Snyder, 1992). However, the analysis of NADPHd-stained and/or bNOS-immunoreactive neurons in the spinal cord of different animal species (Anderson, 1992; Valtschanoff et al., 1992; Dun et al., 1993; Vizzard et al., 1994a, 1995) revealed a morphologically heterogeneous pattern of NADPHd-stained neuronal pools ranging from bipolar, poorly branched NADPHd-stained neurons in the superﬁcial dorsal horn to highly differentiated neurons in the pericentral region (lamina X), deep dorsal horn (laminae IV–V), and dorsal gray commissure

containing widely branching NADPHd-stained neurons (Anderson, 1992; Valtschanoff et al., 1992; Dun et al., 1993; Lee et al., 1993; Saito et al., 1994; Vizzard et al., 1994b; Wetts and Vaughn, 1994; Burnett et al., 1995; Marsˇ ala et al., 1997, 1998). A study from our laboratory aimed at segmental and laminar distribution of NADPHd-stained neurons in the rabbit revealed a large number of NADPHd-positive neurons in the spinal cord falling into six categories, N1–N6, while others could not be classiﬁed (Marsˇ ala et al., 1999). Large numbers of NADPHd-stained neurons were identiﬁed in the superﬁcial dorsal horn and around the central canal at all spinal levels and in the intermediolateral cell column at thoracic and upper lumbar levels. NADPHd-stained somata of the pericentral region were divided into a thin subependymal cell column containing longitudinally arranged, small bipolar neurons with processes penetrating deeply into the intermediolateral cell column and/or running rostrocaudally in the subependymal layer. The second pericentral cell column located more laterally in lamina X contains NADPHd-stained neurons with long dendrites radiating in the transverse plane. In the pericentral region (lamina X), close association of NADPHdstained somata and ﬁbers, and mostly longitudinally oriented blood vessels was detected. Somata of the sacral parasympathetic nucleus seen in segments S1–S3, exhibited prominent NADPHd staining accompanied by heavily stained ﬁbers extending from Lissauer’s tract through lamina I along the lateral edge of the dorsal horn to lamina V. A massive dorsal gray commissure, highly positive in NADPHd staining, was found in segments

173

Sl–S3. Scattered positive cells were also found in the deeper dorsal horn, ventral horn, and white matter. Fiber-like NADPHd staining was found in the superﬁcial dorsal horn and pericentral region in all cervical, thoracic, and lumbosacral segments. Dense, punctate, non-somatic NADPHd staining was detected in the superﬁcial dorsal horn, in the pericentral region all along the rostrocaudal axis and in the phrenic nucleus (segments C4–C5), dorsal nucleus (segments Th2–L2), Onuf’s nucleus (segments S1–S3), and the dorsal part of the dorsal gray commissure (segments S1–S3; Marsˇ ala et al., 1999).

The distribution of NADPH diaphorase-positive and/or NOS immunolabeled ﬁbers in spinal cord white matter The funicular distribution of NADPHd-stained axons was examined in the white matter of the rabbit spinal cord by using horizontal, parasagittal, and transverse sections. Four morphologically distinct kinds of NADPHd-stained axons 2.5–3.5 mm in diameter were found in the sulcomarginal fasciculus as a part of the ventral funiculus in the cervical and upper thoracic segments and in the long propriospinal bundle of the ventral funiculus in Th3–L3 segments. Varicose NADPHdstained axons of the sympathetic preganglionic neurons, characterized by widely spaced varicosities, were found in the ventral funiculus of Th2–L3 segments. A third kind of NADPHd-stained ultraﬁne axons, 0.3–0.5 mm in diameter with numerous varicosities mostly spherical in shape, was identiﬁed in large number within Lissauer’s tract. The last group of NADPHd-stained axons, 1.0–1.5 mm in diameter, occurred in the Lissauer tract. Most of these axons were traceable for considerable distances and generating spherical and elliptical forms of varicosities. The majority of NADPHd-stained axons identiﬁed in the cuneate and gracile fascicles were concentrated in the deep portion of the dorsal funiculus. An extremely reduced number of NADPHd-stained axons, conﬁrmed by a computer-assisted image-processing system were found in the dorsal half of the gracile fascicle. Axonal

NADPHd-staining could not be detected in the lateral funiculus consistent with the location of the dorsal spinocerebellar tract. Numerous, mostly thin NADPHd-stained axonal proﬁles could be detected in the dorsolateral funiculus in all segments studied and in juxtagriseal portion of the lateral funiculus in the extent of the cervical and lumbar enlargement. A massive occurrence of axonal NADPHd-staining was detected in the juxtagriseal layer of the ventral funiculus along the rostrocaudal axis of the spinal cord. The prominent NADPHd-stained bundles containing smooth, thick, non-varicose axons were identiﬁed in the mediobasal and central portion of the ventral funiculus. First, the sulcomarginal fasciculus was found in the basal and medial portion of the ventral funiculus in all cervical and upper thoracic segments. Second, more caudally, a long propriospinal bundle displaying prominent NADPHdstaining was localized in the central portion of the ventral funiculus throughout the Th3–L3 segments (Marsˇ ala et al., 2003).

NO/cGMP signaling in the spinal cord For a long time it was believed that interneuronal communication was realized exclusively via synaptic contacts. However, since the discovery of NO as a neuronal transmitter (Snyder and Bredt, 1991; Dawson et al., 1992), an increasing interest has focused on its ability to modulate synaptic functions. NO passes through cell membranes without speciﬁc release or uptake mechanisms and acts directly on surrounding neural tissue in extended spatial limits of approximately 300–400 mm (Wood and Garthwaite, 1994; Vincent, 1994). After its synthesis in the soma of neurons, NO induces changes in signal-related functions in several ways. In particular, the activation of soluble guanylyl cyclases (sGC), the formation of cyclic guanosine 30 ,50 -monophosphate (cGMP), and the action of cGMP-dependent protein kinases have been identiﬁed as the main signal transduction pathway of NO in the nervous system (Bredt and Snyder, 1992; Wang and Robinson, 1997; Smolenski et al., 1998).

174

It is known that the intracellular level of cGMP is strictly controlled by its rate of synthesis via GCs and/or by the rate of its degradation via 30 ,50 -cyclic nucleotide phosphodiesterases (PDEs; Beavo, 1995; Lucas et al., 2000; Francis et al., 2001). GC can be divided into two main groups, i.e., the membrane-bound or pGC and the cytosolic or sGC (Schmidt, 1992; Murad, 1994; Wedel and Garbers, 1997). PDEs comprise a large group of enzymes that hydrolyze cGMP to its inactive 50 derivate. PDE families (PDE1–11) and a large number of their isoforms are characterized with distinct localization pattern in the brain. However, only three PDE families, i.e., PDE5, PDE9, and a photoreceptor-speciﬁc PDE6, identiﬁed in retina, specifically use cGMP as a substrate (Beavo, 1995; van Staveren et al., 2003). cGMP can be degraded by dual-substrate PDEs (PDE1, PDE2, PDE10, and PDE11) as well. pGC is a large transmembrane molecule that has a receptor domain at the outside and the catalytic site at the inside of the cell (Vles et al., 2000). This enzyme is activated through interaction of peptide hormones with the receptor domain. In the spinal cord, the activation of pGC has only been demonstrated for atrial natriuretic peptide (ANP), which stimulates cGMP accumulation in GABAergic structures of laminae I–III of the rat cervical spinal cord (Vles et al., 2000); these neurons are involved in controlling the action of the spinal locomotor circuit. The GABA-ergic agonist, Baclofen, a well-known drug in the treatment of spasticity, has been shown to express an effect on ANPmediated cGMP synthesis in the superﬁcial dorsal horn layers of the cervical spinal cord (de Louw et al., 2002). These data propose that the clinical effect of Baclofen on the reduction of spasticity may be involved in the triggering of a phosphorylation/dephosphorylation cycle of the intracellular part of the ANP receptor and, subsequently, a downregulation of the pGC activity (Potter and Hunter, 2001). sGC is a heterodimeric enzyme composed of four different subunits (a1, a2, b1, b2; Russwurm et al., 1998); however, all catalytically active GC isoforms are built of the b1 subunit and either the a1 or the a2 subunit (Gupta et al., 1997; Koesling and Friebe, 2000; Koglin et al., 2001). Activation

of different subunits of sGC is associated with the binding of NO to a heme group, leading to formation of a nitrosyl–heme complex and consequent conformational change (Humbert et al., 1990). The activation of sGC and the formation of cGMP have been identiﬁed as the main NO/cGMP signal transduction pathway in the CNS (Bredt and Snyder, 1992; Wang and Robinson, 1997; Smolenski et al., 1998). In view of the abundance of NO-responsive structures in the brain, it is proposed that NO-cGMP signaling will be part of neuronal information processing at many levels. To date, little is known about the role of NOcGMP signaling in the spinal cord. Morris et al. (1994) showed that the NO-cGMP pathway is located primarily in the superﬁcial layers of the dorsal horn in the neonatal rat spinal cord. The NO-mediated cGMP response in immature spinal cord was also conﬁrmed by Vles et al. (2000). The increased level of cGMP in the dorsal horn of the spinal cord seems to be associated with hyperalgesia through activation of cGMP dependent protein kinase Ia (Aley et al., 1998; Tao et al., 2000). NO-mediated cGMP was also identiﬁed in parvalbumin-immunoreactive (GABA-ergic) neurons and in the axon terminals of the ventral horn probably apposing the motor neurons (Vles et al., 2000).

NO-mediated cGMP synthesis in the spinal cord after traumatic injury Recent data from our laboratory have shown that surgically induced Th7 constriction of 24 h duration stimulated both the constitutive NOS activity and cGMP level by 120 and 131%, respectively, in non-compartmentalized white matter of Th8–Th9 segments, located just caudally to the site of injury (Luka´cˇova´ et al., 2002). NO-mediated cGMP formation was only slightly increased in the dorsal funiculus of Th5–Th9 segments, composed entirely of long ascending axons. However, a strong increase of cGMP formation was noted in the lateral funiculus of segments located below the site of Th7 constriction. These results suggest a decrease of the axoplasmic transport in ascending NOSimmunoreactive (NOS-IR) axons, which clearly

175

prevail in the lateral funiculus in comparison with descending axons. The NOS-IR neurons of the intermediolateral cell column sending many NOS-IR dendrites into the lateral funiculus (Marsˇ ala et al., 1998, 1999), together with the NOS-IR axons of the lateral spinothalamic tract may, upon midthoracic constriction, provide the substrate for NO/cGMP formation. A significant decrease in the level of cGMP found in the ventral funiculus of Th5–Th6 segments correlates with a significant decrease of both the constitutive NOS activity, attributable to a slight constriction of the superﬁcial portion of the ventral funiculus at Th7 level, and the axoplasmic transport of the NOS affecting thick, highly NOS-IR axons (Luka´cˇova´ et al., 2000). These axons form an ascending and, to a lesser extent, a descending propriospinal bundle, connecting NOS-IR neurons in lumbosacral enlargement with the ventral motor nucleus at cervicothoracic level (Sherrington and Laslett, 1903; Barilari and Kuypers, 1969; Yezierski et al., 1980; Marsˇ ala et al., 2004; Luka´cˇova´ et al., 2006). The detection of a considerably higher radioassay-assessed NOS activity in the ventral funiculus of the lumbosacral segments, in comparison to cervical segments, supports the conclusions based on immunocytochemical and histochemical studies (Luka´cˇova´ et al., 1999; Luka´cˇova´ and Pavel, 2000). There are some other sources that may inﬂuence the NO-mediated cGMP formation in spinal cord injury (SCI). A high level of glutamate produced at the site of the lesion in the spinal cord (Mc Adoo et al., 1997) and an excessive accumulation of intracellular Ca2+ (Lehotsky´ et al., 1999) may stimulate NOS activity and create suitable conditions for NO synthesis and its adverse effect on white matter. An extreme production of NO and superoxide, appearing early after SCI in its close vicinity, may stimulate the formation of peroxynitrite (ONOO) that in turn may break up, thus giving nitrogen dioxide and a compound having a hydroxyl-like reactivity leading to oxidation of protein, DNA, and membrane-damaging lipid peroxidation (Luka´cˇova´ et al., 1994). Furthermore, the release of free fatty acids, particularly arachidonic acid, in SCI (Horrocks et al., 1984; Hala´t et al., 1987; Nakano et al., 1990) may lead to an overexpression

of NOS accompanied by an increase in cellular cGMP (Strosznajder and Chalimoniuk, 1996; Toborek et al., 2000). Morton and Bredt (1998) have suggested that in primary cell cultures taken from the cerebellum of nNOS knockout mice, norepinephrine is able to stimulate cGMP formation independently from sGC. In addition, a selective inhibitor of sGC points to a possibility that extracellular cGMP released under basal conditions in the brain might come from both NO-sensitive and NO-insensitive GCs (Vallebuona and Raiteri, 1993; Luo et al., 1994; Fedele et al., 1996). The elevation of cGMP levels could be elicited by atrial natriuretic factor, possibly localized in astroglial cells (de Vente et al., 1990). These non-neuronal elements form dense perinodal processes approaching the node of Ranvier (Ransom et al., 1993) and, as such, are located at sites known for a high rate of transmembrane ion trafﬁc between the axolemma and the extracellular space (Stys et al., 1991; Stys and Steffensen, 1996). Based on biochemical and immunocytochemical experiments we can conclude that the NO/cGMP signaling pathway is implicated in the pathology of SCI.

NOS in the spinal cord: effect of traumatic injury in the epicenter of injury Spinal cord neurons are susceptible to toxic effects of excitatory amino acids because of the excitatory amino acid glutamate, which is markedly released from glutamatergic neurons to the extracellular space, stimulates N-methyl-D-aspartate (NMDA) receptors located on nitrergic neurons, and consecutively enhances the release of NO (Prast and Philippu, 2001). Both NO and glutamate, when overstimulated, can initiate a neurotoxic cascade. An increasing interest has focused on the role of NO at the site of SCI and in areas located close to the epicenter of the impact site. Acute changes in NO production after SCI have been studied by Liu et al. (2000), demonstrating a rapid increase of NO immediately after injury followed by a rapid decline during approximately the ﬁrst 60 min after injury. A microdialysis study by Nakahara et al. (2002) has also demonstrated a

176

second wave of NO production after SCI in the rat — the second wave was observed 24 h and 3 days after injury. Long-term changes in the production of parenchymal NO after SCI have been studied by detecting activities of NOS at various posttraumatic periods. Recent data obtained by Trudrung et al. (2000) have shown the upregulation of NOS expression in neurons adjacent to the lesion on both sides of a complete transverse lesion of the spinal cord. Yick et al. (1998) have described the increase of nNOS expression in Clarke’s column 3 days after hemisection, reaching its maximum 20 days after injury. The upregulation of NOS expression was also seen in interneurons located rostrally and caudally to the lesion (Wu et al., 1994; Sharma et al., 2000; Luka´cˇova´ et al., 2005). Surgically induced Th7 constriction of 24 h duration caused an increase of calcium-dependent nitric oxide synthase (cNOS) activity in lateral funiculus just below the site of injury and a strong decrease in the ventral funiculus of two segments, located above the injured site (Luka´cˇova´ et al., 2000, 2002). Diaz-Ruiz et al. (2002) studied the changes in constitutive and inducible NOS (iNOS) activities after severe SCI in rats that survived 2–72 h. They observed a significant increase in cNOS after 4 and 8 h, followed by a decline to normal levels. In contrast, iNOS levels gradually increased after injury, and the difference reached statistical significance at 72 h posttrauma. Similarly, time-dependent decrease in cNOS activity and the expression of iNOS were noted in traumatized Th10 segment and in segments located in close vicinity of spinal cord trauma (Chatzipanteli et al., 2002). This study noted a significant decrease in cNOS activities at 3 h postinjury, and a gradual return to control values within 24 h. Again, iNOS activities displayed contrasting changes. Activity was increased in all spinal cord segments and time points studied (3 h-3 days). The most robust increase was detected at 24 h at rostral and caudal segments relative to injury site. Data obtained recently from our laboratory have shown a remarkable decrease of cNOS activity in the injured site analyzed 1 day after Th9–Th10 spinal cord hemisection (Luka´cˇova´ et al., 2006). At 7th day of survival the enzyme activity returned to approximately half of the control value. The

delayed enhancement in cNOS activity observed at day 7 in the hemisected region may result from the alterations in the neuronal circuitry (Xu et al., 2000), subsequent proliferation of ﬁbers containing NOS, and the expression of NOS observed in axonal swelling after SCI (Guizar-Sahagun et al., 1998). A strong increase of iNOS activity was noted after thoracic spinal cord hemisection in both postoperative times studied. These data suggest that an increase of iNOS expression within polymorphonuclear leucocyte and macrophages (Xu et al., 2001; Chatzipanteli et al., 2002), appearing obvious from 3 h up to 14 days, is consistent with the inﬂammatory response. In CNS inﬂammation, inﬁltrated monocytes and macrophages are the primary source of iNOS, but the presence of endotoxins and proinﬂammatory cytokines induce similar response in astrocytes and microglia, the resident CNS immune cells (Murphy et al., 1993). After SCI, iNOS positivity has been observed and colocalized with macrophages, neurons, astrocytes, and oligodendrocytes (Kwak et al., 2005). Obviously, it is difﬁcult to specify unambiguously all the cell types that express iNOS in the damaged and hemorrhaged tissue invaded by leucocytes. Chatzipanteli et al. (2002) have identiﬁed polymorphonuclear leucocytes as a significant cellular source of iNOS protein (Chatzipanteli et al., 2002). Satake et al. (2000) has identiﬁed macrophages and perivascular cells as predominant iNOS positive cells after SCI. A rapid formation and release of NO by cNOS activity early after SCI (Hamada et al., 1996) and later, increased formation of NO via iNOS in the affected spinal cord region (Chatzipanteli et al., 2002) may accelerate neurodestructive events and lead to a secondary inactivation of cNOS enzyme activity, possibly by NO binding to the heme iron of the enzyme (Rogers and Ignarro, 1992) and/or by the phosphorylation of cNOS by proteinkinases (Dalkara and Moskowitz, 1997). Similarly, the reduction in blood ﬂow and decrease in oxygen tension caused by the structural damage of cell bodies might be responsible for decreased cNOS activity at the site of injury (Tator and Fehlings, 1991; Vanderkooi et al., 1991; Kim et al., 1993; Rengasamy and Johns, 1993). These data provide strong evidence that intrinsic properties of

177

neurons contribute to the different cellular responses. The changes in the constitutive and iNOS activities observed after spinal cord hemisection within the site of injury support the participation of NO in the pathology of SCI and result in a decreased neuronal viability. In principle, the primary injury of the spinal cord evokes an inﬂammatory response marked by generation of tumor necrosis factor-a, the accumulation and adhesion of neutrophils, and the activation of endothelial cells and macrophages at the site of injury (Taoka et al., 1997). These events, in turn, may contribute to neuronal death or injury by inducing an excessive formation of NO and oxygen free radicals (Merril et al., 1993). Another very important mediator of secondary CNS injury is oxidative-nitrosative stress (Hall and Braughler, 1993), induced by antioxidative depletion and/or an excess production of oxygen free radicals and NO at the site of injury. NO is able to promote oxidative damage by reacting with superoxide anion to form a strong oxidant, ONOO, and by perturbing ion metabolism (Beckman et al., 1990; Reif and Simmons, 1990). The rate of ONOO formation is three times faster than the dismutation of the superoxide anion to hydrogen peroxide, the reaction catalyzed by superoxide dismutase (SOD) (Dawson and Dawson, 1996). Thus, at appropriate concentrations NO can effectively compete with SOD for superoxide anion. In addition, NOS may generate ONOO directly by producing both NO and superoxide (Iadecola, 1997). NO can oxidize lipids, proteins, and nucleic acid (Dawson and Dawson, 1994) causing spinal cord cytotoxicity either directly through NO or its derived species. Taken together, such processes point to a possibility that NO represents a critical factor in determining the extent of the functional impairment of the spinal cord.

NADPHd-exhibiting and Fos-like immunolabeling of intrinsic spinal cord neurons in a model of MCEC We found considerable differences in the segmental and laminar distribution of Fos-like immunoreactive and NADPHd-stained neurons in the lower lumbar and sacral segments of the dog spinal cord using the model of MCEC (Orenda´cˇova´

et al., 2000, 2001). NADPHd histochemistry was used as a marker of NOS-containing neurons. The appearance and the time course of Fos-like immunoreactive, NADPHd, and double-labeled neurons was studied at 2 and 8 h postconstriction, characterized as the incipient phase of cauda equina syndrome. An increase in Fos-like immunoreactivity in superﬁcial laminae (I–II) and enhanced NADPHd staining of lamina VIII neurons were found. A statistically significant increase in Fos-like immunoreactive neurons was found in laminae I–II and VIII–X 8 h postconstriction, and in contrast, a prominent decrease in Fos-like immunoreactive neurons was found in laminae I–II, accompanied by a statistically significant increase in Fos-like immunoreactive neurons in more ventrally located laminae VII–X at 5 days postconstriction. A quantitative analysis of laminar distribution of constriction-induced NADPHd-stained neurons revealed a considerable increase in these neurons in laminae VIII–IX 8 h postconstriction and highly statistically significant increase in NADPHd-stained neurons in laminae VII–X 5 days postconstriction. Concurrently, the number of NADPHd-stained neurons in laminae I–II was greatly reduced. While a low number of double-labeled neurons was found throughout the gray matter of lower lumbar and sacral segments at 2 h postconstriction, a statistically significant number of double-labeled neurons was found in lamina X 8 h and in laminae VII–X 5 days postconstriction (Orenda´cˇova´ et al., 2000). A prominent involvement of the spinal cord neurons appearing in the lumbosacral segments at the beginning, and seen early (8 h postconstriction) and in fully developed cauda equina syndrome (5 days postconstriction), results in a Fos-like immunoreactivity and strongly enhanced NADPHd staining of some neuronal pools.

Immunolabeling and cNOS activity in the incipient cauda equina syndrome of the dog Remarkable changes appeared in the distribution and appearance of NOS-IR neurons in the lower lumbar and sacral segments analyzed 5 days postconstriction in a dog model of cauda equina

178

constriction (Marsˇ ala et al., 2003). While NOSimmunolabeled neurons detected in the upper lumbar segments in the intermediate zone and ventral horns were seen dispersed across both regions in a quite small number, passing more caudally, the number of NOS-immunolabeled neurons was steadily growing, reaching in the sacral segments not only to the lateral portion of the intermediate zone, but also encroaching to the lateral and ventralmost portion of the ventral horn (Orenda´cˇova´ et al., 2001). Along with this increasing density of NOS-IR neurons, several more or less loosely dispersed dotted NOS-IR foci appeared in the ventral horn and in the dorsal gray commissure of S1–S3 segments. Highly positive, usually small and middle-sized NOS-IR neurons could be detected around the periphery of the Onuf’s nucleus in S1–S3 segments. Under normal circumstances, such NOS-IR positivity could be seen only scarcely. It should be noted that increased dotted non-somatic NOS-IR positivity was often detected around interneurons in the core region of the ventral horn and in the intermediate zone. In some NOS-IR positive foci, the dark small proﬁles truly matched the outlines of these small neurons having, of course, no somatic NOS-IR positivity. These ﬁndings clearly document a high activation of NOS containing neurons in this compression-induced cauda equina injury. Considerable segmental and regional differences of cNOS activity were found in the gray and white matter regions of the lumbosacral segments of the dog 5 days after MCEC surgery (Luka´cˇova´ et al., 2004). The enzyme activity was assessed by a radioassay (Bredt and Snyder, 1990) in the gray matter, divided into dorsal horn, intermediate zone, ventral horn, and in the white matter, divided into dorsal, lateral, and ventral columns, respectively, comprising the upper (L1–L3) and lower (L4–L7) lumbar segments, then in the noncompartmentalized gray and white matter of S1–S3 segments and cauda equina. While a statistically significant decrease of cNOS activity was found in all gray matter regions of the upper lumbar segments, a statistically significant increase of cNOS activity could be detected in the ventral horn of the lower lumbar segments. With the exception of a statistically significant increase of

cNOS activity in the lateral funiculus of the upper lumbar segments, no changes of cNOS activity could be detected in the white matter funiculus of the upper and lower lumbar segments (Marsˇ ala et al., 2003). An increase of cNOS activity was detected in the non-compartmentalized white matter of S1–S3 segments, a ﬁnding strongly contrasting with a statistically significant decrease of cNOS activity assessed in the cauda equina. Therapeutic interventions and NO reduction after spinal trauma Therapies aimed at limiting the evolution of secondary damage have a long history and are still extensively studied both in experimental and clinical use. Intravenous steroids (methylprednisolone) have been registered for clinical use in acute SCI, although there is still considerable debate whether this therapy was really proved to be safe and efﬁcacious. Importantly, there is strong evidence that secondary damage is associated with invading inﬂammatory cells. Experimental results have shown that depletion of hematogeneous macrophages (Popovich et al., 1999) and inhibiting diapedesis of blood borne leucocytes by antibodies directed against leucocyte adhesion molecules (Mabon et al., 2000; Bao et al., 2004; Gris et al., 2004) reduces secondary damage after SCI, and improves motor, sensitive, and autonomic function recovery in experimental animals. However, as posttraumatic inﬂammation also has a potentially beneﬁcial role, better understanding of the molecular mechanisms and their timing in this process would allow more speciﬁc treatments. These might diminish destructive processes, while preserving the protective and regenerative functions of inﬂammatory cells. Evolving understanding of the role of posttraumatic NO changes represents one example of such therapeutic strategy reﬁnement. Experiments with genetically manipulated animals demonstrated improved recovery with both neuronal NOS-deﬁcient (Farooque et al., 2001) and iNOS-deﬁcient mice (Isaksson et al., 2005). Pharmacological inhibition of NO production by aminoguanidine has been shown to improve functional recovery in various independent studies

179

(Chatzipanteli et al., 2002; Zhang et al., 2003; Soy et al., 2004). The effect of a more speciﬁc iNOS inhibitor, 1400W has been tested in a traumatic brain injury model, where it dramatically reduced lesion volume, even if administered with 18 h delay (Jafarian-Tehrani et al., 2005). In conclusion, SCI is a very complex pathological situation. It is believed that effective therapy will require combined approaches for minimizing adverse reactions in the tissue and maximizing the regenerative capacity of the spinal cord (Ramer et al., 2000; Thuret et al., 2006). Precise understanding of the timing of iNOS expression in speciﬁc models of SCI, and development of new more speciﬁc inhibitors might represent one component of such effective therapy.

Abbreviations ANP cGMP CNS GC iNOS MCEC NADPHd NMDA NO NOS NOS-IR ONOO PDE pGC SCI sGC SOD

atrial natriuretic peptide cyclic guanosine 30 ,50 -monophosphate central nervous system guanylyl cyclase inducible nitric oxide synthase multiple cauda equina constriction nicotinamide adenine dinucleotide phosphate diaphorase N-methyl-D-aspartate nitric oxide nitric oxide synthase NOS-immunoreactive peroxynitrite phosphodiesterase particular guanylyl cyclase spinal cord injury soluble guanylyl cycle superoxide dismutase

Acknowledgments This work was supported by the APVT grant: 51013002 and VEGA grants: 2/5134/25, 2/5135/25, and 2/6212/26.

References Aley, K.O., McCarter, G. and Levine, J.D. (1998) Nitric oxide signaling in pain and nociceptor sensitization in the rat. J. Neurosci., 18: 7008–7014. Anderson, C.R. (1992) NADPH diaphorase-positive neurons in the rat spinal cord include a subpopulation of autonomic preganglionic neurons. Neurosci. Lett., 139: 280–284. Bao, F., Chen, Y., Dekaban, G.A. and Weaver, L.C. (2004) Early anti-inﬂammatory treatment reduces lipid peroxidation and protein nitration after spinal cord injury in rats. J. Neurochem., 88: 1335–1344. Barilari, M.G. and Kuypers, J.M. (1969) Propriospinal ﬁbers interconnecting the spinal enlargements in the cat. Brain Res., 14: 321–330. Beavo, J.A. (1995) Cyclic nucleotide phosphodiesterases: functional implications of multiple isoforms. Physiol. Rev., 75: 725–748. Beckman, J.S., Beckman, T.W., Chen, J., Marshall, P.A. and Freeman, B.A. (1990) Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide and superoxide. Proc. Natl. Acad. Sci. U.S.A., 87: 1620–1624. Blottner, D. and Baumgarten, H. (1992) Nitric oxide synthase (NOS) containing sympathoadrenal cholinergic neurons of the rat IML-cell column: evidence from histochemistry, immunohistochemistry and retrograde labeling. J. Comp. Neurol., 316: 45–56. Bredt, D.S., Hwang, P.M. and Snyder, S.H. (1990) Localization of nitric oxide synthase indicating a neuronal role for nitric oxide. Nature, 347: 768–770. Bredt, D.S. and Snyder, S.H. (1990) Isolation of nitric oxide synthase, a calmodulin-requiring enzyme. Proc. Natl. Acad. Sci. U.S.A., 87: 682–685. Bredt, D.S. and Snyder, S.H. (1992) Nitric oxide: a novel neuronal messenger. Neuron, 8: 3–11. Burnett, A.L., Saito, S., Maguire, M.P., Yamaguchi, H., Chang, T.S.K. and Hanley, D.F. (1995) Localization of nitric oxide synthase in spinal nuclei innervating pelvic ganglia. J. Urol., 153: 212–217. Chatzipanteli, K., Garcia, R., Marcillo, A.E., Loor, K.E., Kraydieh, S. and Dietrich, W.D. (2002) Temporal and segmental distribution of constitutive and inducible nitric oxide synthases after traumatic spinal cord injury: effect of aminoguanidine treatment. J. Neurotrauma, 5: 639–651. Dalkara, T. and Moskowitz, M.A. (1997) Neurotoxic and neuroprotective roles of nitric oxide in cerebral ischaemia. Int. Rev. Neurobiol., 40: 319–336. Dawson, T.M. and Dawson, V.L. (1994) gp120 neurotoxicity in primary cortical cultures. Adv. Neuroimmunol., 4: 167–173. Dawson, T.M., Dawson, V.L. and Snyder, S.H. (1992) A novel neuronal messenger molecule in brain: the free radical, nitric oxide. Ann. Neurol., 32: 297–311. Dawson, V.L. and Dawson, T.M. (1996) Nitric oxide actions in neurochemistry. Neurochem. Int., 29: 97–110. Diaz-Ruiz, A., Ibarra, A., Perez-Severiano, F., Guizar-Sahagun, G., Grijalva, I. and Rios, C. (2002) Constitutive and inducible

180 nitric oxide synthase activities after spinal cord contusion in rats. Neurosci. Lett., 319: 129–132. Dun, N.J., Dun, S.L., Wu, S.Y., Fo¨rstermann, U., Schmidt, H.H.H.W. and Tseng, L.F. (1993) Nitric oxide synthase immunoreactivity in the rat, mouse, cat and squirrel monkey spinal cord. Neuroscience, 54: 845–857. Egberongbe, Y.I., Gentleman, S.M., Falkai, P., Bogerts, B., Polak, J.M. and Roberts, G.W. (1994) The distribution of nitric oxide synthase immunoreactivity in the human brain. Neuroscience, 59: 561–578. Farooque, M., Isaksson, J. and Olsson, Y. (2001) Improved recovery after spinal cord injury in neuronal nitric oxide synthase-deﬁcient mice but not in TNF-alpha-deﬁcient mice. J. Neurotrauma, 18: 105–114. Fedele, E., Jin, Y., Varnier, G. and Raiteri, M. (1996) In vivo microdialysis study of a speciﬁc inhibitor of soluble guanylyl cyclase on the glutamate receptor/nitric oxide/cyclic GMP pathway. Br. J. Pharmacol., 119: 590–594. Francis, S.H., Turko, I.V. and Corbin, J.D. (2001) Cyclic nucleotide phosphodiesterases: relating structure and function. Prog. Nucleic Acid Res. Mol. Biol., 65: 1–52. Gris, D., Marsh, D.R., Oatway, M.A., Chen, Y., Hamilton, E.F., Dekaban, G.A. and Weaver, L.C. (2004) Transient blockade of the CD11d/CD18 integrin reduces secondary damage after spinal cord injury, improving sensory, autonomic, and motor function. J. Neurosci., 24: 4043–4051. Guizar-Sahagun, G., Garcia-Lopez, P., Espitia, A.L., Grijalva, I., Franco-Bourland, R.E. and Madrazo, I. (1998) Transitory expression of NADPH diaphorase (NOS) in axonal swellings after spinal cord injury. NeuroReport, 9: 2899–2902. Gupta, G., Azam, M., Yang, L. and Danziger, R.S. (1997) The beta2 subunit inhibits stimulation of the alpha1/beta1 form of soluble guanylyl cyclase by nitric oxide: potential relevance to regulation of blood pressure. J. Clin. Invest., 100: 1488–1492. Hala´t, G., Luka´cˇova´, N., Chavko, M. and Marsˇ ala, J. (1987) Effects of incomplete ischemia and subsequent recirculation on free palmitate, stearate, oleate and arachidonate levels in lumbar and cervical spinal cord of rabbit. Gen. Physiol. Biophys., 6: 387–399. Hall, E.D. and Braughler, J.M. (1993) Free radicals in CNS injury. Res. Publ. Assoc. Res. Nerv. Ment. Dis., 71: 81–105. Hamada, Y., Ikata, T., Katoh, S., Tsuchiya, K., Niwa, M., Tsutsumishita, Y. and Fukuzawa, K. (1996) Roles of nitric oxide in compression injury of rat spinal cord. Free Radic. Biol. Med., 20: 1–9. Horrocks, L.A., Dorman, R.V. and Porcellati, G. (1984) Fatty acids and phospholipids in brain during ischemia. In: Bes P., Braquet R., Paoletti R. and Siesjo¨ B.K. (Eds.), Cerebral Ischemia Excerpta Medica. Elsevier, Amsterdam, Congress Series 654A, pp. 211–226. Humbert, P., Niroomand, F., Fischer, G., Mayer, B., Koesling, D., Hinsch, K.D., Gausepohl, H., Frank, R., Schultz, G. and Bohme, E. (1990) Puriﬁcation of soluble guanylyl cyclase from bovine lung by a new immunoafﬁnity chromatographic method. Eur. J. Biochem., 190: 273–278.

Iadecola, C. (1997) Bright and dark sides of nitric oxide in ischemic brain injury. Trends Neurosci., 20: 132–139. Isaksson, J., Farooque, M. and Olsson, Y. (2005) Improved functional outcome after spinal cord injury in iNOS-deﬁcient mice. Spinal Cord, 43: 167–170. Jafarian-Tehrani, M., Louin, G., Royo, N.C., Besson, V.C., Bohme, G.A., Plotkine, M. and Marchand-Verrecchia, C. (2005) 1400W, a potent selective inducible NOS inhibitor: improves histopathological outcome following traumatic brain injury in rats. Nitric Oxide, 12: 61–69. Kim, N., Vardi, Y., Padma-Nathan, H., Daley, J., Goldstein, I. and Saenz de Tejada, I. (1993) Oxygen tension regulates the nitric oxide pathway: physiological role in penile erection. J. Clin. Invest., 91: 437–442. Koesling, D. and Friebe, A. (2000) Enzymology of soluble guanylyl cyclase. In: Mayer B. (Ed.), Nitric Oxide. Springer, Berlin, pp. 93–109. Koglin, M., Vehse, K., Budaeus, L., Scholz, H. and Behrends, S. (2001) Nitric oxide activates the beta 2 subunit of soluble guanylyl cyclase in the absence of a second subunit. J. Biol. Chem., 276: 30737–30743. Kowall, N.W., Beal, M.F., Ferrante, R.J. and Martin, J.B. (1985) Topography of nicotinamide adenine dinucleotide phosphate-diaphorase staining neurons in rat striatum. Neurosci. Lett., 59: 61–66. Kowall, N.W., Ferrante, R.J., Beal, M.F., Richardson, Jr., E.P., Sofroniew, M.V., Cuello, A.C. and Martin, J.B. (1987) Neuropeptide Y, somatostatin and reduced nicotinamide adenine dinucleotide phosphate diaphorase in the human striatum: a combined immunocytochemical and enzyme histochemical study. Neuroscience, 20: 817–828. Kowall, N.W. and Mueller, M.O. (1988) Morphology and distribution of nicotinamide adenine dinucleotide phosphate (reduced form) diaphorase reactive neurons in human brainstem. Neuroscience, 26: 645–654. Kwak, E.K., Kim, J.W., Kang, K.S., Lee, Y.H., Hua, Q.H., Park, T.I., Park, J.Y. and Sohn, Y.K. (2005) The role of inducible nitric oxide synthase following spinal cord injury in rat. J. Korean Med. Sci., 20: 663–669. Lee, J.H., Price, R.H., Williams, F.G., Mayer, B. and Beitz, A.J. (1993) Nitric oxide synthase is found in some spinothalamic neurons and in neuronal processes that appose spinal neurons that express Fos induced by noxious stimulation. Brain Res., 608: 324–333. Lehotsky´, J., Kapla´n, P., Racˇay, P., Me´zesˇ ova´, V. and Raeymaekers, L. (1999) Distribution of plasma membrane Ca2+ pump (PMCA) isoforms in the gerbil brain: effect of ischemia-reperfusion injury. Neurochem. Int., 35: 221–227. Liu, D., Ling, X., Wen, J. and Liu, J. (2000) The role of reactive nitrogen species in secondary spinal cord injury: formation of nitric oxide, peroxynitrite, and nitrated protein. J. Neurochem., 75: 2144–2154. de Louw, A.J., de Vente, J., Steinbusch, H.P., Steinbusch, H.W., Troost, J. and Vles, J.S. (2002) Baclofen inhibits ANPmediated cyclic GMP synthesis in the rat cervical spinal cord. Neurosci. Lett., 321: 120–122.

181 Lucas, K.A., Pitari, G.M., Kazerounian, S., Ruiz-Stewart, I., Park, J., Schulz, S., Chepenik, K.P. and Waldman, S.A. (2000) Guanylyl cyclases and signaling by cyclic GMP. Pharmacol. Rev., 52: 375–414. Luka´cˇova´, N., Chavko, M. and Marsˇ ala, J. (1994) Ischemia/ reperfusion-induced lipid peroxidation in the central nervous system. Biologia, 49: 871–879. Luka´cˇova´, N., Cˇı´ zˇkova´, D., Marsˇ ala, M., Jalcˇ, P. and Marsˇ ala, J. (1999) Segmental and laminar distributions of nicotinamide adenine dinucleotide phosphate–diaphorase expressing and neuronal nitric oxide synthase–immunoreactive neurons versus radioassay detection of catalytic nitric oxide synthase activity in the rabbit spinal cord. Neuroscience, 94: 229–237. Luka´cˇova´, N., Cˇı´ zˇkova´, D., Marsˇ ala, M., Pavel, J., Jalcˇ, P., Sˇulla, I., Kafka, J. and Marsˇ ala, J. (2000) Effect of midthoracic spinal cord constriction on catalytic nitric oxide synthase activity in the white matter columns of rabbit. Neurochem. Res., 25: 1139–1148. Luka´cˇova´, N., Kafka, J., Cˇı´ zˇkova´, D., Marsˇ ala, M. and Marsˇ ala, J. (2004) The effect of cauda equina constriction on nitric oxide synthase activity. Neurochem. Res., 29: 429–439. Luka´cˇova´, N., Kolesa´rova´, M., Kucha´rova´, K., Pavel, J., Kolesa´r, D., Radonˇa´k, J., Marsˇ ala, M., Chalimoniuk, M., Langfort, J. and Marsˇ ala, J. (2006) The effect of a spinal cord hemisection on changes in nitric oxide synthase pools in the site of injury and in regions located far away from the injured site. Cell. Mol. Neurobiol., 26: 1365–1383. Luka´cˇova´, N. and Pavel, J. (2000) Catalytic nitric oxide synthase activity in the white and gray matter regions of the spinal cord of rabbits. Physiol. Res., 49: 167–173. Luka´cˇova´, N., Pavel, J., Jalcˇ, P., Cˇı´ zˇkova´, D., Marsˇ ala, M., Silvasˇ iova´, M., Luka´cˇ, I. and Marsˇ ala, J. (2002) The effect of spinal cord constriction on NOS-cGMP signaling pathway in the white matter of rabbit. Biologia, 57: 789–795. Luka´cˇova´, N., Pavel, J., Kucha´rova´, K. and Marsˇ ala, J. (2005) Plasticity of the spinal neural circuitry in nitric oxide synthase immunoreactive pathways after spinal cord hemisection. In: Coleman R.M. (Ed.), Trends in Neurochemistry Research. Nova Publishers, New York, pp. 189–222. Luo, D., Leung, E. and Vincent, S.R. (1994) Nitric oxidedependent efﬂux of cGMP in rat cerebellar cortex: an in vivo microdialysis study. J. Neurosci., 14: 263–271. Mabon, P.J., Weaver, L.C. and Dekaban, G.A. (2000) Inhibition of monocyte/macrophage migration to a spinal cord injury site by an antibody to the integrin alphaD: a potential new anti-inﬂammatory treatment. Exp. Neurol., 166: 52–64. Marsˇ ala, J., Kafka, J., Luka´cˇova´, N., Cˇı´ zˇkova´, D., Marsˇ ala, M. and Katsube, N. (2003) Cauda equina syndrome and nitric oxide synthase immunoreactivity in the spinal cord of the dog. Physiol. Res., 52: 481–496. Marsˇ ala, J., Kluchova´, D. and Marsˇ ala, M. (1997) Spinal cord gray matter layers rich in NADPH diaphorase positive neurons are refractory to ischemia-reperfusion-induced injury: a histochemical and silver impregnation study in rabbit. Exp. Neurol., 145: 165–179.

Marsˇ ala, J., Luka´cˇova´, N., Cˇı´ zˇkova´, D., Luka´cˇ, I., Kucha´rova´, K. and Marsˇ ala, M. (2004) Premotor nitric oxide synthase immunoreactive pathway connecting lumbar segments with the ventral motor nucleus of the cervical enlargement in the dog. J. Chem. Neuroanat., 27: 43–54. Marsˇ ala, J., Marsˇ ala, M., Vanicky´, I. and Taira, Y. (1999) Localization of NADPHd exhibiting neurons in the spinal cord of the rabbit. J. Comp. Neurol., 406: 263–284. Marsˇ ala, J., Vanicky´, I., Marsˇ ala, M., Jalcˇ, P., Orenda´cˇova´, J. and Taira, Y. (1998) NADPH diaphorase in the spinal cord of dogs. Neuroscience, 85: 847–862. Mc Adoo, J., Hughes, M.G., Xu, G.Y., Robak, G. and DeCastro, Jr., R., (1997) Microdialysis studies of the role of chemical agents in secondary damage upon spinal cord injury. J. Neurotrauma, 14: 507–515. Merril, J.E., Ignarro, L., Sherman, M.P., Melinek, J. and Lane, T.E. (1993) Microglial cell cytotoxicity of oligodendrocytes is mediated through nitric oxide. J. Immunol., 151: 2132–2141. Mizukawa, K., McGeer, P.L., Vincent, S.R. and McGeer, E.G. (1989) Distribution of reduced-nicotinamide-adenine-dinucleotide-phosphate diaphorase-positive cells and ﬁbers in the cat central nervous system. J. Comp. Neurol., 279: 281–311. Morris, R., Southam, E., Gittins, S.R., de Vente, J. and Garthwaite, J. (1994) The NO-cGMP pathway in neonatal rat dorsal horn. Eur. J. Neurosci., 6: 876–879. Morton, D.B. and Bredt, D.S. (1998) Norepinephrine increases cyclic GMP levels in cerebellar cells from neuronal nitric oxide synthase knockout mice. J. Neurochem., 71: 440–443. Murad, F. (1994) Regulation of cytosolic guanylyl cyclase by nitric oxide: the NO-cyclic GMP signal transduction system. Adv. Pharmacol., 26: 19–33. Murphy, S., Simmons, M.L., Agullo, L., Garcia, A., Feinstein, D.L., Galea, E., Reis, D.J., Minc-Golomb, D. and Schwartz, J.P. (1993) Synthesis of nitric oxide in CNS glial cells. Trends Neurosci., 16: 323–328. Nakahara, S., Yone, K., Setoguchi, T., Yamaura, I., Arishima, Y., Yoshino, S. and Komiya, S. (2002) Changes in nitric oxide and expression of nitric oxide synthase in spinal cord after acute traumatic injury in rats. J. Neurotrauma, 19: 1467–1474. Nakano, S., Kogure, K., Abe, K. and Yae, T. (1990) Ischemiainduced alterations in lipid metabolism of the gerbil cerebral cortex: I. Changes in free fatty acid liberation. J. Neurochem., 54: 1911–1916. Orenda´cˇova´, J., Cˇı´ zˇkova´, D., Kafka, J., Luka´cˇova´, N., Marsˇ ala, M., Sˇulla, I., Marsˇ ala, J. and Katsube, N. (2001) Cauda equina syndrome. Prog. Neurobiol., 64: 613–637. Orenda´cˇova´, J., Marsˇ ala, M., Sˇulla, I., Kafka, J., Jalcˇ, P., Cˇı´ zˇkova´, D., Taira, Y. and Marsˇ ala, J. (2000) Incipient cauda equina syndrome as a model of somatovisceral pain in dogs: spinal cord structures involved as revealed by the expression of c-fos and NADPH diaphorase activity. Neuroscience, 95: 543–557. Popovich, P.G., Guan, Z., Wei, P., Huitinga, I., van Rooijen, N. and Stokes, B.T. (1999) Depletion of hematogenous macrophages promotes partial hindlimb recovery and

182 neuroanatomical repair after experimental spinal cord injury. Exp. Neurol., 158: 351–365. Potter, L.R. and Hunter, T. (2001) Guanylyl cyclase-linked natriuretic peptide receptors: structure and regulation. J. Biol. Chem., 276: 6057–6060. Prast, H. and Philippu, A. (2001) Nitric oxide as modulator of neuronal function. Prog. Neurobiol., 64: 51–68. Ramer, M.S., Harper, G.P. and Bradbury, E.J. (2000) Progress in spinal cord research: a reﬁned strategy for the International Spinal Research Trust. Spinal Cord, 38: 449–472. Ransom, B.R., Waxman, S.G. and Stys, P.K. (1993) Anoxic injury of central myelinated axons: ionic mechanisms and pharmacology. Res. Publ. Assoc. Res. Nerv. Ment. Dis., 71: 121–151. Reif, D.W. and Simmons, R.D. (1990) Nitric oxide mediates iron release from ferritin. Arch. Biochem. Biophys., 283: 537–541. Rengasamy, A. and Johns, R.A. (1993) Regulation of nitric oxide synthase by nitric oxide. Mol. Pharmacol., 44: 124–128. Rogers, N.E. and Ignarro, L.J. (1992) Constitutive nitric oxide synthase from cerebellum is reversibly inhibited by nitric oxide formed from L-arginine. Biochem. Biophys. Res. Commun., 189: 242–249. Russwurm, M., Behrends, S., Harteneck, C. and Koesling, D. (1998) Functional properties of a naturally occurring isoform of soluble guanylyl cyclase. Biochem. J., 335: 125–130. Saito, S., Kidd, G.J., Trapp, B.D., Dawson, T.M., Bredt, D.S., Wilson, D.A., Traystmann, R.J., Snyder, S.H. and Hanley, D.F. (1994) Rat spinal cord neurons contain nitric oxide synthase. Neuroscience, 59: 447–452. Satake, K., Matsuyama, Y., Kamiya, M., Kawakami, H., Iwata, H., Adachi, K. and Kiuchi, K. (2000) Nitric oxide via macrophage iNOS induces apoptosis following traumatic spinal cord injury. Brain Res. Mol. Brain Res., 85: 114–122. Schmidt, H.H. (1992) NO, CO and OH: endogenous soluble guanylyl cyclase-activating factors. FEBS Lett., 307: 102–107. Sharma, H.S., Westman, J., Gordh, T. and Nyberg, F. (2000) Spinal cord injury induced c-fos expression is reduced by p-CPA, a serotonin synthesis inhibitor: an experimental study using immunohistochemistry in the rat. Acta Neurochir. Suppl., 76: 297–301. Sherrington, C.S. and Laslett, E.E. (1903) Observations on some spinal reﬂexes and the inter connection of spinal segments. J. Physiol. (Lond.), 29: 58–96. Smolenski, A., Burkhardt, A.M., Eigenthaler, M., Butt, E., Gambaryan, S., Lohmann, S.M. and Walter, U. (1998) Functional analysis of cGMP-dependent protein kinases I and II as mediators of NO/cGMP effects. Naunyn Schmiedebergs Arch. Pharmacol., 358: 134–139. Snyder, S.H. and Bredt, D. (1991) Nitric oxide as a neuronal messenger. Trends Pharmacol. Sci., 12: 125–128. Soy, O., Aslan, O., Uzun, H., Barut, S., Igdem, A.A., Belce, A. and Colak, A. (2004) Time-level relationship for nitric oxide and the protective effects of aminoguanidine in experimental spinal cord injury. Acta Neurochir. (Wien.), 146: 1329–1336. van Staveren, W.C., Steinbusch, H.W., Markerink-van Ittersum, M., Repaske, D.R., Goy, M.F., Kotera, J., Omori, K.,

Beavo, J.A. and de Vente, J. (2003) mRNA expression patterns of the cGMP-hydrolyzing phosphodiesterases types 2, 5, and 9 during development of the rat brain. J. Comp. Neurol., 467: 566–580. Strosznajder, J. and Chalimoniuk, M. (1996) Biphasic enhancement of nitric oxide synthase activity and cGMP level following brain ischemia in gerbils. Acta Neurobiol. Exp., 56: 71–81. Stys, P.K. and Steffensen, I. (1996) Na(+)-Ca2+ exchange in anoxic/ischemic injury of CNS myelinated axons. Ann. N.Y. Acad. Sci., 779: 366–378. Stys, P.K., Waxman, S.G. and Ransom, B.R. (1991) Na+-Ca2+ exchanger mediates Ca2+ inﬂux during anoxia in mammalian central nervous system white matter. Ann. Neurol., 30: 375–380. Tao, Y.X., Hassan, A., Haddad, E. and Johns, R.A. (2000) Expression and action of cyclic GMP-dependent protein kinase I alpha in inﬂammatory hyperalgesia in rat spinal cord. Neuroscience, 95: 525–533. Taoka, Y., Okajima, K., Uchiba, M., Murakami, K., Kushimoto, S., Johno, M., Naruo, M., Okabe, H. and Takatsuki, K. (1997) Role of neutrophils in spinal cord injury in the rat. Neuroscience, 79: 1177–1182. Tator, C.H. and Fehlings, M.G. (1991) Review of the secondary injury theory of acute spinal cord trauma with emphasis on vascular mechanisms. J. Neurosurg., 75: 15–26. Thomas, E. and Pearse, A.G.E. (1964) The solitary active cells: histochemical demonstration of damage-resistant nerve cells with a TPN-diaphorase reaction. Acta Neuropathol., 3: 238–249. Thuret, S., Moon, L.D. and Gage, F.H. (2006) Therapeutic interventions after spinal cord injury. Nat. Rev. Neurosci., 7: 628–643. Toborek, M., Garrido, R., Malecki, A., Kaiser, S., Mattson, M.P., Hennig, B. and Young, B. (2000) Nicotine attenuates arachidonic acid-induced overexpression of nitric oxide synthase in cultured spinal cord neurons. Exp. Neurol., 161: 609–620. Trudrung, P., Wirth, U. and Mense, S. (2000) Changes in the number of nitric oxide-synthesizing neurons on both sides of a chronic transection of the rat spinal cord. Neurosci. Lett., 287: 125–128. Vallebuona, F. and Raiteri, M. (1993) Monitoring of cyclic GMP during cerebellar microdialysis in freely-moving rats as an index of nitric oxide synthase activity. Neuroscience, 57: 577–585. Valtschanoff, J.G., Weinberg, R.J. and Rustioni, A. (1992) NADPH diaphorase in the spinal cord of rats. J. Comp. Neurol., 321: 209–222. Vanderkooi, J.M., Erecinska, M. and Silver, I.A. (1991) Oxygen in mammalian tissue: methods of measurement and afﬁnities of various reactions. Am. J. Physiol., 260: C1131–C1150. de Vente, J., Manshanden, C.G., Sikking, R.A., Ramaekers, F.C. and Steinbusch, H.W. (1990) A functional parameter to study heterogeneity of glial cells in rat brain slices: cyclic guanosine monophosphate production in atrial natriuretic factor (ANF)-responsive cells. Glia, 3: 43–54.

183 Vincent, S.R. (1994) Nitric oxide: a radical neurotransmitter in the central nervous system. Prog. Neurobiol., 42: 129–160. Vincent, S.R. and Kimura, H. (1992) Histochemical mapping of nitric oxide synthase in the rat brain. Neuroscience, 46: 755–784. Vizzard, M.A., Erdman, S.L. and de Groat, W.C. (1995) Increased expression of neuronal nitric oxide synthase (NOS) in visceral neurons after nerve injury. J. Neurosci., 15: 4033–4045. Vizzard, M.A., Erdman, S.L., Erickson, V.L., Stewart, R.J., Roppolo, J.R. and de Groat, W.C. (1994a) Localization of NADPH diaphorase in the lumbosacral spinal cord and dorsal root ganglia of the cat. J. Comp. Neurol., 339: 62–75. Vizzard, M.A., Erdman, S.L., Roppolo, J.R., Fo¨rstermann, U. and de Groat, W.C. (1994b) Differential localization of neuronal nitric oxide synthase immunoreactivity and NADPHdiaphorase activity in the cat spinal cord. Cell. Tissue Res., 278: 299–309. Vles, J.S., de Louw, A.J., Steinbusch, H., Markerink-van Ittersum, M., Steinbusch, H.W., Blanco, C.E., Axer, H., Troost, J. and de Vente, J. (2000) Localization and age-related changes of nitric oxide- and ANP-mediated cyclic-GMP synthesis in rat cervical spinal cord: an immunocytochemical study. Brain Res., 857: 219–234. Wang, X. and Robinson, P.J. (1997) Cyclic GMP-dependent protein kinase and cellular signaling in the nervous system. J. Neurochem., 68: 443–456. Wedel, B.J. and Garbers, D.L. (1997) New insights on the functions of the guanylyl cyclase receptors. FEBS Lett., 410: 29–33.

Wetts, R. and Vaughn, J.E. (1994) Choline acetyltransferase and NADPH diaphorase are co-expressed in rat spinal cord neurons. Neuroscience, 63: 1117–1124. Wood, P.L. and Garthwaite, J. (1994) Models for the diffusional spread of nitric oxide: implications for neuronal nitric oxide signalling and its pharmacological properties. Neuropharmacology, 33: 1235–1244. Wu, W., Liuzzi, F.J., Schinco, F.P., Depto, A.S., Li, Y., Mong, J.A., Dawson, T.M. and Snyder, S.H. (1994) Neuronal nitric oxide synthase is induced in spinal neurons by traumatic injury. Neuroscience, 61: 719–726. Xu, J., Kim, G.M., Chen, S., Yan, P., Ahmed, S.H., Ku, G., Beckman, J.S., Xu, X.M. and Hsu, C.Y. (2001) iNOS and nitrotyrosine expression after spinal cord injury. J. Neurotrauma, 18: 523–532. Xu, W., Liu, L., Smith, G.C. and Charles, G. (2000) Nitric oxide upregulates expression of DNA-PKCs to protect cells from DNA-damaging anti-tumour agents. Nat. Cell. Biol., 2: 339–345. Yezierski, R.P., Culberson, J.L. and Brown, P.B. (1980) Cells of origin of propriospinal connections to cat lumbosacral gray as determined with horseradish peroxidase. Exp. Neurol., 69: 493–512. Yick, L.W., Wu, W., So, K.F. and Wong, S.Y. (1998) Time course of NOS expression and neuronal death in Clarke’s nucleus following traumatic injury in adult rat spinal cord. Neurosci. Lett., 241: 155–158. Zhang, X.Y., Zhou, C.S., Jin, A.M., Tian, J., Zhang, H., Yao, W.T. and Zheng, G. (2003) Effect of aminoguanidine on the recovery of rat hindlimb motor function after spinal cord injury. Di Yi Jun Yi Da Xue Xue Bao, 23: 687–689.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 12

Aquaporins: role in cerebral edema and brain water balance$ Zsolt Zador1,2, Orin Bloch1,2, Xiaoming Yao1,2 and Geoffrey T. Manley1,2, 1

Department of Neurological Surgery, University of California, San Francisco, CA, USA Brain and Spinal Injury Center, University of California, San Francisco, CA 94110, USA

2

Abstract: The regulation of water balance in the brain is crucial. A disruption in this equilibrium causes an increase in brain water content that signiﬁcantly contributes to the pathophysiology of traumatic brain injury, hydrocephalus, and a variety of neurological disorders. The discovery of the aquaporin (AQP) family of membrane water channels has provided important new insights into the physiology and pathology of brain water homeostasis. A number of recent studies are described in the review that demonstrated the important role of AQP1 and AQP4 in brain water balance and cerebral edema. Phenotypic analyses of AQP deﬁcient mice have allowed us to explore the role of these membrane water channels in the mechanisms of cytotoxic edema, vasogenic edema, and CSF production. These studies indicate that AQP4 plays signiﬁcant role in the development of cytotoxic edema and the absorption of excess brain water resulting from vasogenic edema. They also have demonstrated the role of AQP1 in CSF production and maintenance of steady-state ICP. The ability to modulate water ﬂux through AQP deletion has provided new insights into brain water homeostasis and suggested a number of new research directions. However, these efforts have not yet translated to the treatment human clinical diseases. These advances will require the development of AQP inhibitors and activators to establish the beneﬁt modulating the function of these water channels. Keywords: water transport; aquaporin; edema; brain injury; stroke; epilepsy; cytotoxic; vasogenic Introduction

water transport (Verkman, 2002). To date, over 12 different members of the AQP family have been identiﬁed, and a number of them have been shown to contribute to rapid water ﬂux in various tissues including the kidney, lung, gastrointestinal tract, and central nervous system (CNS) (Agre et al., 2002; Verkman, 2002). The regulation of water balance in the brain is crucial. Normally, water transport is tightly regulated to maintain a strict homeostatic balance between the cerebral vascular, brain tissue, and cerebrospinal ﬂuid (CSF) compartments. A disruption in this equilibrium causes an increase in brain water content that signiﬁcantly contributes to the pathophysiology of traumatic brain injury,

Water is regarded as the matrix of life: it is an essential solvent for all living organisms. Although it diffuses through tissues relatively freely, the transport of water across cell membranes is facilitated by highly efﬁcient transmembrane channels called aquaporins (AQPs). The selective water ﬂow through AQPs is passive, moving along osmotic gradients, which provide the driving force for $

Supported by NIH NS050173 and the UCSF Brain and Spinal Injury Center.

Corresponding author. Tel.: +415-206-4536; Fax: +415-206-

3948; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61012-1

185

186

hydrocephalus, and a variety of neurological disorders (Fishman, 1975). The rigid nature of the skull provides little capacity to buffer intracranial volume changes, and beyond a limited threshold the intracranial pressure (ICP) rises rapidly. The increased ICP can ultimately lead to the impairment of cerebral blood ﬂow resulting in further brain injury and death. In spite of the clinical importance of brain water balance, the molecular mechanisms remain poorly understood and the therapeutic interventions have improved little over past 80 years (Weed and McKibben, 1919). A number of recent studies have demonstrated that the AQP water channels play a signiﬁcant role in brain water balance and cerebral edema. Thus, they may provide a novel molecular target for the treatment of a number of neurological disorders. The aim of this review is to summarize the current understanding of AQPs in cerebral edema and brain water balance.

Aquaporin expression in the CNS A number of AQPs are expressed in the brain and spinal cord. Two of these water channels, AQP1 and AQP4, are expressed in the brain at tissue– ﬂuid interfaces important for brain water balance. AQP1 is expressed in the apical membrane of the choroid plexus epithelium (Hasegawa et al., 1994; Speake et al., 2003), where it has been shown to facilitate CSF secretion into the cerebral ventricles (Oshio et al., 2005). AQP4 is expressed in the basolateral membrane of the ependymal cells lining the cerebral ventricles, where it may play a role in water transport between the brain parenchyma and CSF compartments. In addition, AQP4 is expressed in the in the astocytic foot processes in direct contact with the cerebral blood vessels that comprise the blood-brain barrier. Highly polarized AQP4 expression is also found in the dense astrocyte cell processes that form the glia limitans, a structure that is adjacent to the CSF-ﬁlled subarachnoid space. The expression of AQP1 and AQP4 at these tissue–ﬂuid interfaces indicate a potential role in maintaining brain water homeostasis by facilitating water ﬂux between these compartments. Furthermore, their

expression pattern suggests they may contribute to the underlying mechanism of cerebral edema.

Aquaporins and brain water balance The brain is composed of nearly 78% water (McIlwain and Bachelard, 1985). While the majority of the brain water is intracellular, it is also distributed in the extracellular space and the CSF compartments (cerebral ventricles and subarachnoid space). Water enters the brain through the blood-brain barrier or across the choroids plexus (Fig. 1). Under normal physiological conditions, the extracellular ﬂuid in the brain is believed to ﬂow into the cerebral ventricles or subarachnoid space and exit the brain through the arachnoid granulations into to the venous system (Abbott, 2004). Recent studies of transgenic mice lacking AQP1 and AQP4 have helped to deﬁne the role of these channels in CSF production and absorbtion. The choroid plexus epithelium expresses a variety of ion transporters and channels that provide the osmotic driving force for para- and transcellular water ﬂux into the cerebral ventricles to form CSF (Wright et al., 1977). The rate of CSF secretion by the choroid plexus is substantial, comparable to the rate of ﬂuid reabsorption in the proximal tubules (Welch, 1963). Earlier studies had demonstrated a high osmotic–diffusion ratio of the choroids plexus, suggesting the presence of a water pore-like pathway (Wright et al., 1977). To explore this hypothesis, our group developed novel methods to measure choroids plexus water permeability and CSF production in wildtype and AQP1-null mice. We found that deletion of AQP1 reduced osmotically induced water transport in the choroid plexus by ﬁvefold. Interestingly, CSF production was signiﬁcantly reduced in AQP1-null mice, but only by 20–25%, indicating a substantial contribution of extrachoroidal ﬂuid production by the brain parenchyma. This result supported previous studies in which removal of the choroid plexus in rhesus monkeys (Milhorat et al., 1971) and dogs (Bering and Sato, 1963) resulted in only a 30–50% decrease in the CSF production.

187

Dura

BRAIN SURFACE

Arachnoid Pia

Arachnoid granulation

Sinus

2

1

AQP1 Ependyma

AQP4 Choroid plexus

Endothelial Astrocyte cell foot processes

EXTRACELLULAR SPACE

VENTRICLE

3 PARENCHYMA

P V

Fig. 1. Aquaporins and brain water balance. The production and circulation of cerebrospinal ﬂuid (CSF) in the brain is facilitated by the highly polarized expression of aquaporin-1 (AQP1) and aquaporin-4 (AQP4). (1) Water transport from the vasculature into the ventricle is facilitated by AQP1 expressed in the apical membrane of the choroid plexus, and via AQP4 in the ependymal lining. (2) Another potential pathway for water ﬂux into the subarachnoid space via the astrocytic processes lining the pial membrane that heavily express AQP4. Fluid from the subarachnoid space is drained through the arachnoid granulations into the low-pressure venous sinus that exits the cranium. (3) Extracellular ﬂuid may alternatively pass through the perivascular processes of astrocytes as another pathway to drain ﬂuid into the vascular compartment.

The rates of CSF production and absorption must be equal in the steady-state. Hydrocephalus and increased ICP can occur when the balance between CSF production and absorption is disturbed. A reversal of CSF ﬂow can be seen in experimental models of obstructive hydrocephalus, where outﬂow-obstruction drives CSF into the brain parenchyma causing extracellular edema (Hiratsuka et al., 1982; Braun et al., 1997). Given the widespread expression of AQP4 at the brain–CSF interface, we examined whether this water channel might facilitate the transparenchymal absorption of CSF when the normal ventricular outﬂow pathways are blocked (Bloch et al., 2006). Using a mouse model of obstructive hyprocephalus, we found that AQP4-null mice had an accelerated course of ventricular enlargement and ICP elevation compared with wildtype mice, indicating a role for AQP4 in CSF absorption. We hypothesized that AQP4 most likely participates in clearance of excess parenchymal ﬂuid seen in this model.

These phenotypic studies of mice lacking AQP1 and AQP4 provide evidence for the role of these water channels in the physiology of brain water homeostasis under normal and abnormal conditions (Oshio et al., 2003; Manley et al., 2004). A key feature in a number of neurological disorders is an imbalance in brain water homeostasis that leads to cerebral edema (Fishman, 1975). The remainder of the review focuses on the compensatory and maladaptive mechanisms linked to these water channels and the discussion of their potential as targets for therapeutic interventions in brain edema.

Cerebral edema Cerebral edema is deﬁned as an abnormal increase in brain water content. Klatzo (1994) described two distinct forms of cerebral edema based their pathogenesis (Fig. 2): vasogenic (extracellular) and cytotoxic (intracellular) brain edema. In most

188

A

B

Cytotoxic Swollen foot processes

Vasogenic Capillary lumen

Astrocyte foot process

TJ

Capillary lumen

Endothelial cell

Endothelial cell

Fig. 2. Principles of cytotoxic and vasogenic edema. (A) The pathologic swelling of astrocytic foot processes is the main contributor to cytotoxic edema. The tight junctions (TJ) help to establish the blood-brain barrier (BBB). (B) The disruption of the BBB allows extravasation of the ﬂuid from the brain capillary into the extracelullar space resulting in vasogenic edema.

human disease, both forms of edema are found, however one type typically dominates depending on the particular disorder. However, these distinct edema types can be purely created in experimental models allowing independent study of the molecular mechanisms involved in their evolution. Vasogenic (extracellular) edema results from extravasation of ﬂuid secondary to defects in the blood-brain barrier, and is seen in experimental models of CNS inﬂammation (infection), tumors (Davies, 2002), and cold injury (Chan et al., 1991; Oury et al., 1993). Another mechanism of extracellular edema is the pressure driven accumulation of CSF in the brain interstitium of the periventricular parenchyma as previously described for obstructive hydrocephalus. In contrast, cytotoxic edema is characterized by acute cellular swelling in the presence of an intact blood-brain barrier. A classic example of this edema type is modeled with water intoxication. In this model, rapid intraperitoneal water infusion causes serum hyponatremia that creates an osmotic gradient driving water entry into the brain, resulting in pure cytotoxic edema. Given the highly polarized expression of AQP4 at the blood-brain barrier, and the bidirectional nature of water ﬂux through this channel, it is not surprising that AQP4 participates in the molecular mechanisms of both types of edema.

Aquaporins and cytotoxic edema Although water intoxication produces a pure cellular edema, in the brain the swelling is predominantly localized to the glial processes around the capillaries with sparing of the neurons (Wasterlain and Torack, 1968; Kimelberg, 1995). Characteristic swelling of astrocytic foot processes is also found in brain tissue from head injured patients (Bullock et al., 1991). The selective swelling of astrocytes following a systemic perturbation in osmolality raised questions about the mechanism of this phenomenon (Kimelberg, 1995). Localization studies of AQP4 suggested that it might play a role in glial-speciﬁc swelling. Using the water intoxication model, our group demonstrated a signiﬁcant reduction in astrocytic foot process swelling in AQP4-null mice (Manley et al., 2000). This was associated with a decrease in brain water content and a profound improvement in survival (Fig. 3). Using a clinically relevant model of ischemic stroke, with its predominantly cytotoxic edema, the AQP4-null mice also had decreased cerebral edema and improved outcome. Similar studies were conducted with the dystrophin null mdx-bgeo transgenic mouse and the a-syntrophin null mouse (Frigeri et al., 2001; Vajda et al., 2002; Amiry-Moghaddam et al., 2003).

189

A

B 100

Survival (%)

80

AQP4+/+

0 min

AQP4−/−

AQP4−/−

60 15 min 40 AQP4+/+

20 0

0

20

40

60 Time (min)

C

80

30 min

100 1440

1.045

1.047

D ICP cmH2O at 60min 80

1.049

1.051

1.053

Specific gravity (g/cm3)

E Infused

% water

Non-Infused

82

Water content (µL) 30

60 81

20

40 80 10 20 79 0

+/+

−/−

+/+

−/−

0

Fig. 3. AQP4 deletion in models of cytotoxic (A, B) and vasogenic edema (C, D, E). AQP4 deletion improves short-term survival following water intoxication produced by intraperitoneal ﬂuid injection (A). Speciﬁc gravity of brain tissue display substantially higher values for AQP4/ mice 30 min after water intoxication (B), indicating reduced brain water content compared with AQP4+/+ littermates. The clearance of edema ﬂuid is impaired in AQP4 null mice following intraparenchimal ﬂuid injection. Higher ICP (C) and brain water content (D) is seen in AQP4-null mice compared with AQP4 controls. Brain water content of AQP4-null mice was also higher than the wildtype controls following brain abscess implantation, a clinically relevant model of vasogenic edema.

These mice have normal levels of AQP4 protein, but because membrane localization of AQP4 requires dystrophin-associated protein complex and a-syntrophin (Frigeri et al., 2001; Neely et al., 2001), AQP4 expression in the astrocytic foot processes is reduced (Vajda et al., 2002; AmiryMoghaddam et al., 2003). As a result, these mice are regarded as alternative models for the AQP4null genotype. In general, these studies have conﬁrmed and extended our observations regarding the role of AQP4 in cytotoxic edema (Vajda et al., 2002; Amiry-Moghaddam et al., 2003).

To further clarify the role of AQP4 in astrocyte swelling, osmotically induced volume changes were assessed in primary astrocyte cultures derived from AQP4-null mice (Solenov et al., 2004). The measurement of rapid changes in cell volume was possible by adapting and validating a method of calcein quenching (Hamann et al., 2002) to measure water permeability in astrocytes. We found that water permeability (Pf) was decreased by sevenfold in the AQP4-null astrocytes compared with wildtype astrocytes. Using small interfering RNA (siRNA) targeted to reduce the level

190

of AQP4 expression, a 50% reduction in apparent water permeability of wildtype astrocytes was found (Nicchia et al., 2003), which also caused a 68% inhibition of cell growth. In contrast, there was no alteration in cell morphology or growth characteristics in the primary cultures of the AQP4-null astrocytes. Astrocytes constitute the largest pool of cells in the human brain, outnumbering neurons by at least 10 to 1. Their numbers and ability to rapidly swell in response to an osmotic gradient highlights their importance in the pathogenesis of cellular edema. Studies of astrocytes in which AQP4 has been deleted or knocked down, conﬁrm that AQP4 is the principal water channel in these cells. From a clinical perspective, the phenotypic studies of AQP4-null mice have clearly demonstrated that deletion or reduction of AQP4 reduces edema and improves neurological outcome in well-established model where cytotoxic edema is in primary pathophysiological mechanism. Together these studies indicate that AQP4 inhibition may be a promising target for drug discovery for the treatment of cytotoxic cerebral edema.

Aquaporins and vasogenic edema The key feature of vasogenic edema is the disruption of the blood-brain barrier and subsequent leakage of the intravascular ﬂuid into the extracellular space of the brain parenchyma. The bloodbrain barrier is essential for the normal function of the brain. It provides a highly selective barrier between the vascular space and the brain tissue that allows for maintenance of the microenvironment that is crucial for overall homeostasis of the brain. The anatomical substrates of the blood-brain barrier are the capillary endothelium and the astocytic foot processes that surround the cerebral capillaries (Reese and Karnovsky, 1967). Unlike the microvasculature in the rest of the body, the cerebral capillaries are not fenestrated and contain few endocytic vesicles (Sage et al., 1998) suggesting limited diffusion and transcellular transport. Molecular tracking studies show that only small lipophilic molecules are allowed to

diffuse passively from the vascular bed into the brain parenchyma, while all remaining molecules are trafﬁcked via active transport across the bloodbrain barrier (Rowland et al., 1992). Surrounding the endothelia are the astrocytic foot processes, or ‘‘end-feet’’ (Janzer and Raff, 1987). These highly specialized processes have multiple ion transporters and channels that are speciﬁcally targeted to the foot processes indicating that they also play a key role in blood-brain barrier function. Membranes of astrocyte foot processes also contain numerous arrays of intramembrane particles in freeze-fracture electron micrographs (FFEM). These particles, which are regular square arrays with a characteristic cobblestone pattern, are referred to as square arrays or orthogonal arrays of particles (OAPs) (Landis and Reese, 1974). Based on the ﬁnding that AQP4 was present in the same cell types in which OAPs were identiﬁed, it was proposed that AQP4 is an OAP protein (Frigeri et al., 1995). Support for this hypothesis came from FFEM studies of brain from AQP4-null mice that were found not to have OAPs (Verbavatz et al., 1997). Subsequent antiAQP4 antibody labeling of OAPs in brain tissue conﬁrmed that OAPs contain AQP4 (Rash et al., 1998). The polarized expression and co-localization of ion and water channels in the astrocyte foot processes suggest that these structures may facilitate ion and water transport across the bloodbrain barrier. Given that the formation of vasogenic edema is primarily due to the disruption of the blood-brain barrier and subsequent leakage of the intravascular ﬂuid into the extracellular space, a role for astrocyte water channels in the formation of this type of edema was not anticipated. However, results from multiple independent experiments suggest a deﬁnite role for AQP4 in the clearance of extracellular ﬂuid, and the resolution of vasogenic edema. As with the CSF, the amount of interstitial ﬂuid present in extracellular space is determined by the balance of ﬂuid production and clearance from the brain parenchyma. Similarly, when the clearance or absorption is impaired, ﬂuid accumulates and ICP increases. To directly examine the role of

191

AQP4 in clearance of extracellular ﬂuid, artiﬁcial CSF was continuously infused into the brain parenchyma of wildtype and AQP4-null mice (Fig. 3). In comparison with the wildtype mice, AQP4-null mice had signiﬁcantly more brain water and elevated ICP (Papadopoulos et al., 2004). Although the exact mechanism is yet to be deﬁned, these results demonstrate that AQP4 is involved in the clearance of extracellular ﬂuid from the brain parenchyma. Our group has used a number of models where vasogenic edema is the predominant form of edema, including cortical freeze injury, tumor implantation, and brain abscess. Cortical freeze injury lead to a marked disruption of the bloodbrain barrier in both wildtype and AQP4-null mice, however AQP4-null mice had a signiﬁcantly greater increase in brain water content and ICP (Papadopoulos et al., 2004). Peritumoral vasogenic edema was generated by implanting melanoma cells into the striatum of AQP4-null and wildtype mice. In accordance with the results from previous models, there was higher ICP and accelerated neurological deterioration in the AQP4 deﬁcient mice (Papadopoulos et al., 2004). A similar result was obtained with implantation of Staphylococcus aureus to create brain abscesses in the striatum of mice. In addition to increased peri-abscess edema and increased ICP, AQP4 deﬁcient mice demonstrated a greater mortality despite equivalent abscess size and cytokine levels (Bloch et al., 2005). In all of the mentioned experiments, the primary lesion size and degree of BBB permeability were similar, indicating that the forces driving the development of vasogenic edema were equivalent in both groups and that differences were due to impaired ﬂuid clearance from the extracellular space in AQP4-null mice. Fluid reabsorption from the brain extracellular space is thought to occur by bulk ﬂow clearance into the CSF and by transport back into cerebral capillaries (Reulen et al., 1978; Marmarou et al., 1994). Although a clear understanding of brain water homeostasis remains to be fully elucidated, several lines of evidence point towards an important role for AQP4. In pathological states that increase extracellular ﬂuid, AQP4 expression has

been shown to be upregulated (Vizuete et al., 1999); perhaps as an adaptive response to edema. From a clinical perspective, the phenotypic studies of AQP4-null mice indicate that deletion of AQP4 reduces the clearance of extracellular ﬂuid and improves outcome in models where vasogenic edema is the primary pathophysiological mechanism. Thus, AQP4 inhibition would not be indicated for the treatment of vasogenic edema. Whether or not strategies aimed at increasing AQP4 expression or enhancing water permeability will improve outcome from vasogenic edema, remain to be investigated.

Aquaporins, ICP, and traumatic brain injury One of the more intriguing and clinically relevant ﬁndings has been the observation that ICP is signiﬁcantly reduced in AQP1-null mice (Oshio et al., 2005). ICP is the dynamic result of the interaction of the CSF, cerebral blood, and brain tissue compartments. The individual effect of each compartment on ICP is difﬁcult to assess due to the complexity of the physiologic relationship between these compartments. However, if the intracranial compliance is assumed to be constant, the steady-state ICP can be expressed using the following equation introduced by Marmarou: ICP ¼ If Rout+Pss, where If is CSF formation rate, Rout is outﬂow resistance, and Pss is sagittal sinus pressure (Marmarou et al., 1978). As previously described, AQP1-null mice have a reduced CSF formation rate that accounts for some of the 50% reduction in ICP seen in these mice. The deletion of AQP1 also leads to polyuria and hypovolemia due to its role in ﬂuid absorption in the kidney (Schnermann et al., 1998). We suspected that this urinary concentrating defect might also decreases the central venous pressure, which in turn, is in continuity with the sagittal sinus venous pressure. This was indeed the case, indicating that the effects of AQP1 deletion on venous pressure accounted for the balance of the reduction in ICP. These results demonstrate that two different AQP1-dependent mechanisms act synergistically to signiﬁcantly decrease in ICP.

192

Fig. 4. The beneﬁcial effect of AQP1 deletion in brain injury. (A) The absence of AQP1 improves survival in focal brain injury. (B) The ICP peak is reduced in AQP1-null mice following cold injury compared with wildtype controls.

ICP management has been the cornerstone of traumatic brain injury care for many years. It is well known that when ICP is elevated and cannot be reduced, nearly all patients with traumatic brain injury will die (Miller et al., 1981). All current surgical and non-surgical treatments of elevated ICP are directed towards decreasing the volume in one of the three major compartments — brain, blood, or CSF (Brain Trauma Foundation, American Association of Neurological Surgeons and Joint Section on Neurotrauma and Critical Care, 1996). The ability of AQP1 deletion to reduce two of the three principal determinants of steady-state ICP, CSF production, and venous pressure, suggested that AQP1deletion might be protective in a model of brain trauma (Fig. 4). In a focal model of brain injury, the peak ICP was 60% lower in AQP1-null mice compared with wildtype littermates. In addition, a dramatic improvement in survival rate was seen in the AQP1 deﬁcient mice: 84% compared with the 25% seen in wild type controls. Currently, there is no consensus on the treatment of elevated ICP. These results provide evidence for the importance of CVP and CSF production on ICP under normal and pathological

conditions and suggest a novel application of AQP1 inhibitors for pharmacological treatment of elevated ICP.

Conclusion Phenotypic analyses of AQP deﬁcient mice have allowed us to explore the role of these membrane water channels in the mechanisms of cytotoxic edema, vasogenic edema, and CSF production. These studies indicate that AQP4 plays signiﬁcant role in the development of cytotoxic edema and the absorption of excess brain water resulting from vasogenic edema. They also have demonstrated the role of AQP1 in CSF production and maintenance of steady-state ICP. The ability to modulate water ﬂux through AQP deletion has provided new insights into brain water homeostasis and suggested a number of new research directions. However, these efforts have not yet translated to the treatment human clinical diseases. These advances will require the development of AQP inhibitors and activators to establish the beneﬁt modulating the function of these water channels.

193

References Abbott, N.J. (2004) Evidence for bulk ﬂow of brain interstitial ﬂuid: signiﬁcance for physiology and pathology. Neurochem. Int., 45(4): 545–552. Agre, P., King, L.S., et al. (2002) Aquaporin water channels: from atomic structure to clinical medicine. J. Physiol., 542 (Pt 1): 3–16. Amiry-Moghaddam, M., Otsuka, T., et al. (2003) An alphasyntrophin-dependent pool of AQP4 in astroglial end-feet confers bidirectional water ﬂow between blood and brain. Proc. Natl. Acad. Sci. U.S.A., 100(4): 2106–2111. Bering, E. and Sato, O. (1963) Hydrocephalus: changes in formation and absorption of cerebrospinal ﬂuid within the cerebral ventricles. J. Neurosurg., 20: 1050. Bloch, O., Auguste, K.I., Manley, G.T. and Verkman, A.S. (2006) Accelerated progression of kaolin-induced hydrocephalus in aquaporin-4-deﬁcient mice. J. Cereb. Blood Flow Metab., 26(12): 1527–1537. Bloch, O., Papadopoulos, M.C., et al. (2005) Aquaporin-4 gene deletion in mice increases focal edema associated with staphylococcal brain abscess. J. Neurochem., 95(1): 254–262. Brain Trauma Foundation, American Association of Neurological Surgeons, Joint Section on Neurotrauma and Critical Care. (1996) Guidelines for the management of severe head injury. J. Neurotrauma, 13(11): 641–734. Braun, K.P., Dijkhuizen, R.M., et al. (1997) Cerebral ischemia and white matter edema in experimental hydrocephalus: a combined in vivo MRI and MRS study. Brain Res., 757(2): 295–298. Bullock, R., Maxwell, W.L., et al. (1991) Glial swelling following human cerebral contusion: an ultrastructural study. J. Neurol. Neurosurg. Psychiatry, 54(5): 427–434. Chan, P.H., Yang, G.Y., et al. (1991) Cold-induced brain edema and infarction are reduced in transgenic mice overexpressing CuZn-superoxide dismutase. Ann. Neurol., 29(5): 482–486. Davies, D.C. (2002) Blood-brain barrier breakdown in septic encephalopathy and brain tumours. J. Anat., 200(6): 639–646. Fishman, R.A. (1975) Brain edema. N. Engl. J. Med., 293(14): 706–711. Frigeri, A., Gropper, M.A., et al. (1995) Localization of MIWC and GLIP water channel homologs in neuromuscular, epithelial and glandular tissues. J. Cell Sci., 108(Pt 9): 2993–3002. Frigeri, A., Nicchia, G.P., et al. (2001) Aquaporin-4 deﬁciency in skeletal muscle and brain of dystrophic mdx mice. FASEB J., 15(1): 90–98. Hamann, S., Kiilgaard, J.F., et al. (2002) Measurement of cell volume changes by ﬂuorescence self-quenching. J. Fluoresc., 12: 139–145. Hasegawa, H., Ma, T., et al. (1994) Molecular cloning of a mercurial-insensitive water channel expressed in selected water-transporting tissues. J. Biol. Chem., 269(8): 5497–5500. Hiratsuka, H., Tabata, H., et al. (1982) Evaluation of periventricular hypodensity in experimental hydrocephalus by metrizamide CT ventriculography. J. Neurosurg., 56(2): 235–240.

Janzer, R.C. and Raff, M.C. (1987) Astrocytes induce bloodbrain barrier properties in endothelial cells. Nature, 325(6101): 253–257. Kimelberg, H.K. (1995) Current concepts of brain edema. Review of laboratory investigations. J. Neurosurg., 83(6): 1051–1059. Klatzo, I. (1994) Evolution of brain edema concepts. Acta Neurochir. Suppl. (Wien.), 60: 3–6. Landis, D.M. and Reese, T.S. (1974) Arrays of particles in freeze-fractured astrocytic membranes. J. Cell Biol., 60(1): 316–320. Manley, G.T., Binder, D.K., et al. (2004) New insights into water transport and edema in the central nervous system from phenotype analysis of aquaporin-4 null mice. Neuroscience, 129(4): 983–991. Manley, G.T., Fujimura, M., et al. (2000) Aquaporin-4 deletion in mice reduces brain edema after acute water intoxication and ischemic stroke. Nat. Med., 6(2): 159–163. Marmarou, A., Hochwald, G., et al. (1994) Brain edema resolution by CSF pathways and brain vasculature in cats. Am. J. Physiol., 267(2 Pt 2): H514–H520. Marmarou, A., Shulman, K., et al. (1978) A nonlinear analysis of the cerebrospinal ﬂuid system and intracranial pressure dynamics. J. Neurosurg., 48(3): 332–344. McIlwain, H. and Bachelard, H.S. (1985) Biochemistry and the Central Nervous System. Churchill Livingstone, Edinburgh. Milhorat, T., Hammock, M., et al. (1971) Cerebrospinal ﬂuid production by the choroid plexus and brain. Science, 173(994): 330–332. Miller, J.D., Butterworth, J.F., et al. (1981) Further experience in the management of severe head injury. J. Neurosurg., 54(3): 289–299. Neely, J.D., Amiry-Moghaddam, M., et al. (2001) Syntrophindependent expression and localization of Aquaporin-4 water channel protein. Proc. Natl. Acad. Sci. U.S.A., 98(24): 14108–14113. Nicchia, G.P., Frigeri, A., et al. (2003) Inhibition of aquaporin4 expression in astrocytes by RNAi determines alteration in cell morphology, growth, and water transport and induces changes in ischemia-related genes. FASEB J., 17(11): 1508–1510. Oshio, K., Song, Y., et al. (2003) Aquaporin-1 deletion reduces osmotic water permeability and cerebrospinal ﬂuid production. Acta Neurochir. Suppl., 86: 525–528. Oshio, K., Watanabe, H., et al. (2005) Reduced cerebrospinal ﬂuid production and intracranial pressure in mice lacking choroid plexus water channel Aquaporin-1. FASEB J., 19(1): 76–78. Oury, T.D., Piantadosi, C.A., et al. (1993) Cold-induced brain edema in mice. Involvement of extracellular superoxide dismutase and nitric oxide. J. Biol. Chem., 268(21): 15394–15398. Papadopoulos, M.C., Manley, G.T., et al. (2004) Aquaporin-4 facilitates reabsorption of excess ﬂuid in vasogenic brain edema. FASEB J., 18(11): 1291–1293. Rash, J.E., Yasumura, T., et al. (1998) Direct immunogold labeling of aquaporin-4 in square arrays of astrocyte and

194 ependymocyte plasma membranes in rat brain and spinal cord. Proc. Natl. Acad. Sci. U.S.A., 95(20): 11981–11986. Reese, T.S. and Karnovsky, M.J. (1967) Fine structural localization of a blood-brain barrier to exogenous peroxidase. J. Cell Biol., 34(1): 207–217. Reulen, H.J., Tsuyumu, M., et al. (1978) Clearance of edema ﬂuid into cerebrospinal ﬂuid. A mechanism for resolution of vasogenic brain edema. J. Neurosurg., 48(5): 754–764. Rowland, L.P., Fink, M.E., et al. (1992) Cerebrospinal ﬂuid: blood-brain barrier, brain oedema and hydrocephalus. In: Kandel E.R., Schwartz J.H. and Jessel T.M. (Eds.), Principles of Neuroscience. Elsevier, New York, pp. 1050–1060. Sage, M.R., Wilson, A.J., et al. (1998) Contrast media and the brain. The basis of CT and MR imaging enhancement. Neuroimaging Clin. N. Am., 8(3): 695–707. Schnermann, J., Chou, C.-L., et al. (1998) Defective proximal tubular ﬂuid reabsorption in transgenic aquaporin-1 null mice. PNAS, 95(16): 9660–9664. Solenov, E., Watanabe, H., et al. (2004) Sevenfold-reduced osmotic water permeability in primary astrocyte cultures from AQP-4-deﬁcient mice, measured by a ﬂuorescence quenching method. Am. J. Physiol. Cell Physiol., 286(2): C426–C432. Speake, T., Freeman, L.J., et al. (2003) Expression of aquaporin 1 and aquaporin 4 water channels in rat choroid plexus. Biochim. Biophys. Acta, 1609(1): 80–86.

Vajda, Z., Pedersen, M., et al. (2002) Delayed onset of brain edema and mislocalization of aquaporin-4 in dystrophin-null transgenic mice. Proc. Natl. Acad. Sci. U.S.A., 99(20): 13131–13136. Verbavatz, J.M., Ma, T., et al. (1997) Absence of orthogonal arrays in kidney, brain and muscle from transgenic knockout mice lacking water channel aquaporin-4. J. Cell Sci., 110(Pt 22): 2855–2860. Verkman, A.S. (2002) Physiological importance of aquaporin water channels. Ann. Med., 34(3): 192–200. Vizuete, M.L., Venero, J.L., et al. (1999) Differential upregulation of aquaporin-4 mRNA expression in reactive astrocytes after brain injury: potential role in brain edema. Neurobiol. Dis., 6(4): 245–258. Wasterlain, C.G. and Torack, R.M. (1968) Cerebral edema in water intoxication. II. An ultrastructural study. Arch. Neurol., 19(1): 79–87. Weed, L.H. and McKibben, P.S. (1919) Experimental alteration of brain bulk. Am. J. Physiol., 48: 531–558. Welch, K. (1963) Secretion of cerebrospinal ﬂuid by choroid plexus of the rabbit. Am. J. Physiol., 205(3): 617–624. Wright, E.M., Wiedner, G., et al. (1977) Fluid secretion by the frog choroid plexus. Exp. Eye Res., 25(Suppl): 149–155.

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 13

Sodium channel expression and the molecular pathophysiology of pain after SCI Bryan C. Hains and Stephen G. Waxman Department of Neurology and Center for Neuroscience and Regeneration Research, Yale University School of Medicine, New Haven, CT 06510, USA; Rehabilitation Research Center, VA Connecticut Healthcare System, West Haven, CT 06516, USA

Abstract: The chronic pain that develops as a result of spinal cord injury (SCI) is extremely debilitating and remains largely unmanageable by current therapeutic strategies. Voltage-gated sodium channels regulate the biophysical properties, and thus ﬁring characteristics, of neurons. After SCI the repertoire of sodium channels produced by dorsal horn nociceptive neurons is altered, enabling neurons to ﬁre at higher than normal rates in response to unchanged peripheral stimuli as well as to generate spontaneous discharges in the absence of stimuli, resulting in the genesis of neuropathic pain. Our results have shown increased expression of the Nav1.3 sodium channel in the spinal cord and thalamus. Nav1.3 upregulation allows dorsal horn neurons to generate ramp currents, enhanced persistent currents, and shifts in steady-state activation and inactivation. Further downstream, Nav1.3 causes increased spontaneous and evoked ﬁring of neurons in the ventroposterior lateral (VPL) nucleus of the thalamus. Nav1.3 also underlies changes in burst ﬁring properties of VPL neurons. The combination of spinal and thalamic generation and ampliﬁcation of pain by Nav1.3 dysregulation contributes to post-SCI chronic pain. If proven to be similar in humans, targeting of this system after SCI may offer hope for treatment of clinical pain. Keywords: spinal cord injury; pain; sodium channels; thalamus; dorsal horn Turner et al., 2001), even to a greater extent than the motor impairment (Mariano, 1992; Stormer et al., 1997). Of patients with low thoracic or lumbrosacral lesions, 37% are willing to trade the possibility of recovery for pain relief (Nepomuceno et al., 1979). Nociceptive signals are transmitted through the dorsal horn of the spinal cord. The spinal cord dorsal horn contains the primary synapses through which afferent somatosensory information, related to touch, pressure, brush, temperature, and noxious stimuli, is received from the periphery. Dorsal horn sensory neurons receive this information primarily from the skin, perform a degree of processing, and transmit signals through distinct tracts within the spinal cord to supraspinal structures

Introduction Although less apparent than paralysis, the prevalence of chronic pain disability after spinal cord injury (SCI) is widespread. Sixty to eighty percent of persons who have sustained SCI, regardless of the level or completeness of lesion or of the type of injury, experience clinically signiﬁcant pain after injury (Finnerup et al., 2001; Siddall et al., 2003). The pain, which can have a pricking, burning, or aching quality (Cairns et al., 1996), can be so severe as to produce a drastic impairment in daily routines and quality of life (Rintala et al., 1998; Corresponding author. Tel.: +203-785-6351; Fax: +203-785-

7826; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61013-3

195

196

where pain is interpreted and perceived. Within each level of the nervous system including the spinal cord, nociceptive signals are subject to a degree of modulation by circuitry that it passes through. Experimental SCI induces electrophysiological changes in dorsal horn and thalamic sensory neurons that contribute to pain-like behaviors in animals. These include shifts in proportions of cells responding to evoked noxious stimulation, increases and irregularity in spontaneous background activity, increased evoked activity to (formerly) innocuous and noxious stimuli, increases in after discharge activity following stimulation, and emergence of abnormal burst ﬁring. Since the nature of the applied stimuli does not change, other mechanisms must account for the alterations in stimulus processing that lead to central pain after SCI. Central changes in expression of molecules such as ion channels, neurotransmitters, and their receptors, contribute to altered sensory processing by changing the electrophysiologic excitability, and therefore output, of spinal sensory neurons. Speciﬁc molecular changes in the expression of voltage-gated sodium channels have recently been shown by our laboratory to contribute to the development and maintenance of central pain following SCI.

Spinal cord sodium channels Action potential generation and propagation by sensory neurons relies on multiple isoforms of voltage-gated sodium channels (termed Nav). The selective expression of ensembles of sodium channels, unique in different types of neurons, tunes the biophysical properties of each cell. Within the normal nervous system, properties fundamental to neuronal function such as activation threshold, inactivation, refractory period, rates of repriming, and the ability to generate and conduct highfrequency trains of action potentials all depend on the type(s) of sodium channels expressed within a given neuron (Waxman, 2000). As might be expected, dysregulation of channel expression can abnormally reconﬁgure neuronal function in disease states.

Ten genes encode molecularly distinct voltagegated sodium channels, at least seven of which are expressed in the rat nervous system. In the adult spinal cord dorsal horn, Nav1.1, Nav1.2, and Nav1.6 are strongly detectable, whereas Nav1.3 expression, while present at early ontogenetic states, decreases with development and is nearly undetectable in adults (Felts et al., 1997).

SCI and dorsal horn ion channel dysregulation We have studied sodium channel expression at several levels along the pain-signaling pathway in a model of contusive SCI in which, as in humans with non-penetrating SCI, there is central necrosis, surrounded by a rim of surviving ascending and descending axons at the level of the injury (Hains et al., 2004). In this rodent model in situ hybridization and immunocytochemistry with an isoform-speciﬁc Nav1.3 antibody show that expression of Nav1.3 is upregulated within dorsal horn neurons caudal to the level of the injury (Hains et al., 2003). These studies show that, 4 weeks after SCI, expression of Nav1.3 is increased, within neurons contained within laminae I–VI in the lumbar dorsal horn (L3–L5) (Fig. 1B). Nav1.3 does not co-localize with OX-42, a marker for activated microglia (which proliferate after SCI and are present in all laminae within the dorsal horn (Hains and Waxman, 2006)), or with GFAP, a marker for astrocytes (which also proliferate in all spinal laminae after SCI). CGRP, restricted to the terminals of primary afferent ﬁbers within laminae I–II, does not co-localize with Nav1.3. The substance-P receptor NK1R, found in second-order nociceptive neurons within laminae I–IV, does however co-localize with Nav1.3, indicating that Nav1.3 is upregulated within nociceptive dorsal horn neurons. In this model, 88% of sampled lumbar dorsal horn units show increased evoked activity to natural peripheral stimuli (brush, press, pinch, graded von Frey ﬁlaments, and 471C thermal stimuli). In comparison to intact animals in which dorsal horn neurons ﬁre at rates of 5–21 Hz, SCI animals demonstrate increased evoked rates of up to 55–60 Hz (Fig. 1E).

197

Fig. 1. Nav1.3 mRNA is expressed at low levels in naı¨ ve animals (A), but is upregulated in lumbar dorsal horn neurons after spinal cord injury (SCI) (B). Treatment with antisense (AS) oligodeoxynucleotides generated against Nav1.3 reduces Nav1.3 transcripts after SCI (C). Corresponding unit recordings show evoked activity to peripheral stimulation (BR, brush; PR, press; PI, pinch, increasing strength von Frey ﬁlament stimulation, and noxious thermal heating (471C)), after SCI (E) compared with controls (D). After Nav1.3 AS delivery, evoked activity of dorsal horn neurons resembles that found in naı¨ ve levels (F). Adapted with permission from Hains et al. (2003).

The Nav1.3 sodium channel plays a role in maintaining neuronal hyperresponsiveness to peripheral stimulation, as well as pain-related behaviors after SCI, as evidenced through selective knock-down of Nav1.3 expression via antisense (AS) oligodeoxynucleotide administration (Hains et al., 2003). Lumbar intrathecal administration of Nav1.3 AS very effectively reduces the expression of Nav1.3 within dorsal horn neurons after SCI (Fig. 1C), and signiﬁcantly

decreases the evoked hyperresponsiveness of dorsal horn multi-receptive neurons (Fig. 1F), whereas administration of Nav1.3 mismatch (MM) sequence, used as a control, has no effect. Electrophysiological recordings show that 4 days after initiation of Nav1.3 AS, high levels of evoked activity of dorsal horn neurons in SCI animals are markedly decreased in response to all peripheral stimuli, and only 20% of sampled units are hyperexcitable.

198

Nav1.3 and neuronal hyperresponsiveness Expression of the Nav1.3 sodium channel after SCI uniquely conﬁgures neurons to ﬁre in an abnormal, heightened manner. Nav1.3 produces a rapidly repriming tetrodotoxin-sensitive sodium current that promotes neuronal ﬁring at higherthan-normal frequencies (Cummins et al., 2001). Since stimulus intensity is encoded in the dorsal horn by the rate of ﬁring, this change in neuronal processing of afferent sensory information serves to amplify incoming signals, such that perceived pain thresholds are lowered after SCI.

We recently examined the biophysical properties of sodium currents in acutely dissociated dorsal horn neurons after chronic SCI, and found a number of changes in sodium current characteristics via whole-cell patch-clamp recordings (Fig. 2A) (Lampert et al., 2006). First, we found shifts in the activation and inactivation properties of the voltage-gated sodium currents to more positive potentials after SCI (Fig. 2B). Second, persistent sodium currents in dorsal horn neurons from SCI animals were increased when compared with cells from intact animals (Fig. 2C). Third, the ramp current elicited in response to slow depolarizing potentials

Fig. 2. Whole-cell patch-clamp recordings of acutely dissociated lumbar dorsal horn neurons 28 days after T9 SCI reveal changes in sodium current properties. Representative current traces from intact and SCI animals, elicited by stepwise depolarizations from a holding potential of 120 mV to voltages ranging form 110 to +40 mV (A). Steady-state inactivation (open squares: intact, ﬁlled squares: SCI) measured as relative available current after a 500 ms prepulse to the indicated potential. Steady-state activation shown as relative conductance (B). Note that SCI shifts steady-state inactivation in dorsal horn neurons toward the curve for axotomized DRG neurons, in which Nav1.3 is upregulated. Representative current traces showing enhancement of persistent current after SCI (C). Vertical lines delimit the region in which the mean persistent current was assessed. Inset: Representative record from a human embryonic kidney cell heterologously transfected with Nav1.3, showing persistent current. Representative current traces in response to a 200 ms voltage ramp from 120 to +30 mV showing an increase in ramp currents in SCI but not intact animals (D). Adapted with permission from Lampert et al. (2006).

199

was increased after SCI (Fig. 2D). The shift in steady-state inactivation and the increase in the persistent and ramp current all would be expected to contribute to dorsal horn neuron hyperresponsiveness, which is observed after SCI. All of these properties, which are consistent with upregulation of Nav1.3, can contribute to hyperresponsiveness of dorsal horn neurons, after SCI.

Thalamic Nav1.3 dysregulation Dorsal horn neurons project rostrally within the spinothalamic tract to third-order neurons of the

ventroposterior lateral (VPL) nucleus of the thalamus (Jones, 1998). VPL neurons are involved in sensory-discriminative aspects of pain processing. Thalamic changes have been associated with pain following SCI in humans (Lenz et al., 1989; Pattany et al., 2002), primates (Weng et al., 2000), and rodent models (Koyama et al., 1993; Gerke et al., 2003), but the underlying molecular mechanisms are still not fully understood. After SCI, expression of the Nav1.3 sodium channel is upregulated in VPL neurons (Fig. 3B), and an increase in the spontaneous activity and responses to natural stimuli is observed (Hains et al., 2005). Lumbar administration of Nav1.3 AS

Fig. 3. Expression of Nav1.3 protein within the ventral posterolateral (VPL) nucleus of the thalamus is absent in intact animals (A), but 28 days after SCI Nav1.3 expression is upregulated within VPL and to some degree the ventral posteromedial (VPM) nuclei. Intrathecal Nav1.3 antisense (AS) administration (C) signiﬁcantly reduced the expression of Nav1.3 within the VPL (D). Long-term recording of a single VPL unit after SCI with a hindpaw receptive ﬁeld revealed high spontaneous discharge activity and evoked responsiveness to brush (BR) stimulation of the receptive ﬁeld (E). Following topical application of lidocaine (lido) and cord transection (tx), evoked responses were abolished however high spontaneous activity persisted. Adapted with permission from Hains et al. (2005).

200

reduced the number of neurons displaying Nav1.3 upregulation within the VPL (Fig. 3C), and reversed thalamic electrophysiologic abnormalities caused by SCI. Importantly, we also observed an increased level of spontaneous background activity in VPL neurons associated with SCI which persists after spinal cord transection (Fig. 3E) that disconnects the site of SCI and ascending projections from the lumbar enlargement from the thalamus, indicating a degree of autonomous hyperexcitability of the thalamus. We cannot exclude the possibility that factors other than Nav1.3 also contribute to the reconﬁguration of the ﬁring properties of thalamic units following SCI since the magnitudes of change of Nav1.3 expression levels did not perfectly match the magnitude of the electrophysiologic changes. However, our results demonstrate changes in sodium channel expression within the thalamus that are associated with abnormal sensory processing and chronic neuropathic pain after SCI. The thalamus also undergoes a change in burst ﬁring properties after SCI. Bursts are involved in normal and pathological perceptual processing; however, rhythmic oscillation of burst ﬁring is observed in pathophysiological conditions, and it has been suggested that abnormal thalamic activity may contribute to the perception of chronic pain (Lenz et al., 1989, 1994; Jeanmonod et al., 1993; McCormick, 1999). Furthermore, the information content of bursts is higher than for single-spikes in the visual system (Reinagel et al., 1999), and the overall probability of generating at least one postsynaptic spike (in the pain-processing sensory cortex) is higher for bursts than for single-spikes (Csicsvari et al., 1998). In neurons of the ventrobasal thalamus, increased gain or transfer ratio induced by burst ﬁring results in increased thalamocortical efﬁcacy, enhancing the postsynaptic response (Swadlow and Gusev, 2001). Thus, it is possible that pathological burst ﬁring after SCI may more potently activate cortical circuits involved in pain perception. We recently showed that after SCI, burst ﬁring intervals become more regular, spike events are reduced within each burst, acceleration in burst duration occurs in bursts containing higher spike counts, and shifts occur among spike ﬁring modes

(Hains et al., 2006). Furthermore, we observed that Nav1.3 AS returned the number of spikes/ burst, burst duration, and interburst interval toward control levels following SCI. Our data are similar to those reported by Lenz et al. (1989, 1994) in humans with post-SCI pain showing that thalamic neurons exhibit oscillatory burst ﬁring characterized by high discharge rates, and deceleration of ﬁring rate throughout the burst period, and suggest that altered sodium channel expression within thalamic neurons contributes to these functional abnormalities.

Multi-tiered alterations in nociceptive processing Our ﬁndings demonstrate changes in excitability and expression of Nav1.3 within dorsal horn and VPL neurons following SCI. We have also observed that hyperexcitability of thalamic neurons following SCI is to at least some degree autonomous, since it persists following spinal cord transection which abolishes ascending barrages from spinal cord neurons near or below the injury site. Together with our earlier results on dorsal root ganglion neurons (Cummins and Waxman, 1997; Cummins et al., 2001), these studies provide evidence for a link between pain after SCI and molecular changes in pain-signaling neurons, suggesting that dysregulation of sodium channel Nav1.3 expression at both spinal and supraspinal levels contribute to altered processing of somatosensory information and chronic neuropathic pain after SCI. Speciﬁcally, our results indicate that there is upregulated expression of Nav1.3 in ﬁrst-, second-, and third-order neurons of the pain pathway after contusive SCI, and suggest that this leads to enhanced neuronal excitability at each of these multiple levels. As originally proposed by Yezierski (2001), SCI can trigger the activity of pain generators and ampliﬁers within the CNS. On the basis of our observations, we propose an integrated model (Fig. 4) in which increased Nav1.3 expression within pain-signaling neurons at multiple levels of the neuraxis causes exaggerated nociceptive signaling by way of spontaneous ﬁring, a lowered threshold for ﬁring in response to synaptic drive,

201

Fig. 4. Model whereby after SCI, multi-tiered dysregulation of sodium channel expression contributes to the pathological ampliﬁcation and generation of aberrant nociceptive information, and ultimately chronic pain. In comparison to intact animals (A), secondorder neurons within the dorsal horn increase the gain with which they respond to innocuous and noxious inputs (10 ) after SCI (B). The exaggerated responses of these dorsal horn neurons in response to peripheral stimulation poise them to act as an ampliﬁer of normal as well as inappropriate pathological nociceptive signals (20 ). After SCI, the thalamus also acts as a pain ampliﬁer by responding in an increased manner to nociceptive information relayed from the dorsal horn, and in addition, thalamic circuitry can act as an autonomous pain signal generator (30 ). Treatment strategies that target one or more components of this pathway may be of value in combating chronic pain following SCI.

and increased synaptic drive due to enhanced excitability of presynaptic elements. According to this model, second-order neurons within the dorsal horn increase the gain with which they respond to innocuous and noxious inputs after SCI. The exaggerated responses of these dorsal horn neurons conﬁgures them to act as an ampliﬁer of normal as well as inappropriate pathological nociceptive signals. Because the thalamus is involved not only in relaying, but also processing incoming information from the spinal cord en route to the cortex, injuryinduced changes in spinal pain generator circuitry may feed aberrant signals into the injured thalamus which further processes and ampliﬁes the signals before relaying messages that are interpreted as signaling pain to suprathalamic structures

(Waxman and Hains, 2006). Abnormal processing at thalamic levels would then be expected to further exaggerate abnormal ﬁring patterns from the spinal cord after SCI. Our observations of increased primary burst ﬁring activity and reduced silence in VPL neurons after SCI leads us to suggest that, following SCI, an increased level of abnormal afferent ﬁring is being forwarded to cortical structures involved in interpreting pain.

Conclusion SCI induces nervous-system-wide changes in neuronal ﬁring properties which contribute to chronic pain, some of which are governed by the pathological expression of the Nav1.3 sodium channel.

202

By prevention of the upregulation of Nav1.3, knock-down of Nav1.3 after it is upregulated, or selective pharmacological targeting of Nav1.3 channels deployed after injury, it is possible that the molecular ampliﬁers and generators of chronic pain associated with SCI may be effectively targeted and muted so that chronic pain can be ameliorated.

Acknowledgments The authors thank Dr. Joel Black for valuable experimental advice. This work was supported in part by grants from the Medical Research Service and Rehabilitation Research Service, Department of Veterans Affairs, and the National Multiple Sclerosis Society. The Center for Neuroscience and Regeneration Research is a Collaboration of the Paralyzed Veterans of America and the United Spinal Association. BCH was funded by The Christopher Reeve Paralysis Foundation (HB10304-2), the NIH/NINDS (1 F32 NS046919-01), and Pﬁzer (Scholar’s Grant in Pain Medicine).

References Cairns, D.M., Adkins, R.H. and Scott, M.D. (1996) Pain and depression in acute traumatic spinal cord injury: origins of chronic problematic pain? Arch. Phys. Med. Rehab., 77: 329–335. Cummins, T.R., Aglieco, F., Renganathan, M., Herzog, R.I., Dib-Hajj, S.D. and Waxman, S.G. (2001) Nav1.3 sodium channels: rapid repriming and slow closed-state inactivation display quantitative differences after expression in a mammalian cell line and in spinal sensory neurons. J. Neurosci., 21: 5952–5961. Cummins, T.R. and Waxman, S.G. (1997) Downregulation of tetrodotoxin-resistant sodium currents and upregulation of a rapidly repriming tetrodotoxin-sensitive sodium current in small spinal sensory neurons after nerve injury. J. Neurosci., 17: 3503–3514. Csicsvari, J., Hirase, H., Czurko, A. and Buzsaki, G. (1998) Reliability and state dependence of pyramidal cell-interneuron synapses in the hippocampus: an ensemble approach in the behaving rat. Neuron, 21: 179–189. Felts, P.A., Yokoyama, S., Dib-Hajj, S., Black, J.A. and Waxman, S.G. (1997) Sodium channel alpha-subunit mRNAs I, II, III, NaG, Na6 and hNE (PN1): different expression patterns in developing rat nervous system. Mol. Brain Res., 45: 71–82.

Finnerup, N.B., Johannesen, I.L., Sindrup, S.H., Bach, F.W. and Jensen, T.S. (2001) Pain and dysesthesia in patients with spinal cord injury: a postal survey. Spinal Cord, 39: 256–262. Gerke, M.B., Duggan, A.W., Xu, L. and Siddall, P.J. (2003) Thalamic neuronal activity in rats with mechanical allodynia following contusive spinal cord injury. Neuroscience, 117: 715–722. Hains, B.C., Klein, J.P., Saab, C.Y., Craner, M.J., Black, J.A. and Waxman, S.G. (2003) Upregulation of sodium channel Nav1.3 and functional involvement in neuronal hyperexcitability associated with central neuropathic pain after spinal cord injury. J. Neurosci., 23: 8881–8892. Hains, B.C., Saab, C.Y., Lo, A.C. and Waxman, S.G. (2004) Sodium channel blockade with phenytoin protects spinal cord axons, enhances axonal conduction, and improves functional motor recovery after contusion SCI. Exp. Neurol., 188: 365–377. Hains, B.C., Saab, C.Y. and Waxman, S.G. (2005) Changes in electrophysiological properties and sodium channel Nav1.3 expression in thalamic neurons after spinal cord injury. Brain, 128: 2359–2371. Hains, B.C., Saab, C.Y. and Waxman, S.G. (2006) Alterations in burst ﬁring of thalamic VPL neurons and reversal by Na(v)1.3 antisense after spinal cord injury. J. Neurophysiol., 95: 3343–3352. Hains, B.C. and Waxman, S.G. (2006) Activated microglia contribute to the maintenance of chronic pain after spinal cord injury. J. Neurosci., 26: 4308–4317. Jeanmonod, D., Magnin, M. and Morel, A. (1993) Thalamus and neurogenic pain: physiological, anatomical, and clinical data. Neuroreport, 4: 475–478. Jones, E.G. (1998) Viewpoint: the core and matrix of thalamic organization. Neuroscience, 85: 331–345. Koyama, S., Katayama, Y., Maejima, S., Hirayama, T., Fujii, M. and Tsubokawa, T. (1993) Thalamic neuronal hyperactivity following transection of the spinothalamic tract in the cat: involvement of N-methyl-D-aspartate receptor. Brain Res., 612: 345–350. Lampert, A., Hains, B.C. and Waxman, S.G. (2006) Upregulation of persistent and ramp sodium current in dorsal horn neurons after spinal cord injury. Exp. Brain Res., 174: 660–666. Lenz, F.A., Kwan, H.C., Dostrovsky, J.O. and Tasker, R.R. (1989) Characteristics of the bursting pattern of action potentials that occurs in the thalamus of patients with central pain. Brain Res., 496: 357–360. Lenz, F.A., Kwan, H.C., Martin, R., Tasker, R., Richardson, R.T. and Dostrovsky, J.O. (1994) Characteristics of somatotopic organization and spontaneous neuronal activity in the region of the thalamic principal sensory nucleus in patients with spinal cord transection. J. Neurophysiol., 72: 1570–1587. Mariano, A.J. (1992) Chronic pain and spinal cord injury. Clin. J. Pain, 8: 87–92. McCormick, D.A. (1999) Are thalamocortical rhythms the Rosetta Stone of a subset of neurological disorders? Nat. Med., 5: 1349–1351.

203 Nepomuceno, C., Fine, P.R., Richards, J.S., Gowens, H., Stover, S.L., Rantanuabol, U. and Houston, R. (1979) Pain in patients with spinal cord injury. Arch. Phys. Med. Rehabil., 60: 605–609. Pattany, P.M., Yezierski, R.P., Widerstrom-Noga, E.G., Bowen, B.C., Martinez-Arizala, A., Garcia, B.R. and Quencer, R.M. (2002) Proton magnetic resonance spectroscopy of the thalamus in patients with chronic neuropathic pain after spinal cord injury. Am. J. Neuroradiol., 23: 901–905. Reinagel, P., Godwin, D., Sherman, S.M. and Koch, C. (1999) Encoding of visual information by LGN bursts. J. Neurophysiol., 81: 2558–2569. Rintala, D.H., Loubser, P.G., Castro, J., Hart, K.A. and Fuhrer, M.J. (1998) Chronic pain in a community-based sample of men with spinal cord injury: prevalence, severity, and relationship with impairment, disability, handicap, and subjective well-being. Arch. Phys. Med. Rehabil., 79: 604–614. Siddall, P.J., McClelland, J.M., Rutkowski, S.B. and Cousins, M.J. (2003) A longitudinal study of the prevalence and characteristics of pain in the ﬁrst 5 years following spinal cord injury. Pain, 103: 249–257. Stormer, S., Gerner, H.J., Gruninger, W., Metzmacher, K., Follinger, S., Wienke, C., Aldinger, W., Walker, N., Zimmermann, M. and Paeslack, V. (1997) Chronic pain/

dysaesthesiae in spinal cord injury patients: results of a multicentre study. Spinal Cord, 35: 446–455. Swadlow, H.A. and Gusev, A.G. (2001) The impact of ‘bursting’ thalamic impulses at a neocortical synapse. Nat. Neurosci., 4: 402–408. Turner, J.A., Cardenas, D.D., Warms, C.A. and McClellan, C.B. (2001) Chronic pain associated with spinal cord injuries: a community survey. Arch. Phys. Med. Rehabil., 82: 501–509. Waxman, S.G. (2000) The neuron as a dynamic electrogenic machine: modulation of sodium-channel expression as a basis for functional plasticity in neurons. Philos. Trans. R. Soc. Lond. B. Biol. Sci., 355: 199–213. Waxman, S.G. and Hains, B.C. (2006) Fire and phantoms after spinal cord injury: Na+ channels and central pain. Trends Neurosci., 29: 207–215. Weng, H.R., Lee, J.I., Lenz, F.A., Schwartz, A., Vierck, C., Rowland, L. and Dougherty, P.M. (2000) Functional plasticity in primate somatosensory thalamus following chronic lesion of the ventral lateral spinal cord. Neuroscience, 101: 393–401. Yezierski, R.P. (2001) Pain following spinal cord injury: pathophysiology and central mechanisms. In: Sandku¨hler J., et al. (Eds.), Progress in Brain Research. Elsevier, Amsterdam, pp. 429–449.

This page intentionally left blank

SECTION IV

Novel Aspects of Clinical Research in CNS Injury

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 14

Monitoring cerebral oxygenation in traumatic brain injury Iain K. Haitsma and Andrew I.R. Maas Department of Neurosurgery, Erasmus Medical Center, ’s-Gravendijkwal 230, 3015 CE Rotterdam, The Netherlands

Abstract: Ischemia is a common problem after traumatic brain injury (TBI) that eludes detection with standard monitoring. In this review we will discuss four available techniques (SjVO2, PET, NIRS and PbrO2) to monitor cerebral oxygenation. We present technical data including strengths and weaknesses of these systems, information from clinical studies and formulate a vision for the future. Keywords: TBI; PbrO2; brain tissue oxygen tension; SjVO2; PET; NIRS; tissue oxygen reactivity; advanced neuromonitoring (dis)advantages, the available clinical data with results of interventions with CPP and respiratory variations and will conclude summarizing our view of the future with these new technologies.

Introduction Treatment of patients with severe traumatic brain injury (TBI) is largely focused on the prevention of secondary insults. Standard monitoring consists of measuring intra cranial pressure (ICP), (mean) arterial blood pressure (MAP) and thus calculating the cerebral perfusion pressure (CPP ¼ MAPICP). This has led to ICP and CPP driven treatment protocols (Rosner et al., 1995; Nordstrom, 2005). However these measurements do not give further information on the existence of ischemia which is common in post mortem studies after TBI (Graham et al., 1978, 1989). Much effort has gone into the development of extra monitoring of the injured brain, so called multi modal monitoring (Unterberg et al., 1997; Meixensberger et al., 1998; Mulvey et al., 2004; De Georgia and Deogaonkar, 2005; Vespa, 2005). This review will focus on monitoring oxygenation in the injured brain. We will discuss the available technologies with special emphasis on brain tissue oxygen tension measurement, their

Monitoring technologies Global vs. focal monitoring The different measurement modalities have different working areas. Some work very focally (e.g. brain tissue oxygen tension measurement) others globally (e.g. jugular bulb oxygenation). This leads to differences in their use and interpretation of their values (Sarrafzadeh et al., 1998; Gopinath et al., 1999; Bellander et al., 2004; Engstrom et al., 2005). The advantage of a global measurement is the information on a large part of the injured brain, more so because most treatments in TBI are systemic. The more focal measurements have been typically used in two ways. Either they have been applied to non-injured areas in essence using them as global monitors or to the penumbra of focal lesions, for instance contusions. The second

Corresponding author. Tel.: +31-10-463-4077; Fax: +31-10-

463-4075; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61014-5

207

208 Table 1. Information regarding measurement technologies Technique

Catheter location

Global vs. focal

Continuous or intermittent

Jugular bulb oximetry PET Near infra red spectroscopy Brain tissue oxygen tension

Intravenous Extracranial Extracranial Intraparenchymal

Global Global and focal Focal Focal

Continuous or intermittent Intermittent Continuous or intermittent Continuous

approach will provide earlier information on the state of the tissue most threatened at that time. However there are practical problems in deﬁning the penumbra zone, in accurately positioning the catheter and ﬁnally in interpreting the absolute values measured. The data regarding the different systems combined with their ability to measure continuously or intermittently are summarized in Table 1.

Jugular bulb oximetry The jugular bulb, the cranial tip of the internal jugular vein, contains pure cerebrovenous blood. Back in the early 1900s the jugular vein was pierced to gain information on cerebrovenous oxygenation (White and Baker, 2002). Later in the 1940s it was used for calculating CBF by Kety and Schmidt (1945). Since the early 1990s literature has accumulated on the value of measuring the oxygen saturation of jugular blood (Robertson et al., 1992; Cruz, 1993a; De Deyne, 1999; White and Baker, 2002). The measurement can be done with intermittent blood gas samples or continuously by inserting a ﬁberoptic catheter. The side of cannulation is debated. Some propose to use the right side, as the right jugular bulb is usually dominant, others test for dominancy of a side, but alternatively the choice can be made to measure on the side that is most injured. Variability exists if both sides are measured continuously (Metz et al., 1998). This of course raises questions on whether the values can really be interpreted as indicative of the whole brain. Technical problems have been reported with continuous measurements with time of good quality ranging from 43 to 91.5% (Cruz,

1993b; Fortune et al., 1994; Kiening et al., 1996; Meixensberger et al., 1998; Gopinath et al., 1999). One article is devoted to possible technical problems with jugular oximetry (Dearden and Midgley, 1993). Others claim to have little difﬁculties with jugular venous oxygen saturation (SjVO2) measurements (Cruz, 1997). Apart from measuring jugular oxygen saturation further information can be gained by calculating the arteriovenous difference in oxygen content (AVDO2) or even cerebral extraction of oxygen (Le Roux et al., 1997; Cruz, 1998). In conclusion, it can be stated that jugular oximetry is a relatively invasive mode of measurement that can be subject to technical difﬁculties, the best side of measurement is the topic of debate. It can give valuable information on oxygen consumption either continuously or intermittent.

PET scan Using 15O2 as an inhalant information can be gained on oxygen consumption using positron emission tomography (PET) scanning. This method is non-invasive, it does however expose the patient to radiation and more importantly depends on the patient being out of his relatively safe ICU environment for a longer period. Other downsides include the intermittent character and the limited availability of this technique. Information can be gained on cerebral blood ﬂow (CBF), oxygen extraction fraction (OEF) and cerebral metabolic rate of O2 (CMRO2). The calculations are regional, but the whole brain can be measured. Using voxel-based analysis the ischemic burden can be quantiﬁed (Coles et al., 2004b).

209

Near infra red spectroscopy (NIRS) NIRS is a non-invasive measurement of Hb oxygenation. As the name implies it uses near infrared light that can cross the skull. Hb is a strong absorber of near infrared light and the amount of absorption depends on the degree of Hb oxygenation. Information can be supplied in the percentage of HbO2 saturation, or absolute changes in HbO2, deoxyhemoglobin and total hemoglobin in micromolar concentrations, referenced to an arbitrary baseline (Cho et al., 2000). Technical problems have precluded wide scale use of this noninvasive measurement. If the source and detector are not far enough apart, measurements are inﬂuenced by subcutaneous blood ﬂow. With complete ischemia near normal values have been found, possibly related to venous blood sequestered in the infracted tissue (Gomersall et al., 1997). Because of these technical problems, especially in TBI patients, NIRS is currently rarely used in adults. With recent modiﬁcations its use might increase again (Al-Rawi, 2005). Technical limitations are less relevant in neonates.

Brain tissue oxygen tension (PbrO2) Following extensive pre-clinical work, the introduction of clinical PbrO2 measurements was pioneered by the Rotterdam group (Maas et al., 1993) and the technique formally introduced during the ﬁrst congress on cerebral oxygenation organized by the Rotterdam, Berlin and Wurzburg groups in 1995. Initial publications followed rapidly (Kiening et al., 1996; van Santbrink et al., 1996; Meixensberger et al., 1998). Many articles including review articles are available on this subject (Andrews, 2001; Haitsma and Maas, 2002; Bader et al., 2003; Littlejohns et al., 2003; De Georgia and Deogaonkar, 2005; Vespa, 2005; Rose et al., 2006). Most reports use one of the two technologies — the Licox system (Integra lifesciences, Plainsboro, NJ) or the Para/Neurotrend system (Charbel et al., 1997) (Diametrics Medical Ltd., UK). Unfortunately the Para/Neurotrend system is no longer available. Both techniques require intraparenchymal insertion of the 0.5–0.8 mm

thick probes, usually via a bolt system. The Licox system uses a Clark type electrode, which generates a current that is dependent on the amount of oxygen near the catheter tip. The Paratrend also uses a Clark electrode and later changes to an optode system in the Neurotrend. The Para/ Neurotrend supplies information on local temperature, pCO2 and pH as well. The Licox now also incorporates a temperature sensor. Complication rates are equally low as in the use of intraparenchymal ICP monitors and data are reliable with very low catheter drift rates (Kiening et al., 1996; Dings et al., 1998; van den Brink et al., 2000; Hoelper et al., 2005). Time of good data quality of up to 95% has been reported (Meixensberger et al., 1998). Catheters are usually inserted in the frontal white matter and give continuous information on the local balance between oxygen supply and uptake. After insertion, a run in time of 1–2 h is advised before reliable data are produced (Dings et al., 1998; van den Brink et al., 2000). We recommend performing an oxygen challenge after catheter insertion to verify its proper placement, as even small hematomas around the catheter tip can lower the oxygen readings, which would result in no or very little increase in PbrO2 with a large increase in inspiratory oxygen fraction (FiO2) or more appropriate arterial oxygen tension (PaO2) (van den Brink et al., 1998). In conclusion, PbrO2 measurement is an invasive, but safe and reliable measurement that gives continuous information on local oxygenation.

Clinical data Jugular oximetry Low SjVO2 values (below 50–55%) are related to poorer outcome (Gopinath et al., 1994; Fandino et al., 2000; Perez et al., 2003). The same holds for values higher than 75% (Cormio et al., 1999; Macmillan et al., 2001). Others have challenged these classic cut-off values after measurements in patients with Cushing syndrome (44.7–69.5%) (Chieregato et al., 2003). A limited improvement as opposed to clear improvement in increased

210

AVDO2 after treatment is also related to poorer prognosis (Le Roux et al., 1997). In two important publications the Houston group compared ICP (CPP50 mmHg, PaCO2 25–30 mmHg) and CBF (CPP70 mmHg, PaCO2 35 mmHg) regimens to each other (Robertson et al., 1999; Contant et al., 2001). In the CBF group jugular desaturations were less frequent (30% vs. 50.6%). No difference in outcome was observed, possibly because beneﬁts of treatment of jugular desaturations were offset by a ﬁvefold increase in ARDS in the CBF group. These publications led to lowering of the recommended CPP level in the guidelines from 70 to 60 mmHg. Hyperventilation leads to a decrease in SjVO2 and increase in AVDO2, these effects however can be offset by hyperoxygenation (Matta et al., 1994; Thiagarajan et al., 1998). A treatment protocol using the cerebral extraction of oxygen as guide for treatment including targeted hyperventilation has shown promising results in adults with TBI and diffuse brain swelling and separately in children with severe TBI and increased ICP (Cruz, 1998; Cruz et al., 2002). In TBI patients with decreased CBF hyperbaric O2 therapy can increase CBF and CMRO2 at 1 and 6 h after treatment, at constant AVDO2 (Rockswold et al., 2001). Increased ICP and CSF lactate levels were also lowered.

PET measurements Even with PET measurements debate exists on the frequency of regional ischemia (Coles et al., 2004a; Vespa et al., 2005). Ischemia was deﬁned as the oxygen extraction fraction at a cerebral venous oxygen (CVO2) content below 3.5 ml/100 ml. One study reports ischemic volume ranging from 1 to 16% (Coles et al., 2004a), the other maximally 1% of brain volume in TBI patients (Vespa et al., 2005). Increasing CPP from 70 to 90 mmHg produces a signiﬁcant fall in ischemic brain volume (IBV) which is most pronounced in patients with a larger IBV (Coles et al., 2004c). A study published in 2000 found a decrease in CBF but no change in CMRO2 with

hyperventilation (Diringer et al., 2000). In a study with 13 TBI patients receiving hyperventilation, CBF was reduced to even below 20 ml/100 g/min, but no changes in oxygen extraction fraction were noted. It was concluded that no energy failure took place during hyperventilation (Diringer et al., 2002). Another study conﬁrmed that hyperventilation lowered CBF but not to the level of ischemia (Coles et al., 2002). NIRS Few clinical studies with regard to oxygenation measurements have been published on NIRS. Interesting is its possible use as a detector of intracranial hematomas. In the presence of a subdural hematoma or bloody contusion, or even in the event of massive blood in the subarachnoid space, absorption is high, thus preventing sufﬁcient light to return to the detectors. This phenomenon may be used to detect delayed traumatic hematomas, leading to better timed follow-up CTs and operations (Gopinath et al., 1993, 1995). PbrO2 Low PbrO2 values are observed in up to 50% of patients during the ﬁrst 24 h with depth and duration of tissue hypoxia being related to outcome and being an independent predictor of unfavorable outcome and death (Bardt et al., 1998; Valadka et al., 1998; van den Brink et al., 2000; Stiefel et al., 2006). No absolute value can be given as a cut-off point for tissue hypoxia but values ranging from 10 to 20 mmHg in uninjured frontal white matter are mentioned based on prognostic analysis or related to CBF values (Zauner et al., 1997; Bardt et al., 1998; Doppenberg et al., 1998b; Valadka et al., 1998; van den Brink et al., 2000). A sudden fall in PbrO2 to zero, without the development of a hematoma around the catheter, is considered an early sign of brain death (van Santbrink et al., 1996; Palmer and Bader, 2005). Recently interest has increased on testing autoregulation (Mascia et al., 2000; Czosnyka et al., 2001; Steiner et al., 2002; Cremer et al., 2004). Patients with mean CPP levels close to their

211

optimal CPP are more likely to have a favorable outcome (Steiner et al., 2002). The response of PbrO2 to changes in CPP is also dependent on the state of autoregulation (Lang et al., 2003; Cremer et al., 2004). With increased ICP a CPP reduction below 77 mmHg can lead to a further ICP increase and decrease in PbrO2 (Cremer et al., 2004). Higher CPP levels might be useful in normalizing tissue oxygenation in ischemic areas (Stocchetti et al., 1998). A historical case control study by the Leipzig group from 2003 with increased CPP for low PbrO2 values showed less frequent cerebral hypoxia, but no difference in 6-month outcome (Meixensberger et al., 2003). One publication by the Pennsylvania group claimed reduced patient death using both ICP and PbrO2 monitoring (Stiefel et al., 2005). This was based on a historical case control study, in which patients with ICP and PbrO2 monitoring ﬁrst were treated according to an ICP/CPP protocol, later patients were entered in a stepwise protocol that also treated PbrO2 values below 25 mmHg. Mortality rate dropped from 44 to 25%. The major critique on this study concerns its retrospective nature. Hyperventilation usually lowers PbrO2 (Schneider et al., 1998), and reactivity is higher in patients with poorer prognosis, statistically signiﬁcant on day 5 after injury (Carmona Suazo et al., 2000). Increases in PbrO2 with hyperventilation in TBI have however also been reported (Dings et al., 1996; Imberti et al., 2000). PbrO2 can be increased by increasing FiO2/ PaO2. Tissue oxygen response (TOR) can be calculated as: TOR ¼

PbrO2 1 PaO2 PbrO2

Higher values of TOR are related to poorer outcome and may reﬂect a loss of oxygen autoregulation (van Santbrink et al., 1996, 2003). Several groups have investigated the effects of increasing FiO2/PaO2 to supranormal levels. In 1999 the Richmond group reported a fall in brain lactate levels measured by microdialysis after administering 100% O2 (Menzel et al., 1999a, b). These results were not without controversy (Rossi

and Stocchetti, 1999; Bressack and Schiffman, 2003). A report by the Milan group showed reduced lactate levels, but not the lactate–pyruvate ratio with hyperoxia which was interpreted as absence of improvement in cerebral metabolism (Magnoni et al., 2003). A more recent report by the Richmond and Bern groups also demonstrated a fall in lactate–pyruvate ratio after hyperoxia compared to a historical control group (Tolias et al., 2004). Nevertheless, even oxygen can be considered a drug and is not without its side effects (Lodato, 1990). Further reports on outcome measures with hyperoxia treatment are eagerly awaited. PbrO2 decreases with hypothermia with an extra reduction below 351C (Gupta et al., 2002) in one study, others report an increase in PbrO2 with hypothermia (Zhi et al., 1999; Jia et al., 2005). Decompressive craniectomies for intractable ICP leads to increased PbrO2 (Jaeger et al., 2003; Stiefel et al., 2004).

Technology comparisons SjVO2 values correlate with IBV, deﬁned as the area with CVO2 content p3.5 ml/100 ml as measured by PET scan, with SjVO2 o50% occurring at IBV at 1375% (Coles et al., 2004a). In a study in 14 patients NIRS detected twice as many events as SjVO2 monitoring although the clinical signiﬁcance of this is not known (Kirkpatrick et al., 1995). Another study found poor correlation between NIRS and SjVO2 monitoring even with SjVO2 desaturations below 55% (Lewis et al., 1996). In a study in 60 children without TBI, SjVO2 and NIRS correlated signiﬁcantly but NIRS had lower sensitivity (Nagdyman et al., 2005). SjVO2 and PbrO2 correlate closely especially when CPP falls below 60 mmHg (Kiening et al., 1996). This correlation is less in areas with focal pathology (Gupta et al., 1999). Correlation between SjVO2 and PbrO2 during CO2 reactivity testing is low (Fandino et al., 1999). In a study in 36 TBI patients moderate hyperventilation reduced SjVO2, but not below 50% with PbrO2 values below 10 mmHg (Imberti et al., 2002), in some tests SjVO2 and PbrO2 changed in

212

different directions, supporting the concept that these technologies are complementary. In a 2002 study PbrO2 did not correlate with cerebral end capillary O2 tension as measured with PET. The change in both values during hyperventilation was however signiﬁcantly correlated. PbrO2 is strongly correlated with local CBF measurement (Doppenberg et al., 1998a; Jaeger et al., 2005).

Acknowledgments The work of Dr. Haitsma is supported by NWO grant: 920-03-130. The Department of Neurosurgery, EMC, has received in natural support of research in neurocritical care from GMS mbh. (Kiel, Germany) and Codman/Johnson & Johnson (Raynham, MA, USA).

References Summary and conclusions ICP and CPP measurements alone are inadequate for detecting ischemia. The above-mentioned techniques can provide valuable information on oxygenation. An ideal technique (non-invasive, continuous, reliable and easy to use on the ICU) does not exist. To gain maximal insight the combination of a local and global monitor should sufﬁce. However the advantages in outcome with extra measurements or targeted treatment have not yet been unequivocally proven. Any effort to do so should be stimulated and probably be part of a multi-center initiative, for instance Brain-IT. Recent reports of improved survival after PbrO2 measurement and treatment (Stiefel et al., 2005) and reports of improved metabolism with hyperoxia (Tolias et al., 2004) seem very promising. The same holds true for the reports on targeted hyperventilation with jugular bulb measurements (Cruz, 1998; Cruz et al., 2002). These studies deserve veriﬁcation in other centers. Because of their broad availability SjVO2 and PbrO2 measurements seem to be the most logical for multi center studies with a slight preference for the Licox system because of its ease of use. After stepwise implementations of new monitoring techniques a software system capable of integrating all values and supplying recommendations for further diagnostic steps or treatment will be crucial. Until then we will eagerly await and participate in further studies on autoregulation guided CPP treatment, effect of hyperoxygenation treatment, targeted hyperventilation and using multimodality monitoring as secondary endpoints in the use of treatments such as barbiturate coma, hypothermia and decompressive craniectomy.

Al-Rawi, P.G. (2005) Near infrared spectroscopy in brain injury: today’s perspective. Acta Neurochir. Suppl., 95: 453–457. Andrews, R.J. (2001) Monitoring for neuroprotection. New technologies for the new millennium. Ann. N.Y. Acad. Sci., 939: 101–113. Bader, M.K., Littlejohns, L.R. and March, K. (2003) Brain tissue oxygen monitoring in severe brain injury, II. Implications for critical care teams and case study. Crit. Care Nurse, 23: 29–38, 40–42, 44. Bardt, T.F., Unterberg, A.W., Hartl, R., Kiening, K.L., Schneider, G.H. and Lanksch, W.R. (1998) Monitoring of brain tissue PO2 in traumatic brain injury: effect of cerebral hypoxia on outcome. Acta Neurochir. Suppl., 71: 153–156. Bellander, B.M., Cantais, E., Enblad, P., Hutchinson, P., Nordstrom, C.H., Robertson, C., Sahuquillo, J., Smith, M., Stocchetti, N., Ungerstedt, U., Unterberg, A. and Olsen, N.V. (2004) Consensus meeting on microdialysis in neurointensive care. Intensive Care Med., 30: 2166–2169. Bressack, M.A. and Schiffman, J. (2003) Hyperoxia: good or bad? J. Neurosurg., 98: 223–224; author reply 224–225. Carmona Suazo, J.A., Maas, A.I., van den Brink, W.A., van Santbrink, H., Steyerberg, E.W. and Avezaat, C.J. (2000) CO2 reactivity and brain oxygen pressure monitoring in severe head injury. Crit. Care Med., 28: 3268–3274. Charbel, F.T., Hoffman, W.E., Misra, M., Hannigan, K. and Ausman, J.I. (1997) Cerebral interstitial tissue oxygen tension, pH, HCO3, CO2. Surg. Neurol., 48: 414–417. Chieregato, A., Calzolari, F., Trasforini, G., Targa, L. and Latronico, N. (2003) Normal jugular bulb oxygen saturation. J. Neurol. Neurosurg. Psychiatry, 74: 784–786. Cho, H., Nemoto, E.M., Sanders, M., Fernandez, K. and Yonas, H. (2000) Comparison of two commercially available near-infrared spectroscopy instruments for cerebral oximetry. Technical note. J. Neurosurg., 93: 351–354. Coles, J.P., Fryer, T.D., Smielewski, P., Chatﬁeld, D.A., Steiner, L.A., Johnston, A.J., Downey, S.P., Williams, G.B., Aigbirhio, F., Hutchinson, P.J., Rice, K., Carpenter, T.A., Clark, J.C., Pickard, J.D. and Menon, D.K. (2004a) Incidence and mechanisms of cerebral ischemia in early clinical head injury. J. Cereb. Blood Flow Metab., 24: 202–211. Coles, J.P., Fryer, T.D., Smielewski, P., Rice, K., Clark, J.C., Pickard, J.D. and Menon, D.K. (2004b) Deﬁning ischemic

213 burden after traumatic brain injury using 15O PET imaging of cerebral physiology. J. Cereb. Blood Flow Metab., 24: 191–201. Coles, J.P., Minhas, P.S., Fryer, T.D., Smielewski, P., Aigbirihio, F., Donovan, T., Downey, S.P., Williams, G., Chatﬁeld, D., Matthews, J.C., Gupta, A.K., Carpenter, T.A., Clark, J.C., Pickard, J.D. and Menon, D.K. (2002) Effect of hyperventilation on cerebral blood ﬂow in traumatic head injury: clinical relevance and monitoring correlates. Crit. Care Med., 30: 1950–1959. Coles, J.P., Steiner, L.A., Johnston, A.J., Fryer, T.D., Coleman, M.R., Smieleweski, P., Chatﬁeld, D.A., Aigbirhio, F., Williams, G.B., Boniface, S., Rice, K., Clark, J.C., Pickard, J.D. and Menon, D.K. (2004c) Does induced hypertension reduce cerebral ischaemia within the traumatized human brain? Brain, 127: 2479–2490. Contant, C.F., Valadka, A.B., Gopinath, S.P., Hannay, H.J. and Robertson, C.S. (2001) Adult respiratory distress syndrome: a complication of induced hypertension after severe head injury. J. Neurosurg., 95: 560–568. Cormio, M., Valadka, A.B. and Robertson, C.S. (1999) Elevated jugular venous oxygen saturation after severe head injury. J. Neurosurg., 90: 9–15. Cremer, O.L., van Dijk, G.W., Amelink, G.J., de Smet, A.M., Moons, K.G. and Kalkman, C.J. (2004) Cerebral hemodynamic responses to blood pressure manipulation in severely head-injured patients in the presence or absence of intracranial hypertension. Anesth. Analg., 99: 1211–1217, table of contents. Cruz, J. (1993a) Cerebral oxygenation. Monitoring and management. Acta Neurochir. Suppl., 59: 86–90. Cruz, J. (1993b) On-line monitoring of global cerebral hypoxia in acute brain injury. Relationship to intracranial hypertension. J. Neurosurg., 79: 228–233. Cruz, J. (1997) Continuous monitoring of partial pressure of brain tissue oxygen in patients with severe head injury. Neurosurgery, 40: 1104. Cruz, J. (1998) The ﬁrst decade of continuous monitoring of jugular bulb oxyhemoglobinsaturation: management strategies and clinical outcome. Crit. Care Med., 26: 344–351. Cruz, J., Nakayama, P., Imamura, J.H., Rosenfeld, K.G., de Souza, H.S. and Giorgetti, G.V. (2002) Cerebral extraction of oxygen and intracranial hypertension in severe, acute, pediatric brain trauma: preliminary novel management strategies. Neurosurgery, 50: 774–779; discussion 779–780. Czosnyka, M., Smielewski, P., Piechnik, S., Steiner, L.A. and Pickard, J.D. (2001) Cerebral autoregulation following head injury. J. Neurosurg., 95: 756–763. Dearden, N.M. and Midgley, S. (1993) Technical considerations in continuous jugular venous oxygen saturation measurement. Acta Neurochir. Suppl., 59: 91–97. De Deyne, C. (1999) Jugular bulb oximetry: the link between cerebral and systemic management of severe head injury. Intensive Care Med., 25: 430–431. De Georgia, M.A. and Deogaonkar, A. (2005) Multimodal monitoring in the neurological intensive care unit. Neuro logist, 11: 45–54.

Dings, J., Meixensberger, J., Amschler, J., Hamelbeck, B. and Roosen, K. (1996) Brain tissue PO2 in relation to cerebral perfusion pressure, TCD ﬁndings and TCD-CO2-reactivity after severe head injury. Acta Neurochir., 138: 425–434. Dings, J., Meixensberger, J., Jager, A. and Roosen, K. (1998) Clinical experience with 118 brain tissue oxygen partial pressure catheter probes. Neurosurgery, 43: 1082–1095. Diringer, M.N., Videen, T.O., Yundt, K., Zazulia, A.R., Aiyagari, V., Dacey Jr., R.G., Grubb, R.L. and Powers, W.J. (2002) Regional cerebrovascular and metabolic effects of hyperventilation after severe traumatic brain injury. J. Neurosurg., 96: 103–108. Diringer, M.N., Yundt, K., Videen, T.O., Adams, R.E., Zazulia, A.R., Deibert, E., Aiyagari, V., Dacey Jr., R.G., Grubb Jr., R.L. and Powers, W.J. (2000) No reduction in cerebral metabolism as a result of early moderate hyperventilation following severe traumatic brain injury. J. Neurosurg., 92: 7–13. Doppenberg, E.M., Zauner, A., Bullock, R., Ward, J.D., Fatouros, P.P. and Young, H.F. (1998a) Correlations between brain tissue oxygen tension, carbon dioxide tension, pH, and cerebral blood ﬂow–a better way of monitoring the severely injured brain? Surg. Neurol., 49: 650–654. Doppenberg, E.M., Zauner, A., Watson, J.C. and Bullock, R. (1998b) Determination of the ischemic threshold for brain oxygen tension. Acta Neurochir. Suppl., 71: 166–169. Engstrom, M., Polito, A., Reinstrup, P., Romner, B., Ryding, E., Ungerstedt, U. and Nordstrom, C.H. (2005) Intracerebral microdialysis in severe brain trauma: the importance of catheter location. J. Neurosurg., 102: 460–469. Fandino, J., Stocker, R., Prokop, S. and Imhof, H.G. (1999) Correlation between jugular bulb oxygen saturation and partial pressure of brain tissue oxygen during CO2 and O2 reactivity tests in severely head-injured patients. Acta Neurochir., 141: 825–834. Fandino, J., Stocker, R., Prokop, S., Trentz, O. and Imhof, H.G. (2000) Cerebral oxygenation and systemic trauma related factors determining neurological outcome after brain injury. J. Clin. Neurosci., 7: 226–233. Fortune, J.B., Feustel, P.J., Weigle, C.G. and Popp, A.J. (1994) Continuous measurement of jugular venous oxygen saturation in response to transient elevations of blood pressure in head-injured patients. J. Neurosurg., 80: 461–468. Gomersall, C.D., Joynt, G.M., Gin, T., Freebairn, R.C. and Stewart, I.E. (1997) Failure of the INVOS 3100 cerebral oximeter to detect complete absence of cerebral blood ﬂow. Crit. Care Med., 25: 1252–1254. Gopinath, S.P., Robertson, C.S., Contant, C.F., Hayes, C., Feldman, Z., Narayan, R.K. and Grossman, R.G. (1994) Jugular venous desaturation and outcome after head injury. J. Neurol. Neurosurg. Psychiatry, 57: 717–723. Gopinath, S.P., Robertson, C.S., Contant, C.F., Narayan, R.K., Grossman, R.G. and Chance, B. (1995) Early detection of delayed traumatic intracranial hematomas using near-infrared spectroscopy. J. Neurosurg., 83: 438–444. Gopinath, S.P., Robertson, C.S., Grossman, R.G. and Chance, B. (1993) Near-infrared spectroscopic localization of intracranial hematomas. J. Neurosurg., 79: 43–47.

214 Gopinath, S.P., Valadka, A.B., Uzura, M. and Robertson, C.S. (1999) Comparison of jugular venous oxygen saturation and brain tissue PO2 as monitors of cerebral ischemia after head injury. Crit. Care Med., 27: 2337–2345. Graham, D.I., Adams, J.H. and Doyle, D. (1978) Ischaemic brain damage in fatal non-missile head injuries. J. Neurol. Sci., 39: 213–234. Graham, D.I., Ford, I., Adams, J.H., Doyle, D., Teasdale, G.M., Lawrence, A.E. and McLellan, D.R. (1989) Ischaemic brain damage is still common in fatal non-missile head injury. J. Neurol. Neurosurg. Psychiatry, 52: 346–350. Gupta, A.K., Al-Rawi, P.G., Hutchinson, P.J. and Kirkpatrick, P.J. (2002) Effect of hypothermia on brain tissue oxygenation in patients with severe head injury. Br. J. Anaesth., 88: 188–192. Gupta, A.K., Hutchinson, P.J., Al-Rawi, P., Gupta, S., Swart, M., Kirkpatrick, P.J., Menon, D.K. and Datta, A.K. (1999) Measuring brain tissue oxygenation compared with jugular venous oxygen saturation for monitoring cerebral oxygenation after traumatic brain injury. Anesth. Analg., 88: 549–553. Haitsma, I.K. and Maas, A.I. (2002) Advanced monitoring in the intensive care unit: brain tissue oxygen tension. Curr. Opin. Crit. Care, 8: 115–120. Hoelper, B.M., Alessandri, B., Heimann, A., Behr, R. and Kempski, O. (2005) Brain oxygen monitoring: in-vitro accuracy, long-term drift and response-time of Licox- and Neurotrend sensors. Acta Neurochir. (Wien.), 147: 767–774. Imberti, R., Bellinzona, G. and Langer, M. (2002) Cerebral tissue PO2 and SjvO2 changes during moderate hyper ventilation in patients with severe traumatic brain injury. J. Neurosurg., 96: 97–102. Imberti, R., Ciceri, M., Bellinzona, G. and Pugliese, R. (2000) The use of hyperventilation in the treatment of plateau waves in two patients with severe traumatic brain injury: contrasting effects on cerebral oxygenation. J. Neurosurg. Anesthesiol, 12: 124–127. Jaeger, M., Soehle, M. and Meixensberger, J. (2003) Effects of decompressive craniectomy on brain tissue oxygen in patients with intracranial hypertension. J. Neurol. Neurosurg. Psychiatry, 74: 513–515. Jaeger, M., Soehle, M., Schuhmann, M.U., Winkler, D. and Meixensberger, J. (2005) Correlation of continuously monitored regional cerebral blood ﬂow and brain tissue oxygen. Acta Neurochir. (Wien.), 147: 51–56; discussion 56. Jia, J., Lin, Y.Q., Liu, W.F., Zhong, T.A., Zhang, J., Ye, Y. and Xu, Y.Q. (2005) Study of the effects of mild hypothermia on cerebral PO2, PCO2 and pH and body temperature in patients with acute severe head injury. Chin. J. Traumatol., 8: 138–141. Kety, S.S. and Schmidt, C.F. (1945) The determination of cerebral blood ﬂow in man by the use of nitrous oxide in low concentrations. Am. J. Physiol., 143: 53–66. Kiening, K.L., Unterberg, A.W., Bardt, T.F., Schneider, G.H. and Lanksch, W.R. (1996) Monitoring of cerebral oxygenation in patients with severe head injuries: brain tissue PO2 versus jugular vein oxygen saturation. J. Neurosurg., 85: 751–757.

Kirkpatrick, P.J., Smielewski, P., Czosnyka, M., Menon, D.K. and Pickard, J.D. (1995) Near-infrared spectroscopy use in patients with head injury. J. Neurosurg., 83: 963–970. Lang, E.W., Czosnyka, M. and Mehdorn, H.M. (2003) Tissue oxygen reactivity and cerebral autoregulation after severe traumatic brain injury. Crit. Care Med., 31: 267–271. Le Roux, P.D., Newell, D.W., Lam, A.M., Grady, M.S. and Winn, H.R. (1997) Cerebral arteriovenous oxygen difference: a predictor of cerebral infarction and outcome in patients with severe head injury. J. Neurosurg., 87: 1–8. Lewis, S.B., Myburgh, J.A., Thornton, E.L. and Reilly, P.L. (1996) Cerebral oxygenation monitoring by near-infrared spectroscopy is not clinically useful in patients with severe closed-head injury: a comparison with jugular venous bulb oximetry. Crit. Care Med., 24: 1334–1338. Littlejohns, L.R., Bader, M.K. and March, K. (2003) Brain tissue oxygen monitoring in severe brain injury, I. Research and usefulness in critical care. Crit. Care Nurse, 23: 17–25; quiz 26–27. Lodato, R.F. (1990) Oxygen toxicity. Crit. Care Clin., 6: 749–765. Maas, A.I., Fleckenstein, W., de Jong, D.A. and van Santbrink, H. (1993) Monitoring cerebral oxygenation: experimental studies and preliminary clinical results of continuous monitoring of cerebrospinal ﬂuid and brain tissue oxygen tension. Acta Neurochir. Suppl., 59: 50–57. Macmillan, C.S., Andrews, P.J. and Easton, V.J. (2001) Increased jugular bulb saturation is associated with poor outcome in traumatic brain injury. J. Neurol. Neurosurg. Psychiatry, 70: 101–104. Magnoni, S., Ghisoni, L., Locatelli, M., Caimi, M., Colombo, A., Valeriani, V. and Stocchetti, N. (2003) Lack of improvement in cerebral metabolism after hyperoxia in severe head injury: a microdialysis study. J. Neurosurg., 98: 952–958. Mascia, L., Andrews, P.J., McKeating, E.G., Souter, M.J., Merrick, M.V. and Piper, I.R. (2000) Cerebral blood ﬂow and metabolism in severe brain injury: the role of pressure autoregulation during cerebral perfusion pressure management. Intensive Care Med., 26: 202–205. Matta, B.F., Lam, A.M. and Mayberg, T.S. (1994) The inﬂuence of arterial oxygenation on cerebral venous oxygen saturation during hyperventilation. Can. J. Anaesth., 41: 1041–1046. Meixensberger, J., Jaeger, M., Vath, A., Dings, J., Kunze, E. and Roosen, K. (2003) Brain tissue oxygen guided treatment supplementing ICP/CPP therapy after traumatic brain injury. J. Neurol. Neurosurg. Psychiatry, 74: 760–764. Meixensberger, J., Jager, A., Dings, J., Baunach, S. and Roosen, K. (1998) Multimodal hemodynamic neuromonitoring–quality and consequences for therapy of severely head injured patients. Acta Neurochir. Suppl., 71: 260–262. Menzel, M., Doppenberg, E.M., Zauner, A., Soukup, J., Reinert, M.M. and Bullock, R. (1999a) Increased inspired oxygen concentration as a factor in improved brain tissue oxygenation and tissue lactate levels after severe human head injury. J. Neurosurg., 91: 1–10.

215 Menzel, M., Doppenberg, E.M., Zauner, A., Soukup, J., Reinert, M.M., Clausen, T., Brockenbrough, P.B. and Bullock, R. (1999b) Cerebral oxygenation in patients after severe head injury: monitoring and effects of arterial hyperoxia on cerebral blood ﬂow, metabolism and intracranial pressure. J. Neurosurg. Anesthesiol., 11: 240–251. Metz, C., Holzschuh, M., Bein, T., Woertgen, C., Rothoerl, R., Kallenbach, B., Taeger, K. and Brawanski, A. (1998) Monitoring of cerebral oxygen metabolism in the jugular bulb: reliability of unilateral measurements in severe head injury. J. Cereb. Blood Flow Metab., 18: 332–343. Mulvey, J.M., Dorsch, N.W., Mudaliar, Y. and Lang, E.W. (2004) Multimodality monitoring in severe traumatic brain injury; The role of brain tissue oxygenation monitoring. Neurocrit. Care, 391–402. Nagdyman, N., Fleck, T., Schubert, S., Ewert, P., Peters, B., Lange, P.E. and Abdul-Khaliq, H. (2005) Comparison between cerebral tissue oxygenation index measured by nearinfrared spectroscopy and venous jugular bulb saturation in children. Intensive Care Med., 31: 846–850. Nordstrom, C.H. (2005) Physiological and biomechanical principles underlying volume-targeted therapy – the "Lund Concept". Neurocrit. Care, 83–96. Palmer, S. and Bader, M.K. (2005) Brain tissue oxygenation in brain death. Neurocrit. Care, 17–22. Perez, A., Minces, P.G., Schnitzler, E.J., Agosta, G.E., Medina, S.A. and Ciraolo, C.A. (2003) Jugular venous oxygen saturation or arteriovenous difference of lactate content and outcome in children with severe traumatic brain injury. Pediatr. Crit. Care Med., 4: 33–38. Robertson, C.S., Contant, C.F., Gokaslan, Z.L., Narayan, R.K. and Grossman, R.G. (1992) Cerebral blood ﬂow, arteriovenous oxygen difference, and outcome in head injured patients. J. Neurol. Neurosurg. Psychiatry, 55: 594–603. Robertson, C.S., Valadka, A.B., Hannay, H.J., Contant, C.F., Gopinath, S.P., Cormio, M., Uzura, M. and Grossman, R.G. (1999) Prevention of secondary ischemic insults after severe head injury. Crit. Care Med., 27: 2086–2095. Rockswold, S.B., Rockswold, G.L., Vargo, J.M., Erickson, C.A., Sutton, R.L., Bergman, T.A. and Biros, M.H. (2001) Effects of hyperbaric oxygenation therapy on cerebral metabolism and intracranial pressure in severely brain injured patients. J. Neurosurg., 94: 403–411. Rose, J.C., Neill, T.A. and Hemphill, III, J.C. (2006) Continuous monitoring of the microcirculation in neurocritical care: an update on brain tissue oxygenation. Curr. Opin. Crit. Care, 12: 97–102. Rosner, M.J., Rosner, S.D. and Johnson, A.H. (1995) Cerebral perfusion pressure: management protocol and clinical results. J. Neurosurg., 83: 949–962. Rossi, S. and Stocchetti, N. (1999) Brain tissue oxygenation. J. Neurosurg., 91: 1065–1067. Sarrafzadeh, A.S., Kiening, K.L., Bardt, T.F., Schneider, G.H., Unterberg, A.W. and Lanksch, W.R. (1998) Cerebral oxygenation in contusioned vs. nonlesioned brain tissue: monitoring of PtiO2 with Licox and Paratrend. Acta Neurochir. Suppl., 71: 186–189.

Schneider, G.H., Sarrafzadeh, A.S., Kiening, K.L., Bardt, T.F., Unterberg, A.W. and Lanksch, W.R. (1998) Inﬂuence of hyperventilation on brain tissue-PO2, PCO2, and pH in patients with intracranial hypertension. Acta Neurochir. Suppl., 71: 62–65. Steiner, L.A., Czosnyka, M., Piechnik, S.K., Smielewski, P., Chatﬁeld, D., Menon, D.K. and Pickard, J.D. (2002) Continuous monitoring of cerebrovascular pressure reactivity allows determination of optimal cerebral perfusion pressure in patients with traumatic brain injury. Crit. Care Med., 30: 733–738. Stiefel, M.F., Heuer, G.G., Smith, M.J., Bloom, S., MaloneyWilensky, E., Gracias, V.H., Grady, M.S. and LeRoux, P.D. (2004) Cerebral oxygenation following decompressive hemicraniectomy for the treatment of refractory intracranial hypertension. J. Neurosurg., 101: 241–247. Stiefel, M.F., Spiotta, A., Gracias, V.H., Garuffe, A.M., Guillamondegui, O., Maloney-Wilensky, E., Bloom, S., Grady, M.S. and LeRoux, P.D. (2005) Reduced mortality rate in patients with severe traumatic brain injury treated with brain tissue oxygen monitoring. J. Neurosurg., 103: 805–811. Stiefel, M.F., Udoetuk, J.D., Spiotta, A.M., Gracias, V.H., Goldberg, A., Maloney-Wilensky, E., Bloom, S. and LeRoux, P.D. (2006) Conventional neurocritical care and cerebral oxygenation after traumatic brain injury. J. Neurosurg., 105: 568–575. Stocchetti, N., Chieregato, A., De Marchi, M., Croci, M., Benti, R. and Grimoldi, N. (1998) High cerebral perfusion pressure improves low values of local brain tissue O2 tension (PtiO2) in focal lesions. Acta Neurochir. Suppl., 71: 162–165. Thiagarajan, A., Goverdhan, P.D., Chari, P. and Somasunderam, K. (1998) The effect of hyperventilation and hyperoxia on cerebral venous oxygen saturation in patients with traumatic brain injury. Anesth. Analg., 87: 850–853. Tolias, C.M., Reinert, M., Seiler, R., Gilman, C., Scharf, A. and Bullock, M.R. (2004) Normobaric hyperoxia–induced improvement in cerebral metabolism and reduction in intracranial pressure in patients with severe head injury: a prospective historical cohort-matched study. J. Neurosurg., 101: 435–444. Unterberg, A.W., Kiening, K.L., Hartl, R., Bardt, T., Sarrafzadeh, A.S. and Lanksch, W.R. (1997) Multimodal monitoring in patients with head injury: evaluation of the effects of treatment on cerebral oxygenation. J. Trauma, 42: S32–S37. Valadka, A.B., Gopinath, S.P., Contant, C.F., Uzura, M. and Robertson, C.S. (1998) Relationship of brain tissue PO2 to outcome after severe head injury. Crit. Care Med., 26: 1576–1581. van den Brink, W.A., Haitsma, I.K., Avezaat, C.J., Houtsmuller, A.B., Kros, J.M. and Maas, A.I. (1998) Brain parenchyma/PO2 catheter interface: a histopathological study in the rat. J. Neurotrauma, 15: 813–824. van den Brink, W.A., van Santbrink, H., Steyerberg, E.W., Avezaat, C.J., Suazo, J.A., Hogesteeger, C., Jansen, W.J., Kloos, L.M., Vermeulen, J. and Maas, A.I. (2000) Brain

216 oxygen tension in severe head injury. Neurosurgery, 46: 868–876; discussion 876–878. van Santbrink, H., Maas, A.I. and Avezaat, C.J. (1996) Continuous monitoring of partial pressure of brain tissue oxygen in patients with severe head injury. Neurosurgery, 38: 21–31. van Santbrink, H., vd Brink, W.A., Steyerberg, E.W., Carmona Suazo, J.A., Avezaat, C.J. and Maas, A.I. (2003) Brain tissue oxygen response in severe traumatic brain injury. Acta Neurochir. (Wien.), 145: 429–438; discussion 438. Vespa, P., Bergsneider, M., Hattori, N., Wu, H.M., Huang, S.C., Martin, N.A., Glenn, T.C., McArthur, D.L. and Hovda, D.A. (2005) Metabolic crisis without brain ischemia is common after traumatic brain injury: a combined microdialysis and positron emission tomography study. J. Cereb. Blood Flow Metab., 25(6): 763–774.

Vespa, P.M. (2005) Multimodality monitoring and telemonitoring in neurocritical care: from microdialysis to robotic telepresence. Curr. Opin. Crit. Care, 11: 133–138. White, H. and Baker, A. (2002) Continuous jugular venous oximetry in the neurointensive care unit – a brief review: [L’oxymetrie continue de la veine jugulaire a l’unite des soins intensifs neurologiques – une breve revue]. Can. J. Anaesth., 49: 623–629. Zauner, A., Doppenberg, E.M., Woodward, J.J., Choi, S.C., Young, H.F. and Bullock, R. (1997) Continuous monitoring of cerebral substrate delivery and clearance: initial experience in 24 patients with severe acute brain injuries. Neurosurgery, 41: 1082–1091; discussion 1091–1093. Zhi, D.S., Zhang, S. and Zhou, L.G. (1999) Continuous monitoring of brain tissue oxygen pressure in patients with severe head injury during moderate hypothermia. Surg. Neurol., 52: 393–396.

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 15

Update on the treatment of spinal cord injury Darryl C. Baptiste1 and Michael G. Fehlings1,2, 1

Division of Cell and Molecular Biology, Toronto Western Research Institute and Krembil Neuroscience Centre, Toronto Western Hospital, Toronto, ON, Canada 2 Division of Neurosurgery, Krembil Neuroscience Centre, Toronto Western Hospital, University of Toronto, Toronto, ON, Canada

Abstract: Acute spinal cord injury (SCI) is a devastating neurological disorder that can affect any individual at a given instance. Current treatment options for SCI include the use of high dose methylprednisolone sodium succinate, a corticosteroid, surgical interventions to stabilize and decompress the spinal cord, intensive multisystem medical management, and rehabilitative care. While utility of these therapeutic options provides modest beneﬁts, there is a critical need to identify novel approaches to treat or repair the injured spinal cord in hope to, at the very least, improve upon the patient’s quality of life. Thankfully, several discoveries at the preclinical level are now transitioning into the clinical arena. These include the Surgical Treatment for Acute Spinal Cord Injury Study (STASCIS) Trial to evaluate the role and timing of surgical decompression for acute SCI, neuroprotection with the semisynthetic second generation tetracycline derivative, minocycline; aiding axonal conduction with the potassium channel blockers, neuroregenerative/neuroprotective approaches with the Rho antagonist, Cethrins; the use of anti-NOGO monoclonal antibodies to augment plasticity and regeneration; as well as cell-mediated repair with stem cells, bone marrow stromal cells, and olfactory ensheathing cells. This review overviews the pathobiology of SCI and current treatment choices before focusing the rest of the discussion on the variety of promising neuroprotective and cell-based approaches that have recently moved, or are very close, to clinical testing. Keywords: acute spinal cord injury; pathobiology; pharmacologic therapy; neuroprotection; rehabilitation; cell-mediated repair Introduction

modest ﬁnancial incentive for the private sector to make and market new medications, SCI can potentially be considered an ‘‘orphan disease.’’ This is troubling since the epidemiological predominance for this debilitating neurological disorder tends to affect individuals aged in their 3rd decade of life; a time typically associated with one’s prime earning potential (Sekhon and Fehlings, 2001). Coupled with the necessity to provide lifelong health care support in the form of therapy and rehabilitation, a signiﬁcant economic strain is therefore placed on society by SCI. Moreover, therapies for SCI have a high likelihood of being translatable for other neurological indications such as stroke and

Acute or traumatic spinal cord injury (SCI) occurs with an annual incidence of 11.5–53.4 cases per million within developed nations, with the causes of these injuries ranging from motor vehicle accidents and community violence to recreational activities and workplace-related injuries (Sekhon and Fehlings, 2001). SCI affects fewer than 200,000 people in the United States and due to the perceived Corresponding author. Tel.: +1 416 603 5229 (lab); +1 416

603 5800 ext. 2973 (ofﬁce); Fax: +1 416 603 5745; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61015-7

217

218

traumatic brain injury. Thus, we would argue that SCI presents excellent opportunities for commercialization of therapies, which target central nervous system disorders characterized by cell death or loss of key neural circuits or pathways. Since the published results from the landmark National Acute Spinal Cord Injury Study (NASCIS) II trial during the early 90 s, steroid use with methylprednisolone has become widely used in the treatment of SCI (Bracken et al., 1990, 1992). However, with methylprednisolone only yielding modest improvements in neurological recovery coupled with the potential risks of impaired wound healing and sepsis the utility of methylprednisolone has been seriously questioned (Coleman et al., 2000; Hurlbert, 2000). Moreover, ﬁndings from the Maryland monosialotetrahexosylganglioside sodium salt (GM-1) clinical trial suggested that GM-1 had the ability to enhance recovery of lower extremities, which had provided strong rationale to further investigate the efﬁcacy of low and high dose GM-1 following standard treatment with methylprednisolone in the Sygen multicenter study (Geisler et al., 1991). However, despite GM-1 demonstrating improved rates of recovery, along with improved bladder/bowel function, sacral sensation, and anal contraction over the ﬁrst 3 months post-injury the prospectively planned analysis at 6 months was negative (Geisler et al., 2001). Hence, with no universally accepted standard therapy in place, novel therapeutic approaches for SCI are urgently required. Surely, such a ﬁnding can only be made through further understanding of the etiological events following traumatic SCI. As such, we owe much of our current state of knowledge regarding the etiology of SCI to those involved in basic scientiﬁc research utilizing clinically relevant animal models of SCI, often involving extradural spinal cord contusion or compression in rodents (Rivlin and Tator, 1978; Gruner, 1992; Stokes et al., 1992; Joshi and Fehlings, 2002a, b).

Pathobiological mechanisms of SCI The pathobiology of SCI follows a biphasic process, involving a ‘‘primary’’ traumatic insult to the spinal cord that initiates a disruption in spinal

cord blood ﬂow dynamics and local ischemiareperfusion. This initial trauma is then followed by ‘‘secondary’’ biochemical events that set the stage for membrane dissolution and cellular death. Some notable ‘‘secondary’’ etiological events include disturbances to ionic homeostasis (i.e., irregular regulation of cellular calcium, sodium, and potassium trafﬁcking), glutamate-mediated excitotoxicity, mitochondrial instability, reactive oxygen species generation, elevated activation of proteolytic enzymes such as caspases and calpains, and inﬂammation. Cell death following SCI likely occurs as a continuum of necrosis and apoptosis. While necrosis predominates immediately following the primary traumatic episode, researchers have observed delayed neuronal and oligodendroglial apoptosis, albeit the former occurs to a lesser degree, relative to the latter (Casha et al., 2001). Applied neuroprotective approaches have been tested in SCI models with the intention that drug-mediated protective effects will lead to preservation of neuronal and glial cell populations via targeting one or more of the aforementioned secondary injury events. However, to date, few pharmacological approaches have been successful in translating therapeutic value in patients.

Current non-pharmacological treatment options for SCI Surgery In SCI, primary mechanical forces together with secondary injurious mechanisms collectively contribute to nervous tissue damage. Experimental evidence exists to support surgical decompression following persistent compression of the spinal cord following SCI (Delamarter et al., 1995). However, despite its widespread use in patients with SCI, the role of surgery in improving neurological recovery remains controversial because of the absence of randomized, controlled clinical trials. Furthermore, the presence and duration of a therapeutic window, during which surgical decompression could mitigate the secondary mechanisms of SCI have yet to be clearly deﬁned.

219

The timing of surgical intervention in the treatment of experimental and clinical SCI was recently reviewed (Fehlings and Perrin, 2006). The authors conducted a MEDLINE search of the literature from 1966 to 2005, dealing with the role of decompression in SCI using the medical subject heading of ‘‘decompression’’ and ‘‘spinal cord injury.’’ The overall conclusions from this extensive search revealed that decompressive surgery performed within the ﬁrst 24 h of injury could be conducted safely. It is the senior authors’ recommendation that urgent decompression be performed within the ﬁrst 24 h of injury following an isolated cervical SCI only if hemodynamic stability can be maintained. The Spinal Cord Injury Committee of the Joint Section on Neurotrauma and Critical Care of the American Association of Neurological Surgeons (AANS) and the Congress of Neurological Surgeons (CNS) formed the Surgical Treatment for Acute Spinal Cord Injury Study (STASCIS) group in 1992. By 1996, STASCIS was joined by the Joint Section on Spinal Disorders and Peripheral Nerves of the AANS/CNS. The principal aim of the STASCIS group was to conduct a randomized prospective controlled trial to examine the role and timing of decompressive surgery performed on the spinal cord and cauda equina of patients after acute spinal injury. A multicenter, retrospective study was performed in 36 North American centers to examine the use and timing of surgery in patients who have sustained SCI (Tator et al., 1999). The study was performed to obtain information required for the planning of a randomized controlled trial in which early and late decompressive surgery are compared. A total of 585 patients aged 16–75 years with acute SCI or cauda equina injury were admitted to participating centers, although approximately half of the patients were excluded due to late admission, age, gunshot wound, or absence of signs of compression on imaging. The timing of surgery varied widely, where 23.5, 15.8, 19, and 41.7% of the patients received surgery within 24, 25–48, 48–96 h, and 5 days post-injury, respectively. The ﬁndings from this study have not led to unequivocal support for varied timing of surgical management in SCI patients. Thus, the beneﬁt of

surgical management following SCI will only be demonstrated clearly with a carefully designed prospective controlled trial to assess the optimum timing of decompressive surgery in SCI. These results spurred the STASCIS to embark on a prospective, multicenter study in North America to determine the effectiveness of early surgical decompression and stabilization procedures for reducing the possibility of further damage to the spinal cord following acute compression of the cervical spinal cord. This study is currently underway and is being converted into a collaborative trial under the auspice of the Spine Trauma Study Group (STSG).

Rehabilitation Gait rehabilitation is a speciﬁc component of physical rehabilitation for persons with subacute or chronic motor-incomplete SCI. One novel method of gait rehabilitation involves the use of an overhead support point and a harness (i.e., body weight supported or BWS). The BWS strategy has been combined with treadmill-based gait training in recent studies with dramatic results (Hicks et al., 2005). This combination of BWS and treadmill rationally evolved from the idea that man possesses a central pattern generator, a circuitry located within the spinal cord capable of driving involuntary locomotor movements, similar to lower mammals (Duysens and Van de Crommert, 1998; Van de Crommert et al., 1998). Indeed, an electrical train of stimuli applied over the second lumbar segment with a frequency of 25–60 Hz and an amplitude of 5–9 V was shown to be effective in inducing rhythmic, alternating stance and swing phases of the lower limbs in paraplegic subjects, providing proof of a central pattern generator in humans (Dimitrijevic et al., 1998). Thus, it is believed that the combination of BWS and treadmill training will enhance the output by the intrinsic central pattern generator. The effects of long-term BWS treadmill training on functional walking ability and perceived quality of life were recently assessed in 14 persons with chronic incomplete SCIs (Hicks et al., 2005). Study subjects, classiﬁed on entrance as being ASIA

220

(American Spinal Injury Association) grade B or C, participated in three weekly sessions of BWS treadmill training over the course of 1 year and were assessed for improvements in functional ability and quality of life throughout the year and up to 8 months thereafter. The effect of this regimen showed that all subjects required less body weight support, and were able to increase both walking speed and distance per session, as well as improved over-ground walking ability. Heightened physical ability was correlated with an enhanced perspective of life. By the 8-month follow-up subjects demonstrated diminished performance in treadmill sessions with stable over-ground performance. Their quality of life reﬂected slight disappointment in their decreased physical function. Thus, overall the BWS treadmill training appeared to be a useful form of rehabilitation. However, the efﬁcacy of BWS treadmill training may not confer additional beneﬁt over other forms of intensive rehabilitation therapies. Recently, the efﬁcacy of step training with BWS treadmill training was compared with over-ground practice to the efﬁcacy of a deﬁned over-ground mobility therapy in 146 patients with incomplete SCI admitted for inpatient rehabilitation in six regional centers (Dobkin et al., 2006a). Following receipt of 12 weeks of BWS treadmill training or controlled over-ground mobility training, primary outcome variables of functional independence measures for locomotion or walking speed and distance for ASIA grade A and B SCI patients, or ASIA grade C and D SCI patients, respectively, obtained at the 6 month endpoint, did not support the expectation that BWS treadmill training would be more efﬁcacious than intensive over-ground mobility therapy.

Improving axonal conduction in the injured spinal cord Fampridines Physical trauma to the spinal cord results in demyelination of preserved spinal tracts. Without insulating sheaths of myelin, these surviving axons become less efﬁcient in their ability to transmit electrical impulses to be conducted. As a result, the

demyelinated axon cannot transmit motor or sensory impulses. When an axon is demyelinated after injury, large numbers of potassium channels are exposed to the extracellular space and permit potassium ion channels to open and release potassium ions. Thus, the result of demyelinated axons and potassium channel exposure is a reduced safety factor of action potential propagation across the demyelinated region of the axon (Nashmi and Fehlings, 2001). As such, pharmacological antagonism of exposed potassium channels on demyelinated axons with Fampridines (4-aminopyridine or 4-AP) had been hypothesized to restore the ability of axons to transmit electrical impulses. In a randomized, double-blind, crossover designed trial, where each patient received 4-AP and a vehicle placebo on different occasions by infusion over 2 h, separated by 2 weeks, 4-AP treatment demonstrated a signiﬁcant temporary neurologic improvement in ﬁve of six patients with incomplete SCI. No effect was detected in two cases of complete paraplegia and one of two severe incomplete cases (Hansebout et al., 1993). More recently, in a prospective, randomized, double-blind, placebo-controlled, crossover trial 15 chronic ambulatory SCI patients were randomized to receive either an initial 2 weeks of 40 mg/day, oral 4-AP medication of placebo or immediate-release, 4-AP before being crossed over to an alternate medication for the next 2 weeks of the study (DeForge et al., 2004). Evaluations for the study were conducted at baseline before initiating treatment, 2 weeks and 4 weeks posttreatment through measurement of dynamometer lower limb isometric muscle force and biomechanical gait assessments. Patients were also required to give their subjective impressions of their received therapy through a survey on exit of the study. The results of the study demonstrated that despite positive feelings towards the treatment no statistical differences were noted between placebo and 4-AP therapies. Information regarding the current status of Fampridines clinical testing for SCI is not yet published, but can be found at the Acorda Therapeutics web site. In 2004, Acorda Therapeutics completed two Phase III clinical trials with their proprietary

221

Fampridines (Sustained Release) formulation. Neither of these studies demonstrated statistical signiﬁcance in their primary endpoints, reduction of spasticity, as measured by the Ashworth score, and improvement of participant’s Subject Global Impression rating. However, in one of the studies, the data showed a positive trend toward improvement on the Ashworth score (http://www.acorda.com/ pipeline_fampridine_sci1.asp). Potassium and sodium channel blockade While the preclinical ﬁndings strongly support the use of potassium channel blockade with 4-aminopyridine to improve axonal conduction following trauma-induced demyelination, the clinical utility of 4-aminopyridine is limited due to its associated adverse effects, which include restlessness, confusion, and infrequent generalized tonic-clonic seizure (Bever, 1994). In an effort to minimize these unwanted side effects Sanoﬁ-Aventis has designed HP184 (N-[Npropyl]-N-[3-ﬂuoro-4-pyridinyl]-1H-3-methylindole1-amine hydrochloride), a pharmacological blocker capable of antagonizing both potassium and sodium channels. Sanoﬁ-Aventis is currently in the process of conducting a Phase II, double-blind, placebo-controlled, multicenter study to assess the efﬁcacy and safety of HP184 at 100, 200, and 400 mg doses administered orally once daily for 24 weeks in 18–65 year old patients with chronic SCIs classed ASIA grade C or D. Sanoﬁ-Aventis will also assess the effect of HP184 treatment on recovery of walking function and spasticity. Neuroprotective/neuroregenerative approaches to treating the injured spinal cord Methylprednisolone sodium succinate Preclinical research originally identiﬁed methylprednisolone as a potent inhibitor of lipid peroxidation following traumatic injury (Hall and Braughler, 1981, 1982). These ﬁndings led to the initiation of NASCIS I, which compared the effectiveness of treating SCI patients with an administered intravenous (i.v.) standard or high dose

methylprednisolone sodium succinate (MPSS) consisting of 100 or 1000 mg bolus followed by 25 or 250 mg MPSS daily thereafter for 10 days, respectively (Bracken et al., 1984). However the 6-month follow-up data for NASCIS I revealed no difference in neurological recovery of motor function or pinprick and light touch sensation between the two treatment arms of the study. Moreover, increased fatality, wound infections of both trauma and operative sites were associated to a greater extent with high-dose MPSS (Bracken et al., 1984). Further animal testing later generated data to suggest that the NASCIS I MPSS dose used was not sufﬁcient to elicit notable therapeutic efﬁcacy related to improved protection from ischemic insults and calcium-dependent degradation of neuroﬁlament cytoskeletal proteins (Braughler and Hall, 1984). Thus, the second NASCIS trial was designed to evaluate the efﬁcacy of high-dose MPSS administered as a 30 mg/kg bolus over the ﬁrst hour followed by an infusion of 5.4 mg/kg/h over the next 23 h to placebo, and to a group of patients receiving the opioid receptor antagonist naloxone (Bracken et al., 1990). The trial was further designed to investigate the effects of differing times for drug administration following SCI. An analysis of all patients who entered into the trial failed to demonstrate a signiﬁcant difference among the treatments. However, upon stratiﬁcation of the data on the basis of time to loading dose (less than 8 h vs. greater than 8 h from SCI) and adjustment for the severity of SCI (complete vs. incomplete), analysis of the results from this second study at the 6-week and 6-month follow-up revealed that patients treated with MPSS within the ﬁrst 8 h of injury had signiﬁcantly improved motor and sensory function in comparison to patients receiving placebo, naloxone, or MPSS at later times. Furthermore, these differences remained signiﬁcant 1 year following SCI (Bracken et al., 1992). None of the differences in patients taking naloxone or in patients treated with MPSS more than 8 h after SCI were statistically signiﬁcant. Despite the beneﬁcial therapeutic effects demonstrated with MPSS treatment, the results from the NASCIS II trial have not been universally

222

accepted (Casha et al., 2001; Hurlbert, 2001; AANS/CNS, 2002). Concerns surrounding the small sample population size for the groups showing beneﬁcial effects, performance of medical and surgical protocols by participating centers in a nonstandardized manner, and the fact that there were no functional outcome measures deﬁned to assess whether the improvements noted with MPSS treatments were correlated with clinical signiﬁcance all reduced enthusiasm for the indication of MPSS for acute SCI. In an attempt to resolve some of the issues raised from the second NASCIS trial, a third NASCIS trial was organized to include a functional independence measure (Bracken et al., 1997). A total of 499 patients suffering a SCI within 8 h were enrolled into the study and received 30 mg/kg bolus of MPSS before randomization. Patients in the 24 h regimen group received a MPSS infusion of 5.4 mg/kg/h over 24 h, those in the 48 h regimen group received a MPSS infusion of 5.4 mg/kg/h for 48 h and those in the tirilazad group received a 2.5 mg/kg bolus infusion of tirilazad mesylate (TM; a ‘‘lazaroid’’ with inhibitory effects on lipid peroxidation without glucocorticoid side effects) every 6 h for 48 h. The results from this 3rd study concluded that no beneﬁt was associated with extending MPSS treatment beyond 24 h if MPSS had been administered within the ﬁrst 3 h of SCI, while those that started MPSS therapy between 3 and 8 h demonstrated improvements in motor capabilities if drug infusion was continued for 48 h in comparison to 24 h infusion. It should be noted that the guidelines committee of the AANS/CNS Joint Section on Disorders of the Spine and Peripheral Nerves on reviewing the evidence regarding the use of MPSS in the treatment of acute SCI in adults concluded that its use could only be supported at the level of a treatment option (AANS/CNS, 2002). However, it is the authors’ view that the intense criticism directed at the NASCIS II and III trials must be balanced by the current lack of alternative neuroprotective strategies for acute SCI. Moreover, the modest therapeutic beneﬁt demonstrated by MPSS may potentially prove to impart a major beneﬁt on cervical SCI patient’s functional independence and quality of life (Sekhon and Fehlings, 2001). Still,

despite these concerns the NASCIS II and III trials effectively demonstrated the validity for targeting secondary injurious pathological events in SCI, while also emphasizing the importance of timing with applied interventions.

Monosialotetrahexosylganglioside Monosialotetrahexosylganglioside GM-1 sodium salt is a naturally occurring compound, which resides within cell membranes of the mammalian CNS (Geisler et al., 2001). Several preclinical studies reporting neuroprotective and neuroregenerative actions by GM-1 in experimental models of ischemia and injury have been reported (Agnati et al., 1983; Fass and Ramirez, 1984; Bose et al., 1986). These ﬁndings have led to the initiation of the Maryland GM-1 Ganglioside Study, a small randomized, placebo-controlled, double-blind trial designed to investigate the efﬁcacy of GM-1 in patients with cervical and thoracic SCIs (Geisler et al., 1991). In this trial, 34 patients completed the test-drug protocol that consisted of either 100 mg GM-1 or placebo i.v. per day for 18–32 doses, with the ﬁrst dose beginning within 72 h of the onset of SCI. The results from this study demonstrated the effectiveness for GM-1 to improve the recovery of neurologic function after 1 year and provided the necessary rationale to support the recruitment of SCI patients for a larger GM-1 study. The randomized double-blind Sygens Multicenter Acute Spinal Cord Injury Study was completed in the late 1990s and reported in 2001. In this trial, all 797 patients were randomized to receive a standard 30 mg/kg bolus of methylprednisolone followed by 5.4 mg/kg/h infusion of methylprednisolone for 23 h within the ﬁrst 8 h of injury. Placebo, low-dose GM-1 (300 mg loading dose followed by 100 mg/day for 56 days) or highdose GM-1 (600 mg loading dose followed by 200 mg/day for 56 days) was then randomly assigned following the completion of corticosteroid administration to avoid unwanted drug interactions between methylprednisolone and GM-1. The ﬁndings from this larger trial demonstrated that patients receiving either low- or high-dose GM-1 had enhanced recovery of neurologic

223

function by 3 months post-injury regardless of their baseline severity, and showed a trend for improved bowel/bladder function, sacral sensation, and anal contraction (Geisler et al., 2001). However, the primary outcome variable, the percentage of marked recovery, at the designated 6 month time-point was negative. While no future clinical trials with Sygen are currently planned, the sponsoring company Fidia is evaluating this option (Corporate Communications Fidia Farmaceutici SpA, Personal communications).

biochemical monitoring in the acute setting of SCI; to monitor the safety and tolerance of i.v. minocycline administration; to assess systemic and cerebral spinal ﬂuid (CSF) absorption characteristics of i.v. minocycline administration; and to assess serial CSF interleukin 1 beta (IL-1b) and nitric oxide concentration, which correspond to the activities of caspase-1 and -3, respectively, along with tumor necrosis factor alpha (TNF-a) levels to reﬂect the levels of a major pro-inﬂammatory cytokine.

Minocycline

Cethrins

The broad-spectrum antibiotic, minocycline, is a lipophilic derivative of tetracycline that has demonstrated ability to provide neuroprotection via intraperitoneal (Wells et al., 2003; Teng et al., 2004) or i.v. (Xu et al., 2004) routes of administration. Mechanisms attributed to the protective actions elicited by minocycline include ability to overcome glutamate-mediated excitotoxicity (Tikka and Koistinaho, 2001; Baptiste et al., 2004), antiinﬂammatory effects through blocking the activation of microglial cells (Dommergues et al., 2003; Baptiste et al., 2005), inhibiting both cytochrome c release and caspase-dependent apoptotic neuronal death (Zhu et al., 2002), as well as a capacity to antagonize the activities of matrix metalloproteinases-2 (Cho et al., 2006) and -9 (Brundula et al., 2002). Additionally, minocycline was shown to reduce oligodendrocyte apoptosis, corticospinal tract dieback, and lesion size, while improving functional outcome following a C7–C8 dorsal column transection SCI in rats (Stirling et al., 2004). Considering, these ﬁndings together with the relatively safe clinical track record of minocycline has made this pharmacological approach a very promising candidate for the SCI. A Phase I/II pilot study to investigate the efﬁcacy of intravenously administered minocycline in patients with acute SCI is currently underway in Calgary, Alberta. The study has aimed to enroll patients above 16 years of age possessing a SCI between C1 and T11. The major objectives of this trial will be to assess the feasibility of both minocycline administration and clinical and

CNS myelin inhibits axon growth due to the expression of several extracellular growth-inhibitory proteins, including myelin-associated glycoprotein (MAG), myelin oligodendrocyte glycoprotein (MOG), and neurite outgrowth (Nogo). It is now evident that the growth-inhibitory effects of these proteins are exerted through the regulatory actions of small intracellular GTPase-associated signaling proteins, Rho and Rac (Niederost et al., 2002; Winton et al., 2002). Rho and Rac can exist in one of two possible forms — an active GTP-bound state, and an inactive GTP-bound state. In the active conformation, these small GTPase-associated proteins work to inhibit neurite outgrowth. For example, puriﬁed cerebellar granule cells grown on Nogo-A, Nogo-66, or MAG containing substrates have all been shown to result in inhibited neurite outgrowth (Niederost et al., 2002). Furthermore, increased levels of Rho-GTP were detected in pheochromocytoma cell line cultures that had been plated on myelin, MAG, or poly-L-lysine containing substrates by measuring the precipitate from cell homogenates mixed with Rho binding domain isolated from rhotekin, a GTP-interacting protein (Dubreuil et al., 2003). Moreover, MAG, MOG, or Nogo-mediated neurite outgrowth inhibition can be relieved with the addition of Rho-Aassociated kinase ROCK inhibitor (Niederost et al., 2002). Preclinical studies now support the hypothesis that SCI is associated with an upregulation of GTP-bound Rho. Dubreuil et al. (2003) demonstrated that Rho-GTP levels are robustly elevated

224

following transection or contusion cord injuries in comparison to uninjured control animals. Rho activation following transection-induced injury began as early as 1.5 h post-SCI and was sustained for 7 days (Dubreuil et al., 2003). Addition of the Rho-selective antagonist, C3–05 transferase, injected in a ﬁbrin matrix to the lesion site following cord transection, or C3–05 alone into the contused cord demonstrated ability to reduce RhoA activation levels back to physiological conditions (Dubreuil et al., 2003). Also, treatment with C3–05 reduced neuronal and glial cell apoptosis that would pathologically occur in secondary injury processes of SCI (Dubreuil et al., 2003). Thus, this novel approach has demonstrated multifaceted therapeutic actions, namely neuroregenerative and neuroprotective properties. BioAxone Inc., a biopharmaceutical company based in Montreal, Quebec, recently sponsored a Phase I/IIa multinational clinical trial (lead site in Toronto, Ontario) designed to evaluate the safety and pharmacokinetics of rising dosages of Cethrin (BA-210), a recombinant fusion protein composed of C3 transferase and a transport sequence that aids the protein’s ability to breach cellular membranes together with a ﬁbrin sealant, following a single extradural administration during surgery for acute thoracic and cervical SCI. The trial was conducted in nine centers located in Canada and the USA. In December 2005, the U.S. Food & Drug Administration granted Orphan Drug Status to Cethrins, which will likely hasten clinical evaluation of BioAxone’s lead product through reducing clinical development costs and facilitating regulatory ﬁlings in other countries. It is anticipated that the results of this Phase I/IIa trial will become available in early 2007.

Nogo It has long been thought that the capacity for the CNS to repair lesions through regeneration was limited. However, hope in this context has been invigorated since the landmark discoveries by Richardson et al. (1980), which have provided the ﬁrst indication that CNS neurons possess an intrinsic ability to regenerate following injury that

can be encouraged further upon manipulation of the local tissue microenvironment. Years later, Caroni et al. (1988) reported that nonpermissive substrates within the CNS were actually associated with oligodendrocytes and myelin. Moreover, monoclonal antibody or pharmacological treatment directed against white matter nonpermissive substrates allowed neuron outgrowth (Caroni and Schwab, 1988; Savio and Schwab, 1989). By 2000, the gene nogo, which can be transcribed to yield Nogo-A, -B, and -C mRNA products, was cloned (Chen et al., 2000). Nogo-A is a high molecular weight transmembrane protein and potent inhibitor of neurite outgrowth produced by oligodendrocytes. Inhibition of Nogo-A with the selective monoclonal antibody IN-1 or with the peptide antagonist NEP1-40, in spinally injured rats demonstrates neutralization of myelindependent neurite outgrowth blockade, improved axonal regeneration, and enhanced rates of functional locomotor recovery (Chen et al., 2000; Li and Strittmatter, 2003). Novartis Institutes for BioMedical Research has since taken a focused interest in developing antagonists for Nogo-A and its corresponding receptor, Nogo-66, to promote neuroregeneration (Wiessner et al., 2003; Walmsley et al., 2004). Currently, Novartis is sponsoring a Phase I clinical trial to evaluate the safety of anti-Nogo-A monoclonal antibody therapy.

Autologous macrophages The CNS was once considered as an ‘‘immuneprivileged’’ environment that was neither susceptible to nor could evoke an inﬂammatory response (Lucas et al., 2006). Consequently, the hypothesis that inﬂammation may be harnessed to heal the injured spinal cord has not been received without skepticism. Although evidence exists to support a pathological role for inﬂammation in the etiology of many CNS disorders (Stoy et al., 2005; Man et al., 2007; Miklossy et al., 2006; Noorbakhsh et al., 2006), it is very likely that inﬂammatory mediators possess dual roles, with detrimental acute effects together with healing effects in longer-term repair and recovery (Lucas et al., 2006).

225

Preclinical investigations involving rats that had received a single complete spinal cord transection together with local implantation of ex vivoactivated homologous macrophages incubated with homologous peripheral nerves demonstrated that stimulation of the host immune response can promote partial recovery of motor function, as demonstrated through hind limb movement, plantar paw placement, and weight-bearing steps, along with electrophysiological assessments via cortically evoked hind limb muscle response (Rapalino et al., 1998). Additionally, macrophages incubated with autologous skin were found to be just as efﬁcacious (Rapalino et al., 1998). The mechanism of action associated with autologous ex vivo-activated macrophages is believed to be both neuroprotective and neuroregenerative. Neuroprotective properties likely arise due to the ability of activated macrophages to synthesize and secrete concurrently protective cytokines IL-1b and IL-6, together with the trophic factor, brain derived neurotrophic factor, and the chemokine, IL-8, while reducing levels of the neurotoxic cytokine, TNF-a. In addition, regenerative properties may result due to macrophage-mediated phagocytosis of myelin debris, an event that would effectively make the environment of the injured spinal cord less cytotoxic and more responsive to axonal regeneration (Bomstein et al., 2003). From 2000 to 2003, Proneuron Biotechnologies conducted a Phase I, open-label nonrandomized safety study in patients diagnosed with a complete SCI (ASIA grade A) within C5–T11 within 14 days of injury at centers located in Israel and Belgium (Knoller et al., 2005). ASIA grade A patients were the only SCI population admitted into the ProCords study, in an effort to limit the confounding interpretation of spontaneous recovery that would result from incomplete injuries (ASIA grades B, C, or D). Proneuron Biotechnologies’ proprietary procedure for obtaining activated autologous macrophages consists of isolating white lymphocytes from the patient’s blood and incubating them ex vivo with autologous dermis. A single dose of 4 million activated autologous macrophages was administered by four microinjections into the spinal cord parenchyma at the caudal border of the spinal

cord lesion. The results from this study at 1-year follow-up demonstrated that the therapy was safe and warranted further investigation in a randomized Phase II, multicenter clinical trial within the USA and Israel. However, despite the FDA granting an orphan drug status to ProCord on September 13, 2004, recruitment into this trial is currently suspended due to presumed ﬁnancial challenges related to the conduct of such a large complex trial. Clinical follow-up is currently being conducted for those patients that have already been admitted.

Cell-mediated repair of the injured spinal cord Neural stem and progenitor cells A variety of cell types have been investigated for their potential use in cell-mediated tissue repair following SCI. Theses include neural stem cells (NSCs), neural progenitor cells (NPCs), Schwann cells, oligodendroglial precursor cells (OPCs), bone marrow stem cells (BMSCs), and olfactory ensheathing cells (OECs). NSCs and NPCs have functional differences, where the former represents an uncommitted cell type that can divide repeatedly while maintaining potency to generate differentiated cell types (neurons and glia), whereas the latter represents a kind of cell that exhibits a limited number of cell divisions before it differentiates into a speciﬁc cell type (Pollard et al., 2006). Despite the fact that stem/ precursor cells may be derived from a variety of sources, research with these cells raises both political and ethical concerns. As a result, stem cell technology is trapped within an ethical dilemma of balancing heightened emotions regarding embryo/ fetal research with the potentially therapeutic beneﬁts to patients with desperate hopes. NSCs have been isolated from the adult human hippocampus and lateral ventricle wall (Johansson et al., 1999b). Additionally, postmortem analysis of tissue derived from the hippocampus and subventricular zone of the caudate nucleus revealed that NPCs reside in the adult human brain (Eriksson et al., 1998). Thus, these ﬁndings show that neurogenesis occurs throughout life and

226

supports the hypothesis that the mature CNS possesses an innate ability to repair damaged tissue. In light of these ﬁndings researchers have attempted to demonstrate the utility harnessing the capabilities of endogenous stem/progenitor cells following traumatic injuries (Johansson et al., 1999a; Kernie et al., 2001). However, the results of these studies demonstrated that the endogenous stem/progenitor supply tend to move toward an astrocytic lineage (Johansson et al., 1999a; Kernie et al., 2001). Studies involving transplantation of NSCs/ NPCs have shown greater promise for traumatic injuries than attempts to stimulate or control endogenous neurogenesis. For instance Cummings et al. (2005) demonstrated that human CNS stem cells grown as neurospheres survive, migrate, and express differentiation markers for neurons and oligodendrocytes after long-term engraftment in the injured spinal cord. Furthermore, the stem cell engraftment was associated with improved locomotor recovery with diminished astrocytic differentiation, remyelination in both the spinally injured and myelin-deﬁcient mice, along with electron microscope evidence consistent with synapse formation between stem cells and host neurons (Cummings et al., 2005). However, despite signiﬁcant advances in this, adult NSCs remain an elusive cell for study, and researchers are facing many challenges to the development of therapeutic applications from adult NSC research. Among these challenges are the identiﬁcation and characterization of NSCs, the understanding of the physiology of newly generated neuronal cells in the adult brain, and the isolation and culture of homogenous populations of NSCs/NPCs from the adult CNS for cell-based therapy. Furthermore, measures must be taken to improve on the survival of transplanted stem/progenitor cells in the hostile environment of the injured spinal cord (KarimiAbdolrezaee et al., 2006).

Schwann cells Schwann cells are a type of glial cell responsible for forming myelin sheaths around peripheral nerves. Researchers are excited about the use of Schwann cells in neural repair because these cells

may act as a physical bridge, allowing the normal functioning of undamaged or regenerated axons. Moreover, these cells are capable of secreting numerous trophic factors including brain derived neurotrophic factor, glial derived neurotrophic factor, nerve growth factor, and ciliary neurotrophic factor (Pellitteri et al., 2006), while also being able to remyelinate (Kohama et al., 2001) and guide regenerating CNS axons to their intended tracts (Zheng and Kufﬂer, 2000). The transplant procedure would involve removing a small amount of the patient’s own peripheral nerve tissue, isolating Schwann cells from the tissue, growing them in plastic culture dishes in an incubator, and then implanting the cultured Schwann cells into the site of SCI. A recent report has demonstrated the efﬁcacy of subarachnoid space transplantation of Schwann cells into the contused epicenter in adult rats 7 days post-SCI resulting in signiﬁcantly enhanced locomotor recovery and signiﬁcantly greater axonal density at the lesion epicenter (Firouzi et al., 2006). Furthermore, this route for transplantation was without notable reduction in tissue integrity or augmented levels of pro-inﬂammatory and proapoptotic mediators (Firouzi et al., 2006). However, in light of these advances with Schwann cells, researchers have still not found a solution to permit regenerating axons to extend beyond the transplant due to the inhibitory nature of the surrounding glial scar (Oudega and Xu, 2006). This problem likely calls for a combinatorial approach, which makes the local microenvironment more receptive for regeneration.

Bone marrow stromal cells Bone marrow stromal cells (BMSCs) provide yet another source for obtaining pluripotent stem cells. Bone marrow contains a population of pluripotent cells that have the capacity to migrate towards a lesion and release various growth factors to facilitate regeneration and lesion repair. Under physiological settings BMSCs function to regulate hematopoiesis; however, if these cells are grown away from their natural environment these cells can be readily and effectively propagated and

227

manipulated genetically to yield a variety of cell types including phenotypes that closely resemble neurons (Clark and Keating, 1995). The advantage of using bone marrow as a source for stem cells is that they are relatively easy to isolate, the cells grow well in tissue culture, BMSCs may be used in autologous transplantation protocols, and these BMSCs have already received approval for hematopoietic diseases (Sykova´ et al., 2006). Hofstetter et al. (2002) differentiated BMSCs in vitro into neuronal-like cells and investigated their fate following immediate or delayed injection into the contused spinal cord of adult rats. Their ﬁndings demonstrated that BMSCs administered 1week post-SCI had better rates of survival than BMSCs administered at the time of injury. Furthermore, the delayed treatment was associated with signiﬁcantly improved locomotor function. Despite their success in improving neurological recovery rates, electrophysiological and immunohistochemical analysis of these differentiated BMSCs demonstrated that the transplanted cells were likely rather immature, which lacked functional sodium and potassium channels. Still the effectiveness of this approach administered within a clinically relevant therapeutic window of opportunity makes BMSCs a promising strategy for SCI. To date, two clinical trials have been preformed in patients with SCI. In South Korea, Park et al. (2005) evaluated the therapeutic efﬁcacy of combining autologous BMSC transplantation, administered directly into the spinal cord lesion site, with granulocyte macrophage-colony stimulating factor (GM-CSF), given subcutaneously, in six patients with complete SCIs. At the 6- and 18-month follow-up periods, four of the six patients showed neurological improvements by two ASIA grade (from ASIA A to ASIA C), while another improved from ASIA A to ASIA B. Moreover, BMSC transplantation together with GM-CSF was not associated with increased morbidity or mortality. The second clinical trial, which was conducted in Prague, Czech Republic, is a Phase I study that has enrolled 20 patients with transversal SCIs to test the safety of autologous bone marrow cell implantation (Sykova´ et al., 2006). The investigators compared intra-arterial versus i.v. routes of

administration and a group of acute SCI patients (10–30 days post-SCI) against patients sustaining chronic SCIs. Motor evoked potential, somatosensory evoked potential, magnetic resonance imaging, and ASIA scores were used in patient follow-up. The results of this study demonstrated that the BMSC-mediated repair is associated with modest improvements in acute patients and that BMSC transplantation is a relatively safe procedure. Thus, it is anticipated that a Phase II clinical trial designed to test the efﬁcacy will be initiated in the near future.

Oligodendrocyte precursor cells Since the neurological disability in SCI is associated with oligodendrocyte death, resulting in demyelination and axonal degeneration, one would anticipate that OPC-mediated repair would produce signiﬁcant improvements in neural recovery. In preclinical research OPC transplant alone or in combination with the glycoprotein molecule Sonic hedgehog (Shh), to further drive OPCs toward an oligodendroglial lineage, was administered to adult rats 5 days following a moderate thoracic SCI (Bambakidis and Miller, 2004). The results of this treatment regimen demonstrated that transplantation of OPCs improves axonal conduction and spinal cord function in the injured spinal cord. Furthermore, the therapeutic beneﬁts were more pronounced when OPC treatment was given in combination with Shh, while treatment of injured rats with Shh on its own resulted in the proliferation of an endogenous population of neural precursor cells (Bambakidis and Miller, 2004). The most frequently investigated OPC is O-2A, which can differentiate into oligodendroglial cells and type-2 astrocytes in vitro, but only into oligodendrocytes following in vivo transplantation into brain or spinal cord (Cao et al., 2002). O-2A cell transplant demonstrates ample capacity to migrate to regions adjacent to the site of injury, differentiate into oligodendrocytes, remyelinate axons, and improve on functional, behavioral, electrophysiological, and morphological measurable outcome variables (Lee et al., 2005). Still, a major shortcoming associated with

228

O-2A-mediated repair is their relatively slow proliferation rates, which effectively limits their ability to self-renewal.

Olfactory ensheathing cells OECs are specialized glial cells that surround the olfactory sensory axons of the ﬁrst cranial nerve within the nose, which possess neuroregenerative properties akin to Schwann cells of the peripheral nervous system (Doucette, 1995). OECs are suitable candidates for cell-mediated repair following SCI because this cell type can support axonal outgrowth and regrowth through the adult CNS; they are able to remyelinate the axons that have undergone primary demyelination; and they can migrate within the CNS and coexisting in an astrocyte-rich environment and therefore fully integrating within a lesion (Barnett, 2004). This latter characteristic of OECs to integrate with astrocytes sets them apart from Schwann cells. There are now several reports documenting the robust therapeutic efﬁcacy of OEC transplantation in animal models of SCI. For instance, Li et al. (1998) have demonstrated that cultured OECs injected into focal lesions within the adult rat cortical spinal tract begin to regenerate within the ﬁrst week following transplantation. Moreover, the cut corticospinal axons extended caudally and ensheathed by P0-positive peripheral myelin had accumulated into parallel bundles, which appropriately extended across the full length of the lesioned area and reentered the caudal part of the host corticospinal tract (Li et al., 1998). Ramon-Cueto et al. (2000) showed that OEC transplantation can provide improved rates of functional and structural recovery following a complete spinal cord transection from 3 to 7 months post-transplant and corticospinal, raphespinal, and coeruleospinal tracts all regenerated for long distances within caudal cord stumps. With these and other ﬁndings promoting the utility for OECs to be used in neural repair following SCI clinical trials have arisen around the world. A Phase I clinical trial was initiated in 2001, in Brisbane, Australia, to investigate the safety of OEC transplantation into the injured thoracic

(T4-T10) spinal cord, 6 months to 3 years postinjury (Feron et al., 2005). Only eight patients were admitted into the trial (four patients to receive OEC transplantation and four sham-operated controls). All participants were closely monitored before the trial was started and for 3 years following the trial admittance. All patients underwent regular MRI examination, along with physical, psychiatric, neurological, and neurophysiological assessments by medical staff and physiotherapy and occupational therapy assessments by respective therapists. By far the largest of clinical trials testing the efﬁcacy of OEC transplants to treat SCI patients comes from Beijing, China, where 171 subjects have participated (Huang et al., 2003). Patients aged 2–64 years, had traumatic or acute compressive injuries due to epidural hematoma. Outcome variables included motor score, light touch, and pin prick scores according to the International Standards for Neurological and Functional Classiﬁcation of SCI by ASIA and the International Society of Paraplegia before and 2 and 8 weeks post-intervention. The conclusions made from this study were that OEC transplantation can partially preserve spinal functions regardless of patient age. However, the recent publication by Dobkin et al. (2006b) warns the SCI community of the procedure of OEC transplantation as performed by Dr. Huang as not meeting international standards of a Phase I safety or Phase II efﬁcacy trial. Considering the rather poorly deﬁned inclusion and exclusion criteria, coupled with the inconsistencies in OEC injection sites, a lack of observed functional beneﬁt, and perioperative morbidity caution must be taken when assessing these anecdotal claims.

Conclusion In addition to methylprednisolone, GM-1 ganglioside, and Fampridine a number of novel pharmacological and cell-based approaches appear to be on the horizon for the therapeutic management of SCI (Table 1). Considering the complexity and multifactorial nature of pathobiological mechanisms that contribute to SCI, it is very likely that combinatorial approaches will have the greatest opportunity to improve on the quality of life of

229 Table 1. Select pharmacological and cell-based therapies for spinal cord injury Approach

Proposed mechanism of action

Methylprednisolone

A glucocorticosteroid that has demonstrated ability to inhibit posttraumatic lipid peroxidation and inﬂammation in animal models of spinal cord injury (SCI), and has shown ability to improve neurological recovery in spinal cord injured humans

GM-1 ganglioside

A naturally occurring compound, residing within cell membranes of the mammalian CNS that has been shown to promote neuritic sprouting, and has demonstrated ability to oppose excitotoxicity and apoptosis

Minocycline

A second generation tetracycline derivative with demonstrated ability to inhibit excitotoxicity, oxidative stress, microglial cell activation, and caspase-dependent and caspase-independent-mediated pathways of neuronal death

Fampridines

A speciﬁc voltage-dependent potassium channel blocker shown to restore action potential conduction, and enhance synaptic transmission in damaged demyelinated nerve ﬁbers via blockade of exposed potassium channels within the internodal membrane of injured axons

HP 184

A use- and frequency-dependent antagonist of sodium channels, and voltage-dependent blocker of potassium channels designed to have less associated adverse effects than Fampridines

Cethrins

A Rho antagonist, administered directly into the injured spinal cord in combination with a ﬁbrin sealant for improved drug delivery, to allow reversal of abnormal Rho activation in neurons and glial cells, and promotion of neuroprotection and neuroregeneration following acute SCI

Anti-Nogo-A monoclonal antibodies

A monoclonal antibody targeted at Nogo-A to neutralize neurite outgrowth inhibition, and allow enhanced regeneration, compensatory sprouting, structural reorganization, plasticity, and functional recovery following SCI

Procords

Autologous macrophages, activated through co-incubation with regenerative autologous skin tissue, are implanted into the injured spinal cord to elevate the secretion of protective cytokines, interleukin-1 beta (IL-1b) and IL-6, the trophic factor, brain derived neurotrophic factor (BDNF), and the chemokine IL8, while reducing the secretion of the proinﬂammatory cytokine, tumor necrosis factor alpha (TNF-a); all of which helps to make the injured environment of the lesioned spinal cord more feasible for axonal regeneration, reduce cyst formation, and improve motor recovery

spinally injured individuals. Moreover, preclinical analysis view SCI in a larger view with a thorough examination of potentially beneﬁcial effects in addition to ambulation. For example, neuroprotective/regenerative approaches that yield modest effects in enhancing rates of locomotor recovery would be of bladder/bowel function and relieve chronic pain. With regard to cell-based approaches, in addition to political and religious conﬂicts, it is clear that careful planning and design is required to obtain meaningful data.

References AANS/CNS. (2002) Pharmacological therapy after acute cervical spinal cord injury. Neurosurgery, 50: S63–S72.

Agnati, L.F., Fuxe, K., Calza, L., Benfenati, F., Cavicchioli, L., Toffano, G. and Goldstein, M. (1983) Gangliosides increase the survival of lesioned nigral dopamine neurons and favour the recovery of dopaminergic synaptic function in striatum of rats by collateral sprouting. Acta Physiol. Scand., 119: 347–363. Bambakidis, N.C. and Miller, R.H. (2004) Transplantation of oligodendrocyte precursors and sonic hedgehog results in improved function and white matter sparing in the spinal cords of adult rats after contusion. Spine J., 4: 16–26. Baptiste, D.C., Hartwick, A.T., Jollimore, C.A., Baldridge, W.H., Seigel, G.M. and Kelly, M.E. (2004) An investigation of the neuroprotective effects of tetracycline derivatives in experimental models of retinal cell death. Mol. Pharmacol., 66: 1113–1122. Baptiste, D.C., Powell, K.J., Jollimore, C.A., Hamilton, C., LeVatte, T.L., Archibald, M.L., Chauhan, B.C., Robertson, G.S. and Kelly, M.E. (2005) Effects of minocycline and tetracycline on retinal ganglion cell survival after axotomy. Neuroscience, 134: 575–582.

230 Barnett, S.C. (2004) Olfactory ensheathing cells: unique glial cell types? J. Neurotrauma, 21: 375–382. Bever Jr., C.T. (1994) The current status of studies of aminopyridines in patients with multiple sclerosis. Ann. Neurol., 36(Suppl): S118–S121. Bomstein, Y., Marder, J.B., Vitner, K., Smirnov, I., Lisaey, G., Butovsky, O., Fulga, V. and Yoles, E. (2003) Features of skin-coincubated macrophages that promote recovery from spinal cord injury. J. Neuroimmunol., 142: 10–16. Bose, B., Osterholm, J.L. and Kalia, M. (1986) Gangliosideinduced regeneration and reestablishment of axonal continuity in spinal cord-transected rats. Neurosci. Lett., 63: 165–169. Bracken, M.B., Collins, W.F., Freeman, D.F., Shepard, M.J., Wagner, F.W., Silten, R.M., Hellenbrand, K.G., Ransohoff, J., Hunt, W.E., Perot Jr., P.L., et al. (1984) Efﬁcacy of methylprednisolone in acute spinal cord injury. JAMA, 251: 45–52. Bracken, M.B., Shepard, M.J., Collins Jr., W.F., Holford, T.R., Baskin, D.S., Eisenberg, H.M., Flamm, E., Leo-Summers, L., Maroon, J.C., Marshall, L.F., et al. (1992) Methylprednisolone or naloxone treatment after acute spinal cord injury: 1-year follow-up data. Results of the second national acute spinal cord injury study. J. Neurosurg., 76: 23–31. Bracken, M.B., Shepard, M.J., Collins, W.F., Holford, T.R., Young, W., Baskin, D.S., Eisenberg, H.M., Flamm, E., LeoSummers, L., Maroon, J., et al. (1990) A randomized, controlled trial of methylprednisolone or naloxone in the treatment of acute spinal-cord injury. Results of the second national acute spinal cord injury study. N. Engl. J. Med., 322: 1405–1411. Bracken, M.B., Shepard, M.J., Holford, T.R., Leo-Summers, L., Aldrich, E.F., Fazl, M., Fehlings, M., Herr, D.L., Hitchon, P.W., Marshall, L.F., Nockels, R.P., Pascale, V., Perot Jr., P.L., Piepmeier, J., Sonntag, V.K., Wagner, F., Wilberger, J.E., Winn, H.R. and Young, W. (1997) Administration of methylprednisolone for 24 or 48 hours or tirilazad mesylate for 48 hours in the treatment of acute spinal cord injury. Results of the third national acute spinal cord injury randomized controlled trial. National acute spinal cord injury study. JAMA, 277: 1597–1604. Braughler, J.M. and Hall, E.D. (1984) Effects of multi-dose methylprednisolone sodium succinate administration on injured cat spinal cord neuroﬁlament degradation and energy metabolism. J. Neurosurg., 61: 290–295. Brundula, V., Rewcastle, N.B., Metz, L.M., Bernard, C.C. and Yong, V.W. (2002) Targeting leukocyte MMPs and transmigration: minocycline as a potential therapy for multiple sclerosis. Brain, 125: 1297–1308. Cao, Q., Benton, R.L. and Whittemore, S.R. (2002) Stem cell repair of central nervous system injury. J. Neurosci. Res., 68: 501–510. Caroni, P., Savio, T. and Schwab, M.E. (1988) Central nervous system regeneration: oligodendrocytes and myelin as nonpermissive substrates for neurite growth. Prog. Brain Res., 78: 363–370. Caroni, P. and Schwab, M.E. (1988) Antibody against myelin-associated inhibitor of neurite growth neutralizes

nonpermissive substrate properties of CNS white matter. Neuron, 1: 85–96. Casha, S., Yu, W.R. and Fehlings, M.G. (2001) Oligodendroglial apoptosis occurs along degenerating axons and is associated with FAS and p75 expression following spinal cord injury in the rat. Neuroscience, 103: 203–218. Chen, M.S., Huber, A.B., van der Haar, M.E., Frank, M., Schnell, L., Spillmann, A.A., Christ, F. and Schwab, M.E. (2000) Nogo-A is a myelin-associated neurite outgrowth inhibitor and an antigen for monoclonal antibody IN-1. Nature, 403: 434–439. Cho, K.O., La, H.O., Cho, Y.J., Sung, K.W. and Kim, S.Y. (2006) Minocycline attenuates white matter damage in a rat model of chronic cerebral hypoperfusion. J. Neurosci. Res., 83: 285–291. Clark, B.R. and Keating, A. (1995) Biology of bone marrow stroma. Ann. N.Y. Acad. Sci., 770: 70–78. Coleman, W.P., Benzel, D., Cahill, D.W., Ducker, T., Geisler, F., Green, B., Gropper, M.R., Gofﬁn, J., Madsen III, P.W., Maiman, D.J., Ondra, S.L., Rosner, M., Sasso, R.C., Trost, G.R. and Zeidman, S. (2000) A critical appraisal of the reporting of the national acute spinal cord injury studies (II and III) of methylprednisolone in acute spinal cord injury. J. Spinal Disord., 13: 185–199. Cummings, B.J., Uchida, N., Tamaki, S.J., Salazar, D.L., Hooshmand, M., Summers, R., Gage, F.H. and Anderson, A.J. (2005) Human neural stem cells differentiate and promote locomotor recovery in spinal cord-injured mice. Proc. Natl. Acad. Sci., U.S.A., 102: 14069–14074. DeForge, D., Nymark, J., Lemaire, E., Gardner, S., Hunt, M., Martel, L., Curran, D. and Barbeau, H. (2004) Effect of 4aminopyridine on gait in ambulatory spinal cord injuries: a double-blind, placebo-controlled, crossover trial. Spinal Cord, 42: 674–685. Delamarter, R.B., Sherman, J. and Carr, J.B. (1995) Pathophysiology of spinal cord injury. Recovery after immediate and delayed decompression. J. Bone Joint Surg. Am., 77: 1042–1049. Dimitrijevic, M.R., Gerasimenko, Y. and Pinter, M.M. (1998) Evidence for a spinal central pattern generator in humans. Ann. N.Y. Acad. Sci., 860: 360–376. Dobkin, B., Apple, D., Barbeau, H., Basso, M., Behrman, A., DeForge, D., Ditunno, J., Dudley, G., Elashoff, R., Fugate, L., Harkema, S., Saulino, M. and Scott, M. (2006a) Weightsupported treadmill vs. over-ground training for walking after acute incomplete SCI. Neurology, 66: 484–493. Dobkin, B.H., Curt, A. and Guest, J. (2006b) Cellular transplants in China: observational study from the largest human experiment in chronic spinal cord injury. Neurorehabil. Neural Repair, 20: 5–13. Dommergues, M.A., Plaisant, F., Verney, C. and Gressens, P. (2003) Early microglial activation following neonatal excitotoxic brain damage in mice: a potential target for neuroprotection. Neuroscience, 121: 619–628. Doucette, R. (1995) Olfactory ensheathing cells: potential for glial cell transplantation into areas of CNS injury. Histol. Histopathol., 10: 503–507.

231 Dubreuil, C.I., Winton, M.J. and McKerracher, L. (2003) Rho activation patterns after spinal cord injury and the role of activated Rho in apoptosis in the central nervous system. J. Cell Biol., 162: 233–243. Duysens, J. and Van de Crommert, H.W. (1998) Neural control of locomotion: the central pattern generator from cats to humans. Gait Posture, 7: 131–141. Eriksson, P.S., Perﬁlieva, E., Bjork-Eriksson, T., Alborn, A.M., Nordborg, C., Peterson, D.A. and Gage, F.H. (1998) Neurogenesis in the adult human hippocampus. Nat. Med., 4: 1313–1317. Fass, B. and Ramirez, J.J. (1984) Effects of ganglioside treatments on lesion-induced behavioral impairments and sprouting in the CNS. J. Neurosci. Res., 12: 445–458. Fehlings, M.G. and Perrin, R.G. (2006) The timing of surgical intervention in the treatment of spinal cord injury: a systematic review of recent clinical evidence. Spine, 31: S28–S36. Feron, F., Perry, C., Cochrane, J., Licina, P., Nowitzke, A., Urquhart, S., Geraghty, T. and Mackay-Sim, A. (2005) Autologous olfactory ensheathing cell transplantation in human spinal cord injury. Brain, 128: 2951–2960. Firouzi, M., Moshayedi, P., Saberi, H., Mobasheri, H., Abolhassani, F., Jahanzad, I. and Raza, M. (2006) Transplantation of Schwann cells to subarachnoid space induces repair in contused rat spinal cord. Neurosci. Lett., 402: 66–70. Geisler, F.H., Coleman, W.P., Grieco, G. and Poonian, D. (2001) The Sygen multicenter acute spinal cord injury study. Spine, 26: S87–S98. Geisler, F.H., Dorsey, F.C. and Coleman, W.P. (1991) Recovery of motor function after spinal-cord injury — a randomized, placebo-controlled trial with GM-1 ganglioside. N. Engl. J. Med., 324: 1829–1838. Gruner, J.A. (1992) A monitored contusion model of spinal cord injury in the rat. J. Neurotrauma, 9: 123–128. Hall, E.D. and Braughler, J.M. (1981) Acute effects of intravenous glucocorticoid pretreatment on the in vitro peroxidation of cat spinal cord tissue. Exp. Neurol., 73: 321–324. Hall, E.D. and Braughler, J.M. (1982) Effects of intravenous methylprednisolone on spinal cord lipid peroxidation and Na++K+-ATPase activity. Dose-response analysis during 1st hour after contusion injury in the cat. J. Neurosurg., 57: 247–253. Hansebout, R.R., Blight, A.R., Fawcett, S. and Reddy, K. (1993) 4-Aminopyridine in chronic spinal cord injury: a controlled, double-blind, crossover study in eight patients. J. Neurotrauma, 10: 1–18. Hicks, A.L., Adams, M.M., Martin Ginis, K., Giangregorio, L., Latimer, A., Phillips, S.M. and McCartney, N. (2005) Long-term body-weight-supported treadmill training and subsequent follow-up in persons with chronic SCI: effects on functional walking ability and measures of subjective wellbeing. Spinal Cord, 43: 291–298. Hofstetter, C.P., Schwarz, E.J., Hess, D., Widenfalk, J., El Manira, A., Prockop, D.J. and Olson, L. (2002) Marrow stromal cells form guiding strands in the injured spinal cord and promote recovery. Proc. Natl. Acad. Sci. U.S.A., 99: 2199–2204.

Huang, H., Chen, L., Wang, H., Xiu, B., Li, B., Wang, R., Zhang, J., Zhang, F., Gu, Z., Li, Y., Song, Y., Hao, W., Pang, S. and Sun, J. (2003) Inﬂuence of patients’ age on functional recovery after transplantation of olfactory ensheathing cells into injured spinal cord injury. Chin. Med. J. (Engl.), 116: 1488–1491. Hurlbert, R.J. (2000) Methylprednisolone for acute spinal cord injury: an inappropriate standard of care. J. Neurosurg., 93: 1–7. Hurlbert, R.J. (2001) The role of steroids in acute spinal cord injury: an evidence-based analysis. Spine, 26: S39–S46. Johansson, C.B., Momma, S., Clarke, D.L., Risling, M., Lendahl, U. and Frisen, J. (1999a) Identiﬁcation of a neural stem cell in the adult mammalian central nervous system. Cell, 96: 25–34. Johansson, C.B., Svensson, M., Wallstedt, L., Janson, A.M. and Frisen, J. (1999b) Neural stem cells in the adult human brain. Exp. Cell Res., 253: 733–736. Joshi, M. and Fehlings, M.G. (2002a) Development and characterization of a novel, graded model of clip compressive spinal cord injury in the mouse: Part 1. Clip design, behavioral outcomes, and histopathology. J. Neurotrauma, 19: 175–190. Joshi, M. and Fehlings, M.G. (2002b) Development and characterization of a novel, graded model of clip compressive spinal cord injury in the mouse: Part 2. Quantitative neuroanatomical assessment and analysis of the relationships between axonal tracts, residual tissue, and locomotor recovery. J. Neurotrauma, 19: 191–203. Karimi-Abdolrezaee, S., Eftekharpour, E., Wang, J., Morshead, C.M. and Fehlings, M.G. (2006) Delayed transplantation of adult neural precursor cells promotes remyelination and functional neurological recovery after spinal cord injury. J. Neurosci., 26: 3377–3389. Kernie, S.G., Erwin, T.M. and Parada, L.F. (2001) Brain remodeling due to neuronal and astrocytic proliferation after controlled cortical injury in mice. J. Neurosci. Res., 66: 317–326. Knoller, N., Auerbach, G., Fulga, V., Zelig, G., Attias, J., Bakimer, R., Marder, J.B., Yoles, E., Belkin, M., Schwartz, M. and Hadani, M. (2005) Clinical experience using incubated autologous macrophages as a treatment for complete spinal cord injury: phase I study results. J. Neurosurg. Spine, 3: 173–181. Kohama, I., Lankford, K.L., Preiningerova, J., White, F.A., Vollmer, T.L. and Kocsis, J.D. (2001) Transplantation of cryopreserved adult human Schwann cells enhances axonal conduction in demyelinated spinal cord. J. Neurosci., 21: 944–950. Lee, K.H., Yoon do, H., Park, Y.G. and Lee, B.H. (2005) Effects of glial transplantation on functional recovery following acute spinal cord injury. J. Neurotrauma, 22: 575–589. Li, S. and Strittmatter, S.M. (2003) Delayed systemic Nogo-66 receptor antagonist promotes recovery from spinal cord injury. J. Neurosci., 23: 4219–4227. Li, Y., Field, P.M. and Raisman, G. (1998) Regeneration of adult rat corticospinal axons induced by transplanted olfactory ensheathing cells. J. Neurosci., 18: 10514–10524.

232 Lucas, S.M., Rothwell, N.J. and Gibson, R.M. (2006) The role of inﬂammation in CNS injury and disease. Br. J. Pharmacol., 147(Suppl 1): S232–S240. Man, S.M., Ma, Y.R., Shang, D.S., Zhao, W.D., Li, B., Guo, D.W., Fang, W.G., Zhu, L. and Chen, Y.H. (2007) Peripheral T cells overexpress MIP-1 alpha to enhance its transendothelial migration in Alzheimer’s disease. Neurobiol. Aging, 28: 485–496. Miklossy, J., Doudet, D.D., Schwab, C., Yu, S., McGeer, E.G. and McGeer, P.L. (2006) Role of ICAM-1 in persisting inﬂammation in Parkinson disease and MPTP monkeys. Exp. Neurol., 197: 275–283. Nashmi, R. and Fehlings, M.G. (2001) Mechanisms of axonal dysfunction after spinal cord injury: with an emphasis on the role of voltage-gated potassium channels. Brain Res. Brain Res. Rev., 38: 165–191. Niederost, B., Oertle, T., Fritsche, J., McKinney, R.A. and Bandtlow, C.E. (2002) Nogo-A and myelin-associated glycoprotein mediate neurite growth inhibition by antagonistic regulation of RhoA and Rac1. J. Neurosci., 22: 10368–10376. Noorbakhsh, F., Tsutsui, S., Vergnolle, N., Boven, L.A., Shariat, N., Vodjgani, M., Warren, K.G., Andrade-Gordon, P., Hollenberg, M.D. and Power, C. (2006) Proteinase-activated receptor 2 modulates neuroinﬂammation in experimental autoimmune encephalomyelitis and multiple sclerosis. J. Exp. Med., 203: 425–435. Oudega, M. and Xu, X.M. (2006) Schwann cell transplantation for repair of the adult spinal cord. J. Neurotrauma, 23: 453–467. Park, H.C., Shim, Y.S., Ha, Y., Yoon, S.H., Park, S.R., Choi, B.H. and Park, H.S. (2005) Treatment of complete spinal cord injury patients by autologous bone marrow cell transplantation and administration of granulocyte-macrophage colony stimulating factor. Tissue Eng., 11: 913–922. Pellitteri, R., Russo, A. and Stanzani, S. (2006) Schwann cell: a source of neurotrophic activity on cortical glutamatergic neurons in culture. Brain Res., 1069: 139–144. Pollard, S.M., Conti, L., Sun, Y., Goffredo, D. and Smith, A. (2006) Adherent neural stem (NS) cells from fetal and adult forebrain. Cereb. Cortex, 16: i112–i120. Ramon-Cueto, A., Cordero, M.I., Santos-Benito, F.F. and Avila, J. (2000) Functional recovery of paraplegic rats and motor axon regeneration in their spinal cords by olfactory ensheathing glia. Neuron, 25: 425–435. Rapalino, O., Lazarov-Spiegler, O., Agranov, E., Velan, G.J., Yoles, E., Fraidakis, M., Solomon, A., Gepstein, R., Katz, A., Belkin, M., Hadani, M. and Schwartz, M. (1998) Implantation of stimulated homologous macrophages results in partial recovery of paraplegic rats. Nat. Med., 4: 814–821. Richardson, P.M., McGuinness, U.M. and Aguayo, A.J. (1980) Axons from CNS neurons regenerate into PNS grafts. Nature, 284: 264–265. Rivlin, A.S. and Tator, C.H. (1978) Effect of duration of acute spinal cord compression in a new acute cord injury model in the rat. Surg. Neurol., 10: 38–43.

Savio, T. and Schwab, M.E. (1989) Rat CNS white matter, but not gray matter, is nonpermissive for neuronal cell adhesion and ﬁber outgrowth. J. Neurosci., 9: 1126–1133. Sekhon, L.H. and Fehlings, M.G. (2001) Epidemiology, demographics, and pathophysiology of acute spinal cord injury. Spine, 26: S2–S12. Stirling, D.P., Khodarahmi, K., Liu, J., McPhail, L.T., McBride, C.B., Steeves, J.D., Ramer, M.S. and Tetzlaff, W. (2004) Minocycline treatment reduces delayed oligodendrocyte death, attenuates axonal dieback, and improves functional outcome after spinal cord injury. J. Neurosci., 24: 2182–2190. Stokes, B.T., Noyes, D.H. and Behrmann, D.L. (1992) An electromechanical spinal injury technique with dynamic sensitivity. J. Neurotrauma, 9: 187–195. Stoy, N., Mackay, G.M., Forrest, C.M., Christoﬁdes, J., Egerton, M., Stone, T.W. and Darlington, L.G. (2005) Tryptophan metabolism and oxidative stress in patients with Huntington’s disease. J. Neurochem., 93: 611–623. Sykova´, E., Jendelova, P., Urdzikova, L., Lesny, P. and Hejcl, A. (2006) Bone marrow stem cells and polymer hydrogels: two strategies for spinal cord injury repair. Cell. Mol. Neurobiol., 26: 1111–1127. Tator, C.H., Fehlings, M.G., Thorpe, K. and Taylor, W. (1999) Current use and timing of spinal surgery for management of acute spinal cord injury in North America: results of a retrospective multicenter study. J. Neurosurg., 91: 12–18. Teng, Y.D., Choi, H., Onario, R.C., Zhu, S., Desilets, F.C., Lan, S., Woodard, E.J., Snyder, E.Y., Eichler, M.E. and Friedlander, R.M. (2004) Minocycline inhibits contusiontriggered mitochondrial cytochrome c release and mitigates functional deﬁcits after spinal cord injury. Proc. Natl. Acad. Sci. U.S.A., 101: 3071–3076. Tikka, T.M. and Koistinaho, J.E. (2001) Minocycline provides neuroprotection against N-methyl-D-aspartate neurotoxicity by inhibiting microglia. J. Immunol., 166: 7527–7533. Van de Crommert, H.W., Mulder, T. and Duysens, J. (1998) Neural control of locomotion: sensory control of the central pattern generator and its relation to treadmill training. Gait Posture, 7: 251–263. Walmsley, A.R., McCombie, G., Neumann, U., Marcellin, D., Hillenbrand, R., Mir, A.K. and Frentzel, S. (2004) Zinc metalloproteinase-mediated cleavage of the human Nogo-66 receptor. J. Cell Sci., 117: 4591–4602. Wells, J.E., Hurlbert, R.J., Fehlings, M.G. and Yong, V.W. (2003) Neuroprotection by minocycline facilitates signiﬁcant recovery from spinal cord injury in mice. Brain, 126: 1628–1637. Wiessner, C., Bareyre, F.M., Allegrini, P.R., Mir, A.K., Frentzel, S., Zurini, M., Schnell, L., Oertle, T. and Schwab, M.E. (2003) Anti-Nogo-A antibody infusion 24 hours after experimental stroke improved behavioral outcome and corticospinal plasticity in normotensive and spontaneously hypertensive rats. J. Cereb. Blood Flow Metab., 23: 154–165.

233 Winton, M.J., Dubreuil, C.I., Lasko, D., Leclerc, N. and McKerracher, L. (2002) Characterization of new cell permeable C3-like proteins that inactivate Rho and stimulate neurite outgrowth on inhibitory substrates. J. Biol. Chem., 277: 32820–32829. Xu, L., Fagan, S.C., Waller, J.L., Edwards, D., Borlongan, C.V., Zheng, J., Hill, W.D., Feuerstein, G. and Hess, D.C. (2004) Low dose intravenous minocycline is neuroprotective after middle cerebral artery occlusion-reperfusion in rats. BMC Neurol., 4: 7.

Zheng, M. and Kufﬂer, D.P. (2000) Guidance of regenerating motor axons in vivo by gradients of diffusible peripheral nerve-derived factors. J. Neurobiol., 42: 212–219. Zhu, S., Stavrovskaya, I.G., Drozda, M., Kim, B.Y., Ona, V., Li, M., Sarang, S., Liu, A.S., Hartley, D.M., Wu du, C., Gullans, S., Ferrante, R.J., Przedborski, S., Kristal, B.S. and Friedlander, R.M. (2002) Minocycline inhibits cytochrome c release and delays progression of amyotrophic lateral sclerosis in mice. Nature, 417: 74–78.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 16

Cerebral contusion: a role model for lesion progression Tatsuro Kawamata1,2, and Yoichi Katayama1,2 2

1 Department of Neurological Surgery, Nihon University School of Medicine, Tokyo, Japan The Japan Neurotrauma Data Bank Committee (The Japan Society of Neurotraumatology, The Japanese Council of Traffic Science)

Abstract: The early massive edema caused by severe cerebral contusion results in progressive intracranial pressure (ICP) elevation and clinical deterioration within 24–72 h post-trauma. Surgical excision of the necrotic brain tissue represents the only therapy, which can provide satisfactory control of the elevated ICP and clinical deterioration. In this chapter, we review the results of our clinical studies regarding the pathophysiology of contusion edema and evaluate the effects of surgical treatment, i.e. contusion necrotomy, by analyzing the data from the Japan Neurotrauma Data Bank. Keywords: cerebral contusion; brain edema; diffusion-weighted image; vascular permeability; tissue osmolality; contusion necrotomy therapy, which can provide satisfactory control of the elevated ICP. The precise mechanism underlying such an early massive edema is not yet clearly understood. We review the results of our clinical studies (Katayama et al., 1990, 1992, 1998; Kushi et al., 1994; Kawamata et al., 2000; Katayama and Kawamata, 2003), which have provided several lines of evidence to suggest that a large amount of edema ﬂuid is accumulated in the necrotic brain tissue within the central area of contusion, and this contributes to the early massive edema. The clinical results of surgical treatment, i.e. contusion necrotomy, are also discussed.

Introduction In patients with cerebral contusions, two types of edema can be clinically recognized. One is the early massive edema that occurs within the period of 24–72 h post-trauma. This type of edema creates strong mass effect resulting in progressive elevation of the intracranial pressure (ICP) and clinical deterioration known as talk-and-deteriorate (Katayama et al., 1990). The other is the delayed peri-contusion edema, which is typically seen in the white matter adjacent to the cerebral contusion, at several days post-trauma by T2-weighted magnetic resonance (MR) imaging. This type of edema rarely causes ICP elevation leading to fatal deterioration. Despite intensive medical therapy, the elevated ICP in patients with early massive edema is often uncontrollable and fatal. In such instances, surgical excision of the necrotic brain tissue is the only

Histopathology of cerebral contusion The classical histopathological study of Lindenberg and Freytag (1957) demonstrated the presence of two components of cerebral contusion; one is the central (core) area in which cells undergo necrosis as the primary consequence of mechanical injury

Corresponding author. Tel.: +81-3-3972-8111 ext. 2481;

Fax: +81-3-3554-0425; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61016-9

235

236

(contusion necrosis proper), and the other is the peripheral (rim) area in which cellular swelling occurs as a consequence of ischemia. A clear demarcation line separates these two components. The area of contusion necrosis is histopathologically evident as early as at 3 h post-trauma, as a primary brain damage (Eriskat et al., 1994). The cellular elements in the central area, both neuronal as well as glial cells, uniformly undergo shrinkage, and then disintegration, homogenation and cyst formation eventually. In contrast, cell swelling is predominant in the peripheral area (Lindenberg and Freytag, 1957). The ischemia in the peripheral area is largely attributable to microthrombosis, which is a main cause of secondary brain damages. Numerous clinical studies have demonstrated a decrease in cerebral blood ﬂow in the central as well as the peripheral areas of contusion (e.g., Alexander et al., 1994).

MR diffusion study (ADC mapping) We have investigated the evolution of cerebral contusion by diffusion-weighted MR imaging and apparent diffusion co-efﬁcient (ADC) mapping in head trauma patients (Kawamata et al., 2000). Following conventional T1- and T2-weighted MR imaging, diffusion-weighted MR images are obtained, and ADC mapping is computed from various b factor diffusion images. The diffusion-weighted MR images demonstrate a low intensity core in the central area, beginning at approximately 24 h post-trauma. The ADC value within this central area clearly increases during the period of 24–72 h post-trauma (Fig. 1; Kawamata et al., 2000). This elevated ADC value appears to represent the contusion necrosis proper, since cellular disintegration and homogenization in this area would result in an expansion of the extracellular space. The diffusion-weighted MR images demonstrate a high intensity rim in the peripheral area of contusion beginning at approximately 24 h post-trauma, which corresponds with the timing of the elevated ICP and neurological deterioration. The combination of a low intensity core and a high intensity rim is a consistent ﬁnding in the cerebral contusion

Fig. 1. A representative case of cerebral contusion in the acute phase post-trauma. Upper: A CT scan revealed low-density area in bilateral frontal lobe. Lower: Diffusion-weighted image demonstrated low intensity core with surrounding high intensity rim, which was typical pattern of diffusion-weighted image in the acute phase (o48 h post-trauma) of cerebral contusion without massive hemorrhage.

during this period, which can be termed a halo appearance (Fig. 2). The ADC value decreases in the peripheral area during the period of 24–72 h posttrauma (Kawamata et al., 2000). This decreased ADC value appears to represent the cellular swelling, which would result in shrinkage of the extracellular space. The ADC values between the central and peripheral areas are maximally dissociated during the period of 24–72 h post-trauma. The ADC value in the peripheral area shifts from a decrease to an increase after 72 h posttrauma. At the same time, an increase in ADC value becomes evident in the adjacent white matter. This increase in ADC value in the adjacent white matter appears to represent the delayed

237

Fig. 2. Bilateral contusion demonstrating a halo appearance with a combination of a low intensity core and a high intensity rim on diffusion-weighted image (left). On ADC mapping (right), the peripheral rim showed low ADC value representing the cellular swelling resulting from shrinkage of the extracellular space.

peri-contusion edema, which can commonly be seen on T2-weighted MR images as vasogenic edema.

Increased cerebrovascular permeability Since gadolinium (Gd)-DTPA, like plasma protein, does not normally cross the blood-brain barrier, enhancement with Gd-DTPA, if observed in association with cerebral contusion, implies an increased cerebrovascular permeability. In a previous study (Lang et al., 1991), Gd-DTPA was administered intravenously by bolus injection and MR imaging was undertaken soon after the administration. Such a procedure failed to detect any increase in cerebrovascular permeability associated with cerebral contusions during the initial few days post-trauma. Enhancement with Gd-DTPA has been reported to become detectable at 6–9 days post-trauma. Furthermore, an immunohistochemical study of the post-mortem brain in such patients failed to reveal any plasma protein leakage around the contused brain areas during this early period (Todd and Graham, 1990). Evaluations of the vascular permeability by 99mTc pertechnetate single photon emission tomography failed to reveal any evidence of an increased vascular permeability within the area of contusion during the initial few days post-trauma (Bullock et al., 1990). Such ﬁndings

appear to contradict the widely held view that an increased cerebrovascular permeability is responsible for the development of contusion edema (Marmarou, 2003). We examined the changes in cerebrovascular permeability employing intravenous slow infusion of Gd-DTPA and delayed MR imaging (Kushi et al., 1994). In general, increases in cerebrovascular permeability can be detected more clearly by intravenous slow infusion of high dose contrast medium and delayed neuroimaging (Ito et al., 1988). This technique revealed that cerebral contusions can be enhanced at as early as 24–48 h post-trauma. In MR imaging undertaken at 2 h after Gd-DTPA administration, enhancement on T1-weighted images was observed in either the central area or the peripheral area, or both. It is evident that water supply from the blood vessels is not completely interrupted even in the central area of contusion. The delayed peri-contusion edema has commonly been attributed to an increased cerebrovascular permeability. As mentioned above, enhancement with Gd-DTPA has been reported to become detectable at 6–9 days post-trauma. It is also possible, however, that resolution of the cellular swelling in the peripheral area of contusion might permit propagation of the edema ﬂuid accumulated within the central area to the adjacent white matter. This would lead to edema appearance on CT and MR images which resembles that induced by an

238

increased cerebrovascular permeability, i.e. vasogenic edema.

Edema ﬂuid accumulation in the central area In approximately 50% of patients with cerebral contusions, a crescent-shaped zone of very high ADC value develops at the border between the central and peripheral areas beginning at approximately 24 h post-trauma (Fig. 3; Kawamata et al., 2000). This crescent-shaped zone was always located within the central area of contusion. The

crescent-shaped zone still existed at 2 weeks posttrauma in some cases (Fig. 4). The very high ADC value appears to represent edema ﬂuid accumulation within the central area of contusion. Edema ﬂuid accumulation within the necrotic brain tissue is also suggested by the formation of a ﬂuid–blood interface within cerebral contusions (Fig. 5). The ﬂuid–blood interface can be formed without a ﬂuid cavity. When we carry out surgery, we ﬁnd no ﬂuid cavity but softened and water-rich necrotic brain tissue is present (Katayama et al., 1992). Within the contusion necrosis, hemorrhage undergoes enlargement within the initial 6 h

Fig. 3. Cerebral contusion in the temporal tip at 45 h post-trauma. Left: CT scan showed small low-density area in the temporal tip. Middle: T2-weighted MRI revealed mixed intensity area in the core of contusion. Right: ADC mapping demonstrated a crescentshaped zone (arrows) of very high ADC value, as high as that of CSF, representing edema ﬂuid accumulation within the central area of contusion.

Fig. 4. Cerebral contusion in the subacute phase, 2 weeks post-trauma. Left: T2-weighted image. Middle: Diffusion-weighted image. Right: ADC mapping. A crescent shaped zone still existed (arrows). ADC value in the peripheral area of contusion increased, indicating that vasogenic edema turned to be predominant.

239

Fig. 5. CT scan and its scheme of cerebral contusion in the temporal tip (10 h post-trauma), showing ﬂuid–blood interface in the area of contusion necrosis. (A) edema ﬂuid; (B) hematoma; (C) ﬂuid–blood interface; (D) layering of red blood cells.

post-trauma (Katayama et al., 1990), and red blood cells diffusely permeate the softened necrotic brain tissue (Lindenberg and Freytag, 1957). The ﬂuid–blood interfaces observed within the central area of contusion may represent layering of red blood cells in the softened necrotic brain tissue, which has accumulated voluminous edema ﬂuid.

Osmotic potential of the contusion necrosis We have reported that necrotic brain tissue sampled from the central area of contusion during surgery shows a very high osmolality, reaching 350–400 mOsm (Katayama et al., 1998). It is uncertain whether or not such a marked increase in osmolality is osmotically active and causes edema ﬂuid accumulation. It appears that expansion of the extracellular space in the central area increases the capacitance for edema ﬂuid accumulation. In contrast, shrinkage of the extracellular space in the peripheral area increases the resistance for edema ﬂuid propagation or resolution. We hypothesize that the barrier formed by swollen cells in the peripheral area may prevent edema ﬂuid propagation and also help to generate osmotic potentials across the central and peripheral areas. Since blood ﬂow is greatly reduced but is not completely interrupted in the contused brain tissue, water is supplied from the blood vessels into the central area. We suggest that a combination of these events may facilitate

edema ﬂuid accumulation in the central area and contribute to the early massive edema of cerebral contusion. Surgical treatment We investigated the effects of surgical excision of the necrotic brain tissue in patients with severe cerebral contusion. The data from the Japan Neurotrauma Data Bank (JNTDB) in which a total of 1002 patients suffering severe traumatic brain injury registered during the period between 1998 and 2001 were analyzed (Nakamura et al., 2002). Among these patients, 182 (18%) demonstrated severe cerebral contusions as the major lesions contributing to their clinical status. Results Among the 182 patients with severe cerebral contusion, 121 (66%; Group I) were treated conservatively and the remaining 61 (34%; Group II) underwent surgery. The ratio of selecting surgical management was far lower in patients with cerebral contusion (34719%), as compared with those with acute epidural hematoma (88711%) or acute subdural hematoma (68718%). There was a huge variation in the ratio of selecting surgical management (9–77%) among the contributing centers (n ¼ 10) of this data bank. While older patients,

240

especially those between 40 and 60 years old, tended to undergo surgery more frequently, younger patients tended to be treated conservatively. There was, however, no signiﬁcant difference in age between Groups I and II (47.8723.8 vs. 54.4719.5 years). The surgical management involved internal decompression (complete excision of the necrotic brain tissue and evacuation of clots) with or without external decompression in most patients (90%) of Group II. The remaining patients underwent external decompression alone. Surgery was performed at 1.886.1 (19.5724.2) h, mostly (73%) within 24 h post-trauma. Group I demonstrated a clearly poorer outcome on the Glasgow outcome scale (GOS) at 6 months post-trauma (Table 1). The mortality was clearly higher in Group I, as compared with Group II (48% vs. 23%; p ¼ 0.0001; n ¼ 182). A difference in mortality between the two groups was noted in patients who were scored at 8 or less on the GCS at the time of admission, but did not reach a statistically signiﬁcant level (Table 1). The most striking difference was observed in patients who were scored at 9 or better on the GCS at the time of admission (Table 1). The mortality was clearly higher in Group I, as compared with Group II (56% vs. 17%; p ¼ 0.017; n ¼ 45). A clear difference in Table 1. Outcome (6 months post-trauma) GOS (%) n

GR

MD

SD

VS

D

GCS on admission: 3–5 Conservative 47 Surgical 11

7 9

2 9

11 27

11 0

70 55

GCS on admission: 6–8 Conservative 58 Surgical 21

29 24

21 29

10 24

10 10

29 14

GCS on admission: 9–15 Conservative 16 19 Surgical 29 28

13 28

13 17

0 10

56 17

Total Conservative Surgical

12 25

12 21

9 8

48 23

121 61

19 23

mortality between the two groups was observed even when the analysis was restricted to patients who deﬁnitely demonstrated ‘‘talk-and-deteriorate’’ (64% vs. 22%; p ¼ 0.026; n ¼ 29; Table 2). The indications for surgical intervention in case of severe cerebral contusion remain controversial (Bullock et al., 1989). The huge variation in ratio of selecting surgical management among the present centers reﬂects a diversity in management policy and an absence of consensus regarding the indications for surgery. The higher mortality in Group I, as compared with Group II, suggests that the surgery performed in the Group II patients helped to prevent their clinical deterioration and death. Such an effect of surgery was most striking in those patients who were scored at 9 or better at the time of admission. It is by no means certain whether patients in Group II who were scored at 9 or better before surgery would have deteriorated or not if surgery had not been carried out. An effect of surgery on mortality is evident, however, since a difference in mortality was clearly observed even when the analysis was restricted to patients who deﬁnitely demonstrated ‘‘talk-and-deteriorate’’. This ﬁnding strongly suggests that the surgery itself was the major reason for the improved mortality in Group II. In other words, death was probably prevented by the surgery in many patients of Group II. The present ﬁndings support our hypothesis that early massive edema is caused by cerebral contusion through the presence of necrotic brain tissue, and indicate that surgical excision of the necrotic brain tissue is the only therapy which can provide satisfactory control of the progressive elevation of the ICP and clinical deterioration in many cases. Surgical intervention should be considered in Table 2. Outcome (6 months post-trauma) in patients demonstrating ‘‘talk-and-deteriorate’’ GOS (%)

Conservative Surgical

n

GR

MD

SD

VS

D

11 18

18 11

9 22

9 39

0 6

64 22

Note: GCS, Glasgow coma scale; GOS, Glasgow outcome scale. p ¼ 0.017. p ¼ 0.0001.

Note: GCS, Glasgow coma scale; GOS, Glasgow outcome scale. p ¼ 0.026.

241

patients with severe cerebral contusion who demonstrate ‘‘talk-and-deteriorate’’. Conclusion There is at present no established medical treatment which can effectively inhibit edema ﬂuid accumulation within cerebral contusions. The most effective therapy for ameliorating the potentially fatal edema is surgical excision of the necrotic brain tissue. The effects of surgical excision of necrotic brain tissue have commonly been accounted for on the basis of an increased space compensation for mass lesions. It is possible, however, that excision of the necrotic brain tissue does mean elimination of the cause of edema ﬂuid accumulation. If the ICP is elevated by early massive edema due to cerebral contusion and the elevated ICP is medically uncontrollable, surgical excision of the necrotic brain tissue would appear to represent the therapy of choice, regardless of the size of the associated hemorrhages. Surgery should be considered in patients who are scored at 9 or better at the time of admission, as soon as they show clinical deterioration. References Alexander, M.J., Martin, N.A., Khanna, M., Caron, M. and Becker, D.P. (1994) Regional cerebral blood ﬂow trends in head injured patients with focal contusions and cerebral edema. Acta Neurochir., s60: 479–481. Bullock, R., Golek, J. and Blake, G. (1989) Traumatic intracerebral hematoma Which patients should undergo surgical evacuation? CT scan features and ICP monitoring as a basis for decision making. Surg. Neurol., 32: 181–187. Bullock, R., Statham, J., Patterson, D., Wyper, D., Hadley, D. and Teasdale, E. (1990) The time course of vasogenic oedema after focal human head injury: evidence from SPECT mapping of blood brain barrier defects. Acta Neurochir., s51: 286–288.

Eriskat, J., Schurer, L., Kempski, O. and Baethmann, A. (1994) Growth kinetics of a primary brain tissue necrosis from focal lesion. Acta Neurochir., s60: 425–427. Ito, U., Reulen, H.-J., Tomita, H., Ikeda, J., Saito, J. and Maehara, T. (1988) Formation and propagation of brain oedema ﬂuid around human brain metastasis: a CT study. Acta Neurochir., 90: 35–41. Katayama, Y. and Kawamata, T. (2003) Edema ﬂuid accumulation within necrotic brain tissue as a cause of the mass effect of cerebral contusion in head trauma patient. Acta Neurochir., s86: 323–327. Katayama, Y., Mori, T., Maeda, T. and Kawamata, T. (1998) Pathogenesis of the mass effect of cerebral contusions: a rapid increase in osmolality within the contusion necrosis. Acta Neurochir., s71: 289–292. Katayama, Y., Tsubokawa, T., Kinoshita, K. and Himi, K. (1992) Intra-parenchymal ﬂuid-blood levels in traumatic intracerebral hematomas. Neuroradiology, 34: 381–383. Katayama, Y., Tsubokawa, T., Miyazaki, S., Kawamata, T. and Yoshino, A. (1990) Oedema ﬂuid formation within contused brain tissue as a cause of medically uncontrollable elevation of intracranial pressure in head trauma patients. Acta Neurochir., s51: 308–310. Kawamata, T., Katayama, Y., Aoyama, N. and Mori, T. (2000) Heterogeneous mechanisms of early edema formation in cerebral contusion: diffusion MRI and ADC mapping study. Acta Neurochir., s76: 9–12. Kushi, H., Katayama, Y., Shibuya, T., Tsubokawa, T. and Kuroha, T. (1994) Gd-DTPA enhanced magnetic resonance imaging of cerebral contusions. Acta Neurochir., s60: 472–474. Lang, D.A., Hadley, D.M., Teasdale, G.T., Macpherson, P. and Teasdale, E. (1991) Gadolinium DTPA enhanced magnetic resonance imaging in acute head injury. Acta Neurochir., 109: 5–11. Lindenberg, R. and Freytag, E. (1957) Morphology of cortical contusions. AMA Arch. Pathol., 63: 23–42. Marmarou, A. (2003) Pathophysiology of traumatic brain edema: current concepts. Acta Neurochir., s86: 7–10. Nakamura, N., Yamaura, A., Shigemori, M., Ono, J., Kawamata, T., Sakamoto, T. and Japanese Data Bank Committee for Traumatic Brain Injury. (2002) Epidemiology, prevention and countermeasures against severe traumatic brain injury in Japan and abroad. Neurol. Res., 24: 45–53. Todd, N.V. and Graham, D.I. (1990) Blood-brain barrier damage in traumatic brain contusion. Acta Neurochir., s51: 296–299.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 17

Ethical implications of time frames in a randomized controlled trial in acute severe traumatic brain injury Erwin J.O. Kompanje1,2,, Andrew I.R. Maas2, Franc- ois J.A. Slieker2 and Nino Stocchetti3 1

Department of Intensive Care, Erasmus MC University Medical Center Rotterdam, P.O. Box 2040, 3000 CA Rotterdam, The Netherlands 2 Department of Neurosurgery, Erasmus MC University Medical Center Rotterdam, P.O. Box 2040, 3000 CA Rotterdam, The Netherlands 3 Ospedale Policlinico IRCCS, Milan University, Milan, Italy

Abstract: Objectives: To analyze factors determining the time between injury and study drug administration (SDA) in a randomized controlled trial (RCT) of acute severe traumatic brain injury (TBI) and to discuss the ethical implications. Methods: Time frames prior to SDA, differentiated per country, were analyzed in a recently conducted RCT in severe TBI. Per protocol, the time window for SDA was 6 h after injury. We selected patients for whom written proxy consent (PC) was obtained prior to SDA (n ¼ 631). Results: The time between injury and admission to the neurotrauma center (NTC) varied per country from 1.16 to 2.35 h, but CT scan was obtained on average within 1 h of admission. The median time between injury and CT scan was within 3 h in all but one country. The broadest time window was observed between CT scan and obtaining required PC (1.71–2.74 h). The median time between injury and PC varied between countries from 3.75 to 5.00 h. After consent had been obtained, almost all patients subsequently received study drug within 1 h. In 85.3% of all cases time between injury and SDA exceeded 4 h, in 60% 5 h. Conclusions: The requirement of written PC causes a signiﬁcant delay in SDA in TBI. With deferred consent, the ﬁrst dose of an investigational drug could potentially be administered directly after completion of the admission CT scan, which reduce the time to SDA by 50%. We argue that randomization under deferred consent is ethically defendable for emergency research in severe TBI. Recommendations for patient protection are proposed. Keywords: traumatic brain injury; informed consent; emergency trial; deferred consent TBI is about 30% and a signiﬁcant disability persists in a further 35–40%. These data signify an ethical imperative to develop and test neuroprotective agents and other new therapeutic strategies. Various neuroprotective agents, mainly targeting speciﬁc pathophysiologic mechanisms, have been tested in TBI, but convincing beneﬁt has not been shown (Maas et al., 1999). Randomized controlled trials (RCTs) in emergency and intensive care medicine pose complex ethical and methodological

Introduction Severe traumatic brain injury (TBI) remains a major cause of death and disability afﬂicting mostly young adult males and elderly people, and results in high economic costs to society (McGarry et al. 2002). The fatality rate for severe Corresponding author. Tel.: +31-6-53837655; Fax: +31-10-

4634075; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61017-0

243

244

challenges (Maas et al., 2004; Kompanje et al., 2005). Speciﬁc ethical issues pertaining to clinical testing of neuroprotective agents in TBI include the emergency nature of the research, the incapacity of the patients to informed consent before inclusion, short therapeutic time windows, and a risk–beneﬁt ratio based on concept that in relation to the severity of the trauma, signiﬁcant adverse side effects may be acceptable for treatments with proven beneﬁt. The relevance of these issues and implications for trial design are, however, not fully recognized in- and outside the expert ﬁeld. Legislation in countries in the European Union is being amended to comply with the European Union Directive 2001/20/EC (European Union, 2001). In this new European legislation, emergency research under deferred or waiver of consent is not permitted. This will impede or even obviate emergency research phase III trials in TBI in the European countries (Kompanje and Maas, 2004; Silverman et al., 2004; Liddell et al., 2006a, b). We aimed to analyze the implications of critical time frames (time between injury and admission to a neurotrauma center (NTC), between admission and the ﬁrst cranial CT scan, between CT scan and

proxy consent (PC), and between PC and study drug administration (SDA)) in a recently completed RCT in TBI.

Materials and methods We analyzed critical time frames in a multi-center placebo-controlled phase III trial, investigating the efﬁcacy and safety of a single dose of dexanabinol in severe TBI. This RCT was conducted from January 2001 to March 2004 in Europe, Israel, Australia, and the United States. Dexanabinol is a synthetic cannabinoid analogue with strong neuroprotective potential and devoid of psychotropic activity. The protocol stipulated SDA within 6 h after injury. The study recruited 861 patients with severe TBI. No beneﬁcial effect of dexanabinol was found. Full details and results of the study have been published (Maas et al., 2006). In total, 7164 patients were screened for participation in the trial. Enrollment criteria were not met in 6303 patients. Of these, the sole reason for exclusion was the time window in 671 patients and in a further 944 patients inability to give study medication within 6 h was one of the reasons for exclusion (Fig. 1).

Time window as main reason for exclusion

4688 5000 4000 3000 944

671

2000 1000 0

patients

time window as only reason for exclusion

time window + other reason for exclusion

other reason for exclusion

671

944

4688

Fig. 1. Reasons for exclusion (n ¼ 6303).

245

In view of the severity of the brain injury, informed consent could not be obtained from the patients themselves. PC was accepted in all countries. Deferred consent was allowed in Australia, Austria, Finland, and in some centers in France and Germany. Consent by an independent physician was allowed in Israel, Italy, Spain, and the United Kingdom. In all cases of deferred consent, subsequent written assent by patient or proxy was obtained. For the analysis of time frames we selected patients enrolled in Europe and Israel, in whom PC was obtained before SDA (n ¼ 631). We excluded patients from the United States for reason of

incomplete screening logs due to HIPAA regulations (Maas et al., 2005; Kompanje and Maas, 2006), and patients from Australia seeing the different infrastructure of this country in comparison to European countries.

Results The median age of our study population was 32 years (IQR 23–44); 520 (82.4%) were male, 111 (17.6%) female. The time between injury and admission to the NTC varied between countries from 1.16 to 2.35 h (Table 1, Fig. 2). A total of 501

Table 1. Time windows per country (median+IQR) Country (N)

Hours between injury and admission NTC median (IQR)

Hours between injury and CT scan median (IQR)

Hours between injury and obtained consent median (IQR)

Hours between injury and SDA median (IQR)

Belgium (23) Netherlands (73) Israel (116) Spain (75) Germany (109) Italy (146) France (34) Other countriesa (55)

0.93 1.00 0.93 1.33 1.20 1.25 2.17 1.47

1.80 1.65 1.91 2.07 1.65 1.77 3.08 1.82

3.75 4.53 4.01 4.17 4.08 4.92 5.00 4.00

4.60 5.53 4.67 5.17 5.25 5.50 5.75 5.25

a

(0.65–1.27) (0.75–1.33) (0.72–1.40) (0.97–1.67) (0.88–2.00) (0.83–2.60) (1.42–3.00) (1.00–2.67)

(1.28–2.27) (1.32–2.00) (1.58–2.47) (1.65–2.53) (1.30–2.13) (1.40–2.35) (1.97–3.53) (1.33–2.50)

(2.75–4.75) (3.95–5.05) (3.20–4.83) (3.33–5.00) (3.42–4.98) (4.08–5.28) (4.50–5.38) (3.08–4.75)

(3.98–5.42) (5.07–5.75) (4.00–5.33) (4.30–5.58) (4.25–5.67) (4.98–5.75) (5.17–5.83) (4.33–5.75)

Countries with small patient populations (United Kingdom, Denmark, Austria, Poland, and Turkey) were combined.

other france italy germany spain israel netherlands belgium injury

1h

2h

3h

4h

5h

6h

time between injury and admission NTC time between admission and ct-scan time between ct-scan and obtained consent time between consent and SDA Fig. 2. Time between injury and admission neurotrauma center, time between admission and ﬁrst CT scan, time between ﬁrst CT scan and informed consent for inclusion in trial and time between consent and start study drug admission.

246 379

400

number of patients

350 300

informed consent before inclusion, and a risk–beneﬁt ratio based on the concept that in relation to the severity of the trauma, signiﬁcant adverse side effects may be acceptable for treatments with proven beneﬁt (Kompanje et al., 2005).

250 168

200

The emergency nature of research and short therapeutic time windows in TBI

150 70

100 50

1

13

0 1-2hours 2-3hours 3-4hours 4-5hours >5hours time between injury and study drug admission Fig. 3. Distribution of number of patients. Time between injury and study drug administration (n ¼ 631).

(79.4%) patients were directly admitted to the NTC, and 130 (20.6%) concerned secondary referrals. In 71 secondarily referred patients the CT scan was made before admission to the NTC. The time between admission and the ﬁrst diagnostic CT scan was within 1 h in all patients. With the exception of France, in all countries the median time between injury and CT scan was within 3 h (Table 1). The broadest time window was found between the admission CT scan and obtaining the required PC (between 1.71 and 2.74 h). The median time between injury and obtained PC varied between 3.75 and 5.00 h (IQR 2.75–5.38 h) (Table 1). After consent had been given, almost all patients subsequently received the study drug within 1 h (Table 1, Fig. 2). In 85.3% of all cases the time between injury and SDA exceeded 4 h, in 60% of the cases it even exceeded 5 h (Fig. 3). In total, 139 patients were randomized with deferred consent. The median time between injury and SDA in this group was 4.75 h (IQR 3.90–5.67). These 139 patients are not included in our analysis. Discussion Speciﬁc ethical issues pertaining to clinical evaluation of neuroprotective agents in TBI include the emergency nature of the research, short therapeutic windows, the incapacity of the patients to

Severe TBI is an emergent and life-threatening condition and existing therapy is unsatisfactory given the high morbidity and mortality. Neurological damage does not only occur at the moment of impact, but can evolve over the following hours and days. Deleterious effects of this progressive damage are determined at clinical and biochemical levels. This has led to the development of new pharmaceuticals with promising potential to limit secondary damage and to improve outcome. One of these new pharmaceuticals was dexanabinol, but efﬁcacy in the treatment of severe TBI was not demonstrated (Maas et al., 2006). Recruitment in this study was relatively slow and the majority of patients enrolled toward the end of the 6 h time window stipulated in the protocol. We hypothesized that the relatively slow recruitment and lack of demonstrable efﬁcacy in the clinical situation may have been inﬂuenced by informed consent procedures and resulting delays in SDA. Early intervention would appear crucial to the effect of neuroprotective agents. Experimental data have consistently shown better protection the sooner an agent is administered after TBI (Hoff, 1986). For dexanabinol, experimental studies have shown beneﬁcial effects with administration within 3 h after injury, demonstrating protection against breakdown of the blood-brain barrier, reduction of edema formation, and improved outcome (Shohami, 1995). No signiﬁcant reduction of cerebral edema was noted if the drug was administered between 4 and 6 h after injury, but some improvement in neurological symptoms was found. Based on these ﬁndings, it may be concluded that in the experimental model the pathophysiologic time window can be determined at 3 h. Whether this experimental time window may be extrapolated to the clinical situation in patients with TBI remains

247

uncertain. Prespeciﬁed subgroup (patients who received the SD within 4 h and patients who received the SD after 4 h) analysis showed no signiﬁcant differential treatment effect (Maas et al., 2006). Nonetheless, it may be concluded in general that chances of efﬁcacy are greater if treatment is provided earlier. We found that in almost all of the studied cases the time between injury and completion of the primary diagnostic CT scan remained within 3 h post injury, which corresponds to the therapeutic time window in the animal model. In 60% of the cases, however, the time between injury and SDA was more than 5 h, and in 85.3% of all cases more than 4 h (Fig. 3). We have shown that the main determinant of the time to SDA is formed by the time required to obtain informed (proxy) consent, and these results indicate that delays in SDA could be reduced by 50% with the adoption of deferred consent or waiver of consent. We could, however, not demonstrate shorter inclusion times in the cohort of patients randomized with deferred consent. The reason for the delay in this cohort was that most investigators waited for PC before randomization, and only used deferred consent when at the end of the inclusion boundary. These observations favor adopting deferred consent procedures in trials in acute TBI as primary approach, rather than as ‘‘ultimum refugium’’, only to be undertaken if PC cannot be obtained. Other studies have demonstrated advantages in using the deferred consent and waiver of consent in emergency research. In the National Acute Brain Injury Study: Hypothermia (NABIS-H) the adoption of waiver of consent resulted in higher enrollment and reduced the time between injury and treatment by approximately 45 min (Clifton et al., 2002). In this study, relatives of only 11 out of 113 patients arrived within 6 h after the injury. In a septic shock trial the investigators could not contact the proxies within the inclusion time in 74% of the cases, and these were included under waiver of consent (Annane et al., 2004). In the CRASH trial, mean time to randomization was signiﬁcantly longer in those hospitals where consent was required compared with those it was not (4.4 h [SE ¼ 0.21] versus 3.2 h [SE ¼ 0.16]), the difference in the mean time to randomization was 1.2 h [95% CI 0.7–1.8 h]

(CRASH trial management group, 2004). The observation in the cohort of patients enrolled in the dexanabinol trial with deferred or waiver of consent, that (proxy) assent was obtained later in all cases supports the concept of accepting deferred or waiver of consent for emergency research in TBI trials.

The acute incapacity of the patients and informed consent All patients with severe TBI are unconscious, and consequently informed consent can never be obtained from the patients themselves. As substitute for informed consent, a legal representative must give consent for inclusion in research, or in the absence thereof, by proxies; alternatively consent may be deferred or waived. We have argued that deferred consent is preferable. However, the new European Union Directive (2001/20/EC) (European Union, 2001; Liddell et al., 2006a, b) stipulates a requirement for informed (proxy) consent. This is motivated by respect for the autonomy of patients, and to ensure that that patient’s wishes are guaranteed as far as possible. The requirement for PC assumes that relatives are available in emergency situations, and that these relatives can be fully informed and given sufﬁcient time to make a balanced, ethically valid decision in a relatively short time period under emotional distress. Even when proxies are available, many are not aware of the patient’s wishes (Luce, 2003). Surrogate decision makers for critical care research resulted in falsepositive consent rates of 16–20.3% (Coppolino and Ackerson, 2001). The emotional nature of an emergency situation limits the reliability of PC for clinical research (Mason and Allmark, 2000; Coppolino and Ackerson, 2001; Hsieh et al., 2001). Under emergency circumstances, PC does not seem to secure proper patient/subject protection. In our experience, the validity of informed consent and PC given in an emergency situation is at least troubling. When consent for clinical research is sought during an emergency situation, comprehension is generally less than optimal (Cuttini, 2000; Sugarman, 2000; Williams et al., 2003). A small minority realizes that pharmacological

248

trials are designed to assess not only efﬁcacy but safety as well (Harth and Thong, 1995). Are patients willing to be represented by their close relatives? Roupie et al. (2000) found that only 40.6% of 1089 patients wanted their spouse/partner to be their surrogate, 28% wanted to be represented by the physician in charge of their care. One study searching for public views on emergency exception to informed consent found that most of 530 people (88%) believed that research subjects should be informed prior to being enrolled, while 49% believed enrolling patients without prior consent would be acceptable in an emergency situation and 70% (369) would not object to be entered into such a study without providing prospective informed consent (McClure et al., 2003). In another study 11 of 12 stroke patients stated that, if the patient or family was not able to consent, the treating physician should make the decision for inclusion in an emergency trial (Blixen and Agich, 2005). Furthermore, the requirement for all patients to give written informed consent before enrollment can result in major selection biases, such that enrolled patients may not be representative of the typical patient (Tu et al., 2004). Although respect for a patient’s autonomy is a guiding principle in human rights, we doubt very much whether this can be guaranteed by mandatory PC in acute TBI research. A balance should be sought between optimal patient protection and research to advance the standards of clinical care in emergency situations to the beneﬁt of society.

Risk– benefit ratio and patient protection In our opinion the balance between risk and beneﬁt should be the guiding principle in emergency research in severe TBI. This also applies to the nature and the type of consent procedures. The ethical principle of respect for the autonomy of the patient underpinning the informed consent procedures is not valid for acutely incapacitated patients as TBI victims (Kompanje et al., 2005). Signiﬁcant concerns have been raised on the validity and ethics of PC in acute emergency situations, and the required written consent causes a signiﬁcant delay

in treatment initiation, as we have shown by our analysis. The possible therapeutic beneﬁt, with short therapeutic time windows in experimental models, forms the moral justiﬁcation for randomizing patients under deferred consent or waiver of consent within a sufﬁcient period of time. The risks should however be acceptable in relation to the severity of the disease or injury. For an effective agent in life-threatening condition adverse effects may be expected and accepted in some patients. Careful monitoring and followup of such adverse events is mandatory. For trials under deferred consent or waiver of consent in acute emergency situations we propose to institute an independent safety committee, under the auspices of regulatory authorities. Such an independent safety committee, without (ﬁnancial) ties to industry or investigators, offers the best safeguard for patient protection. The obligation to such a committee is based on the experience of a dramatically harmful outcome in some trials under waiver of consent in other ﬁelds of medicine (Freeman, 2001; Lewis et al., 2001). Clinical research in emergency situations without prospective informed or PC is ethically challenging. Severe TBI is without doubt an emergent and life-threatening condition and existing therapy is unsatisfactory. This should qualify severe TBI as emergency exception to informed consent or deferred consent for randomized clinical controlled trials involving pharmacological agents with promising therapeutic beneﬁt facing short therapeutic time windows. Randomized placebo-controlled investigations are necessary to determine the safety and efﬁcacy of new developed agents under these circumstances. The requirement for prior written PC causes signiﬁcant delays in SDA. With deferred consent or waiver of consent the ﬁrst dose of the experimental drug can be administered directly after completion of the ﬁrst diagnostic CT scan, which is very close to the experimental therapeutic time window. Randomized controlled phase III trials investigating the safety and efﬁcacy of agents with promising beneﬁt, conducted in acute emergency situations with short therapeutic time windows, should allow randomization under deferred consent or waiver of consent. This is ethically defendable, off course

249

with proper safeguard from an independent safety committee. Making progress in knowledge of treatment in acute neurological and other intensive care conditions is only possible if national regulations and legislations allow waiver of consent or deferred consent for clinical trials (Lemaire, 2005). As two of us have said before: ‘treat first, ask later’ is ethically defendable in emergency and intensive care medicine research (Kompanje and Maas, 2004).

References Annane, D., Outlin, H., Fisch, C. and Bellissant, E. (2004) The effect of waiving consent on enrolment in a sepsis trial. Intensive Care Med., 30: 632–637. Blixen, C.E. and Agich, G.J. (2005) Stroke patients preferences and values about emergency research. J. Med. Ethics, 31: 608–611. Clifton, G.L., Knudson, P. and McDonald, M. (2002) Waiver of consent in studies of acute brain injury. J. Neurotrauma, 19: 1121–1126. Coppolino, M. and Ackerson, L. (2001) Do surrogate decision makers provide accurate consent for intensive care research? Chest, 119: 603–612. CRASH trial management group. (2004) Research in emergency situations: with or without relatives consent. Emerg. Med. J., 21: 703. Cuttini, M. (2000) Proxy informed consent in pediatric research: a review. Early Hum. Dev., 60: 89–100. European Union. (2001) Directive 2001/20/EC of the European Parliament and the Council of 4 April 2001 on the approximation of the laws, regulations and administrative provisions of the Member States relating to the implementation of good clinical practice in the conduct of clinical trials on medicinal products for human use. Off. J. Eur. Comm., L121/34. Freeman, D.B. (2001) Safeguarding patients in clinical trials with high mortality rates. Am. J. Crit. Care Med., 164: 190–192. Harth, S.C. and Thong, Y.H. (1995) Parental perceptions and attitudes about informed consent in clinical research involving children. Soc. Sci. Med., 40: 1573–1577. Hoff, J.T. (1986) Cerebral protection. J. Neurosurg., 65: 579–591. Hsieh, M., Dailey, M.W. and Callaway, C.W. (2001) Surrogate consent by family members for out-of-hospital cardiac arrest research. Acad. Emerg. Med., 8: 851–853. Kompanje, E.J.O., Maas, A.I.R., Hilhorst, M.T., Slieker, F.J.A. and Teasdale, G.M. (2005) Ethical considerations on consent procedures for emergency research in severe and moderate traumatic brain injury. Acta Neurochir. (Wien.), 147: 633–640.

Kompanje, E.J.O. and Maas, A.I.R. (2004) ‘Treat ﬁrst, ask later?’ Emergency research in acute neurology and neurotraumatology in the European Union. Intensive Care Med., 30: 168–169. Kompanje, E.J.O. and Maas, A.I.R. (2006) Is the Glasgow Coma Scale score protected health information? The effect of new United States regulations (HIPAA) on completion of screening logs in emergency research trials. Intensive Care Med., 32: 313–314. Lemaire, F. (2005) Waiving consent for emergency research. Eur. J. Clin. Invest., 35: 287–289. Lewis, R.J., Berry, D.A., Cryer, H., Fost, N., Krome, R., Washington, G.R., Houghton, J., Blue, J.W., Bechhofer, R., Cook, T. and Fisher, M. (2001) Monitoring a clinical trial conducted under the Food and Drug Administration regulations allowing a waiver of prospective informed consent: the diaspirin cross-linked hemoglobin traumatic hemorrhagic shock efﬁcacy trial. Ann. Emerg. Med., 38: 397–404. Liddell, K., Chamberlain, D., Menon, D.K., Bion, J., Kompanje, E.J.O., Lemaire, F., Druml, C., Vrhovac, B., Wiedermann, C.J. and Sterz, F. (2006a) The European Clinical Trial Directive revisited: the VISEAR recommendations. Resuscitation, 69: 9–14. Liddell, K., Kompanje, E.J.O., Lemaire, F., Vrhovac, B., Menon, D.K., Bion, J., Chamberlain, D., Wiedermann, C.J. and Druml, C. (2006b) Recommendations in relation to the EU Clinical Trials Directive and medical research involving incapacitated adults. Wien. Klin. Wochenschr., 118: 183–191. Luce, J.M. (2003) Is the concept of informed consent applicable to clinical research involving critically ill patients? Crit. Care Med., 31(Suppl): S153–S160. Maas, A.I.R., Kompanje, E.J.O., Slieker, F.J.A. and Stocchetti, N. (2005) Differences in completion of screening logs between Europe and the United States in an Emergency phase III trial resulting from HIPAA requirements. Ann. Surg., 241(2): 382–383. Maas, A.I.R., Marmarou, A., Murray, G.D. and Steyerberg, E.W. (2004) Clinical trials in traumatic brain injury: current problems and future solutions. Acta Neurochir., 89: 113–118. Maas, A.I.R., Murray, G., Henney, H., Kassem, N., Legrand, V., Mangelus, M., Muizelaar, J.P., Stocchetti, N., Knoller, N. and on behalf of the Pharmos TBI Investigators. (2006) Efﬁcacy and safety of dexanabinol in sever traumatic brain injury: results of a phase III randomized placebo controlled clinical trial. Lancet Neurol., 5: 38–45. Maas, A.I.R., Steyerberg, E.W. and Murray, G.D. (1999) Why have recent trials of neuroprotective agents in head injury failed to show convincing efﬁcacy? A pragmatic analysis and theoretical considerations. Neurosurgery, 44: 1286–1298. Mason, S.A. and Allmark, P.J. (2000) Obtaining informed consent to neonatal randomised controlled trials: interviews with parents and clinicians in the Euricon study. Lancet, 356: 2045–2051. McClure, K.B., Delorio, N.M., Gunnels, M.D., Ochsner, M.J., Biros, M.H. and Schmidt, T.A. (2003) Attitudes of emergency department patients and visitors regarding emergency exception from informed consent in resuscitation research,

250 community consultation and public notiﬁcation. Acad. Emerg. Med., 10: 352–359. McGarry, L.J., Thompson, D., Millham, F.H., Cowell, L., Snyder, P.J., Lenderking, W.R. and Weinstein, M.C. (2002) Outcomes and costs of acute treatment of traumatic brain injury. J. Trauma, 53: 1152–1159. Roupie, E., Santin, A., Boulme, R., Wartel, J.S., Lepage, E., Lemaire, F., Lejonc, J.L. and Montagne, O. (2000) Patients preferences concerning medical information and surrogacy: results of a prospective study in a French emergency department. Intensive Care Med., 26: 52–56. Shohami, E. (1995) Long-term effect of HU-211, a novel competitive NMDA antagonist, on motor and memory functions after closed head injury in the rat. Brain Res., 674: 55–62.

Silverman, H.J., Druml, C., Lemaire, F. and Nelson, R. (2004) The European Union Directive and the protection of incapacitated subjects in research: an ethical analysis. Intensive Care Med., 30: 1723–1729. Sugarman, J. (2000) Is the emperor really wearing new clothes? Informed consent for acute coronary syndromes. Am. Heart J., 140: 2–3. Tu, J.V., Willison, D.J., Silver, F.L., Fang, J., Richards, J.A., Laupacis, A. and Kapral, M.K. (2004) Impracticability of informed consent in the registry of the Canadian stroke network. NEJM, 350: 1414–1421. Williams, B.F., French, J.K. and White, H.D. (2003) Informed consent during the clinical emergency of acute myocardial infarction (HERO-2 consent substudy): a prospective observational study. Lancet, 361: 918–922.

SECTION V

Emerging Topics in CNS Trauma

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 18

Experimental models of repetitive brain injuries John T. Weber Department of Neuroscience, Erasmus Medical Centre, Rotterdam, The Netherlands

Abstract: Repetitive traumatic brain injury (TBI) occurs in a signiﬁcant portion of trauma patients, especially in speciﬁc populations, such as child abuse victims or athletes involved in contact sports (e.g. boxing, football, hockey, and soccer). A continually emerging hypothesis is that repeated mild injuries may cause cumulative damage to the brain, resulting in long-term cognitive dysfunction. The growing attention to this hypothesis is reﬂected in several recent experimental studies of repeated mild TBI in vivo. These reports generally demonstrate cellular and cognitive dysfunction after repetitive injury using rodent TBI models. In some cases, data suggests that the effects of a second mild TBI may be synergistic, rather than additive. In addition, some studies have found increases in cellular markers associated with Alzheimer’s disease after repeated mild injuries, which demonstrates a direct experimental link between repetitive TBI and neurodegenerative disease. To complement the ﬁndings from humans and in vivo experimentation, my laboratory group has investigated the effects of repeated trauma in cultured brain cells using a model of stretch-induced mechanical injury in vitro. In these studies, hippocampal cells exhibited cumulative damage when mild stretch injuries were repeated at either 1-h or 24-h intervals. Interestingly, the extent of damage to the cells was dependent on the time between repeated injuries. Also, a very low level of stretch, which produced no cell damage on its own, induced cell damage when it was repeated several times at a short interval (every 2 min). Although direct comparisons to the clinical situation are difﬁcult, these types of repetitive, low-level, mechanical stresses may be similar to the insults received by certain athletes, such as boxers, or hockey and soccer players. This type of in vitro model could provide a reliable system in which to study the mechanisms underlying cellular dysfunction following repeated injuries. As this area of TBI research continues to evolve, it will be imperative that models of repetitive injury replicate injuries in humans as closely as possible. For example, it will be important to model appropriately concussive episodes versus even lower level injuries (such as those that might occur during boxing matches). Suitable interinjury intervals will also be important parameters to incorporate into models. Additionally, it will be crucial to design and utilize proper controls, which can be more challenging than experimental approaches to single mild TBI. It will also be essential to combine, and compare, data derived from in vitro experiments with those conducted with animals in vivo. These issues, as well as a summary of ﬁndings from repeated TBI research, are discussed in this review. Keywords: hippocampus; in vivo models; in vitro; mild injury; repeated injury; TBI; trauma

Introduction

Corresponding author. Present address: School of Pharmacy,

Memorial University of Newfoundland, Health Sciences Centre, St. John’s, NL, Canada. Tel.: +1-709-777-7022; Fax: +1-709-777-7044; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61018-2

Repetitive injuries occur in a considerable portion of individuals who experience a traumatic brain 253

254

injury (TBI). Child abuse victims, and in particular, victims of shaken baby syndrome, are often subjected to multiple injuries to the head (Shannon et al., 1998). Spousal abuse also contributes to the number of repeated brain injuries (Roberts et al., 1990). However, many injuries of these types go unreported, and it is difﬁcult to assess how many insults a patient may have suffered. Arguably, athletes represent the largest group of patients that are at risk for experiencing repeated brain injuries (Kelly and Rosenberg, 1997; Kelly, 1999; Powell and Barber-Foss, 1999). In fact, sports and recreational activities contribute to approximately 10% of all mild injuries, and these individuals are at high risk of experiencing a second TBI (Woo and Thoidis, 2000). Also, in comparison to child or spousal abuse victims, there is generally better documentation of how many brain injuries an individual has sustained due to recreational or sports related activities, making this population easier to study. The idea that multiple head injuries in athletes could lead to clinical problems has long been suggested. For example, the development of dementia pugilistica in professional boxers is believed to be caused by the multiple hits to the head that boxers endure over the course of their career (Jordan, 2000). Also, recent evidence has shown that the number of concussions is inversely related to performance on several neuropsychological tests in soccer players (Matser et al., 1999, 2001) and jockeys who have experienced multiple concussions generally display more cognitive dysfunctions than those who have had a single injury (Wall et al., 2006). When studying repetitive brain trauma in athletes, we can gain much information about the pathology and progress of such injuries from the injured athletes themselves, for example by measuring changes in cognitive and motor performance. However, these injuries are generally at a mild level, and therefore, except in rare cases when athletes die as a result of the insult, we cannot assess the changes that have occurred in the brain at the cellular and sub-cellular levels. In order to compile this type of information, we must turn to experimental models of TBI.

Experimental studies of repeated mild TBI in vivo It should be made clear that when discussing experimental studies of repetitive TBI in vivo, this does not include studies of secondary insults, such as a mechanical insult to the head followed by a duration of ischemia or glutamate exposure. Repeated TBI experimentation can be deﬁned as an initial mechanical injury to the head followed by another mechanical insult to the head of the same or different degree. Based on these criteria, little attention was given to these types of experiments before the year 2000, with only a few repetitive injury studies being published (Olsson et al., 1971; Weitbrecht and Noetzel, 1976; Kanayama et al., 1996). Several additional in vivo studies of repeated injuries in rodents have now been conducted in the past decade (Allen et al., 2000; Laurer et al., 2001; DeFord et al., 2002; Uryu et al., 2002; Conte et al., 2004; Creeley et al., 2004; Raghupathi et al., 2004; Dobrowolska and Gibson, 2005; Gibson and Schalles, 2005; Longhi et al., 2005; Yoshiyama et al., 2005). All of these repeated mild injury studies were conducted using rodent models of TBI with the exception of the study by Raghupathi et al. (2004), which reported a pediatric model of repeated injury conducted in pigs. Repetitive TBI generally occurs at a mild level, therefore experimental models have been used which are minimally invasive and do not require a craniotomy, such as weight drop models. The models must also be administered at a level that produces minimal, or preferably, no fatality. Individuals who have suffered from a mild TBI often complain of cognitive difﬁculties post-injury. Therefore, repeated injury studies usually evaluate cognitive function, for example using the Morris water maze (MWM) test, as well as the extent of cell abnormalities in the cortex and hippocampus. The hippocampus in particular has received signiﬁcant attention in the study of repeated mild TBI, because it plays a critical role in certain types of learning and aspects of memory storage. Experimental and clinical data have demonstrated not only the importance of this brain region in learning and memory, but also that the hippocampus is uniquely vulnerable to injury, even after

255

mild brain trauma (Lyeth et al., 1990; Lowenstein et al., 1992). In a study by DeFord et al. (2002), repeated mild injuries were administered to mice (four times every 24 h), followed by MWM testing and histological analysis. Signiﬁcant learning deﬁcits were found after repeated injuries, which were not evident after a single injury. These deﬁcits occurred even in absence of cell death within the cortex and hippocampus. These studies were extended by Gibson and Schalles (2005) using the same injury model; they demonstrated impaired function in the Barnes circular maze after multiple, but not single injuries. Cognitive deﬁcits after multiple mild TBIs (using MWM analysis) have also been demonstrated in a similar study using a weight drop model (Creeley et al., 2004). In order to most closely mimic the type of insult that athletes may receive, Laurer et al. (2001) developed an injury regimen that they described as ‘‘concussive’’; this model was used for subsequent studies as well (Uryu et al., 2002; Conte et al., 2004; Longhi et al., 2005). In an assessment of cognitive and motor function after repeated injury in mice, Laurer et al. (2001) found that the brain was more vulnerable to a second insult if the second injury occurred 24 h after the ﬁrst. Even though no cognitive deﬁcits were demonstrated in mice receiving repeated injuries, there was a decrease in motor function and some neuronal loss. The authors also stated that the effects of a second mild TBI may be synergistic, rather than additive. To analyze further the effects of lengthening the inter-injury interval, Longhi et al. (2005) investigated repetitive injuries three, ﬁve, and seven days apart. Animals that received repeated injuries three or ﬁve days apart exhibited cognitive dysfunction not evident in sham animals or those injured only once. However, no deﬁcits were observed when the injury interval was extended to seven days. This experimental evidence — that the brain can recover from a ﬁrst injury, given a sufﬁcient amount of time — is appealing, especially in relation to establishing ‘‘return-toplay’’ guidelines for athletes. Overall, the evidence from in vivo experimental models suggests that repetitive mild TBI causes more cognitive and cellular dysfunction than a single injury, if the brain is not given a sufﬁcient amount of time to recover.

Repetitive injury and neurodegenerative disease It has been noted for some time that there is a correlation between the occurrence of TBI and the further development of neurodegenerative disease later in life. In fact, TBI is considered to be one of the most robust risk factors for developing Alzheimer’s disease (AD) (Szczygielski et al., 2005). There is also evidence that genetic predisposition may increase one’s risk of developing AD, such as possession of the apolipoprotein E e4 allele. A phenomenon known as chronic TBI occurs in a signiﬁcant amount of professional boxers (Jordan, 2000), with the most serious form, dementia pugilistica, resulting in severe motor and cognitive dysfunctions. It is now known that the pathology of AD and dementia pugilistica are quite similar (Geddes et al., 1999; Schmidt et al., 2001). Although the epidemiological data are quite strong, very little has been done to make the mechanistic link between repeated mild TBI and the development of either AD or dementia pugilistica. A few studies have found increases in cellular markers associated with AD after repeated mild injuries (Kanayama et al., 1996; Uryu et al., 2002; Conte et al., 2004; Dobrowolska and Gibson, 2005). For example, Kanayama et al. (1996) showed an increase in tau immunoreactivity in neurons and Dobrowolska and Gibson (2005) reported increased amyloid precursor protein following multiple injuries. Uryu et al. (2002) and Conte et al. (2004) used a transgenic mouse model of AD-like amyloidosis, and demonstrated that amyloid levels and deposits increased in the brain after repeated injuries, but not after a single insult. Yoshiyama et al. (2005) utilized the same transgenic mouse model and analyzed the effects of four injuries given per day, once a week, for a period of four weeks, in order to simulate dementia pugilistica. After nine months, only one mouse showed pathology consistent with this syndrome. Although the ﬁndings were not robust, this is perhaps the only study to attempt to model dementia pugilistica, and more studies are warranted. The overall ﬁndings of these studies are quite important, because they can demonstrate a direct

256

experimental link between repeated mild TBI and the development of AD-like pathology. Generally, it takes years before the onset of symptoms of neurodegenerative disorders after an individual has experienced a TBI. Therefore, it requires an extremely long amount of time to gather this type of epidemiological data from the human population. This area of research, in particular, is where experimental models could truly help decipher the mechanisms by which neurodegenerative disease may be triggered by repetitive brain injury.

Studying repeated injuries in vitro Several in vitro approaches have now been developed to study traumatic injury, which utilize cells grown in culture (Morrison et al., 1998; Weber, 2004; LaPlaca et al., 2005; also see Chapter 3 by Spaethling et al., in this volume). My laboratory group utilizes a model of stretch-induced mechanical injury in vitro originally developed by Ellis et al. (1995). We have characterized this stretch injury model in cell cultures composed of neurons and glia from murine hippocampus (Slemmer et al., 2002; Slemmer and Weber, 2005), cortex (Engel et al., 2005), and cerebellum (Slemmer et al., 2004). To complement the ﬁndings from humans and from in vivo experimentation, we have recently conducted a series of studies investigating the effects of repeated trauma on hippocampal cells (Slemmer et al., 2002; Slemmer and Weber, 2005). In these studies, we utilized a mild level of stretch injury that produces some measurable damage to cells when administered a single time. When mild stretch injuries were repeated at either 1-h or 24-h intervals, cells exhibited cumulative damage. For example, cultures that received a second insult displayed a signiﬁcant loss of neurons not evident in cultures that received only one injury. Additionally, cultures injured twice released a signiﬁcant level of neuron speciﬁc enolase (NSE), which was not observed in cultures injured a single time. Interestingly, the extent of damage to the cells was dependent on the time between repeated injuries. For example, cultures that received a second insult 1 h after the ﬁrst injury released more S-100B

protein (a biomarker of injury commonly employed in the clinic) than cultures that received a second injury at 24 h. Cultures injured 24 h apart also exhibited less staining with the intravital dye, propidium iodide, than those injured 1 h apart. As shown in some in vivo studies, these ﬁndings suggest that a level of injury, which produces measurable damage or dysfunction on its own, may cause cumulative damage if repeated within a certain time frame (Laurer et al., 2001; Longhi et al., 2005). We also investigated the effects of a very low level of stretch, which produced no overt cell damage (Slemmer and Weber, 2005). This ‘‘subthreshold’’ level of stretch did not cause signiﬁcant damage or death, even when it was repeated at a 1-h interval. However, this low level of stretch did induce cell damage when it was repeated several times at a short interval (every 2 min), indicated by increased propidium iodide staining, neuronal loss, and an increase in NSE release. Although direct comparisons to the clinical situation are difﬁcult, these types of repetitive, low-level, mechanical stresses may be similar to the insults received by certain athletes, such as boxers and hockey and soccer players (Matser et al., 1998, 1999; Jordan, 2000; Webbe and Ochs, 2003; Wennberg and Tator, 2003). This type of in vitro model could comprise a reliable system in which to study the mechanisms underlying cellular dysfunction following repeated injuries. In addition, this approach could provide a means for relatively rapid screening of potential therapeutic strategies for both single and repeated mild TBI.

The phenomenon of preconditioning Several studies have indicated that an initial, very mild insult to either cultured cells or to the brain itself, may provide some protection from a second, more severe insult, a ﬁnding that has been termed ‘‘preconditioning’’. Ischemic preconditioning, in which a brief exposure to ischemia renders the brain more resistant to subsequent longer periods of ischemia, has been well described (for review, see Schaller and Graf, 2002). There is also evidence of preconditioning cross-tolerance. For example,

257

brief ischemia lessens damage following TBI in vivo (Pe´rez-Pinzo´n et al., 1999). Another interesting phenomenon is that heat acclimation (chronic exposure to moderate heat) can also provide resistance to TBI (see Chapter 25 by Shein et al., in this volume). In our in vitro studies, we observed a novel form of mechanical preconditioning. When hippocampal cultures were administered a subthreshold level of stretch 24 h prior to a mild stretch, there was a signiﬁcant decrease in released S-100B protein compared to cultures that were injured at a mild level alone (Slemmer and Weber, 2005). This observation suggests some form of protection initiated by this low level of stretch. A similar ﬁnding in vivo was reported by Allen et al. (2000). In their study, rats received a series of mild injuries spaced three days apart using a weight drop model. Some of these animals received a severe injury after the repetitive mild injuries. Motor function deﬁcits were evident in severely injured animals, but not in animals that received repeated mild injuries or repeated mild injuries followed by a severe injury. This last observation suggests a preconditioning effect. The most important question now is how do we utilize this information? One can imagine the ethical implications of suggesting to people that a mild insult to their brains may in fact protect them from worse insults in the future. We still have much to learn about preconditioning. For example, what is the threshold for mechanical insults between initiating protective versus damaging mechanisms in the brain? A clearer understanding of the mechanisms by which this protection is elicited holds potential for the management of mild TBI. The fact that different stressors can protect the brain from TBI (i.e. cross-tolerance) suggests that the same or similar mechanisms are responsible for the endogenous protection. Increasing the expression of these protective systems could not only be a reliable way for managing mild TBI, but also could provide resistance in individuals who may be at risk of sustaining an additional head injury, such as athletes. Both in vivo and in vitro models could provide reliable systems in which to study the mechanisms underlying the preconditioning phenomenon.

Future directions Experimental evidence suggests that animals can demonstrate cognitive deﬁcits and cellular dysfunction after repetitive mild TBI, even though the injury may not necessarily lead to cell death (Kanayama et al., 1996; DeFord et al., 2002). Therefore, rather than trying to prevent cells from dying after repeated injuries, it is probably more important to learn how to restore normal cellular physiology after a traumatic episode. Combining studies at the cellular and behavioral levels is essential, and one particular area of interest is the evaluation of the effects of repeated TBI on synaptic plasticity in the hippocampus. This ability of neurons to undergo changes in synaptic strength, such as long-term potentiation (LTP), is postulated to be a cellular correlate of learning and memory (Bliss and Collingridge, 1993; Malenka and Nicoll, 1999). Several studies have reported impaired hippocampal LTP after TBI in vivo (Albensi, 2001; Weber, 2004). One area of future research could focus on restoring mechanisms of synaptic plasticity after injury (such as LTP), as well as correlated hippocampal-mediated behavioral tasks. The hippocampus shares neuronal projections with areas of the cerebral cortex, which undoubtedly also contributes to memory formation and storage. Indeed, alterations in synaptic plasticity may also occur directly in the cortex after repeated mild TBI. Therefore, although the hippocampus may play a central role in the cognitive dysfunction observed after mild TBI, it is important not to overlook contributions from other brain areas as well. Along these lines, since some repeated injury studies demonstrate motor dysfunction, it may also be appropriate to investigate cellular physiology and synaptic plasticity in the cerebellum (Hansel et al., 2001; Weber et al., 2003; Slemmer et al., 2005) after repetitive TBI. The utilization of proper parameters for repeated injury studies may be even more crucial than deciding on appropriate research directions. For example, what are the best inter-injury interval, or intervals, to use? Although 24 h between injuries is the most common (and perhaps practical) interval in the laboratory (Weitbrecht and

258

Noetzel, 1976; Laurer et al., 2001; DeFord et al., 2002; Uryu et al., 2002; Conte et al., 2004; Creeley et al., 2004; Dobrowolska and Gibson, 2005; Gibson and Schalles, 2005; Yoshiyama et al., 2005), is it the most appropriate in mimicking what occurs in humans? Also, how many injuries should a researcher administer? If one is attempting to model concussive episodes, then two, three, or four may be enough, as this may closely mimic a true situation, especially with athletes. However, when attempting to recreate dementia pugilistica

(Yoshiyama et al., 2005), the number of injuries should certainly be increased. What are the proper controls and endpoints to use for repeated injury studies? For in vivo studies analyzing the effects of a single TBI, the issue of controls is fairly straightforward. Sham animals are treated at an equivalent time as injured animals, and the analysis, cellular or behavioral, is also performed at the same timepoint. However, when comparing uninjured animals to animals that have received more than one injury, what is

Fig. 1. Issues for consideration when designing repeated TBI experiments (i.e. choosing proper timepoints for controls and behavioral/ tissue analysis). (A) Hypothetical timeline for two injuries 24 h apart. (B) Timeline for greater than two injuries. T ¼ time.

259

the proper comparison? For example, if an animal receives an injury on day one, an additional injury on day two, and analysis takes place on day three, does one compare the data with sham animals from day one, or from day two (or both, see Fig. 1A). The issue is further complicated when comparing repeatedly injured animals to animals that have received a single TBI. If the comparison concerns animals that undergo four injuries or a single injury, are the single insult animals injured at the same time as injury one in the repeated group, or at the same time as the fourth injury (see Fig. 1B)? This decision will also affect the endpoint as well. For example, if animals or tissue are analyzed one day after the fourth injury, then four days will have passed for the single injury group if those animals were injured on day one. This difference in time could affect the observations. We have struggled with these issues when designing our in vitro experiments. One could argue that if a long enough period of time passes after the injuries, such as weeks or months, then the effect of when the single insult animals were injured will be negligible. Admittedly, this would be more proper for comparison to the human situation in which the effects of mild TBI can be manifested for weeks, months, or even years. However, this is often not practical for many laboratories, as the costs of housing animals for months can at times be prohibitive. Also, conducting long-term experiments in vitro is limited, since the cells generally remain viable for only a few weeks. This raises a critical point as to the relevance of repeated injury studies in vitro. I strongly believe that in vitro experiments can deliver information about the cellular mechanisms of repeated injury that are difﬁcult to obtain in vivo, and that it is essential to combine data derived from in vitro experiments with those conducted with animals in vivo. However, I am unsure how to directly compare the data. For example, is a 24 h injury interval in vitro equivalent to 24 h in vivo? The more consensus that exists on these issues with individuals who conduct repeated injury studies, the easier it will be to compare the data, and the stronger a case can be made for showing unequivocally, that repeated mild TBI could lead to long-term dysfunction in humans.

Abbreviations AD LTP MWM NSE TBI

Alzheimer’s disease long-term potentiation Morris water maze neuron speciﬁc enolase traumatic brain injury

Acknowledgments The in vitro work discussed in this manuscript was supported by grants from the Research Council for Earth and Life Sciences (ALW) and Research Council for Medical Sciences (MW) with ﬁnancial aid from the Netherlands Organization for Scientiﬁc Research (NWO), a Breedtestrategie subsidie from Erasmus Medical Centre, and funding from Hersenstichting Nederland. I would also like to thank Jennifer E. Slemmer for editorial assistance.

References Albensi, B.C. (2001) Models of brain injury and alterations in synaptic plasticity. J. Neurosci. Res., 65: 279–283. Allen, G.V., Gerami, D. and Esser, M.J. (2000) Conditioning effects of repetitive mild neurotrauma on motor function in an animal model of focal brain injury. Neuroscience, 99: 93–105. Bliss, T.V. and Collingridge, G.L. (1993) A synaptic model of memory: long-term potentiation in the hippocampus. Nature, 361: 31–39. Conte, V., Uryu, K., Fujimoto, S., Yao, Y., Rokach, J., Longhi, L., Trojanowski, J.Q., Lee, V.M.-Y., McIntosh, T.K. and Pratico, D. (2004) Vitamin E reduces amyloidosis and improves cognitive function in Tg2576 mice following repetitive concussive brain injury. J. Neurochem., 90: 758–764. Creeley, C.E., Wozniak, D.F., Bayly, P.V., Olney, J.W. and Lewis, L.M. (2004) Multiple episodes of mild traumatic brain injury result in impaired cognitive performance in mice. Acad. Emerg. Med., 11: 809–819. DeFord, S.M., Wilson, M.S., Rice, A.C., Clausen, T., Rice, L.K., Barabnova, A., Bullock, R. and Hamm, R.J. (2002) Repeated mild brain injuries result in cognitive impairment in B6C3F1 mice. J. Neurotrauma, 19: 427–438. Dobrowolska, J.A. and Gibson, C.J. (2005) Immunohistochemical analysis of amyloid precursor protein in the hippocampus and cortex of mice following repeated mild traumatic brain injuries. Soc. Neurosci. Abst., 31: Program Number 334.7.

260 Ellis, E.F., McKinney, J.S., Willoughby, K.A., Liang, S. and Povlishock, J.T. (1995) A new model for rapid stretchinduced injury of cells in culture: characterization of the model using astrocytes. J. Neurotrauma, 12: 325–339. Engel, D.C., Slemmer, J.E., Vlug, A.S., Maas, A.I. and Weber, J.T. (2005) Combined effects of mechanical and ischemic injury to cortical cells: secondary ischemia increases damage and decreases effects of neuroprotective agents. Neuropharmacology, 49: 985–995. Geddes, J.F., Vowles, G.H., Nicoll, J.A. and Revesz, T. (1999) Neuronal cytoskeletal changes are an early consequence of repetitive head injury. Acta Neuropathol. (Berl.), 98: 171–178. Gibson, C.J. and Schalles, M.D. (2005) Barnes circular maze deﬁcits in repeated mildly brain injured mice. Soc. Neurosci. Abst., 31: Program Number 334.6. Hansel, C., Linden, D.J. and D’Angelo, E. (2001) Beyond parallel ﬁber LTD: the diversity of synaptic and non-synaptic plasticity in the cerebellum. Nat. Neurosci., 4: 467–475. Jordan, B.D. (2000) Chronic traumatic brain injury associated with boxing. Semin. Neurol., 20: 179–185. Kanayama, G., Takeda, M., Niigawa, H., Ikura, Y., Tamii, H., Taniguchi, N., Kudo, T., Miyamae, Y., Morihara, T. and Nishimura, T. (1996) The effects of repetitive mild brain injury on cytoskeletal protein and behavior. Methods Find. Exp. Clin. Pharmacol., 18: 105–115. Kelly, J.P. (1999) Traumatic brain injury and concussion in sports. JAMA, 282: 989–991. Kelly, J.P. and Rosenberg, J.H. (1997) Diagnosis and management of concussion in sports. Neurology, 48: 575–580. LaPlaca, M.C., Cullen, D.K., McLoughlin, J.J. and Cargill II, R.S. (2005) High rate shear strain of three-dimensional neural cell cultures: a new in vitro traumatic brain injury model. J. Biomech., 38: 1093–1105. Laurer, H.L., Bareyre, F.M., Lee, V.M., Trojanowski, J.Q., Longhi, L., Hoover, R., Saatman, K.E., Raghupathi, R., Hoshino, S., Grady, S.M. and McIntosh, T.K. (2001) Mild head injury increasing the brain’s vulnerability to a second concussive impact. J. Neurosurg., 95: 859–870. Longhi, L., Saatman, K.E., Fujimoto, S., Raghupathi, R., Meaney, D.F., Davis, J., McMillan, B.S.A., Conte, V., Laurer, H.L., Stein, S., Stocchetti, N. and McIntosh, T.K. (2005) Temporal window of vulnerability to repetitive experimental concussive brain injury. Neurosurgery, 56: 364–374. Lowenstein, D.H., Thomas, M.J., Smith, D.H. and McIntosh, T.K. (1992) Selective vulnerability of dentate hilar neurons following traumatic brain injury: a potential mechanistic link between head trauma and disorders of the hippocampus. J. Neurosci., 12: 4846–4853. Lyeth, B.G., Jenkins, L.W., Hamm, R.J., Dixon, C.E., Phillips, L.L., Clifton, G.L., Young, H.F. and Hayes, R.L. (1990) Prolonged memory impairment in the absence of hippocampal cell death following traumatic brain injury in the rat. Brain Res., 526: 249–258. Malenka, R.C. and Nicoll, R.A. (1999) Long-term potentiation — a decade of progress? Science, 285: 1870–1874.

Matser, E.J.T., Kessels, A.G.H., Lezak, M.D., Jordan, B.D. and Troost, J. (1999) Neuropsychological impairment in amateur soccer players. JAMA, 282: 971–973. Matser, J.T., Kessels, A.G.H., Jordan, B.D., Lezak, M.D. and Troost, J. (1998) Chronic traumatic brain injury in professional soccer players. Neurology, 51: 791–796. Matser, J.T., Kessels, A.G.H., Lezak, M.D. and Troost, J. (2001) A dose-response relation of headers and concussions with cognitive impairment in professional soccer players. J. Clin. Exp. Neuropsychol., 23: 770–774. Morrison III, B., Saatman, K.E., Meaney, D.F. and McIntosh, T.K. (1998) In vitro central nervous system models of mechanically induced trauma: a review. J. Neurotrauma, 15: 911–928. Olsson, Y., Rinder, L., Lindgren, S. and Stalhammar, D. (1971) Studies on vascular permeability changes in experimental brain concussion. 3. A comparison between the effects of single and repeated sudden mechanical loading of the brain. Acta Neuropathol. (Berl.), 19: 225–233. Pe´rez-Pinzo´n, M.A., Alonso, O., Kraydieh, S. and Dietrich, W.D. (1999) Induction of tolerance against traumatic brain injury by ischemic preconditioning. Neuroreport, 10: 2951–2954. Powell, J.W. and Barber-Foss, K.D. (1999) Traumatic brain injury in high school athletes. JAMA, 282: 958–963. Raghupathi, R., Mehr, M.F., Helfaer, M.A. and Margulies, S.S. (2004) Traumatic axonal injury is exacerbated following repetitive closed head injury in the neonatal pig. J. Neurotrauma, 21: 307–316. Roberts, G.W., Whitwell, H.L., Acland, P.R. and Bruton, C.J. (1990) Dementia in a punch-drunk wife. Lancet, 335: 918–919. Schaller, B. and Graf, R. (2002) Cerebral ischemic preconditioning. An experimental phenomenon or a clinical important entity of stroke prevention? J. Neurol., 249: 1503–1511. Schmidt, M.L., Zhukareva, V., Newell, K.L., Lee, V.M.-Y. and Trojanowski, J.Q. (2001) Tau isoform proﬁle and phosphorylation state in dementia pugilistica recapitulate Alzheimer’s disease. Acta Neuropathol. (Berl.), 101: 518–524. Shannon, P., Smith, C.R., Deck, J., Ang, L.C., Ho, M. and Becker, L. (1998) Axonal injury and the neuropathology of shaken baby syndrome. Acta Neuropathol. (Berl.), 95: 625–631. Slemmer, J.E., De Zeeuw, C.I. and Weber, J.T. (2005) Don’t get too excited: mechanisms of glutamate-mediated Purkinje cell death. Prog. Brain Res., 148: 367–390. Slemmer, J.E., Matser, E.J.T., De Zeeuw, C.I. and Weber, J.T. (2002) Repeated mild injury causes cumulative damage to hippocampal cells. Brain, 125: 2699–2709. Slemmer, J.E. and Weber, J.T. (2005) The extent of damage following repeated injury to hippocampal cells is dependent on the severity of insult and inter-injury interval. Neurobiol. Dis., 18: 421–431. Slemmer, J.E., Weber, J.T. and De Zeeuw, C.I. (2004) Cell death, glial protein alterations and elevated S-100b protein in cerebellar cell cultures following mechanically induced trauma. Neurobiol. Dis., 15: 563–572.

261 Szczygielski, J., Mautes, A., Steudel, W.I., Falkai, P., Bayer, T.A. and Wirths, O. (2005) Traumatic brain injury: cause or risk of Alzheimer’s disease? A review of experimental studies. J. Neural Transm., 112: 1547–1564. Uryu, K., Laurer, H., McIntosh, T., Pratico, D., Martinez, D., Leight, S., Lee, V.M.-Y. and Trojanowski, J.Q. (2002) Repetitive mild brain trauma accelerates Ab deposition, lipid peroxidation, and cognitive impairment in a transgenic mouse model of Alzheimer amyloidosis. J. Neurosci., 22: 446–454. Wall, S.E., Williams, W.H., Cartwright-Hatton, S., Kelly, T.P., Murray, J., Murray, M., Owen, A. and Turner, M. (2006) Neuropsychological dysfunction following repeat concussions in jockeys. J. Neurol. Neurosurg. Psychiatry, 77: 518–520. Webbe, F.M. and Ochs, S.R. (2003) Recency and frequency of soccer heading interact to decrease neurocognitive performance. App. Neurol., 10: 31–41. Weber, J.T. (2004) Calcium homeostasis following traumatic neuronal injury. Curr. Neurovasc. Res., 1: 151–171. Weber, J.T., De Zeeuw, C.I., Linden, D.J. and Hansel, C. (2003) Long-term depression of climbing ﬁber-evoked

calcium transients in Purkinje cell dendrites. Proc. Natl. Acad. Sci. U.S.A., 100: 2878–2883. Weitbrecht, W.U. and Noetzel, H. (1976) Autoradiographic investigations in repeated experimental brain concussion (author’s transl.) [Article in German]. Arch. Psychiatr. Nervenkr., 223: 59–68. Wennberg, R.A. and Tator, C.H. (2003) National Hockey League reported concussions, 1986–87 and 2001–02. Can. J. Neurol. Sci., 30: 206–209. Woo, B.H. and Thoidis, G. (2000) Epidemiology of traumatic brain injury. In: Woo B.H. and Nesathurai S. (Eds.), The Rehabilitation of People with Traumatic Brain Injury. Blackwell Science, Malden, pp. 13–17. Yoshiyama, Y., Uryu, K., Higuchi, M., Longhi, L., Hoover, R., Fujimoto, S., McIntosh, T., Lee, V.M.-Y. and Trojanowski, J.Q. (2005) Enhanced neuroﬁbrillary tangle formation, cerebral atrophy, and cognitive deﬁcits induced by repetitive mild brain injury in a transgenic tauopathy mouse model. J. Neurotrauma, 22: 1134–1141.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 19

Minor traumatic brain injury in sports: a review in order to prevent neurological sequelae Nicola Biasca1, and William L. Maxwell2 1

Clinic of Orthopaedic, Sports Medicine and Traumatology, Department of Surgery, Spital Oberengadin, CH-7503 Samedan/St. Moritz, Switzerland 2 Institute of Biomedical and Life Sciences, University of Glasgow, Glasgow G12 8QQ, UK

Abstract: Minor traumatic brain injury (mTBI) is caused by inertial effects, which induce sudden rotation and acceleration forces to and within the brain. At less severe levels of injury, for example in mTBI, there is probably only transient disturbance of ionic homeostasis with short-term, temporary disturbance of brain function. With increased levels of severity, however, studies in animal models of TBI and in humans have demonstrated focal intra-axonal alterations within the subaxolemmal, neuroﬁlament and microtubular cytoskeletal network together with impairment of axoplasmic transport. These changes have, until very recently, been thought to lead to progressive axonal swelling, axonal detachment or even cell death over a period of hours or days, the so-called process of ‘‘secondary axotomy’’. However, recent evidence has suggested that there may be two discrete pathologies that may develop in injured nerve ﬁbers. In the TBI scenario, disturbances of ionic homeostasis, acute metabolic changes and alterations in cerebral blood ﬂow compromise the ability of neurons to function and render cells of the brain increasingly vulnerable to the development of pathology. In ice hockey, current return-to-play guidelines do not take into account these new ﬁndings appropriately, for example allow returning to play in the same game. It has recently been hypothesized that the processes summarized above may predispose brain cells to assume a vulnerable state for an unknown period after mild injury (mTBI). Therefore, we recommend that any confused player with or without amnesia should be taken off the ice and not be permitted to play again for at least 72 h. Keywords: minor traumatic brain injury, mTBI; neuromechanics; neuropathology; neurobiology and associated neurometabolic changes of mTBI; process of delayed axotomy; vulnerability of cells in the brain; return to play guidelines; ice hockey

nervous system. This is especially true in ice hockey, which has many inherent features that may predispose players to injury, including high acceleration–deceleration, rapid changes of direction, shooting, body checking and a low friction ice surface. Before the introduction of full-face masks and helmets, head injuries, including injuries to the face, scalp and brain, accounted for at least 50% of all serious injuries in ice hockey (Pashby, 1993). Indeed, before the use of any type of helmet and before the mandatory use of

Introduction Ice hockey is one of the speediest, competitive contact team sports involving both rapid changes of direction of the whole body and controlled aggression. Athletes are therefore exposed to an increased risk of a high-velocity collision that may result in a sports-related injury to the central Corresponding author. Tel.: +41-81-851-85-15; Fax: +41-81-

851-85-16; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61019-4

263

264

standardized helmets, in the early 1960s, fatalities secondary to head injuries have been reported in Sweden, Canada and the United States (Pashby, 1993). Since 1963, however, there have been no fatalities from head injury reported in Sweden (Odelgard, 1989). However, an alarming and persistently high rate of minor traumatic brain injuries or mTBI, often termed cerebral concussion worldwide over the last decade and a half has been reported (Clayton, 1997; Dick, 1997; Biasca, 2000; Tegner, 2000; www.hockeyinjuries). Although the majority of cases of traumatic brain injury (TBI) do occur after Motor Vehicle Accidents and falls, it is notable that sport activities are responsible for approximately 15–25% of all cases (Gennarelli, 1991; McGehee, 1996). Recent evidence, obtained over the last 15 years, has suggested that mTBI may be more common and more serious than previously thought worldwide (Denny and Russel, 1941; Gennarelli, 1991; Castaldi, 1993; Tegner and Lorentzon, 1996; Clayton, 1997; Dick, 1997; Gennarelli et al., 1998; Biasca et al., 1999; Roberts et al., 1999; Biasca, 2000; Ranalli, 2000; Tegner, 2000, 2001; www.hockeyinjuries; Burke, 2001; Laprade, 2001; Pashby et al., 2001). According to the Center for Disease Control in the United States, there are more than 300,000 sports-related mTBI each year and certainly this number underestimates the true level of incidence, because many of these injuries are not reported (MMWR, 1997). There has long been controversy concerning the structural basis of mTBIs, like concussion. However, the importance of acceleration forces was already established by the early 1940s (Denny and Russel, 1941; Holbourn, 1943; Gennarelli, 1991; Elson and Ward, 1994). Historically, many investigators of the early 1990s were of the belief that mTBI of the brain could occur without evidence of cellular or vascular lesions of the cerebrum, for example a number of studies have been cited by Denny and Russel (1941), Gennarelli (1991) and Elson and Ward (1994). Further, areas of petechiae, in the absence of other lesions, have been reported after fatal brain injury (Elson and Ward, 1994). Holbourn (1943) stated that the relative motion between brain constituents is insignificant when

the brain is exposed to a linear acceleration, but not so in angular acceleration. This set the stage for further inquiry into the effects of acceleration versus impact trauma to the head (Denny and Russel, 1941; Holbourn, 1943; Gennarelli, 1991; Elson and Ward, 1994). It is now generally accepted that in patients who experience blunt head injury, nerve ﬁbers/axons are particularly susceptible to damage. It is also now well recognized that diffuse axonal injury (DAI) is a consistent feature of severe human TBI, particularly those injuries involving rapid acceleration/deceleration of the brain (Denny and Russel, 1941; Gennarelli, 1991; Gennarelli et al., 1998; Kelly, 1999; Graham et al., 2000). Recently, there have been considerable advances in our understanding of the nature and time course of axonal injury (AI) after TBI (Graham et al., 2000). There is also increasing evidence that an mTBI may represent the mildest form of a diffuse brain injury. The descriptor ‘‘minor Traumatic Brain Injury’’ (mTBI) has been used to refer to such an injury (Gennarelli et al., 1998; Kelly, 1999; Graham et al., 2000). The purpose of this manuscript is to review recent studies in the epidemiology, neuromechanics, neuropathology, neurobiology and associated neurometabolic changes in mTBI in order to understand why, following a ‘‘mild’’ insult to the head, brain cells are more susceptible and are more ‘‘vulnerable’’ to a second and perhaps more damaging accident.

Epidemiology of mTBI Injuries to the head constitute a significant proportion of the total number of injuries in many, different contact sports (Kelly et al., 1991; Cantu, 1992; Biasca et al., 1995; Tegner and Lorentzon, 1996; Clayton, 1997; Dick, 1997; Roberts et al., 1999; Biasca, 2000; Tegner, 2000; Tegner et al., 2000; Laprade, 2001; Pashby et al., 2001; SportsRelated Recurrent Brain Injuries–United States, 1997). National and international analyses report that the proportion of mTBI to the overall number of injuries ﬂuctuates in American football between 3 and 24%, and in soccer between 4 and 22% (Castaldi, 1993; Tegner and Lorentzon, 1996;

265

Clayton, 1997; Dick, 1997; Gennarelli et al., 1998; Biasca et al., 1999; Roberts et al., 1999; Biasca, 2000; Ranalli, 2000; Tegner, 2000, 2001; Tegner et al., 2000; Burke, 2001; Laprade, 2001; Pashby et al., 2001). In ice hockey, more specifically, the proportion of mTBI to the overall number of injuries ﬂuctuates between 2 and 20%. More detailed analysis has also shown that the frequency of mTBI has risen markedly over the last 15 years. In the Canadian Hockey League (CHL) Clayton (1997) reported that the proportion of mTBI rose from 4% of all injuries in the period 1991–1996, to 8% in 1997, 14% in 1998 and 1999 and 17% in the period 1999–2000. The National Collegiate Athletic Association (NCAA) Injury Surveillance System recorded a similar increase in North America Junior Hockey where the incidence of mTBI almost doubled between 1986 (0.7 injuries per 1000 AE in 1986) and 1996 (1.5 injuries per 1000 AE in 1996) (Dick, 1997). Similarly, in a study in Scandinavia, one Swedish ice hockey team reported a rising incidence of mTBI over the last 15 years from 2% of all injuries in the season 1986–1987 to 18% in the season 2000–2001 (Tegner, 2000; Tegner et al., 2000). In parallel, in the professional Ice Hockey

League (NHL), John Powell (2001) reported an increased incidence of mTBI from 2% in the season 1989–1990, to 4.9% in 1995–1996, to 8% in each of the last two Seasons 1999–2001 for which data is available (www.hockeyinjuries). Lastly, mTBI represented 20% of all head injuries in the Swiss National Ice Hockey Leagues A and B during the seasons 1996–1998 (Fig. 1) (Biasca, 2000). However, the diagnosis of an mTBI is subject to widespread controversy both in definition and the degree of severity of injury. This controversy has contributed to the current paucity of reliable epidemiological data. It has recently been suggested that it would be helpful to have a worldwide-standardized assessment for mTBI and a centralized data pool using the International Sports Injury System ‘‘ISIS’’. Such would greatly facilitate international comparison through use of an Internet-based database system (Biasca and Tegner, 2001).

Neuromechanics of mTBI The most common mechanical input to the head is dynamic loading of either impact or impulsive type

Fig. 1. Percent of mTBI, i.e. cerebral concussion, to all injuries in Sweden, Canada, North America and Switzerland from 1986 to 2001 (Clayton, 1997; Biasca et al., 2000; Tegner et al., 2000; Tegner, 2000; www.hockeyinjuries).

266

(Elson and Ward, 1994; Gennarelli, 1991; Bailes and Cantu, 2001; Graham et al., 2000). Impact loading requires a direct blow to the head occurring in a period of less than 200 ms (Elson and Ward, 1994; Gennarelli, 1991) and, in most cases, in less than 20 ms (Graham and Gennarelli, 1997). It occurs when a blunt object strikes the head and usually initiates a combination of contact and inertial forces that result in a series of events. The forces vary with the size of the impacting object and the magnitude of the force delivered to the contact point, the so-called contact phenomena (Elson and Ward, 1994; Graham et al., 2000). These contact phenomena are a complex group of mechanical events that occur both locally and distant from the point of the impact. They may cause laceration of the scalp, skull deformations, fractures, cerebral contusions and propagation of shock waves that travel through the skull and brain. Protective headgear is designed to reduce the severity of these injuries (i.e., skull deformation, fracture and hematoma). Impulsive loading, in contrast, occurs when the head is put into motion and accelerated as a result of either an impact to another part of the body, or as a result of a secondary response to a direct impact. Impulsive loading occurs also when the moving head is stopped without it striking anything or is arrested by impact (Elson and Ward, 1994; Gennarelli, 1991; Graham et al., 2000). Under these conditions the resulting head injuries are solely caused by inertia resulting from the manner (translation, rotation, angulation) in which the head is moved. In biomechanical terms this results in shearing stresses within the brain substance, resulting in the incidence of foci of strain. Strain is the amount of deformation of brain or other tissue as a result of the application of mechanical force. There are three types of strain: compression, tension and shear. The brain is highly incompressible but will deform in shear most easily when subjected to rotational loads (reviewed by Margulies and Thibault, 1989). Individual axons can sustain only tensile strain, which is the amount of elongation that occurs when a material is stretched. Biological materials withstand strain if they are deformed slowly. But in head-injury impulsive and inertial loading, giving rise to dynamic strain, is

both more frequent and more damaging to the substance of the brain because the substance of the brain is exposed to mechanical loading over a very short time span — 2–20 ms — and has a very low tolerance of tensile and shear strain. Thus, the highest levels of damage to the brain result from exposure to shearing stresses where individual axons are stretched. Inertial forces are responsible for the two most important types of damage encountered in blunt head injury: First, acute subdural hematoma (SDH) resulting from tearing of subdural bridging veins. Such injury occurs when the rise time and duration of inertial loading to the head is relatively short over 2–12 ms (Gennarelli and Thibault, 1982). SDH occurs most frequently in patients that have experienced a fall or been exposed to an assault (Graham and Gennarelli, 1997). Second, widespread damage in the white matter resulting in diffuse axonal injury DAI (Elson and Ward, 1994; Gennarelli, 1991; Graham et al., 2000) where inertial injury to the head occurs over a longer time span — 11–22 ms (Gennarelli et al., 1982). Adams et al. (1989) deﬁned DAI and distinguished three levels or grades of severity of injury to white matter — grade I in which there is widespread damage to axons within the cerebral hemispheres; grade II in which there is, in addition to the above, the occurrence of focal lesions in the corpus callosum often related to small hemorrhages (petechia) or tissue tears and grade III in which there is also damage to axons within the rostral brainstem. The brain, which is housed in the protective bony cranium and bathed by a layer of cerebrospinal ﬂuid (CSF), has freedom of movement before it abuts the cranium (i.e., smooth intracranial surfaces, CSF interfaces). The CSF provides a natural shock-absorbing system, converting focally applied external stress to compressive stress, following the contours of the sulci and gyri, and distributing the force in a uniform fashion. The CSF, however, does not provide complete protection against shearing forces being induced in the brain, especially when rotational forces are applied to the head and shearing forces occur at those sites in the brain where rotational gliding is hindered — for

267

example by the relatively rigid membranes forming the falx cerebri and the tentorium cerebelli. The incidence of DAI is greatest when the brain is exposed to rotational forces arising during a horizontal/coronal displacement of the head. Inertial effects result in the skull moving laterally before the brain does and the brain may continue to move after the head has stopped moving. The falx and tentorium are ﬁrmly adherent to the internal surface of the skull. The brain, to the contrary, ﬂoats in CSF. The mid-line falx will therefore move within a different time frame to the brain. The medial side of one cerebral hemisphere may contact the surface of the falx while the other hemisphere continues moving. Structures that cross the midline, for example nerve ﬁbers in the corpus callosum will therefore be exposed to tensile strain as one hemisphere moves away from the other. Thus nerve ﬁbers passing through the corpus callosum or through the cerebral peduncles have a greater risk of being exposed to tensile strain and thus to injury. In the last several years a widely held consensus has developed that there is a spectrum of levels of injury to nerve ﬁbers when they are exposed to tensile strain. When an athlete or patient experiences either mild concussion — (disturbance of neurological function without loss of consciousness), classical cerebral concussion — (a temporary, reversible loss of neurological function with temporary loss of consciousness for up to 6 h) or more severe head injury when the patient is unconscious for more than 6 h — is now thought that nerve ﬁbers are injured in all of the above clinical scenarios. The distinction between the above levels of injury is thought to reﬂect the number and distribution of injured ﬁbers within the white matter of the brain. The less severe forms of injury in which only a small number of ﬁbers are injured are now thought to be the principal causes of an mTBI (Elson and Ward, 1994; Gennarelli et al., 1998; Kelly, 1999; Graham et al., 2000). The situation is further complicated in that an inertial loading of the brain and head, whether caused by an impact or by an impulsive loading is inﬂuenced by the direction of movement of the head during the injury episode. If the acceleration causes a straight movement of the center of the head, then we have a linear acceleration; if the center of gravity of the

head travels in an arc then we have a rotational acceleration (Bishop, 1997, 2000; Bishop and Arnold, 1993; Bailes and Cantu, 2001). Focal brain injuries are more common after a linear acceleration and diffuse brain injuries after a rotational acceleration (Bishop, 1997, 2000; Bishop and Arnold, 1993). However, in humans, almost any injury event involves a combination of these two and separation of brain responses due to rotation and linear acceleration is almost impossible (Bishop, 1997, 2000; Bishop and Arnold, 1993). Often the mechanism of a head injury is complex involving a combination of focal and diffuse components. Direct head contact may induce both a rotational acceleration to cause both diffuse and focal injury (Bishop, 1997, 2000; Bishop and Arnold, 1993; Gennarelli et al., 1998; Kelly, 1999; Graham et al., 2000). Further, the direction in which the head moves plays an important role in determining the amount and distribution of axonal damage in a given situation (Gennarelli et al., 1998). mTBI in ice hockey and correlation with the type of blow to the head Specifically in relation to ice hockey, following examination of 11 cases of mTBI in the professional NHL, in the Canadian Amateur Hockey League and in the Swiss Ice Hockey League, it was proposed that three different possible mechanisms may be responsible for causing mTBI in ice hockey (Bishop, 2001): 1. a direct eccentric blow to the head, 2. a direct blow to the face and 3. a blow directed to the chin. A direct eccentric blow to the head, not passing through the center of mass of the head, may expose brain tissue to many forces. These may be simpliﬁed as a force Ft with translatory components (with transversal and axial forces) and a force Fr with rotatory components in the coronal, sagittal and transverse (horizontal) planes (Fig. 2). The rotatory component in the coronal plane, after a blow to the side of the head, causes the most severe injuries within the brain while a blow in the

268

Fig. 2. Example of a direct eccentric blow to the head: a protective headgear, with its padding, helps to reduce the effect of the ﬁnal force F and therefore may reduce the risk of a focal head injury. (Mt: Moving force in the transversal plane; Ms: moving force in the sagittal plane; Mf: moving force in the frontal plane; Ft1: force 1 in the transversal plane; Ft2: force 2 in the transversal plane; Fa: force in the axial plane.) Reprinted with permission and courtesy of the player.

sagittal plane (ﬂexion/extension of the cervical vertebral joints) is tolerated best (Gennarelli et al., 1998). Protective headgear including padding, helps to reduce the overall effect of the ﬁnal force F, and also reduces the contributory effects of contact and rotational phenomenon. Hockey helmets were originally developed to reduce the risk of serious head injuries, precipitated by direct blunt traumas and provided protection against brain injury death and/or intracranial hematoma (Odelgard, 1989; Bishop, 1997, 2000; Bishop and Arnold, 1993; Dixon and Brodie, 1993; Pashby, 1993; Stoner and Kreating, 1993; Pashby et al., 2001; Tegner, 2001). The two most critical functions of helmets are to reduce impact energy and load distribution to the head. These properties reduce the magnitude of the forces applied to the head, reduce the stress and strain to which the skull and brain are exposed and thus reduce the severity of mTBI and head injury. Thus, protective headgear ﬁrst of all reduces the risk of a focal head

Fig. 3. Example of a direct blow to the face or jaw: There is no protective equipment, such as helmets and mouth guards, which may be effective in reducing any rotatory component. (Mt: Moving force in the transversal plane; Mf: moving force in the frontal plane; Ft: force in the transversal plane.) Reprinted with permission and courtesy of the player.

injury, but although development of padding materials has resulted in improvement of the energy absorption characteristics of helmets, the crucial factor causing mTBI is widely acknowledged to be the effect of rotational acceleration of the brain. A direct blow to the face may cause both a force Ft with translatory and force Fr with rotatory components. In ice hockey contacts to the jaw and face, a similar mechanism to the knockout punch in boxing, are numerous. As a result there is often a significant rotational acceleration of the head and brain (Fig. 3). Gennarelli et al. (1982) found that the severity of experimental TBI in a non-human primate appeared to be related to the plane of rotational acceleration. For equivalent levels of angular acceleration, nerve ﬁbers in the white matter of the brain were most vulnerable to injury if the head was rotated laterally (vide supra). When the brain moved laterally tensile strain occurred in a coronal plane parallel to the long axis of nerve ﬁbers crossing the sagittal plane. Lesser degrees of injury were obtained when the brain moved either in the sagittal or horizontal planes (Gennarelli et al., 1982). Recently Smith et al. (2000) have

269

suggested that the severity of experimental TBI may be more of a reﬂection of the severity of axonal damage in speciﬁc regions of the brain, most notably the brainstem, rather than the total sum of AI distributed throughout the organ (Smith et al., 2000). Following rotation of the head in the axial plane, substantially more axonal damage was found in the brainstem, compared with that incurred when the head was rotated in the coronal plane (Smith et al., 2000). But care needs to be taken in the interpretation of these results and their comparability with injury in the human. In the pig the brain stem lies in the same approximate longitudinal axial plane as the cerebrum. In the human, however, the brain stem lies at an angle of some 1201 to the cerebrum. Thus, in the pig an axial injury which occurs in the horizontal plane of the head (Fig. 1 in Smith et al., 2000) will provide the greatest mechanical strain to nerve ﬁbers at the lateral limits of the brain stem. Indeed, Smith et al. (2000) illustrate such data in Fig. 4 of their paper. In humans, however, because of the different planes of orientation of the longitudinal axes of the cerebrum and brain stem mechanical strain will be placed largely on those nerve ﬁbers lying parallel to the direction of displacement of the head. That is in coronal and parasagittal ﬁbers within the cerebral hemispheres. Nonetheless, Smith et al. (2000) are the ﬁrst to suggest (1) a relationship between the plane of head rotational acceleration and the distribution of AI and (2) that injury to axons in the brainstem plays a major role in the induction of an immediate post-traumatic coma. With speciﬁc regard to sports-related injuries, no protective equipment, which may reduce the effects of this rotatory force (Fr), is presently available. For this reason it is necessary to eliminate the practice by players of checking against the head and neck of other players and which may thereby expose the brain to rotatory acceleration/deceleration. The Rules Committee of the International Ice Hockey Federation IIHF, under the chairmanship of Mr. Philippe Lacarrie`re, has therefore introduced in 2002 new ‘‘Head-Checking Rules’’ (IIHF Rule No. 540: http://www.iihf.com// hockey/rules/rules.htm) in order to eliminate every intentional or unintentional check or blow to the head and neck. This rule has been made

Fig. 4. Example of a direct blow to the chin: mouth guards may have the ability to absorb the impact loading from the chin and to distribute the remaining energy, throughout the resilience of its material combined with its design, over a much larger surface area, thereby reducing the ﬁnal force on the brain. (Ft: Force in the transversal plane.) Reprinted with permission and courtesy of the player.

mandatory, worldwide since June 2002. This means that every player making contact to the head and neck with the body, elbow, shoulder, knee or stick of an opposing player will be penalized, at the discretion of the referee, with either a Minor (2 min) or Major (5 min) Penalty plus automatic Game Misconduct or even Match Penalty. Prevention strategies, such as the introduction of ‘‘Checking from behind’’ Rules in 1994, have become effective during the last years in decreasing the number of severe spinal injuries worldwide. Therefore, it is hoped and expected that these new ‘‘Head-Checking’’ rules should reduce the incidence and severity of mTBI. A blow directed to the chin may cause a translatory force Ft, which is transmitted from the chin through the lower jaw, through the temporomandibular joint at the base of the skull and then to the brain. These forces may also propagate shock waves through the brain, which may induce mTBIs (Fig. 4). Mouth guards alone or in conjunction with additional face protection, have been demonstrated

270

to be effective in preventing and/or reducing the incidence of dental and orofacial injuries (Piccininni, 2001). Numerous anecdotal reports have indicated a correlation between the use of mouth guards and prevention of mTBI. Piccininni (2001) focused on two primary mechanisms to explain this correlation. The ﬁrst mechanism was dissipation of forces delivered to the maxilla, skull and temporomandibular joint complex when the mandible received a blow from the front whereby the force applied to the mandible is transmitted to both temporomandibular joints, absorbed in part by the articular disc and ﬁbrous capsule of those joints before being transmitted to the temporal bone and the base of the skull. The second mechanism was the stabilization of the skull through the increased contraction of the muscles of mastication, principally the massetur and temporalis muscles, when clenching the lower jaw. This may be enhanced with the presence of a mouth guard (Piccininni, 2001). For this reason Piccininni (2001) stated that mouth guards can and should be promoted as effective devices for the prevention of dental and orofacial injuries and that properly designed mouth guards may also play a role in the reduction of incidence or severity of mTBI. However, evidence of any injury protection using mouth guards is derived from case series and retrospective injury surveys. Further epidemiological research using the ISIS will surely contribute to an increased understanding of the incidence and mechanism of any injury and allow a reduction of the risk of injury. The use of properly ﬁtted mouth guards would no doubt reap numerous beneﬁts in term of reducing medical, ﬁnancial, cognitive, psychological and social consequences of any dental and cerebral injuries, like mTBI, at a minimal cost. As a result, the Rules Committee of the IIHF has made in 2002 some rule changes. This rule will be mandatory, as a ﬁrst step, for all junior ice hockey players younger than 20 years and not wearing full-face masks. As a next step, this rule should be mandatory for all players born after December 31, 1975 and for all players wearing full-face masks as well, because a full-face mask alone can not reduce the forces transmitted to the brain from any blow to the jaw or chin (Biasca et al., 2002).

Neuropathology of mTBI The pathology of human TBI is complex and heterogeneous, comprising either single or a combination of focal and diffuse lesions (Gennarelli, 1991; Elson and Ward, 1994; Graham et al., 2000; Tegner, 2001). Trauma to the skull and brain manifests in a variety of pathological hallmarks reﬂecting the distribution of the impact energy. Sharp objects at high velocity tend to cause perforating focal cortical contusions (FCC). Collisions against sharp edges lead to non-penetrating, mostly focal head injuries (FHI). Blunt trauma — where the head does not necessarily hit anything but is exposed to rapid acceleration and deceleration — often leads to diffuse trauma, that is DAI. FCC, FHI and DAI often occur together at any combination of lesion severity (Elson and Ward, 1994; Graham et al., 2000). FHI are caused by forces of contact and head acceleration from direct blunt trauma, such as being struck with a hockey stick or a puck, or falling on the ice surface and striking the head. These types of injuries may produce a focal cortical contusion with well-demarcated damage to the parenchyma of the brain and associated disruption of arterial and venous vessels in the region of the contusion. Such gross lesions may readily be observed in CT scans, MRI or macroscopically at autopsy. FHI often result in fracture of the skull, with or without an associated epidural hematoma, in cortical contusions and in intracranial hemorrhage (Gennarelli, 1991; Bishop, 1997; Bishop and Arnold, 1993; Elson and Ward, 1994; Kelly, 1999; Graham et al., 2000; Tegner, 2001). Tissue damage occurs either by direct mechanical destruction of brain parenchyma, compression through interstitial bleeding or by secondary infarction following disrupted circulation. Wearing an internationally approved, standardized helmet will clearly reduce the number and severity of these focal injuries (Bishop, 1997; Bishop and Arnold, 1993). Diffuse head injuries, on the other hand, are caused by the inertial effect of either a mechanical blow to the head or rapid rotation acceleration/ deceleration phenomena as a result of a blow to another part of the body or a fall (Gennarelli,

271

1991; Elson and Ward, 1994; Graham et al., 2000; Tegner, 2001). In ice hockey, such injuries may involve axonal damage due to excessive angular acceleration (e.g., after an abrupt body check) caused by either direct or indirect inertial loading of the head and are characterized by a global disruption of neurological function (Gennarelli, 1991; Elson and Ward, 1994; Graham et al., 2000; Tegner, 2001). It is now well recognized that (1) traumatic axonal injury (TAI) is a consistent feature of human TBI, particularly those injuries involving rapid acceleration/deceleration of the brain (Elson and Ward, 1994; Graham et al., 2000) and (2) that these injuries represent a continuous spectrum of the same pathology: a progressive widespread but heterogeneous damage to axons (Gennarelli, 1991; Elson and Ward, 1994; Graham et al., 2000; Tegner, 2001). The majority of studies of TAI have long been focused upon human DAI and animal models thereof. Here TAI has been linked to the morbidity associated with that condition (Strich, 1961, 1970; Nevin, 1967; Peerless and Rewcastle, 1967; Oppenheimer, 1968; Clark, 1974; Adams et al., 1982; Gennarelli et al., 1982; Povlishock et al., 1983; Pettus et al., 1994; Maxwell et al., 1997). Historically, it was suggested that forces acting during TBI and TAI tear or shear axons, causing them to retract from the site of injury and expel a ball of axoplasm. This forms a reactive swelling or retraction ball, the histological structural feature then used to diagnose DAI (Adams et al., 1982). More recent studies, however, have not provided support for the concept that large numbers of axons are sheared at the time of injury. Rather experimental, animal models have demonstrated that traumatic injuries elicit a delayed or secondary axotomy wherein the trauma initially injures an axon, which, however, remains in continuity, causing it, over a period of at least several hours, to fail at one or more foci, and disconnect. A reactive swelling of classic description then forms (Povlishock et al., 1983; Povlishock, 1986, 1992, 1993; Gennarelli, 1991; Pettus et al., 1994; Povlishock and Christman, 1995; Povlishock and Jenkins, 1995; Povlishock and Pettus, 1996; Maxwell et al., 1997; Gennarelli et al., 1998; Graham et al., 2000). Further, studies within the last decade have

provided considerable advances in the understanding of the nature and time course of TAI in and following TBI (Povlishock et al., 1983; Grady et al., 1993; Blumbergs et al., 1994, 1995; Christman et al., 1994; Elson and Ward, 1994; Pettus et al., 1994; Sherriff et al., 1994a, b; Blumbergs, 1997; Maxwell et al., 1997; Gennarelli et al., 1998; Graham et al., 2000). Graham et al. (2000) reported that over the last several years a consensus has arisen that the prime site of injury in an injured axon in TBI is the axolemma. There is now a consensus that pathology in the injured axon then occurs because of loss of homeostatic mechanisms that maintain differential ionic gradients necessary for the normal electrical activity of the axon. TAI has been shown in a variety (ﬂuid percussion, cortical impact, rotational acceleration and stretch-injury) of experimental animal models of human head injury (Povlishock et al., 1983; Pettus et al., 1994; Graham et al., 2000). Povlishock et al. (1983) reported TBI after ﬂuid percussion injury and controlled cortical impact in rats, cats and swine. Gennarelli et al. (1982) reported TBI after rotational head acceleration in a non-human primate, and Ross et al. (1995) in miniature swine. Lastly, Gennarelli et al. (1989) reported TAI after controlled stretch-injury to the guinea pig optic nerve. A consensus concerning the sequence of pathologies within injured axons has developed from the above studies of moderate-to-severe TAI generated in numerous, different laboratories. This sequence may be summarized very brieﬂy as follows. First, injury to the axolemma; second, swelling of axonal mitochondria as a result of speciﬁc damage to the internal mitochondrial membrane; third, loss of axonal microtubules and alterations of the intra-axonal relationship of neuroﬁlaments; fourth, the occurrence of so-called ‘‘axonal swellings’’ (foci of increased axonal diameter at which there is still continuity of the axon on both proximal and distal sides of the region of increased axonal diameter) and ﬁnally axonal disconnection to form so-called degeneration bulbs or ‘‘axonal bulbs’’ (Povlishock et al., 1983, 1997; Pettus et al., 1994; Povlishock and Pettus, 1996; Maxwell et al., 1997, 1999; Okonkwo and Povlishock, 1999).

272

Povlishock and colleagues initiated such study in traumatically induced axons using the anterograde tracer horseradish peroxidase (HRP) (Povlishock et al., 1983; Erb and Povlishock, 1988). Their approach was based on the premise that axons tear at the moment of injury. At injury, axons laden with anterogradely transported protein would rupture and expel a mass of protein (peroxidase)-containing axoplasm that could be easily visualized using histochemical methods at by both light and electron microscopes (Povlishock et al., 1983). But evidence for tearing or shearing — that is loss of the integrity of the axolemma and the presence of label free within the periaxonal space — was not obtained (Erb and Povlishock, 1988). Rather at 1 h and later after injury a local, lobular pooling of intra-axonal peroxidase still in continuity with both proximal and distal parts of the peroxidase containing axon occurred. Moreover, the diameter of the focal pools of HRP increased in diameter — from 10 mm at 1 h to 18 mm at 6 h after injury (Erb and Povlishock, 1988). In addition, multilobulated proﬁles interconnected by thin HRP labeled bands were observed. Evidence for loss of continuity of labeled axons was only described at 2 h and later after injury. The pooling of peroxidase was suggested to occur because of a focal impairment of axonal transport. Continued anterograde tracer delivery to the site, resulted in focal expansion of diameter of the axon, swelling and ultimately detachment from its distal counterpart (Figs. 5A, B) (Povlishock et al., 1983; Povlishock, 1992; Pettus et al., 1994; Maxwell et al., 1997). A second approach using HRP, as a marker of AI (Pettus et al., 1994; Pettus and Povlishock, 1996), was to infuse peroxidase into the CSF of anaesthetized animals prior to TBI (Pettus and Povlishock, 1996). The intact axolemma normally excludes molecules as large as HRP from the axoplasm. But, within 5 min after TBI, axons were obtained that contained peroxidase and such labeled axons were obtained up to 6 h, the end of the experimental period examined, after injury. Thus TBI resulted in damage to the axolemma of some axons such that peroxidase was able to enter the axoplasm (Pettus et al., 1994; Pettus and Povlishock, 1996). In the aforesaid axons there was also

Fig. 5. Light micrograph (A) illustrates the changes in intraaxonal anterograde horseradish peroxidise passage following ﬂuid percussion injury. Note that within several hours of the traumatic insult the axon shows a lobulated swelling with a site of axonal narrowing (red arrow). Reprinted with permission and courtesy from Maxwell et al. (1997). Light micrograph (B) illustrates a lobulated terminal swelling after axonal disconnection to form a conspicuous proximal bulb. Human material labelled with antibodies for beta amyloid precursor protein.

loss of axonal microtubules and a reduced spacing between, or compaction, of neuroﬁlaments. However, such alterations were not representative of all axons demonstrating pathology. In some axons, despite the fact that peroxidase had not entered the axoplasm, the components of the cytoskeleton no longer ran parallel to the long axis of the axon but adopted a spiral course through the axoplasm (Pettus and Povlishock, 1996). Thus in mild experimental TBI neuroﬁlament misalignment and axonal swelling were not associated with the passage of peroxidase from the extracellular to the intra-axonal compartment (Pettus et al., 1994, 1996). The latter ﬁndings provided support for the concept that there is a spectrum of AI in TAI. More recently, this concept has been strengthened by the ﬁndings of other investigators who have

273

shown that TAI does not affect only large axons, as had been suggested in the mid 1980s, but also injures small, myelinated ﬁbers that occur, for example in the optic nerve (Jafari et al., 1997, 1998). Indeed, these latter studies showed that alterations for numbers of and spacing between microtubules and neuroﬁlaments differed between nerve ﬁbers of different axonal diameter. Most recently, a further complication of the study of developing pathology in TAI has been indicated in that immunocytochemical labeling of injured axons for either b-amyloid precursor protein, a normal component of axoplasm, or compacted neuroﬁlaments (Stone et al., 2001) has suggested that one subtype of injured axons undergo a progressive focal swelling resulting in disconnection or secondary axotomy between 2 and 6 h while others show compaction of neuroﬁlaments within minutes of injury but do not undergo any further change over the following several hours. Over the last decade use of antibodies against amyloid precursor protein (APP) (Sherriff et al., 1994a, b; Gentleman et al., 1995; McKenzie et al., 1996) has been adopted as the marker of choice for detecting TAI in the diagnostic laboratory. APP is a membrane-spanning glycoprotein originating from a gene on chromosome 21 and of unknown function (Hartmann, 2001). It is transported by fast axoplasmatic transport, is located at both the pre- and post-synaptic sites and accumulates within axonal swellings in TBI very quickly in humans (Blumbergs et al., 1994; Blumbergs, 1997) as a result of loss of fast axonal transport. However, the recent ﬁndings by Stone et al. (2001) now question the sensitivity of a-APP immunohistochemistry alone as a tool for demonstrating TAI. Rather, a range of immonocytochermical markers should be used because use of APP labeling probably underestimates the true number of axons undergoing pathology. Indeed, use of the antibody Ab38 that labels sites of calpain-mediated spectrin proteolysis (Buki et al., 1999), shows increased density of axons/mm2 labeled for Ab38 (calpainmediated spectrin proteolysis) in the pyramidal tracts and medial lemniscus of rats after impact acceleration injury at 60–120 compared with 15–30 min after injury. In addition, a-APP labeling techniques have also shown accumulation of

b-APP in relatively normal-looking axons (Gentleman et al., 1995) and may simply represent changes in axoplasmic transport of a temporary nature in axons which, hypothetically, have a potential for recovery (Gennarelli, 1991; Gentleman et al., 1995; Gennarelli et al., 1998). The time course of recovery is unclear, but it is interesting to note that Blumbergs et al. (1994) reported that bAPP accumulation can be shown in humans as long as 99 days after the initial injury (Blumbergs et al., 1994, 1995; Blumbergs, 1997). Overall, these immunocytochemical alterations have provided further conﬁdence that, after TAI, there is a focal impairment of axoplasmic transport, leading to progressive axonal swelling and detachment/secondary axotomy over a period of several hours or days after injury (Grady et al., 1993; Christman et al., 1994; Pettus et al., 1994; Povlishock and Christman, 1995; Maxwell et al., 1997; Graham et al., 2000). However, patients with mTBI rarely come to neuropathological examination. Nonetheless, there is increasing evidence that some permanent damage may be present (Christman et al., 1994; Blumbergs, 1997; Gennarelli et al., 1998; Graham et al., 2000). Initially it was thought that mTBI might produce only temporary disturbances of brain function without gross structural changes. Now there is increasing evidence that mTBI, in both animals and humans, can lead to an impairment of axoplasmic transport, which results in a progressive pathology comparable with that described earlier in moderate or severe TBI (vide supra). Thus axonal swellings form in mTBI, followed by axonal detachment in ‘‘delayed axotomy’’ (Blumbergs et al., 1995; Povlishock and Christman, 1995; Maxwell et al., 1997; Gennarelli et al., 1998; Graham et al., 2000). This process of delayed axotomy is a complex network of interacting functional, structural, cellular and molecular changes (Gennarelli et al., 1998). Fairly recent evidence (Maxwell et al., 1999) has provided further conﬁdence for the concept that tensile strain to an axon leads to injury to the axolemma. Cytochemical techniques which label for activity of Ca2+-ATPase and Na+/K+-ATPase (pNNP-ase) show loss of activity of these two membrane pumps after injury. Loss of such membrane pump

274

activity would result in a loss of transmembrane homeostasis for ions, and in particular for Ca2+. The resulting inﬂux of calcium down the normal diffusion gradient, which maintains Ca2+ levels 10,000 times higher in the extra-axonal compartment in the normal axon, will result in elevation of intra-axonal concentrations of Ca2+ and allow activation of calpain-mediated spectrin proteloysis (CMSP). Spectrin is an integral component of the subaxolemma cytoskeleton. Buki et al. (1999) showed that CMSP breakdown products appear ﬁrst within the subaxolemma compartments of injured axons at 15 min and only extend throughout the whole diameter of the axoplasm at 1 h after injury. Unfortunately CMSP immunocytochemistry was not examined beyond 2 h after injury. But, quantitative freeze-fracture studies provided morphological evidence of damage to cell membranes (Maxwell et al., 1999), both the axolemma and myelin sheath, up to 24 h after TAI. In parallel, excess axoplasmic calcium leads to a dysfunction and a breakdown of neuroﬁlaments and microtubules. Further, release of acetylcholine contributes not only to the depolarization of neurons but also to calcium-associated excitoxic tissue damage. Thus there is increasing evidence that the initial site of damage is the axolemma. Such damage may then allow abnormal ionic or molecular inﬂux into the injured nerve ﬁbers. Over time, these changes lead to foci of disruption of axonal structure and impaired axonal transport culminates in secondary or delayed axotomy. Thus, experimental studies of TAI, over the last decade, have lead to the now widely held consensus that there is a spectrum of pathological axonal changes reﬂecting both the severity of injury (Povlishock et al., 1983; Pettus and Povlishock, 1996) and size of axons (Maxwell et al., 1997, 2003). The situation is complicated because there is increasing evidence that the spectrum of axonal response to TAI may reﬂect either or a combination of (1) the level of tensile strain applied to individual axons at the time of injury, (2) the degree of damage to the axolemma leading to the injured axon entering the ‘‘pathological cascade’’ which results in secondary axotomy, (3) the size of the axon before injury in that large and small axons respond differently to TAI and (4) whether the axonal response is

manifested as either long-term compaction of neuroﬁlaments or loss of fast axonal transport leading to secondary axotomy. APP as a marker for TAI, is of particular importance with regard to mTBI following the very important observation of multifocal AI within the fornices in human beings with a recorded loss of consciousness of as little as 60 s (Blumbergs et al., 1994). Thus TAI has been documented after relatively mild head injury and in milder grades of DAI in humans (Christman et al., 1994; Blumbergs et al., 1995; Povlishock and Christman, 1995; Povlishock and Jenkins, 1995). There is also a growing consensus that mTBI may represent the mildest form of TAI (Povlishock et al., 1983; Elson and Ward, 1994; Pettus et al., 1994; Gennarelli et al., 1998). At mild levels of AI, a primary misalignment of the components of the axonal cytoskeleton occurs in which neuroﬁlaments and microtubules assume a spiral orientation within the axoplasm independent of damage to or an increase in permeability of the axolemma as may be indicated through use of HRP tracers within the CSF (Sherriff et al., 1994b; Gentleman et al., 1995; Pettus and Povlishock, 1996; Maxwell et al., 1997). In damaged axons regions of increased axonal diameter are termed axonal swellings. Here continuity of the axon is maintained on both the proximal and distal sides of the swelling (Fig. 6) (Povlishock et al., 1983; Pettus et al., 1994; Sherriff et al., 1994b; Gentleman et al., 1995; Maxwell et al., 1997).

Fig. 6. This light micrograph of human material labelled for beta amyloid precursor protein, illustrates, early in the posttraumatic course, the occurrence of two regions of increased axonal caliber where the two swellings are still linked.

275

In parallel with moderate or severe TAI, in mTBI in sports-related injuries the magnitude, duration and rate of onset of angular acceleration as well as the direction of head motion during the injury episode will determine the amount of axonal damage, the severity and the reversibility of the clinical syndromes and the resulting neurological deficits (Gennarelli et al., 1998; Graham et al., 2000). Although, hypothetically, both the number and overall anatomical distribution of injured axons is lower and less widespread compared with the severe traumatic state, the very presence of immunocytochemically labeled, damaged axons is indicative of AI and thought to reﬂect the lower end of the spectrum of TAI (Povlishock, 1992, 1993). It is now becoming accepted that there are a range of categories of diffuse brain injury, which have the same pathophysiology within the individual, injured axons; but a different clinical outcome. The latter, hypothetically, relates to the severity of axonal elongation resulting from mechanical strain placed upon axons at the time of injury (Elson and Ward, 1994; Graham et al., 2000). Current thinking with regard to the spectrum of axonal responses between momentary periods of unconsciousness on the playing ﬁeld, mild concussion and concussion may be summarized as follows in terms of the degree of mechanical strain placed upon axons at the time of injury and their physiological/pathological responses. (1) At the lowest level of strain (5% or less elongation) (Stage I) (Fig. 7), ionic inﬂux of various species occurs and causes temporary failure of the generation and propagation of action potentials (Gennarelli et al., 1998). Inﬂux of Ca2+, Na+ and Cl and the efﬂux of K+ at this level of injury are fully restored in a matter of minutes and the damaged axon recovers its function completely. With increasing strain applied to the axon, a series of increasingly severe mechanisms and pathological changes occur. If the level of axonal stretch is in the 5–10% range (Stage II) (Fig. 7) then additional ionic perturbations are associated with either ﬂuid ﬂuxes in an attempt to maintain osmotic balance (Gennarelli et al., 1998) or

saturation of the mechanism of mitochondrial sequestration of excess cytoplasmic calcium. This causes damage to and swelling of mitochondria (Okonkwo and Povlishock, 1999), together with enlargement of the injured axon. Hypothetically, this mild impairment of axoplasmic transport is reﬂected, ultrastructurally, as a non-linear alignment of the components of the axonal cytoskeleton. It is currently hypothesized that most of these axons recover and thus restitution of structure and function may be expected. But, if the level of stretch is between 10 and 20% (Stages III and IV respectively) (Fig. 7), then the axon cannot fully restore ionic and osmotic homeostasis and, as a consequence, numerous pathological changes are initiated (Gennarelli et al., 1998). Recent evidence (Stone et al., 2001) suggests that there may be two separate pathological pathways followed by different axons (Fig. 7). As the level of clinical injury progresses from less to more severe, greater amounts of damage to individual axons and greater numbers of axons are thought to be involved. In injuries from which the patient apparently recovers, for example mTBI, there is presently no evidence for Stages III and IV damage. This is due to the fact that the patient resumes normal activities with no detectable change in their performance and that Stages III and IV may only be determined post-mortem. Therefore, it is suggested that Stages I/II represent the great majority of axonal injuries experienced, for example by ice hockey and other contact sports players. (2) In the context of mTBI, the inﬂuence of the degree of damage to the axolemma upon subsequent axonal responses has not been investigated in any systematic or controlled, experimental manner. But there is now good evidence that the integrity of and the functional state of the axolemma are related to axonal responses after TAI. The structural and physiological integrity of the axolemma is essential for the generation and propagation of action potentials. It is suggested that perhaps electrophysiological assessment of membrane function would possibly provide the best means of analysis of the functional

276 WALLERIAN DEGERATION (24 hours - 7 days) PROXIMAL REMNANT OF AXON reduced calibre of axon neurofilament compaction over 30-50 µm of axonal length

(1 to 12 weeks) Proteolysis of axonal components but myelin figures remain DISTAL REMNANT OF AXON

SECONDARY AXOTOMY (minimum of 4 hours - 7 days) (up to 12 weeks)

PRIMARY AXOTOMY (less than 1 hour)

15 - 20% strain (4 hours - 99 days) axonal disconnection disruption of the myelin sheath enlarged periaxonal space focal compaction of neurofilaments disruption of the axolemma proteolysis/destruction of neurofilaments

Compaction of neurofilaments reduced axonal cross-sectional area increased g ratio

% strain currently unknown (30 min-7 days) multiple foci of neurofilament compaction membrane renting/fragmentation swelting of/damage to mitochondria proteolysis of axonal cytoskeleton neurofilament compaction in zone of axotomy marked infolding of the exolemma 10-15% strain (1.5 - 6 hours) axonal swelling loss of axonal transport loss of microtubules µM calpain/phosphatase/kinase activation collapse/loss of neurofilament sidearms compaction of neurofilaments more than 20% strain

5 - 10% strain (15 min - 24 hours) functional and structural damage to the axolemma depolarisation and loss of ionic homeostasis loss of and/or de-activation of ion channels and ATP dependent membrane pumps, mitochondrial damage KEY Effects of injury Mechanical strain (Time scale) mechanism and ultrastructural changes

5% or less strain (seconds) translent depolarisation Recovery INJURY

DESCRIPTOR

Fig. 7. Schematic overview of current thinking with regards to AI in human DAI and animal diffuse traumatic brain injury. (Modiﬁed from Maxwell et al., 1977.)

integrity of an injured axon. The authors are aware of only two experimental, electrophysiological studies concerning axonal responses to the application of tensile strain (Tomei et al., 1990; Galbraith et al., 1993) and only the former involved experiments on mammalian axons. With increasing loading from 0 to 20% elongation there is rapid depolarization following by a slow repolarization. But with an increased loading, the

time required for repolarization increases in a non-linear manner. At loadings resulting in more than 20% elongation the axolemma is irreparably damaged (Galbraith et al., 1993). There is thus an apparent, critical threshold at or around 20% elongation at which the axolemma is irreparably damaged. In mammalian myelinated axons, analysis of visual evoked potentials in the visual cortex after TAI to optic nerve showed (1) a

277

reduction in signal amplitude to 33% of control values immediately after injury, to 66% at 24 h and 77% at 1 week. (2) An increased signal latency to 126% of control values after injury, 119% at 24 h and 97% at 1 week in an animal model (guinea pig optic nerve stretch-injury) in which the current literature documents that only 20% of ﬁbers provide morphological evidence for injury (Gennarelli et al., 1989). Thus, at the functional level, axons are rapidly depolarized at TAI (Stage I, Fig. 7). However, in this experimental paradigm, recovery of activity occurs relatively slowly, will vary with the initial loading, and total recovery is not obtained even at 1 week after injury (Tomei et al., 1990). Two research groups have provided morphological evidence for damage to the axolemma. First, after ﬂuid percussion injury, inﬂux of HRP from the CSF into damaged axons was shown (Pettus and Povlishock, 1996). But, in molecular terms HRP, is a large molecule and thus inﬂux into the axoplasm of an injured axon would necessitate the appearance of relatively large ‘‘holes’’ in the membrane. Hypothetically, such damage would probably result from elongation of axons by more that 20% of their pre-injury length. It is noteworthy that in the same study the authors noted axons in which there was axonal pathology but that HRP had not entered the axoplasm. It is currently hypothesized that such damage represents a mild level of injury to those axons. Second, using the same model as Tomei et al. (1990), Maxwell et al. (1999) provided both cytochemical evidence for loss of membrane pump activity — vide supra — and quantitative data for an alteration in the number and distribution of intramembranous particles (IMPs) visualized through freeze-fracture techniques. It is currently believed that the altered distribution of IMPs reﬂects changes in the distribution and number of transmembrane proteins some of which are membrane pumps and ion channels. Hypothetically, such changes may also reﬂect damage to the axolemma in ﬁbers

which are not damaged to the degree that large molecules such as HRP may enter the axoplasm. (3) Evidence obtained over the last several years has shown that neuroﬁlaments and microtubules demonstrate a spectrum or range of responses after TAI. Current results suggest that at lower or mild levels of injury, there is focal disarray of neuroﬁlament and microtubule alignment within the axoplasm, but no loss of microtubules (Povlishock et al., 1983; Pettus and Povlishock, 1996; Maxwell et al., 1997, 2003). But, in moderate or severe TAI there is (1) rapid loss of microtubules (Maxwell and Graham, 1997) and (2) a reduction in the spacing between adjacent neuroﬁlaments, resulting in both an increased density of neuroﬁlaments per unit area of the axoplasm (Pettus and Povlishock, 1996; Povlishock et al., 1997) and reduced interneuroﬁlament spacing (Jafari et al., 1997, 1998; Maxwell et al., 2003). The term applied to this latter change in the organization of the axonal cytoskeleton is that the neuroﬁlaments are said to be ‘‘compacted’’. More recently, quantitative analysis has shown that alterations in the organization of the axonal cytoskeleton after TAI differ (1) in the time course of the response by microtubules and neuroﬁlaments and (2) in different sized axons. After TAI in the guinea pig optic nerve compaction of neuroﬁlaments is most marked in smaller axons. Indeed, the reduced spacing between neuroﬁlaments has been correlated with a reduction in the calibre of axons (Jafari et al., 1997; Maxwell et al., 2003). But in larger axons of the optic nerve there is an increased spacing between components of the cytoskeleton throughout the length of the nerve. In these larger ﬁbers, nonetheless, there are focal sites of compacted neuroﬁlaments. But these sites are associated with loss of the axolemma and possibly are sites at which the axon is undergoing the ﬁnal stages of secondary axotomy (Jafari et al., 1998; Maxwell et al., 2003). In other animal models, however, for example cortical impact and ﬂuid percussion injury, there is compaction

278

of neuroﬁlaments in relatively large ﬁbers (Pettus and Povlishock, 1996; Povlishock et al., 1997; Stone et al., 2001). (4) Recent, novel work has suggested that compaction of neuroﬁlaments may occur in axons other than those which progress to the formation of axonal swellings and secondary axotomy (Stone et al., 2001). However, only one research group has so far described such a ﬁnding and thus there is presently no other material that would allow discussion. But, it is included here as a pointer to the fact that attention should be paid to this phenomenon in later investigations of TAI. Clearly, much further work needs to be done to fully understand the significance of differing cytoskeletal responses to (1) axonal loading, (2) different responses by axons within different CNS tracts and (3) between axons that are undergoing secondary axotomy and those that are not. However, in summary, three different types of evidence support the hypothesis that damage to the axolemma is key to the development of pathology in injured axons after TAI. Overall, there is variation in the sites of axonal damage within the brain of each and any patient and axonal cytoskeletal responses may differ between axons in different CNS tracts. It is intriguing as to whether these detailed differences in axonal response to TAI are important in consideration of the question as to why some ﬁbers within a particular ﬁeld are injured while other, spatially closely related ﬁbers are not in cases of TAI. It is suggested that the above should be rigorously considered in the formulation of any future experimental studies of TAI. Neurobiology and associated neurometabolic changes of mTBI In recent years extensive knowledge concerning pathophysiological processes of mTBI has been gleaned from in vivo and in vitro experiments (Yaghmai and Povlishock, 1992; Hovda, 1996; Pettus and Povlishock, 1996; Cecil et al., 1998; Hovda et al., 1994, 1999; Giza and Hovda, 2000; Maroon et al., 2000). Alterations in ionic ﬂuxes,

cerebral blood ﬂow (CBF) and cerebral glucose metabolism have all been documented in braininjured patients (Giza and Hovda, 2000). Utilizing animal models, investigators have been able to better characterize trauma-induced ionic ﬂux and provide quantitative data in support of the corresponding metabolic dysfunction that has become the clinically, accepted diagnostic descriptor of mTBI. Microdialysis probes have detected ionic disturbances (Katayama et al., 1990; Kawamata et al., 1992; Hovda et al., 1999; Cantu, 2000; Reinert et al., 2000; Klatky et al., 2002) and nuclear magnetic resonance spectroscopy (MRS) has recently been used to demonstrate neurochemical alterations associated with pathological conditions after TBI (McIntosh et al., 1987, 1988; Pettus and Povlishock, 1996; Hovda, 1996; Cecil et al., 1998; Hovda et al., 1999; Bigler, 1999; Garnett et al., 2000; Giza and Hovda, 2000; Brooks et al., 2001; Sinson et al., 2001). Positron emission tomography (PET) has provided quantitative data for alterations in cerebral glucose metabolism (Nedd et al., 1993; Ruff et al., 1994; Gross et al., 1996; AbdelDayem et al., 1998; Bigler, 1999; Ricker et al., 2001; Levine et al., 2002; Umile et al., 2002) while both radioisotope studies (Hovda et al., 1999; Ricker et al., 2001) and transcranial Doppler imaging (Ng et al., 2000; Muller et al., 2001) have been used to follow changes in CBF. Fairly recently, application of the above investigative techniques have been applied to human head-injured patients. These techniques have conﬁrmed and extended the experimental data derived from in vivo animal models (Pettus and Povlishock, 1996; Cecil et al., 1998; Hovda et al., 1999) by providing good evidence that exactly comparable pathophysiological responses occur in head-injured patients. Thus, there is increasingly convincing evidence that mTBI triggers a complex and interwoven sequence of ionic and metabolic events from which damaged cells may eventually either recover or degenerate and die. The metabolic cascade is a multidimensional process (Hovda, 1996; Giza and Hovda, 2000). To brieﬂy summarize, after mTBI, there is a significant K+ efﬂux from injured cells, owing to mechanical deformation of the axolemma and opening of voltage-dependent K+ channels. These lead to neuronal ﬁring, depolarization

279

and release of excitatory neurotransmitters. In the normal brain, excess extracellular K+ is subject to reuptake by surrounding glial cells. This compensatory mechanism can maintain physiologic extracellular K+ levels even after mild TBI or ongoing seizure activity (Hovda et al., 1999; Matsushita et al., 2000; Reinert et al., 2000). But more severe TBI, where greater numbers of axons and neurons are injured however, will overcome this glial safety valve. Initially there is a slow rise in extracellular K+. But as the physiological ceiling for K+ balance is breached, an abrupt increase occurs, leading to depolarization, release of excitatory amino acids (EAAs) and a further massive K+ ﬂux though EAA/ligand-gated ion channels (Hovda et al., 1999; Yamamoto et al., 1999). Following this wave of excitation neurons are hyperpolarized and there is consequent suppression of neuronal activity, referred to as ‘‘spreading depression’’, which is thought responsible for the neurologic dysfunction across the cortical surface noted in the clinical setting of seizure propagation and migraine aura. When the extracellular potassium concentration increases beyond the normal, physiological upper limit (4–5 mmol/L) to levels of 20–50 mmol/L and greater, then inhibition of the action potential and loss of consciousness may occur (Maroon et al., 2000). The EAAs, especially glutamate, activated kainate and a-amino-3-hydroxy-5-methylisoxazole4-propionic acid (AMPA) receptors located on the cell body which open channels permeable to both Na+ and K+. EAAs also activate N-methyl-D-aspartate (NDMA) receptors permeable to Ca2+ as well as Na+ and K+. The release of glutamate permits a rapid and sustained inﬂux of Ca2+, which leads to dysfunction of the oxidative metabolism as a result of damage to mitochondria (Fineman et al., 1993; Maxwell et al., 1999) and which is ameliorated by therapy using cyclosporin A (Okonkwo et al., 1999). The cell therefore becomes increasingly dependent on ATP generated through glycolysis. Alterations in brain Ca2+ homeostasis through loss/inactivation of membrane pump Ca2+ ATPase and receptors/channels associated with Ca2+ entry (voltage sensitive channels or ionophore-associated glutamate receptors, such as NMDA receptors) have been also associated with regional cerebral edema, vasospasm and delayed cell death (Graham

et al., 2000). Traumatic, ischemic or anoxic injury to axons is associated with widespread neuronal depolarization and release of EAA neurotransmitters such as glutamate, leading to the opening of NMDA receptor-associated ion channels and inﬂux of Ca2+ (Graham et al., 2000). The resulting posttraumatic Ca2+ storm has been documented using 45 Ca-autoradiography after a lateral ﬂuid percussion injury (Fineman et al., 1993), and indirectly, through use of cytochemical techniques, as a result of redistribution of membrane pump calcium-ATPase and Ca2+ inﬂux into myelinated nerve ﬁbers of the guinea pig optic nerve after stretch-injury (Maxwell et al., 1995). Finally, the injured axon may lose the capacity to buffer, extrude or to bind the intracellular calcium. When the calcium load exceeds the capacity of axoplamic buffers, there is a consensus that activation of calpains and phospholipases initiate processes of cytoskeletal and membrane degeneration. Several recent studies have documented both acute calpain activation and regional calpaininduced cytoskeletal proteolysis (Buki et al., 1999; Graham et al., 2000; Saatman et al., 2003). This initiates a cascade of structural changes in injured axons beginning with breakdown of the subaxoplasmic cytoskeleton (Buki et al., 1999), parts of the cytoskeleton related to swollen, damaged mitochondria, and later (between 30 and 120 min after injury) compaction of cytoskeletal elements throughout the remnants of the axoplasm. This pathology results in focal impairment of axoplasmic transport leading to focal accumulation of transported materials and loss of delivery of molecules required for repair of the damaged axolemma. This leads to more widespread damage to the axolemma (Buki et al., 1999; Maxwell et al., 1999), generation of numerous injurious compounds, including free radicals and various inﬂammatory mediators, for example cytokines, prostaglandins and leukotrienes. At this stage, since the cell body is now also damaged, the potential for axonal recovery is lost and pathological changes in the injured axon continue until the axon undergoes secondary axotomy (Maxwell et al., 1997; Gennarelli et al., 1998). Calcium inﬂux may also activate breakdown/ loss of microtubules (Maxwell and Graham, 1997;

280

Maxwell et al., 1999) since calcium both prevents assembly of microtubules and causes rapid disassembly of pre-existing polymers (Matsumura and Hayashi, 1976; Langford, 1978). In addition to calcium activating calpains (Buki et al., 1999), accumulation above normal axoplasmic levels may activate, cystein proteases, spectrin proteases, calcineurin and phospholipase A2 proteases. Further, post-traumatic depolarization will activate phospholipase C. Collectively these act to degrade a wide range of cytoskeletal and membrane components and eventually lead to cell damage or death. More specifically, in axons after TAI, posttraumatic elevated concentrations of Ca2+ within the axoplasm leads to compaction (i.e., a reduction in the spacing between neighboring) of neuroﬁlaments. This has recently become a major pathological marker of TAI (Pettus and Povlishock, 1996; Jafari et al., 1997, 1998; Povlishock et al., 1997; Buki et al., 1999; Stone et al., 2001; Maxwell et al., 2003). Experimental studies following a lateral ﬂuid percussion brain injury using 45Ca autoradiography have shown that Ca2+ accumulation continues for several days, only returning to control levels by the 4th day after the injury (Hovda et al., 1999; Samii et al., 1999). In certain situations, it may require several seconds or longer for ion concentrations in injured axons to rise above the threshold required for initiation of the development of pathology. In mTBI, the presence of such a time course may explain why some athletes can walk from the ﬁeld and then collapse unconscious on the sideline (Maroon et al., 2000). An injured axon will, probably, attempt to restore membrane potential and recover homeostasis. Membrane pumps and transporters (e.g., NPP-ase, Na+/ K+-ATPase) and membrane pump Ca2+ATPase may work overtime, consuming increased amounts of ATP. To meet these elevated ATP requirements, there is a marked up-regulation of cellular glycolysis, which occurs within minutes after TBI. Correlated with this period of hyperglycolysis, there is an increased synthesis of lactate and other toxic products (i.e., free fatty acids) in cells of the brain and CSF which leads to further damage to neurons. In parallel with this, CBF is reduced to oligaemic levels. The resultant mismatch between oxygen, glucose delivery and consumption and the

physiological stress to which cells in the brain are thereby exposed may make those cells much more susceptible to secondary injury. Indeed, CBF may remain depressed for several days after TBI, thereby limiting the ability of the brain to respond adequately to subsequent perturbations in energy demand (Cantu, 2000). Bergsneider et al. (2000) suggested that there are three distinct periods of altered tissue/brain metabolism following almost all types of animal experimental TBI — ﬂuid percussion, cortical impact and SDH — and head-injury in humans. In summary, the initial response is a total-body increase in glucose utilization governed by pathophysiologic processes, which lead to the endogenous release of EAAs and adrenergic neurotransmitters (Bergsneider et al., 2000), which are independent of the functional state of neuronal activation and, we suggest, may reﬂect responses to release of the above by traumatized neurons. This total-body hypermetabolic state, reﬂects the activation of energy-consuming, ATP-dependent, ion pumps which attempt to redress abnormal ion ﬂuxes (Hovda, 1996; Hovda et al., 1999; Bergsneider et al., 2000; Giza and Hovda, 2000). Secondly, related regions of the brain undergo a period of reduced glucose utilization, which, over the course of 4 days, return to normal values (Fineman et al., 1993). Little is known of the processes affecting glucose utilization during this second phase of metabolic depression although it has been suggested that Ca2+ sequestration by mitochondria leads to opening of the permeability transition pore of the inner mitochondrial membrane (Okinkwo and Povlishock, 1999) and inhibition of oxidative phosphorylation and energy transduction (Fineman et al., 1993; Bergsneider et al., 2000). This leads to cells utilizing the anaerobic, glycolytic pathway to synthesize high energy compounds needed to fuel membrane pumps. Indeed, one experimental model utilizing cytochemical labeling for activity of two ATP-dependent ionic pumps has shown that there is a time course of at least 1 h after injury for the loss of pump activity of both p-NPPase and membrane pump Ca2+ -ATPase and that this loss of activity extends for at least 24 h after injury (Maxwell et al., 1999). The correlated inﬂux of Ca2+ into injured axons has been

281

suggested to stimulate both synthesis of glycogen and reduce axonal transport of phosphofructokinase leading to increased content of glycogen in injured axons with 15 min of injury. Moreover, there is loss of glycogen deposits in injured axons between 7 and 14 days after injury. This loss, it is suggested, may correlate with the reported return to normal glucose utilization by 4 days (Fineman et al., 1993) and recovery of metabolic function (Pappius, 1981; Fineman et al., 1993; Hovda, 1996) which showed that neurological deficits persist both as long as there is metabolic depression, and that the rate of recovery of behavioral function parallels the recovery of metabolic function in the relevant regions of the brain. However, at present, the mechanisms leading to the third phase (recovery) are not well understood although they do appear to be spontaneous (Bergsneider et al., 2000). Overall, these studies have demonstrated that the brain is not metabolically quiescent or stagnant after TBI, but instead exhibits a metabolic response that is both dynamic and stereotypic (Fineman et al., 1993). Recent studies in head-injured humans (Bergsneider et al., 2000) further raise the possibility that the extent of the cerebral metabolic rate of glucose (CMRglc) and oxygen (CMRO2) depression after TBI is governed by the severity of injury rather than the level of consciousness. Hypothetically, it is possible that the relative demand placed upon the glycolytic pathway will vary with the severity of the injury, as a function of CMRO2. At this time there is too little evidence to be able to suggest that in mTBI patients the CMRglc may be a better indicator of neurological status. For example, Bergsneider et al. (2000) cite two head-injured patients with comparable CMRglc where one patient was ambulatory with a Glasgow Coma Score (GCS) of 15 while the other patient was comatose and had to be mechanically ventilated. Nonetheless, it is suggested that, with data obtained from a larger aliquot of patients, the CMRglc may prove to be a more sensitive indicator of neurological status in the mTBI patient. Furthermore, evidence that alteration of glucose metabolism is not restricted to the ﬁrst hours following mTBI (Hovda, 1996; Hovda et al., 1999;

Giza and Hovda, 2000), does not normalize until between 5 and 10 days after injury (Kawamata et al., 1992) and may result in reduction of both protein synthesis and oxidative phosphorylation. The term ‘‘energy crisis’’ has been coined to refer to this state of metabolic disequilibrium. The magnitude of this ‘‘energy crisis’’ has important implications for cellular recovery and for increased vulnerability to secondary insults. As a result cells in the brain are more susceptible and are more ‘‘vulnerable’’ to damage or even death should the patient experience a second insult of, even, lesser intensity (Kawamata et al., 1992; Hovda, 1996; Buki et al., 1999; Hovda et al., 1999; Giza and Hovda, 2000). Overall, the above ﬁndings suggest that loss of consciousness after head trauma, the development of secondary brain damage and the enhanced vulnerability of cells in the brain to a second insult can be explained largely on the basis of abnormal ionic ﬂuxes, acute metabolic changes and loss of autoregulation of CBF immediately after mTBI (Maroon et al., 2000). However, following a traumatic insult to the head, other cells, perhaps the most important of which are smooth muscle and endothelial cells in the walls of the microvasculature of the brain, are also damaged. This results in changes in the extracellular milieu and compromise of the ability of neurons to function and initiate a number of processes that might reestablish the normal pericellular environment (Hovda et al., 1999). It is now evident that after mTBI the stoichiometric relationship of metabolism and the coupling to CBF is fundamentally compromised. There is also evidence that such breakdown is thought to be responsible for the degree and extent of vulnerability following mTBI (Hovda et al., 1999). However, the concept of injury-induced vulnerability to even minor changes in CBF, increased intracranial pressure and apnea (Hovda, 1996; Hovda et al., 1999; Giza and Hovda, 2000) has risen from studies of patients severely enough injured to have been admitted to hospital and, most probably, in intensive care. That is a patient that has most likely experienced moderate or severe head-injury. A basic tenet of this review is that the concept of injury-induced vulnerability should also be a major concern in the management

282

of head-injured athletes who have experienced mTBI but are not hospitalized. However, there is now an increasing acceptance that injured axons (1) undergo or demonstrate a wide spectrum of physiopathological responses with a time course extending to at least several hours, (2) may be physiologically and functionally stressed after injury and (3) as a result, may be more susceptible to any second, even a milder, insult for at least several tens of hours. Therefore, after any type of head-injury, which may potentiate the development of pathology within axons and their cell bodies of the injured sportsman, should be advised not to resume play in the same game. But rather, those sportsmen should be encouraged to rest. This rest could, we suggest, provide mildly injured neurons and other cells the opportunity to respond to post-traumatic increased demands for energy and would allow re-establishment of normal chemical and ionic cellular peri- and intracellular environments.

Biochemical markers of mTBI During the past decade neurobiochemical markers of brain damage have gained increasing interest in experimental and clinical neurotraumatology. Proteins synthesized in astrogial cells or neurons have been proposed as markers of cell damage in the central nervous system. The most common include ﬁbrillary acid protein, myelin basic protein, creatine kinase isoenzyme BB, neuron speciﬁc enolase (NSE) and protein serum-100b. Increased levels of these proteins in the CSF and serum have been observed in patients with various neurological diseases (Skogseid et al., 1992; Ingebrigsten, 1995; Ingebrigsten et al., 1995, 1997; Yamazaki et al., 1995; Ross et al., 1996; Woertgen et al., 1997; McKeating et al., 1998; Raabe et al., 1998; Hermann et al., 1999, 2000, 2001; De Kruijk et al., 2001; Ringger et al., 2004). Structural proteins of astroglial (glial ﬁbrillary acidic protein) or neuronal (neuroﬁlament protein) brain tissue have mostly been used in experimental settings. Alternatively, and with a much greater applicability in the clinical environment, protein serum S-100b and NSE are two markers which, due to present high

sensitivity of commercially available detection kits (Sangtec 100s, AB Sangtec Medical, S-161 02 Bromma, Sweden) to analyze NSE and S-100b concentrations in serum samples, have recently attracted growing attention in clinical research and diagnostic practice. Both proteins are now accepted as speciﬁc neurobiochemical markers of brain damage in clinical and experimental ischemic brain infarction and TBI (Moore, 1965; Skogseid et al., 1992; Ingebrigsten, 1995; Ingebrigsten et al., 1995, 1997, 1999, 2000; Yamazaki et al., 1995; Ross et al., 1996; Woertgen et al., 1997; McKeating et al., 1998; Raabe et al., 1998; Hermann et al., 1999, 2000, 2001; Romner et al., 2000; Biberthaler et al., 2001; De Kruijk et al., 2001; Pleines et al., 2001). The serum S-100b protein (S-100b) is a member of a large family of Ca2+-binding proteins, the cellular synthesis of which has been localized predominantly in astroglia and Schwann cells. When Moore discovered the S-100 protein, he suggested that it was located only in brain tissue (Moore, 1965). Later studies, however, demonstrated its presence in the body in different forms, and that the b form predominates in the brain (Moore, 1965; Ingebrigsten et al., 1995, 1997, 1999, 2000; Woertgen et al., 1997; Ingebrigsten, 1998; McKeating et al., 1998; Raabe et al., 1998; Hermann et al., 1999, 2001; Romner et al., 2000; Biberthaler et al., 2001; Pleines et al., 2001). Increased serum S100b protein concentrations in peripheral blood are considered as a marker for dysfunction of the blood-brain barrier and therefore, protein S-100b release into peripheral blood may indicate functional brain dysfunction without the necessary presence of a pathology visible through use of CT imaging (Moore, 1965; Ingebrigsten et al., 1995, 1997, 1999, 2000; Woertgen et al., 1997; Ingebrigsten, 1998; McKeating et al., 1998; Raabe et al., 1998; Hermann et al., 1999, 2001; Romner et al., 2000; Biberthaler et al., 2001; Pleines et al., 2001). Little is known regarding the mechanism by which S-100b passes through the blood-brain barrier and enters the blood. However, it may be posited that where there is loss of autoregulation and stripping of endothelium, as has, for example been noted in ischemia or severe TBI (Maxwell, 1999; Maxwell et al., 1992) that damage to the microvasculature

283

of the brain that has been described in DAI may provide a pathway for passage of S-100b from injured glial cells into blood. The detection of protein S-100b in peripheral blood may therefore reﬂect either an ongoing pathophysiological cascade that leads to secondary brain damage or result from necrosis (apoptosis) of previously damaged cells (Hermann et al., 2000). However, a recent study showed that the sensitivity of S-100b measurement for the identiﬁcation of patients with a pathological CT scan was high (90%), but the speciﬁcity was only 65% with a cutoff level of 0.5 ng/ml (Ingebrigtsen et al., 2000). Biberthaler et al. (2001) found an even higher sensitivity of S-100b measurement (100%) than that of Ingebrigtsen et al., but that the speciﬁcity was reduced (40.5%). Therefore, measurement of this parameter may be helpful as an additional screening tool to identify risk groups in a cohort of mTBI patients. Biberthaler et al. (2001) suggested using a cutoff point at a concentration of 0.1 ng/ml plasma since this concentration was the lowest obtained in the serum of patients with mTBI but lacking signs of intracerebral injury on cranial computed tomography. Further, it has recently been demonstrated that the early post-traumatic elevated serum levels of S100b declined rapidly during the ﬁrst 6 h after the trauma. Long lasting post-traumatic elevated S100b values probably indicate either an ongoing pathophysiological cascade that leads to secondary brain damage or the egress of S-100b either as a result of necrosis or apoptosis of previously damaged cells (Romner et al., 2000) or focal damage to the microvasculature. Further, it has recently been suggested that lack of detectable concentrations of S-100b protein in serum predicts a normal CT scan provided the blood sample is collected within the ﬁrst 2 or 3 h after injury (Ingebrigtsen et al., 2000). A comparative analysis of the predictive value of the neurological status, CT data, and NSE and S-100b serum concentrations has shown that the initial (i.e., within 2–3 h of injury) serum protein concentration of S-100b is the best predictor of outcome for long-term neuropsychological disorders (Herrmann et al., 2001). Advanced MRI techniques, such as quantitative MRI analysis or diffusion weighted or

magnetization transfer imaging, allow for the detection of diffuse white matter brain damage or axonal brain injury with a greater sensitivity than conventional CT procedures. Correspondingly these probably result in a better prediction of outcome to the patient (Ingebrigsten et al., 1999). NSE is an isoenzyme of enolase and is predominantly found in the cytoplasm of neurons, in neuroendocrine cells, smooth muscle ﬁbers and adipose tissue (Skogseid et al., 1992; Yamazaki et al., 1995; Ross et al., 1996; McKeating et al., 1998; Hermann et al., 1999, 2000, 2001; De Kruijk et al., 2001; Pleines et al., 2001). The cytoplasmic enzyme is liberated by cell destruction. It has been suggested, therefore, that increased levels of NSE in CSF and peripheral blood indicate neuronal damage. Herrmann et al. (1999, 2000) demonstrated an association between the release of both NSE and S-100b protein and the severity of a TBI. Notably, patients with mTBI showed a different temporal proﬁle of NSE in their serum to moderate or severe TBI patients (Hermann et al., 2000). A significant difference for the concentration of NSE in serum from mild and moderate to severe head injury patients was only obtained up to 24 h after injury. Thereafter serum values for either mTBI or moderate to severe injury patients were not elevated and did not differ from control values (Hermann et al., 2000). The release patterns of NSE and S-100b is thought to reﬂect different types and dynamics of underlying pathophysiology between moderate to severe and mild TBI patients. Overall, patients with moderate to severe head injury show a significantly higher and longer release of both markers (Herrmann et al., 2000). In addition, recent studies have shown a correlation between the clinical outcome and CSF or serum concentrations of protein S-100b and NSE in patients with severe TBI (Skogseid et al., 1992; Ross et al., 1996; Woertgen et al., 1997). Raabe et al. (1998) showed that excessive secondary increase in S-100b serum concentration relates to severe brain damage associated with a fatal outcome. The same group reported a significant correlation between early S-100b values and the volume of cerebral contusions in patients with severe TBI but no association between NSE concentration and contusion volume. However, there

284

is still some controversy concerning the clinical significance of such data. For example, Skogseid et al. (1992) found a significant correlation between maximum NSE concentration and contusion volume. Alternatively, Woertgen et al. (1997) found no correlation of either NSE or S-100b and the severity of intracranial pathology as deﬁned by the Traumatic Coma Data Bank (TCDB) CT classiﬁcation. Further, Hermann et al. (1999, 2000) showed that the severity of TBI is associated (1) with early post-traumatic release of both protein S100b and NSE and (2) that the early kinetics of neurobiochemical markers of brain damage after TBI do reﬂect intracranial pathology as demonstrated in cranial CT (Hermann et al., 1999, 2000). More recently, Pleines et al. have indicated that the elevation of S-100b and NSE in CSF depends on the extent of the injury and that S-100bmay be a predictor of outcome after TBI, whereas NSE is a better indicator of inﬂammatory responses (Pleines et al., 2001). In this latter study, release patterns of both proteins did not differ between patients with or without signs of focal neurological deficits. Herrmann et al. (2001) found that post-traumatic serum concentrations of NSE and S-100b not only reﬂected the overall severity of brain trauma, as deﬁned by GCS scores, but also subtle neuropsychological dysfunction. Neuropsychological disorders, identiﬁed 2 weeks after injury, were correlated with a significantly higher and longer lasting release of both brain proteins. Patients with neuropsychological disorders at 6 months after head trauma, however, also exhibited significantly higher NSE and S-100b serum concentrations during the ﬁrst 3 days after TBI (Herrmann et al., 2001). De Kruijk et al. (2001) investigated whether serum concentrations of NSE and S-100b in a group of patients with clearly deﬁned mTBI were higher than in serum of healthy controls. The mean serum concentrations of NSE did not provide a sensitive marker between mTBI patients and controls. On the contrary, the median serum S-100b concentration from 104 mTBI patients was significantly elevated compared with control, healthy patients (n ¼ 92). Neither did trauma to other parts of the body induce an elevation of either S-100b or NSE. Thus, current opinion is that elevated serum

marker concentrations of S-100b are a consequence of brain injury including mTBI (De Kruijk et al., 2001). A major conclusion drawn by this review is that determination of S-100b serum levels within 2–3 h of head-injury may prove to be a useful indicator for damage to the brain after mTBI. But, importantly, serum levels of S-100b also rise after fracture of bones, thoracic contusion, burns and even minor bruising (Anderson et al., 2001). On the other hand, increased serum NSE levels do not seem to be a speciﬁc sensitive marker for mTBI (De Kruijk et al., 2001). Given that S-100b is a marker of damage to glial and/or Schwann cells and NSE indicates damage to neurons, elevated plasma levels of S-100b after mTBI is a good indicator that damage which results in release of S100b is more likely to be localized in the white than the grey matter of the brain. This conclusion is supported by the literature related to the effects of rotational and acceleration/deceleration injuries to the brain and allows development of the hypothesis that TAI is an important pathophysiological cellular response during and following mTBI. However, since S-100b is released by glia than directly from injured neurons it is implicit that the time course of release of detectable levels of S-100b in CSF is longer than that of a direct marker of neuronal injury. Siman et al. (2004) have recently reported, both in vitro and in vivo, elevated levels of a-spectrin and calpain/caspase mediated aminoand carboxy-terminal breakdown products thereof. Moreover, 150-kDa a-spectrin fragments and proteins 14-3-z and 14-3-b were released by 6 h after both mild, moderate TBI and transient global ischemia in rat (Siman et al., 2004). Although aspectrin is widely distributed in many types of cell, spectrin breakdown products have been reported early after TAI in the subaxolemma region (Buki et al., 1999), throughout the axoplasm within 2 h of injury, and has a biphasic presentation with a second peak at 4–5 days after TBI in mouse (Saatman et al., 2003). In humans peak levels occur in CSF at 2–3 days after severe TBI (Farkas et al., 2005) and may fall over 4 days in patients with a good outcome (D’Avella et al., 2004). Thus, there is still a major requirement to deﬁne a marker of injury to neurons in mTBI.

285

Conclusions In mTBI 1. Primary axotomy, or shearing of axons, does not occur. 2. There is no support for the hypothesis that there is a rapid post-traumatic proteolysis of the entire axonal cytoskeleton in a signiﬁcant number of injured axons except in the most severe cases of diffuse head injury. 3. mTBI may produce only a temporary electrochemical disturbance of brain function caused by neuronal, chemical, and neuroelectrical changes and without development of overt pathology. 4. The great majority of injured axons in cases of mTBI, concussion and more severe levels of injury enter a ‘‘pathological cascade’’ of events that culminate in ‘‘secondary axotomy’’ at least 4 h and extend over many days after the initial trauma. 5. The use of modern, sensitive, techniques for analysis of plasma components (e.g., plasma levels of S-100b and NSE) allows demonstration of injury to axons in humans even after mTBI and detection of damaged axons at earlier post-traumatic intervals than have been possible until very recently. 6. The time course of formation of axonal swellings is longer in humans than in the majority of animal experimental models — there is no evidence for recruitment at present — it is just that reactive axons may be demonstrated up until 99 days after injury. 7. There is now a clear consensus that axonal changes leading to the development of pathology are initiated as a result of damage to the axolemma. This damage leads to disruption of ionic homeostatic mechanisms and/or a changed permeability of the axolemma. 8. Responses by microtubules and neuroﬁlaments differ in their time course both within a single axon, between axons in different tracts within the CNS and between species. The time course for loss of microtubules, compaction of neuroﬁlaments and secondary axotomy is longest in humans.

9. Recent evidence has suggested that not all injured axons proceed to secondary axotomy but that some demonstrate long-term compaction of neuroﬁlaments within their cytoskeleton. 10. There is increasing evidence, from both experimental, animal and clinical human studies that a mild or severe sudden acceleration–deceleration head trauma may result in cells of the brain assuming a vulnerable state for an unknown period. The fact that cells of the brain are in such a vulnerable state may predispose the affected athlete to an increased risk of a more serious and dangerous head injury if the athlete is exposed to a second head injury, for example a Second Impact Syndrome, if allowed to return to the ﬁeld of play during the same game. There are many Return-to-play Guidelines to help sport-physicians decide upon the initial management and determine when an injured athlete should return to play. However, considerable controversy exists and none of these guidelines has been developed on the basis of speciﬁc, scientiﬁc knowledge regarding the processes of either white matter injury or the recovery thereof. These guidelines rely ﬁrst of all upon self-reported symptoms. If the athlete is confused, the validity of the former has to be questioned. Generally, most current guidelines may allow the athlete to return to play after mTBI, even by the deepest grade I, in the same competition, if his/her clinical examination results are apparently normal at rest and with exertion. It is acknowledged that experts agree that no athlete who is still symptomatic should be allowed to return to competition and that ongoing and repeated examination should be conducted after the injury. However, the present review provides strong evidence that cells of the brain after mTBI may remain alive but in a vulnerable state for an, as yet, undetermined period. It is a major conclusion of the present study that the brain of the injured athlete may be in a vulnerable state, which may predispose the athlete to an increased risk of serious injury to the brain from a second head injury

286

(i.e., Second Impact Syndrome). In addition, since brain responses may be blunted resulting in a longer reaction time to changes in the environment, the injured athlete may be at increased risk of serious injury both to the brain and other parts of the body. Therefore it is recommended that any confused player, with or without amnesia, after receiving either a blow or any significant acceleration–deceleration force to the head, a grade I mTBI lesion, should be taken off the ice and should not be permitted to return to play for at least 72 h.

Acknowledgements We would like to express our deep appreciation to Prof. Patrick Bishop, Department of Kinesiology, University of Waterloo, Canada, for helping analyzing the video mechanisms of mTBI by the ice hockey players. We would also like to thank Mr. Yelverton Tegner, MD, Associated Professor at the Institution of Health Sciences at the Lulea˚ University of Technology, SE 961 36 Boden, Sweden, and Mr. Reto Agosti, MD, Swiss Specialist of Neurology, Director of the Headache Center Clinic Hirslanden, Zurich, Switzerland, who assisted us with suggestions and input to further improve the quality of this manuscript. Special thanks go to my resident Mr. Stephan Wirth, MD, Orthopedic University Hospital Zurich-Balgrist, Zurich, Switzerland, who helped me editing the text and the ﬁgures.

References Abdel-Dayem, H.M., Abu-Judeh, H., Kumar, M., et al. (1998) SPECT brain perfusion abnormalities in mild or moderate traumatic brain injury. Clin. Nucl. Med., 23(5): 309–317. Adams, J.H., Doyle, D., Ford, I., et al. (1989) Diffuse axonal injury in head injury: definition, diagnosis and grading. Histopathology, 15: 49–59.

Adams, J.H., Graham, D.I., Murray, L.S., et al. (1982) Diffuse axonal injury due to non-missile head in humans: an analysis of 45 cases. Ann. Neurol., 12: 557–563. Anderson, R.E., Hansson, L.O., Nilsson, O., et al. (2001) High serum S100B levels for trauma patients without head injuries. Neurosurgery, 48: 1255–1260. Bailes, J.E. and Cantu, R.C. (2001) Head injury in athletes. Neurosurgery, 48(1): 26–42. Bergsneider, M., Hovda, D.A., Lee, S.M., et al. (2000) Dissociation of cerebral glucose metabolism and level of consciousness during the period of metabolic depression following human traumatic brain injury. J. Neurotrauma, 17(5): 389–401. Biasca, N. (2000) Review and consequences of the typical ice hockey related injuries in the Swiss Ice Hockey Association. (SIHA) 1996–1998. In: Biasca N., Montag W.D. and Gerber Ch. (Eds.), Safety in Ice Hockey, Eighth International Symposium of the International Ice Hockey federation IIHF. International Ice Hockey Federation IIHF, Zurich, pp. 36–52. Biasca, N., Battaglia, H. and Simmen, H.P. (1995) Analyse typischer Eishockey-Verletzungen. SUVA Monographie Nr. 6, Schweizerische Unfallversicherungsanstalt SUVA, Abteilung Unfallmedizin, CH-6002 Luzern, Switzerland. Biasca, N., Battaglia, H. and Tegner, Y. (1999) Diagnose und Behandlungsvorschla¨ge der Commotio cerebri bei Kontaktund Kampf-sportarten: beispiel im Eishockey. Schweiz. A¨rtzezeit., 80(10): 595–598. Biasca, N. and Tegner, Y. (2001) International Sports Injury System — ISIS. Sports Medicine and Hockey; A summit for the NHL and Beyond, American Orthopedic Society for Sports Medicine (AOSSM) and National Hockey League Team Physician Society (NHLTPS), Hilton Toronto, Canada, August 24–26, pp. 15–36. Biasca, N., Wirth, S. and Tegner, Y. (2002) The avoidability of head and neck injuries in ice hockey: a historical review. Br. J. Sports Med., 36: 410–427. Biberthaler, P., Mussack, T., Wiedemann, E., et al. (2001) Evaluation of S-100b as a speciﬁc marker for neuronal damage due to minor head trauma. World J. Surg., 25: 93–97. Bigler, E.D. (1999) Neuroimaging in pediatric traumatic head injury: diagnostic considerations and relationships to neurobehavioral outcome. J. Head Trauma Rehabil., 14(4): 406–423. Bishop, P.J. (1997) Impact performance characteristics of hockey helmets with liners of differing thickness. In: Ashare A. (Ed.), Third International Symposium on Safety in Ice Hockey. American Society for Testing and Materials, Philadelphia, pp. 112–117. Bishop, P.J. (2000) Biomechanics of sports protective equipment in ice hockey to prevent injuries. In: Biasca N., Montag W.D. and Gerber Ch. (Eds.), Safety in Ice Hockey, Eight International Symposium of the International Ice Hockey federation IIHF. International Ice Hockey Federation IIHF, Zurich, pp. 36–52. Bishop, P.J. (2001) Protective equipment standards as a means of reducing injury risk. Sports Medicine and Hockey; A summit for the NHL and Beyond, American Orthopedic

287 Society for Sports Medicine (AOSSM) and National Hockey League Team Physician Society (NHLTPS), Hilton Toronto, Canada, August 24–26, p. 57. Bishop, P.J. and Arnold, J. (1993) The effectiveness of hockey helmets in limiting localized loading on the head. In: Castaldi C.R., Bishop P.J. and Hoerner E.F. (Eds.), Safety in Ice Hockey: Second Volume, ASTM STP 1212. American Society for Testing and Materials, Philadelphia, pp. 175–182. Blumbergs, P.C. (1997) Pathology, Head Injury. In: Reilly P. and Bullock R. (Eds.). Chapman and Hall, London, pp. 39–70. Blumbergs, P.C., Scott, G., Manavis, J., et al. (1994) Staining of amyloid precursor protein to study axonal damage in mild head injury. Lancet, 344: 1055–1056. Blumbergs, P.C., Scott, G., Manavis, J., et al. (1995) Topography of axonal injury as deﬁned by amyloid precursor protein and the sector scoring method in mild and severe closed head injury. J. Neurotrauma, 12: 565–572. Brooks, W.M., Frieman, S.D. and Gasparovic, C. (2001) Magnetic resonance spectroscopy in traumatic brain injury. J. Head Trauma Rehabil., 16(2): 149–164. Buki, A., Siman, R., Trojanowski, J.Q., et al. (1999) The role of calpain-mediated spectrin proteolysis in traumatically induced axonal injury. J. Neuropathol. Exp. Neurol., 58: 365–375. Burke, C.J. (2001) Definition, incidence and initial assessment of concussion. Sports Medicine and Hockey; A summit for the NHL and Beyond, American Orthopedic Society for Sports Medicine (August 24–26) (AOSSM) and National Hockey League Team Physician Society (NHLTPS), Hilton Toronto, Canada, pp. 107–108. Cantu, R. (1992) Cerebral concussion in sport: management and prevention. Sports Med., 14: 64–74. Castaldi, C.R. (1993) The sports mouthguard: its use and misuse in ice hockey. In: Castaldi C.R., Bishop P.J. and Hoerner E.F. (Eds.), Safety in Ice Hockey: Second Volume, ASTM STP 1212. American Society for Testing and Materials, Philadelphia, pp. 164–174. Cecil, K.M., Hills, E.C., Sandel, M.E., et al. (1998) Proton magnetic resonance spectroscopy for detection of axonal injury in the splenium of the corpus callosum of brain-injured patients. J. Neurosurg., 88: 795–801. Christman, C.W., Grady, M.S., Walker, S.A., et al. (1994) Ultrastructural studies of diffuse axonal injury in humans. J. Neurotrauma, 11: 173–186. Clark, J.M. (1974) Distribution of microglial clusters in the brain after head trauma injury. J. Neurol. Neurosurg. Psychiatry, 37: 463–474. Clayton, P. (1997) A review of head injury data in Canadian hockey league with an emphasis on the incidence of concussion. In: Ashare A. (Ed.), Third International Symposium on Safety in Ice Hockey. American Society for Testing and Materials, Philadelphia Abstract. D’Avella, D., Aguennouz, M., Angileri, F.F., et al. (2004) Detection of a-II spectrin and breakdown products in humans after severe traumatic brain injury. Acad. Emerg. Med., 11: 515–520.

De Kruijk, J.R., Leffers, P., Menheere, P.P., et al. (2001) S100B and neuron-speciﬁc enolase in serum of mild traumatic brain injury patients. Acta Neurol. Scand., 103: 175–179. Denny, B. and Russel, W.R. (1941) Experimental cerebral concussion. Brain, 64: 93–164. Dick, R.W. (1997) Injury in collegiate ice hockey with an emphasis on injuries to the head. In: Ashare A. (Ed.), Third International Symposium on Safety in Ice Hockey. American Society for Testing and Materials, Philadelphia Abstract. Dixon, J.L. and Brodie, I.K.R. (1993) The new ISO Standards for ice hockey helmets and face protectors: moving toward International Standards harmonization and conformity. In: Castaldi C.R., Bishop P.J. and Hoerner E.F. (Eds.), Safety in Ice Hockey: Second Volume, ASTM STP 1212. American Society for Testing and Materials, Philadelphia, pp. 192–213. Elson, L.M. and Ward, C.C. (1994) Mechanisms and pathophysiology of mild head injury. Semin. Neurol., 14(1): 8–18. Erb, D.E. and Povlishock, J.T. (1988) Axonal damage in severe traumatic brain injury: an experimental study in the cat. Acta Neuropathol., 76: 347–358. Farkas, O., Polgar, B., Szekeres-Bartho, J., et al. (2005) Spectrin breakdown products in the cerebrospinal ﬂuid in severe head injury — preliminary observations. Acta Neurochir., 147: 855–861. Fineman, I., Hovda, D.A., Smith, M., et al. (1993) Concussive brain injury is associated with a prolonged accumulation of calcium: a 45Ca autoradiographic study. Brain Res., 624: 94–102. Galbraith, J.A., Thibault, L.E. and Matheson, R.A. (1993) Mechanical and electrical responses of the squid giant axon to simple elongation. J. Biomech. Eng., 115: 13–22. Garnett, M.R., Blamire, A.M., Corkill, R.G., et al. (2000) Early proton magnetic resonance spectroscopy in normal-appearing brain correlates with outcome in patients following traumatic brain injury. Brain, 123: 2046–2054. Gennarelli, T.A. (1991) Cerebral concussion and diffuse brain injuries. In: Torg J.S. (Ed.), Athletic Injuries to the Head, Neck and Face (2nd ed.). Mosby Year Book, St. Louis, MO, pp. 270–282. Gennarelli, T.A. and Thibault, L.E. (1982) Biomechanics of acute subdural hematoma. J. Trauma, 22: 680–686. Gennarelli, T.A., Thibault, L.E., Adams, J.H., et al. (1982) Diffuse axonal injury and traumatic coma in the primate. Ann. Neurol., 12: 564–574. Gennarelli, T.A., Thibault, L.E. and Graham, D.I. (1998) Diffuse axonal injury: an important form of traumatic brain damage. Neuroscientist, 4(3): 202–215. Gennarelli, T.A., Thibault, L.E., Tipperman, R., et al. (1989) Axonal injury in the optic nerve: a model simulating diffuse axonal injury in man. J. Neurosurg., 71: 244–253. Gentleman, S.M., Roberts, G.W., Gennarelli, T.A., et al. (1995) Axonal injury: a universal consequence of fatal closed head injury? Acta Neuropathol., 89: 537–543. Giza, C.C. and Hovda, D.A. (2000) Ionic and metabolic consequences of concussion. In: Cantu R.C. (Ed.), Neurologic Athletic Head and Spine Injuries. WB Saunders Company, New York, pp. 80–100.

288 Grady, M.S., Mclaughlin, M.R., Christman, C.W., et al. (1993) The use of antibodies targeted against the neuroﬁlaments subunits for the detection of diffuse axonal injury in humans. J. Neuropathol. Exp. Neurol., 52: 143–152. Graham, D.I. and Gennarelli, T.A. (1997) Trauma. In: Graham D.I. and Lantos P.L. (Eds.), Chapter 5 in Greenﬁeld’s Neuropathology (6th ed.). Arnold, London, pp. 197–262. Graham, D.I., Mcintosh, T.K., Maxwell, W.L., et al. (2000) Recent advances in neurotrauma. J. Neuropathol. Exp. Neurol., 59(8): 641–651. Gross, H., Kling, A., Henry, G., et al. (1996) Local cerebral glucose metabolism in patients with long-term behavioral and cognitive deficits following mild traumatic brain injury. J. Neuropsychiatry Clin. Neurosci., 8(3): 324–334. Hartmann, T. (2001) Cholesterol, Ab and Alzheimer’s disease. TINS, 24: S45–S48. Hermann, M., Curio, N., Jost, S.T., et al. (1999) Protein S-100B and neuron speciﬁc enolase as early neurobiochemical markers of the severity of traumatic brain injury. Restor. Neurol. Neurosci., 14: 109–114. Herrmann, M., Curio, N., Jost, S., et al. (2001) Release of biochemical markers of damage to neuronal and glial brain tissue is associated with short and long term neuropsychological outcome after traumatic brain injury. J. Neurol. Neurosurg. Psychiatry, 70(1): 95–100. Hermann, M., Jost, S.T., Kutz, S., et al. (2000) Temporal proﬁle of release of neurobiochemical markers of brain damage after traumatic brain injury is associated with intracranial pathology as demonstrated in cranial computerized tomography. J. Neurotrauma, 17(2): 113–122. Holbourn, A.H.S. (1943) Mechanics of head injury. Lancet, 2: 438–441. Hovda, D.A. (1996) Metabolic dysfunction. In: Narayan R.K., Wiberger J.E. and Povlishock J.T. (Eds.), Neurotrauma. McGraw-Hill, Inc, New York, pp. 1459–1478. Hovda, D.A., et al. (1994) Long-term changes in metabolic rates for glucose following mild, moderate and severe concussive head injuries in adult rats. Soc. Neurosci. Abstract, 20: 845. Hovda, D.A., Prins, M., Becker, D.P., et al. (1999) Neurobiology of concussion. In: Bailes J.E., Lovell M.R. and Maroon J.C. (Eds.), Sports-related Concussion. Quality Medical Publishing, Inc., Pittsburgh, PA, pp. 13–51. Ingebrigsten, T. (1998) Aspect on the management of minor head injury. Habilitation Work, Department of neurosurgery, Institute of clinical medicine. University of Tromso, Norway. Ingebrigsten, T., Kongstad, B.R.P. and Langbakk, B. (1995) Increased serum concentrations of protein S-100 after minor head injury: a biochemical serum marker with prognostic value? J. Neurol. Neurosurg. Psychiatry, 59: 103–104. Ingebrigtsen, T., Romner, B., Marup-Jensen, S., et al. (2000) The clinical value of serum S-100 protein measurements in minor head injury: a Scandinavian multicentre study. Brain Inj., 14(12): 1047–1055. Ingebrigsten, T., Rommer, B. and Trumpy, J.H. (1997) Management of minor head injury: the value of early computed

tomography and serum proteion S-100 measurements. J. Clin. Neurosci., 4: 29–33. Ingebrigtsen, T., Waterloo, K., Jacobsen, E.A., et al. (1999) Traumatic brain damage in minor head injury: relation of serum S-100 protein measurements to magnetic resonance imaging and neurobehavioral outcome. Neurosurgery, 45: 468–476. Jafari, S.S., Maxwell, W.L., Neilson, M., et al. (1997) Axonal cytoskeletal changes after non-disruptive axonal injury. J. Neurocytol., 26: 207–221. Jafari, S.S., Neilson, M., Graham, D.I., et al. (1998) Axonal cytoskeletal changes after non-disruptive axonal injury. II. Intermediate sized axons. J. Neurotrauma, 15: 955–966. Katayama, Y., Becker, D.P., Tamura, T. and Hovda, D.A. (1990) Massive increases in extracellular potassium and the indiscriminate release of glutamate following concussive brain injury. J. Neurosurg., 73: 889–900. Kawamata, T., Katayama, Y., Hovda, D.A., et al. (1992) Administration of excitatory amino acid antagonists via microdialysis attenuated the increased in glucose utilization seen following concussive brain injury. J. Cereb. Blood Flow Metab., 12: 12–24. Kelly, J.P. (1999) Traumatic brain injury and concussion in sports. JAMA, 282: 989–991. Kelly, J.P., Nichols, J.S., et al. (1991) Concussions in sports: guidelines for the prevention of catastrophic outcome. JAMA, 266: 2867–2869. Klatky, R., Furuya, Y., Valadka, A.B., Goodman, J.C. and Robertson, C.S. (2002) Comparison of microdialysate arginine and glutamate levels in severely head-injured patient. Acta Neurochir. Suppl., 81: 347–349. Langford, G.M. (1978) In vitro assembly of dogﬁsh brain tubulin and induction of coiled ribbon polymers by calcium. Exp. Cell Res., 111: 139–151. Laprade, R.F. (2001) Ice hockey injuries at the collegiate level. Sports Medicine and Hockey; A summit for the NHL and Beyond, American Orthopedic Society for Sports Medicine (AOSSM) and National Hockey League Team Physician Society (NHLTPS), Hilton Toronto, Canada, August 24–26, pp. 37–45. Levine, B., Cabezy, R., Mcintosh, A.R., et al. (2002) Functional reorganization of memory after traumatic brain injury: a study with H(2)(15)O positron emission tomography. J. Neurol. Neurosurg. Psychiatry, 73(2): 173–181. Margulies, S.S. and Thibault, L.E. (1989) An analytical model of traumatic diffuse brain injury. J. Biomech. Eng., 111: 241–249. Maroon, J.C., Lovell, M.R., Norwig, J., et al. (2000) Cerebral concussion in athletes: evaluation and neuropsychological testing. Neurosurgery, 47(3): 659–672. Matsumura, F. and Hayashi, M. (1976) Polymorphism of tubulin assembly. In vitro formation of sheet, twisted ribbon and microtubule. Biochim. Biophys. Acta, 453: 162–175. Matsushita, Y., Shima, K., Katoh, H. and Nawashiro, H. (2000) The synergistic effect of high K+ and hypoxia on extracellular concentrations of neuroactive amino acid in hippocampus. Acta Neurochir. Suppl., 76: 247–249.

289 Maxwell, W.L. (1999) Cellular responses to ischaemic CNS injury. In: Berry M. and Logan A. (Eds.), Pharmacological strategies in the treatment of CNS injuries. CRC Press, Baton Rouge, pp. 19–41. Maxwell, W.L., Domleo, A., McColl, G., Jafari, S.S. and Graham, D.I. (2003) Post-acute alterations in the axonal cytoskeleton after traumatic axonal injury. J. Neurotrauma, 20: 151–168. Maxwell, W.L. and Graham, D.I. (1997) Loss of axonal microtubules and neuroﬁlaments after stretch-injury to guinea pig optic nerve ﬁbers. J. Neurotrauma, 14: 603–614. Maxwell, W.L., Kosanlavit, R., McCreath, B.J., et al. (1999) Freeze-fracture and cytochemical evidence for structural and functional alteration in the axolemma and myelin sheath of adult guinea pig optic nerve ﬁbers after stretch injury. J. Neurotrauma, 16: 273–284. Maxwell, W.L., Mccreath, B.J., Graham, D.I., et al. (1995) Cytochemical evidence for redistribution of membrane pump calcium-ATPase and ecto-Ca-ATPase activity, and calcium inﬂux in myelinated nerve ﬁbers of the optic nerve after stretch injury. J. Neurocytol., 24: 925–942. Maxwell, W.L., Povlishock, J.T. and Graham, D.I. (1997) A mechanistic analysis of nondisruptive axonal injury: a review. J. Neurotrauma, 14(7): 419–440. Maxwell, W.L., Whitﬁeld, P., Suzen, B., et al. (1992) The cerebrovascular response to experimental lateral head acceleration. Acta Neuropathol., 84: 289–296. McGehee, D.V. (1996) Head injury in motor vehicle crashes: human factors, effects, and prevention. In: Rizzo M. and Tranel D. (Eds.), Head Injury and Postconcussive Syndrome. Churchill Livingstone, London and New York, pp. 57–69. McIntosh, T.K., Faden, A.I., Bendall, M.R. and Vink, R. (1987) Traumatic brain injury in the rat: alterations in brain lactate and ph as characterized by H and P nuclear magnetic resonance. J. Neurochem., 49: 1530–1540. McIntosh, T.K., Faden, A.I., Yamakami, I. and Vink, R. (1988) Magnesium deﬁciency exacerbates and pretreatment improves outcome following traumatic brain injury in the rat. J. Neurotrauma, 5: 17–31. McKeating, E.G., Andrew, P.J. and Mascia, L. (1998) Relationship of neuron speciﬁc enolase and protein S-100 concentrations in systemic and jugular venous serum to injury severity and outcome after traumatic brain injury. Acta Neurochir., 71: 117–119. McKenzie, K.J., Mclellan, D.R., Gentleman, S.M., et al. (1996) Is b-APP a marker of axonal damage in short-surviving head injury? Acta Neuropathol., 92: 608–613. Moore, B.W. (1965) A soluble protein characteristic of the nervous system. Biochem. Biophys. Res. Commun., 19: 739–744. Muller, M., Schwerdtfegler, K., Maier, B., et al. (2001) Cerebral blood ﬂow velocity and inﬂammatory response after severe traumatic brain injury. Eur. J. Ultrasound, 12(3): 203–208. Nedd, K., Sfakianakis, G., Ganz, W., et al. (1993) 99mTcHMPAO SPECT of the brain in mild to moderate traumatic brain injury patients: compared with CT, a prospective study. Brain Inj., 7(6): 469–479.

Nevin, N.C. (1967) Neurophathological changes in the white matter following head injury. J. Neuropathol. Exp. Neurol., 26: 77–84. Ng, S.C., Poon, W.S., Chan, M.T., et al. (2000) Transcranial Doppler ultrasonography (TCD) in ventilated head injure patients: correlation with stable xenon-enhanced CT. Acta Neurochir. Suppl., 76: 479–482. Odelgard, B. (1989) The development of head, face, and neck protectors for ice hockey players. In: Castaldi C.R. and Hoerner E.F. (Eds.), Safety in Ice Hockey: Second Volume, ASTM STP 1050. American Society for Testing and Materials, Philadelphia, pp. 220–234. Okonkwo, D.O., Buki, A., Siman, R. and Povlishock, J.T. (1999) Cyclosporin A limits calcium-induced axonal damage following traumatic brain injury. Neuroreport, 10: 353–358. Okonkwo, D.O. and Povlishock, J.T. (1999) An intrathecal bolus of cyclosporin a before injury preserves mitochondrial integrity and attenuates axonal disruption in traumatic brain injury. J. Cereb. Blood Flow Metab., 19: 443–451. Oppenheimer, D.R. (1968) Microscopic lesions in the brain following head injury. J. Neurol. Neurosurg. Psychiatry, 31: 299–306. Pappius, H.M. (1981) Local cerebral glucose utilization in thermically traumatized rat brain. Ann. Neurol., 9: 484–491. Pashby, T. (1993) Eye protection in ice hockey: an historical review. In: Castaldi C.R., Bishop P.J. and Hoerner E.F. (Eds.), Safety in Ice Hockey: Second Volume, ASTM STP 1212. American Society for Testing and Materials, Philadelphia, pp. 159–163. Pashby, T., Carson, J.D., Ordogh, D., et al. (2001) Eliminate headchecking in ice hockey. Clin. J. Sport Med., 11(4): 211–213. Peerless, S.J. and Rewcastle, N.B. (1967) Shear injuries of the brain. J. Can. Med. Assoc., 96: 577–582. Pettus, E.H., Christman, C.W., Giebel, M.L., et al. (1994) Traumatically induced altered membrane permeability: its relationship traumatically induced reactive axonal changes. J. Neurotrauma, 11(5): 507–522. Pettus, E.H. and Povlishock, J.T. (1996) Characterization of a distinct set of intra-axonal ultrastructural changes associated with traumatically induced alteration in axolemma permeability. Brain Res., 772: 1–11. Piccininni, P. (2001) Mouthguards and concussion prevention, International Symposium on Concussion in Sport. Br. J. Sports Med., 35(5): 375. Pleines, U.E., Morganti-Kossmann, M.C., Rancan, M., et al. (2001) S-100B reﬂects the extent of injury and outcome, whereas NSE is a better indicator of neuroinﬂammation in patients with severe traumatic brain injury. J. Neurotrauma, 18(5): 491–498. Povlishock, J.T. (1986) Traumatically induced axonal damage without concomitant change in focally related neuronal somata and dendrites. Acta Neuropathol., 70: 53–59. Povlishock, J.T. (1992) Traumatically induced axonal injury: pathogenesis and pathobiological implications. Brain Pathol., 2: 1–12.

290 Povlishock, J.T. (1993) Pathobiology of traumatically induced axonal injury in animals and man. Ann. Emerg. Med., 22(6): 980–986. Povlishock, J.T., Becker, D.P., Cheng, C.L.Z., et al. (1983) Axonal changes in minor head injury. J. Neuropathol. Exp. Neurol., 42(3): 225–242. Povlishock, J.T. and Christman, C.W. (1995) The pathobiology of traumatically induced axonal injury in animals and humans: a review of current thoughts. J. Neurotrauma, 12(4): 555–564. Povlishock, J.T. and Jenkins, L.W. (1995) Are the pathobiological changes evoked by traumatic brain injury immediate and irreversible? Brain Pathol., 5: 415–426. Povlishock, J.T., Marmarou, A., McIntosh, T., et al. (1997) Impact acceleration injury in the rat: evidence for focal axolemmal and related neuroﬁlament sidearm alteration. J. Neuropathol. Exp. Neurol., 56: 347–359. Povlishock, J.T. and Pettus, E.H. (1996) Traumatically induced axonal damage: evidence for enduring changes in axolemmal permeability with associated cytoskeletal changes. Acta Neurochir., 66: 81–86. Raabe, A., Grolms, C., Keller, M., et al. (1998) Correlations of computed tomography ﬁndings and serum brain damage markers following severe head injury. Acta Neurochir., 140: 787–792. Ranalli, D.N. (2000) Prevention of sports-related traumatic dental injuries. Advances in Sports dentistry. Dent. Clin. North Am., 44(1): 35–51. Reinert, M., Hoelper, B., Doppenberg, E., Zauner, A. and Bullock, R. (2000) Substrate delivery and ionic balance disturbance after severe human head injury. Acta Neurochir. Suppl., 76: 439–444. Ricker, J.H., Muller, R.A., Zafonte, R.D., et al. (2001) Verbal recall and recognition following traumatic brain injury: a (0–15)-water positron emission tomography study. J. Clin. Exp. Neuropsychol., 23(2): 196–206. Ringger, N.C., O’Steen, B.E., Brabham, J.G., et al. (2004) A novel marker for traumatic brain injury: CSF aII-spectrin breakdown product levels. J. Neurotrauma, 21: 1443–1456. Roberts, W.O., Brust, J.D. and Leonard, B. (1999) Youth ice hockey tournament injuries: rates and patterns compared to season play. Med. Sci. Sports Exerc., 31(1): 46–51. Romner, B., Ingebrigtsen, T., Kongstad, P., et al. (2000) Traumatic brain damage: serum S-100 protein measurements related to neuroradiological ﬁndings. J. Neurotrauma, 17(8): 641–647. Ross, S.A., Cunningham, R.T., Johnston, C.F., et al. (1996) Neuron-speciﬁc enolase as an aid to outcome prediction in head injury. Br. J. Neurosurg., 10: 471–476. Ross, D.T., Meaney, D.F., Sabol, M.K., et al. (1995) Distribution of forebrain diffuse axonal injury following inertial closed head injury in miniature swine. Exp. Neurol., 126: 291–299. Ruff, R.M., Crouch, J.A., Troster, A.I., et al. (1994) Selected cases of poor outcome following a minor brain trauma: comparing neuropsychological and positron emission tomography assessment. Brain Inj., 8(4): 297–308.

Samii, A., Badie, H., Fu, K., et al. (1999) Effects of an N-Type calcium channel antagonist on calcium-45 accumulation following ﬂuid-percussion injury. J. Neurotrauma, 16(10): 879–892. Sherriff, F.E., Bridges, L.R., Gentleman, S.M., et al. (1994b) Markers of axonal injury in post mortem human brain. Acta Neuropathol., 88: 433–439. Sherriff, F.E., Bridges, L.R. and Sivologanathan, S. (1994a) Early detection of axonal injury after human head trauma using immunocytochemistry for b-amyloid precursor protein. Acta Neuropathol., 87: 55–62. Siman, R., McIntosh, T.K., Soltesz, K.M., et al. (2004) Proteins released from degenerating neurons are surrogate markers for acute brain damage. Neurobiol. Dis., 16: 311–320. Sinson, G., Bagley, L.J., Cecil, K.M., et al. (2001) Magnetization transfer imaging and proton MR spectroscopy in the evaluation of axonal injury: correlation with clinical outcome after traumatic brain injury. Am. J. Neuroradiol., 22(1): 143–151. Skogseid, H.M., Nordby, H.K., Urdal, P., et al. (1992) Increased serum creatine kinase BB and neuron-speciﬁc enolase following head injury indicates brain damage. Acta Neurochir., 115: 106–111. Smith, D.H., Nonaka, M., Miller, R., et al. (2000) Immediate coma following inertial brain injury dependent on axonal damage in the brainstem. J. Neurosurg., 93: 315–322. Sports-Related Recurrent Brain Injuries-United States. (1997) MMWR Morb. Mortal Wkly. Rep., 46: 224–227. Stone, J.R., Singleton, R.H. and Povlishock, J.T. (2001) Intraaxonal neuroﬁlament compaction does not evoke local axonal swelling in all traumatically injured axons. Exp. Neurol., 172: 320–331. Stoner, L.J. and Kreating, M. (1993) Hockey equipment: safety or an illusion? In: Castaldi C.R., Bishop P.J. and Hoerner E.F. (Eds.), Safety in Ice Hockey: Second Volume, ASTM STP 1212. American Society for Testing and Materials, Philadelphia, pp. 183–191. Strich, S.J. (1961) Shearing of nerves ﬁbers as a cause of brain damage due to head injury: a pathological study of twenty cases. Lancet, 2: 443–448. Strich, S.J. (1970) Lesions in the cerebral hemispheres after blunt head injury. J. Clin. Pathol., 23(4): 166–171. Tegner, Y. (2000) Brain concussion in ice hockey: a suggestion for uniform treatment. In: Biasca N., Montag W.D. and Gerber Ch. (Eds.), Safety in Ice Hockey, Eight International Symposium of the International Ice Hockey federation IIHF. International Ice Hockey Federation IIHF, Zurich, pp. 86–89. Tegner, Y. (2001) Concussion experiences: Swedish elite ice hockey league, International Symposium on Concussion in Sport. Br. J. Sports Med., 35(5): 376. Tegner, Y., Back, B., Gilenstam, K. and Holm, A. (2000) Injury pattern in Swedish Elite Ice Hockey. Hygiae, 109(2): 204. Tegner, Y. and Lorentzon, R. (1996) Concussion among Swedish elite ice hockey players. Br. J. Sports Med., 30(3): 252–255.

291 Tomei, G., Spagnoli, D., Ducati, A., et al. (1990) Morphology and neurophysiology of focal axonal injury experimentally induced in the guinea pig optic nerve. Acta Neuropathol., 80: 506–513. Woertgen, C.H., Rothoerl, R.D., Holzschuh, M., et al. (1997) Comparison of serial S-100 and NSE serum measurements after severe head injury. Acta Neurochir., 139: 1161–1165. Yaghmai, A. and Povlishock, J.T. (1992) Traumatically induced reactive change as visualized through the use of monoclonal antibodies targeted to the neuroﬁlament subunits. J. Neuropathol. Exp. Neurol., 51: 158–176.

Yamamoto, T., Rossi, S., Stiefel, M., et al. (1999) CSF and ECF glutamate concentrations in head injured patients. Acta Neurochir. Suppl., 75: 17–19. Yamazaki, Y., Yada, K., Morii, S., et al. (1995) Diagnostic significance of serum neuron-speciﬁc enolase and myelin basic protein assay in patients with acute head injury. Surg. Neurol., 43: 267–270. Umile, E.M., Sandel, M.E., Alavi, A., et al. (2002) Dynamic imaging in mild traumatic brain injury: support for the theory of medial temporal vulnerability. Arch. Phys. Med. Rehabil., 83(11): 1506–1513.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 20

Traumatic brain injury in infants: the phenomenon of subdural hemorrhage with hemispheric hypodensity (‘‘Big Black Brain’’) Ann-Christine Duhaime and Susan Durham Pediatric Neurosurgery, Children’s Hospital at Dartmouth, Dartmouth Hitchcock Medical Center, Lebanon, NH 03756, USA

Abstract: Clinical and experimental studies of traumatic brain injury during immaturity have been far less numerous than those involving adults, and many questions remain about differences in injury responses among patients of different ages. This chapter reviews a distinctive injury pattern common in infants, the so-called ‘‘big black brain’’ response to acute subdural hematoma. The pathophysiology of this injury remains incompletely understood. Insights from both clinical observation and experimental studies have helped to clarify the probable causes of this injury pattern, which appears to require a combination of stressors during a particular period of maturation. Keywords: subdural hematoma; infant; black brain; child abuse; pathophysiology how clinicians and scientists work synergistically in head injury research.

Introduction There are a number of unique phenomena seen in clinical practice that distinguish infant traumatic brain injury from that occurring in older children or adults. One of the most dramatic of these is the so-called ‘‘big black brain,’’ a pattern of tissue loss affecting the entire supratentorial hemisphere in association with acute subdural hematoma. The circumstances required to cause this injury pattern and the pathophysiology of the extensive parenchymal destruction remain incompletely understood. This chapter will address insights gained into this injury by both clinical observations and experimental strategies. Both approaches have contributed to advancing understanding of the response of the infant to brain trauma, and illustrate

Early clinical observations It has long been recognized that acute subdural hematoma in infancy could be associated with profound brain injury. The pioneering observations of Guthkelch (1971) and Caffey (1972, 1974) linked infantile subdural with a range of neurologic deﬁcits including chronic disability, coma, and death. These injuries were hypothesized to occur most often from violent manual shaking. The hypothesis that shaking caused these injuries was based on statements from some perpetrators and witnesses who described shaking, the frequent paucity of visible external cranial injury in affected infants, an association with distal metaphyseal and rib fractures, and contemporaneous experimental ﬁndings on the role of rotational forces in subdural

Corresponding author. Tel.: +1 603-653-9880;

Fax: +1 603-650-0908; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61020-0

293

294

hematoma and diffuse brain injuries (Ommaya and Yarnell, 1962; Ommaya et al., 1968, 1973). The forces and mechanisms necessary to cause acute subdural hematoma in infancy remain uncertain, and many affected infants with accidental or inﬂicted events do, in fact, have evidence of head impact (Hahn et al., 1983; Duhaime et al., 1987, 1998; Alexander et al., 1990; Ghahreman et al., 2005). In this chapter we will focus on a speciﬁc pattern of immature brain response to acute subdural hematoma, regardless of mechanism of injury. Not all infants with acute subdural hematoma develop severe brain injury. The spectrum of clinical outcomes ranges from no deﬁcits to death (Ludwig and Warman, 1984; Duhaime et al., 1987, 1996; Gilles and Nelson, 1998; Ewing-Cobbs et al., 1999; Ghahreman et al., 2005). As CT scanning became widely available, patterns of brain damage emerged in the premorbid state and in survivors. Areas of hypodensity often were seen which might be patchy in distribution or could involve the entire supratentorial compartment unilaterally or bilaterally (Zimmerman et al., 1979; Cohen et al., 1986; Dias et al., 1998; Gilles and Nelson, 1998). Because of the typical dark-appearing, homogeneous, and

extensively distributed low density seen on the CT scan of such infants, the term ‘‘big black brain’’ has been used to describe those with hypodensity involving the entirety of one or both hemispheres (Duhaime et al., 1993; Graupman and Winston, 2006). This nomenclature evolved from the use of the term ‘‘black brain’’ to describe the CT appearance of diffuse bilateral hypodensity and loss of gray-white differentiation, with or without relative sparing of the thalami (‘‘reversal sign’’), seen most often in cases of severe or fatal pediatric head injury (Cohen et al., 1986; Whyte and Pascoe, 1989). In a subset of infant subdurals, hypodensity is seen which extends from the frontal pole to the occipital pole, crosses multiple vascular territories, and stops abruptly at the tentorium. This pattern may be present at presentation to the hospital or evolve over several days (Dias et al., 1998). In unilateral cases, a contralateral frontal wedge-shaped region of low density also typically is seen, but the rest of the supratentorial compartment does not exhibit hypodensity or loss of gray-white differentiation (Fig. 1). The hypodense hemisphere(s) typically go on to sustain rapidly progressive atrophic changes indicative of severe, diffuse brain destruction

Fig. 1. Two-year-old boy with a fall from a bunk bed resulting in acute left subdural hematoma. (A) Acute CT scan, with midline shift. Hemispheres appear to have similar density at this time-point. The patient underwent emergency clot evacuation. (B) Four days postinjury. The child has developed hypodensity and swelling of the entire left hemisphere. He underwent aggressive management of intracranial pressure and survived with a right hemiparesis and hemianopsia.

295

Fig. 2. Twenty-three month old boy who sustained a witnessed accidental injury with an acute right subdural hematoma. He stood up initially but then had a generalized seizure within seconds after impact. He received medical attention within minutes, and there was no apnea or hypoventilation noted. (A) Acute CT scan. (B) CT scan after clot evacuation. There was no brain swelling at surgery. Note lack of hypodensity, and bifrontal air. The patient regained consciousness promptly after surgery. Despite the symmetric radiologic appearance of the hemispheres, the child was noted to have a left hemiparesis postoperatively. (C) MRI scan done on postoperative day 2, FLAIR sequence. Note the high signal in the basal ganglia. (D) MRI, T2 weighted sequence, several months post-injury. Note the widespread signal change and atrophy in the right hemisphere. (E) MRI, T1 sequence with contrast, 10 years after injury. There is marked atrophy of the right hemisphere. The patient had a persistent hemiparesis and hemianopsia, developmental delays, and a seizure disorder. He became seizure-free with no new deﬁcits after hemispherectomy.

(Fig. 2). In children with subdural hematomas, the bilateral pattern is seen about twice as commonly as the unilateral pattern, and occurs more commonly in very young infants, while older infants and toddlers more often develop the unilateral form of ‘‘big black brain’’ (Duhaime et al., 1993; Gilles and Nelson, 1998). The extent of the

hypodensity correlates with worse acute clinical status and with worse prognosis (Duhaime et al., 1996; Ewing-Cobbs et al., 1998; Gilles and Nelson, 1998). Mortality in children with unilateral or bilateral hemispheric involvement is 67% (Duhaime et al., 1993). Even many years after injury, survivors of bilateral ‘‘big black brain’’ remain blind,

296

non-ambulatory, nonverbal, and profoundly developmentally delayed. Those with the predominantly unilateral form may make a better functional recovery but remain severely impaired (Duhaime et al., 1996). While many workers have described the areas of hypodensity seen on CT as ‘‘edema’’ or ‘‘infarction,’’ these terms should be used with caution. This is because the pathophysiologic processes these terms most often connote, in other ages and with adult disorders, may not apply in this context. For instance, excitotoxic lesions, to which the immature brain may be particularly vulnerable, can occur in the setting of normal vascular perfusion. The pathology and radiology of this type of process may look similar to that occurring from vascular occlusion, since both represent a mismatch between metabolic demand and substrate delivery. If such a pathophysiology were in play, it might not be necessary to postulate a mechanism requiring large vessel occlusion, such as strangulation, as some authors have done (Bird et al., 1987). Thus, care should be taken to describe ﬁndings rather than ascribing mechanisms that might be associated with these radiologic ﬁndings in other ages and clinical contexts (Gilles and Nelson, 1998). Older children and adults with acute subdural hematoma may suffer extensive concomitant brain damage, presumably due to the effects of primary mechanical brain injury and/or to secondary insults including hypoxia, ischemia, or elevated intracranial pressure. However, the rapid appearance of extensive, diffuse supratentorial hypodensity with loss of gray-white differentiation in the unilateral or bilateral patterns described is seen routinely only in infants and toddler-aged children. The vast majority of children with the speciﬁc injury type reviewed in this paper are under 3 years of age. For the sake of brevity we most often use the term ‘‘infants’’ in this chapter, but it should be understood that this phenomenon occurs in toddlers as well. More recent studies using magnetic resonance imaging (MRI) have shown changes on diffusionweighted imaging which corroborate the idea that some form of hypoxic-ischemic injury or perfusiondemand mismatch is particularly common in inﬂicted injury (Ichord et al., 2007). Early involvement

of the basal ganglia can also been seen on MRI (Figs. 2 and 3). Therefore, the question has arisen as to whether this striking injury pattern reﬂects a unique mechanism of injury, the presence of a particular type of secondary injury, or a unique response of the immature brain to trauma. Since bilateral ‘‘black brain’’ can occur from diffuse hypoxicischemic insults as well as in the setting of subdural hematomas, we will focus on the unilateral pattern as the traumatic entity requiring a unique pathophysiologic explanation. Both experimental and clinical investigations offer insights, as reviewed below.

Experimental models The rodent subdural hematoma model most widely in use presently was ﬁrst described by Miller et al. (1990). The model created a subdural hematoma by injecting autologous blood into the subdural space of the cerebral convexity through a small burr hole. Histopathology demonstrates a bowl-shaped volume of brain damage underlying the clot. By measuring brain metabolism and cerebral blood ﬂow, investigators have demonstrated that perfusion in the area of the clot is decreased, while metabolism is increased, creating a relative mismatch between substrate delivery and demand. The perfusion abnormality appears to be mediated by microvessel vasospasm rather than by large vessel abnormalities (Inglis et al., 1990; Kuroda and Bullock, 1992). Microdialysis studies with the model have shown an increase in extracellular glutamate, and treatment with glutamate receptor blockers limits the volume of damage, suggesting that the increased metabolism seen under the hematoma is related to excitotoxic stress (Bullock et al., 1990, 1991; Inglis et al., 1992). Similar damage was not seen if a comparable volume of blood was simply layered over the exposed cortical surface in rodents and a clot left in place for several days (Duhaime et al., 1992). This demonstrates that the presence of blood alone is insufﬁcient to cause damage, but that the blood must be injected into a closed space to cause injury. While these studies are intriguing, the pathology using the rodent subdural hematoma model does

297

Fig. 3. Two-year-old girl with injury, suspicious for inﬂicted trauma with acute right subdural hematoma. She had a seizure witnessed in an outside emergency department, and was unresponsive with a dilated right pupil on presentation. The clot was evacuated and the brain was noted to be slightly full at surgery, so a hemicraniectomy was performed. Intracranial pressure was measured and easily controlled, and excellent brain perfusion was maintained. The patient was intubated but responsive and was able to follow commands. (A) MRI, T2 sequence, done 1 day after surgery. Note symmetric appearance of brain parenchyma. (B) Diffusion-weighted sequence. Note bright signal in basal ganglia and temporal and occipital cortex. (C) CT scan, postoperative day 3. Note development of hypodensity of entire right hemisphere, with increased swelling. (D) MRI, diffusion sequence, one week postoperatively. Note bright signal in entire hemisphere, as well as contralateral basal ganglia. (E) MR angiogram. All arteries appear patent. (F) MRI, fast spinecho T2 weighted sequence, 2 months post-injury. Note widespread damage to right hemisphere. The patient has a hemiparesis and hemianopsia.

not reproduce the pathology of the ‘‘big black brain.’’ Injection of blood causes a localized lesion rather than one involving the entire hemisphere. To learn whether the infantile pattern could be reproduced by adapting the rodent model to an immature, gyrencephalic brain, Shaver et al. (1996) injected blood into the subdural space of three-week-old piglets. Their model was modiﬁed by use of a ‘‘cranial window’’ through the bone that allowed conﬁrmation of a thin but extensive subdural clot overlying the hemisphere, comparable to that seen most often in human infants. This

model also failed to reproduce the ‘‘big black brain’’ phenomenon, as it too created a localized injury. However, unlike the rodent lesion, which was largely cortical, the piglet lesion preferentially affected the underlying white matter. The authors hypothesized that the white matter was affected because of its increased vulnerability during early life, when the metabolic rate is relatively high in white matter but the vascular supply is immature and may respond inadequately to increased demand. It is for this reason that the white matter in infants is thought to be preferentially vulnerable to

298

damage from perinatal hypoxic-ischemic insults and prematurity (Rorke, 1992). Both the rodent and the piglet subdural injection models fail to mimic the mechanical trauma that occurs during the injury event in patients. Since the presence of blood in the subdural space was insufﬁcient to cause the hemispheric damage seen in infant subdural injuries, Duhaime et al. (2000) performed a series of studies using piglets to isolate the effect of mechanical trauma on the immature brain. This was done to test the hypothesis that the infant brain might be particularly vulnerable to damage from mechanical trauma relative to other ages, thus explaining the extensive hemispheric injuries seen with acute subdural hematoma. The group developed a model of scaled cortical impact that delivered a mechanically comparable strain load to the same brain structures at different ages. These studies showed that rather than being selectively vulnerable to mechanical trauma, the infant brain was relatively resistant to injury from mechanical deformation (Duhaime et al., 2003). Therefore, another explanation for the unique pattern of extensive brain damage seen with subdurals in infancy was needed. Other workers have attempted to more closely reproduce the mechanical forces that might be responsible for infant subdural hemorrhage using animal models. Because of the long-held concept that the injury might be caused by manual shaking of an infant, Smith et al. (1998) subjected rats to prolonged mechanical shaking. While subarachnoid hemorrhage and surface contusions were found, hemispheric tissue loss was not produced in this model. Anesthetized, immature sheep subjected to 30 min of intermittent, violent manual shaking exhibited similar ﬁnding to the shaken rats, without hemispheric loss (Finnie et al., 2006). Piglets subjected to a single or double high-velocity head rotation showed diffuse axonal injury and focal frontal subdural hemorrhage, but have not to date demonstrated the ‘‘big black brain’’ phenomenon (Raghupathi and Margulies, 2002; Raghupathi et al., 2004). It can be seen that in all these experimental models, elements of injury occur in isolation, whereas in the clinical setting, multiple pathophysiologic stressors may occur in combination.

We will now return to the clinical arena, where additional observations relevant to this issue have been made.

The role of apnea It has long been recognized that infants are susceptible to apnea from a variety of causes, including head injury. That children with subdurals often present with hypoventilation or frank apnea is well known (Johnson et al., 1995; Kemp et al., 2003; Ichord et al., 2007). Therefore, it has been hypothesized that hypoxia/ischemia, rather than trauma itself, is the cause of the diffuse hypodensity pattern seen in the most severe of these injuries. The fact that many of these infants have been the victims of inﬂicted trauma has been thought to contribute to this association. Speciﬁcally, it has been speculated that after the infant is injured, the perpetrator does not seek immediate medical attention, hoping the child will recover spontaneously. This leads to delay in care. If the child hypoventilates during this delay period, diffuse hypoxic brain damage may result. This scenario matches the histories obtained in many cases, and likely contributes to the damage seen in many infants. However, it is not a sufﬁcient explanation for all aspects of the ‘‘black brain’’ phenomenon. This is because there are welldocumented accidental cases of subdural hematoma in infants and toddlers that were witnessed and in which children got immediate medical attention with no apparent apnea or hypoventilation, but in which the phenomenon of hemispheric hypodensity nonetheless appeared (Figs. 1 and 2). The other limitation of apnea as the sole explanation for the phenomenon is that it fails to account for the fact that in one-third of the cases, the hypodensity is unilateral. One would generally expect a diffuse insult like hypoxia to affect the brain symmetrically. The observation that one hemisphere can be totally destroyed while the other remains relatively preserved suggests that more than one factor is in play. Gilles and others have noted that in unilateral cases, the hypodensity occurs on the side of the subdural hematoma (Duhaime et al., 1998; Gilles and Nelson, 1998;

299

Gilles et al., 2003). Therefore, it appears that some synergistic effect of the hemorrhage (or the forces that caused it) and a second insult may be needed. The role of seizures Infants with head injury and other neurologic insults often present with seizures, which also may occur subclinically, especially in very young children. The incidence of clinical seizures in inﬂicted head injuries has been reported in 40–79% of patients (Ludwig and Warman, 1984; Johnson et al., 1995; Gilles and Nelson, 1998; Ghahreman et al., 2005; Bechtel et al., 2006). Clancy et al. (1988) have noted that up to 79% of seizures in neonates are subclinical, that is, are only detected on EEG recordings. This would suggest the possibility that even more seizure events may occur in infants with inﬂicted injury than are apparent clinically, especially in the youngest patients. Pharmacologic agents used for sedation and/or paralysis during head trauma management in the intensive care unit could inﬂuence clinical epileptic events, with some agents (such as midazolam) potentially protecting against seizures, and others (narcotics, paralytics) potentially obscuring their detection. Seizures are known to increase excitotoxic stress in animal models and to be associated with increased metabolic demand and worse outcome in human patients. Therefore, some workers have hypothesized that the frequent occurrence of seizures in infants with traumatic brain injury may contribute to the pathophysiology of an infarct-like picture, such as occurs in hemispheric hypodensity (Duhaime et al., 1998). However, the exact contribution of seizures, clinical or subclinical, to this phenomenon remains unknown. Neuropathology and clinical neurophysiology As immunohistochemical techniques have become available, additional detail about the neuropathology of fatal infant inﬂicted injuries has been determined. Geddes reported that contrary to older studies in which light microscopy was used to assess neuropathology, in her series most patients exhibited little in the way of diffuse axonal

injury as determined by both immunohistochemistry and light microscopic ﬁndings. However, patients frequently did exhibit ‘‘ischemic’’ ﬁndings (Geddes et al., 2001a, b). Additionally, damage at the cervicomedullary junction was often encountered, also raising the question whether apnea due to damage in this location might contribute to the ﬁndings; this has been hypothesized by others as well (Hadley et al., 1989; Johnson et al., 1995). However, Geddes and colleagues did not address the speciﬁc phenomenon of unilateral ‘‘big black brain’’ in their series. Kohanek and colleagues have measured a wide variety of cerebrospinal ﬂuid markers from ventriculostomy samples from infants with traumatic brain injury compared to those found in lumbar puncture samples from uninjured infants. They have reported higher levels of excitatory amino acids and other markers of cellular damage in ﬂuid from infants with inﬂicted injuries and in younger infants compared to older children and those with accidental mechanisms (Ruppell et al., 2001; Berger et al., 2006). Whether these differences reﬂect an age effect or more extensive damage in the inﬂicted injury patients (primary or secondary) remains unclear.

Brain swelling, variability, and decompressive craniectomy Most babies with bilateral or unilateral hemispheric hypodensity develop brain swelling and increased intracranial pressure. Survival is better in the youngest infants, probably because of their ability to split the sutures to relieve some of the pressure. Interestingly, in some cases brain swelling is not problematic, and why some children do not develop signiﬁcant swelling remains unknown (see Figs. 1–3). It may be that acute injury factors or genetic differences determine this variability. In recent years, hemicraniectomy has gained increasing acceptance as a means to alleviate the brainstem compressive effects of brain swelling in infants and young children (Cho et al., 1995). It is not yet clear in which children and at what timepoint this should be undertaken, and the effect of this practice on outcome has been difﬁcult to

300

assess, although it does appear to be effective in lowering intracranial pressure (Taylor et al., 2001). At present it appears that even in those cases in which the hypodensity appears over several days, the processes involved have already begun at presentation and do not appear reversible with current therapies (Duhaime et al., 1996; Dias et al., 1998; Graupman and Winston, 2006). Therefore, treatment strategies are aimed at protecting the undamaged hemisphere. It has been hypothesized that the nearly universal ﬁnding of a wedge of damage in the contralateral medial frontal lobe results from subfalcine herniation with trapping of the callosomarginal branch of the anterior cerebral artery against the falx, resulting in focal infarction (Gilles and Nelson, 1998). In the setting of hemicraniectomy, this internal herniation is ameliorated. However, the authors have observed that some contralateral progression can occur despite early relief of pressure, raising the possibility of a neurochemically-mediated process rather than a purely vascular occlusive one (Fig. 3). Pathophysiology of the ‘‘Big Black Brain’’: a hypothesis Both experimental and clinical evidences support the idea that no single insult is sufﬁcient to explain all cases of unilateral hemispheric hypodensity in association with subdural hematoma in infants and toddlers. Not all children with subdural hemorrhage develop this ﬁnding; thus subdural blood itself appears to be an insufﬁcient cause. Global physiologic insults cannot fully explain the distribution in unilateral cases. Hemispheric hypodensity can be seen in nonaccidental and accidental injuries, with apnea and with no apparent apnea, with clinical seizures and without. Brain swelling usually but not always occurs. Atrophy is rapid and profound. The phenomenon of subdural blood with unilateral hemispheric hypodensity has not yet been reproduced in an experimental model. In many situations, infants and children demonstrate a remarkable ability to compensate for systemic physiologic stresses. A common example is the child’s ability to maintain blood pressure in the face of hypovolemia. However, the immature

organism has a limit to its ability to compensate, and when this limit is reached, the resulting decompensation is often precipitous. It has been demonstrated in a number of contexts that the immature brain’s response to insult varies from the adult response. Its response to global hypoxia-ischemia varies, with some ages and in some outcome measures having increased vulnerability to cell death and in others, increased resistance compared to the mature brain (Painter, 1995; Johnston et al., 2001; Vanucci and Hagberg, 2004). Resistance to other insults seen in the context of trauma has been demonstrated by a number of experimental models cited above, including isolated focal mechanical trauma and injected subdural hematoma with increased intracranial pressure; in these contexts, immaturity confers protection. In cases of infantile subdural hematoma with hemispheric hypodensity, it appears likely that the brain is subjected to a combination of stresses that exceed its capacity to compensate. Most clinical scenarios would point to a combination of local perfusion decrement over the surface of the cerebral hemisphere related to the presence of subdural blood, in combination with a second, more global insult. This could include apnea/hypoxia, hypercarbia, hypotension, or increased metabolic demand from seizures. There may be additional insults that are at present difﬁcult to discern or measure. Why the process involves the entire hemisphere, even beyond the extent of visible subdural blood, remains incompletely understood. It appears to represent a diffuse decompensation resulting in widespread parenchymal death that under certain circumstances may reach threshold only on the side of the larger hemorrhage. Individual differences with respect to cerebrovascular responses, acute injury cascades, inﬂammation, apoptotic pathways, or other neurochemical processes likely inﬂuence why some patients have profound swelling and others do not. Conclusion The unique infant pattern of unilateral hemispheric hypodensity associated with accidental or inﬂicted acute subdural hematoma provides

301

insight into age-related differences in brain injury response. This pattern of extensive and homogeneous tissue destruction appears to require a combination of insults that overwhelm the immature brain’s ability to compensate and affects the entire hemisphere. Because the speciﬁc physiologic stresses vary from case to case, including hypoxia, hypotension, hypercarbia, and seizure activity, different factors that decrease substrate delivery or increase metabolic demand may be at play for different patients. The observation that this pattern occurs uniquely in infants and toddlers more likely reﬂects their age-dependent brain response to combined stresses rather than reﬂecting a single, speciﬁc mechanism or circumstance of injury. References Alexander, R., Sato, Y., Smith, W. and Bennett, T. (1990) Incidence of impact trauma with cranial injuries ascribed to shaking. Am. J. Dis. Child, 144: 724–726. Bechtel, K., Stoessel, K., Leventhal, J.M., Ogle, E., Teague, B., Lavietes, S., Banyas, B., Allen, K., Dziura, J. and Duncan, C. (2006) Characteristics that distinguish accidental from abusive injury in hospitalized young children with head trauma. Pediatrics, 114: 165–168. Berger, P.R., Adelson, P., Richichi, R. and Kochanek, P.M. (2006) Serum biomarkers after traumatic and hypoxemic brain injuries: insight into the biochemical response of the pediatric brain to inﬂected brain injury. Dev. Neurosci., 28: 327–335. Bird, C.R., McMahan, J.R., Gilles, F.H., Senac, M.O. and Apthorp, J.S. (1987) Strangulation in child abuse: CT diagnosis. Radiology, 163: 373–375. Bullock, R., Butcher, S. and Mcculloch, J. (1990) Changes in extracellular glutamate concentration after acute subdural haematoma in the rat: evidence for an ‘‘excitotoxic’’ mechanism? Acta Neurochir. Suppl., 51: 274–276. Bullock, R., Butcher, S.P., Chen, M., Kendall, L. and Mcculloch, J. (1991) Correlation of the extracellular glutamate concentration with extent of blood ﬂow reduction after subdural hematoma in the rat. J. Neurosurg., 74: 794–802. Caffey, J. (1972) On the theory and practice of shaking infants: its potential residual effects of permanent brain damage and mental retardation. AJDC, 124: 161–169. Caffey, J. (1974) The whiplash shaken infant syndrome: manual shaking by the extremities with whiplash-induced intracranial and intraocular bleedings, linked with residual permanent brain damage and mental retardation. Pediatrics, 54: 396–403. Cho, D., Wang, Y. and Chi, C. (1995) Decompressive craniotomy for acute shaken/impact baby syndrome. Pediatr. Neurosurg., 23: 192–198.

Clancy, R.R., Legido, A. and Lewis, D. (1988) Occult neonatal seizures. Epilepsia, 29: 256–261. Cohen, R.A., Kaufman, R.A., Myers, P.A. and Towbin, R.B. (1986) Cranial computed tomography in the abused child with head injury. AJR, 146: 97–102. Dias, M.S., Backstrom, J., Falk, M. and Li, V. (1998) Serial radiography in the infant shaken impact syndrome. Pediatr. Neurosurg., 29: 77–85. Duhaime, A.C., Bilaniuk, L. and Zimmerman, R. (1993) The ‘‘big black brain’’: radiographic changes after severe inﬂicted head injury in infancy. J. Neurotrauma, 10: S59. Duhaime, A.C., Christian, C., Moss, E. and Seidl, T. (1996) Long-term outcome in children with the shaking-impact syndrome. Pediatr. Neurosurg., 24: 292–298. Duhaime, A.C., Christian, C.W., Rorke, L.B. and Zimmerman, R.A. (1998) Nonaccidental head injury in infants: the ‘‘Shaken Baby Syndrome.’’ New Engl. J. Med., 338: 1822–1829. Duhaime, A.C., Gennarelli, L., Yachnis, A.T. and Ross, D.T. (1992) Acute subdural hematoma: is the blood itself toxic? Pediatric Section, American Association of Neurological Surgeons, Annual Meeting. Vancouver, B.C. Duhaime, A.C., Gennarelli, T.G., Thibault, L.E., Bruce, D.A., Margulies, S.S. and Wiser, R. (1987) The shaken baby syndrome: a clinical, pathological, and biomechanical study. J. Neurosurg., 66: 409–415. Duhaime, A.C., Hunter, J.V., Grate, L.L., Kim, A., Demidenko, E., Golden, J.A. and Harris, C. (2003) Magnetic resonance imaging studies of age-dependent responses to scaled focal brain injury in the piglet. J. Neurosurg., 99: 542–548. Duhaime, A.C., Margulies, S.S., Durham, S.R., O’Rourke, M.M., Golden, J.A., Marwaha, S. and Raghupathi, R. (2000) Maturation-dependent response of the piglet brain to scaled cortical impact. J. Neurosurg., 93: 455–462. Ewing-Cobbs, L., Kramer, L., Prasad, M., Canales, D.N., Louis, P.T., Fletcher, J.M., Vollero, H., Landry, S.H. and Cheung, K. (1998) Neuroimaging, physical, and developmental ﬁndings after inﬂicted and noninﬂicted traumatic brain injury in young children. Pediatrics, 102: 300–307. Ewing-Cobbs, L., Prasad, M., Kraer, L. and Landry, S. (1999) Inﬂicted traumatic brain injury: relationship of developmental outcome to severity of injury. Pediatr. Neurosurg., 31: 251–258. Finnie, J.W., Manavis, J. and Blumbergs, P.C. (2006) An ovine model of the shaken baby syndrome. International Neurotrauma Society 8th Annual Neurotrauma Symposium. Rotterdam, The Netherlands, International Neurotrauma Society. Geddes, J.F., Hackshaw, A.K., Vowles, G.H., Nickols, C.D. and Whitwell, H.L. (2001a) Neuropathology of inﬂicted head injury in children: I. Patterns of brain damage. Brain, 124: 1290–1298. Geddes, J.F., Hackshaw, A.K., Vowles, G.H., Nickols, C.D. and Whitwell, H.L. (2001b) Neuropathology of inﬂicted head injury in children: II. Microscopic brain injury in infants. Brain, 124: 1299–1306. Ghahreman, A., Bhasin, V., Chaseling, R., Andrews, B. and Lang, E.W. (2005) Nonaccidental head injuries in children: a Sydney experience. J. Neurosurg., 103: 213–218.

302 Gilles, E.E., Mcgregor, M.L. and Levy-Clark, G. (2003) Retinal hemorrhage asymmetry in inﬂicted head injury: a clue to pathogenesis? J. Pediatr., 143: 494–499. Gilles, E.E. and Nelson Jr., M.D. (1998) Cerebral complications of nonaccidental head injury in childhood. Pediatr. Neurol., 19: 119–128. Graupman, P. and Winston, K.R. (2006) Nonaccidental head trauma as a cause of childhood death. J. Neurosurg. (4 Suppl. Pediatr.), 104: 245–250. Guthkelch, A.K. (1971) Infantile subdural hematoma and its relationship to whiplash injuries. Br. Med. J., 2: 430–431. Hadley, M.N., Sonntag, V.K.H., Rekate, H.L. and Murphy, A. (1989) The infant whiplash-shake syndrome: a clinical and pathological study. Neurosurgery, 24: 536–540. Hahn, Y.S., Raimondi, A.J., Mclone, D.G. and Yamanouchi, Y. (1983) Traumatic mechanisms of head injury in child abuse. Childs Brain, 10: 229–241. Ichord, R.N., Naim, M., Pollock, A.N., Nance, M.L., Margulies, S.S. and Christian, C.W. (2007) Hypoxic-ischemic injury complicates inﬂicted and accidental traumatic brain injury in young children: the role of diffusion-weighted imaging. J. Neurotrauma, 24(1): 106–118. Inglis, F., Kuroda, Y. and Bullock, R. (1992) Glucose hypermetabolism after acute subdural hematoma is ameliorated by a competitive NMDA antagonist. J. Neurotrauma, 9: 75–84. Inglis, F.M., Bullock, R., Chen, D.I., Graham, D.I., Miller, J.D. and Mcculloch, J. (1990) Ischaemic brain damage associated with tissue hypermetabolism in acute subdural haematoma: reduction by a glutamate antagonist. Acta Neurochir. Suppl., 51: 277–279. Johnson, D.L., Boal, D. and Baule, R. (1995) Role of apnea in nonaccidental head injury. J. Ped. Neurosurg., 23: 305–310. Johnston, M.V., Trescher, W.H., Ishida, A. and Nakajima, W. (2001) Neurobiology of hypoxic-ischemic injury in the developing brain. Pediatr. Res., 49: 735–741. Kemp, A.M., Stoodley, N., Cobley, C., Coles, L. and Kemp, K.W. (2003) Apnoea and brain swelling in non-accidental head injury. Arch. Dis. Child., 88: 472–476. Kuroda, Y. and Bullock, R. (1992) Local cerebral blood ﬂow mapping before and after removal of acute subdural hematoma in the rat. Neurosurgery, 30: 687–691. Ludwig, S. and Warman, M. (1984) Shaken baby syndrome: a review of 20 cases. Ann. Emerg. Med., 13: 104–107. Miller, J.D., Bullock, R., Graham, D.I., Chen, M.H. and Teasdale, G.M. (1990) Ischemic brain damage in a model of acute subdural hematoma. Neurosurgery, 27: 433–439. Ommaya, A.K., Corrao, P. and Letcher, F.S. (1973) Head injury in the chimpanzee. Part 1: biodynamics of traumatic unconsciousness. J. Neurosurg., 39: 152–166.

Ommaya, A.K., Faas, F. and Yarnell, P. (1968) Whiplash injury and brain damage: an experimental study. JAMA, 204: 285–289. Ommaya, A.K. and Yarnell, P. (1962) Subdural hematoma after whiplash injury. Lancet, 2: 237–239. Painter, M.J. (1995) Animal models of perinatal asphyxia: contributions, contradictions, and clinical relevance. Semin. Pediatr. Neurol., 2: 37–56. Raghupathi, R. and Margulies, S.S. (2002) Traumatic axonal injury after closed head injury in the neonatal pig. J. Neurotrauma, 19: 843–853. Raghupathi, R., Mehr, M.F., Helfaer, M.A. and Margulies, S.S. (2004) Traumatic axonal injury is exacerbated following repetitive closed head injury in the neonatal pig. J. Neurotrauma, 21: 307–316. Rorke, L.B. (1992) Anatomical features of the developing brain implicated in pathogenesis of hypoxic-ischemic injury. Brain Pathol., 2: 211–221. Ruppell, R.A., Kochanek, P.M., Adelson, P.D., Rose, M.E., Wisniewski, S.R., Bell, M.J., Clark, R.S., Marion, D.W. and Graham, S.H. (2001) Excitatory amino acid concentrations in ventricular cerebrospinal ﬂuid after severe traumatic brain injury in infants and children: the role of child abuse. J. Pediatr., 138: 18–25. Shaver, E., Duhaime, A.C., Curtis, M., Gennarelli, L.M. and Barrett, R. (1996) Experimental acute subdural hematoma in infant piglets. Pediatr. Neurosurg., 25: 123–129. Smith, S.L., Andrus, P.K., Gleason, D.D. and Hall, E.D. (1998) Infant rat model of the shaken baby syndrome: preliminary characterization and evidence for the role of free radicals in cortical hemorrhaging and progressive neuronal degeneration. J. Neurotrauma, 15: 693–705. Taylor, A., Butt, W., Rosenfeld, J., Shann, F., Ditchﬁeld, M., Lewis, E., Klug, G., Wallace, D., Henning, R. and Tibbals, J. (2001) A randomized trial of very early decompressive craniectomy in children with traumatic brain injury and sustained intracranial hypertension. Child Nerv. Syst., 17: 154–162. Vanucci, S.J. and Hagberg, H. (2004) Hypoxia-ischemia in the immature brain. J. Exp. Biol., 207(Pt 18): 3149–3154. Whyte, K.M. and Pascoe, M. (1989) Does ‘‘black’’ brain mean doom? Computed tomography in the prediction of outcome in children with severe head injuries: ‘‘benign’’ vs. ‘‘malignant’’ brain swelling. Australas. Radiol., 33: 344–347. Zimmerman, R.A., Bilaniuk, L.T., Bruce, D., Schut, L., Uzzell, B. and Goldberg, H.I. (1979) Computed tomography of craniocerebral injury in the abused child. Radiology, 130: 687–690.

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 21

Traumatic brain injury and Alzheimer’s disease: a review Corinna Van Den Heuvel1,, Emma Thornton1 and Robert Vink1,2 1 Discipline of Pathology, University of Adelaide, Adelaide, Australia Centre for Neurological Diseases, The Hanson Institute, Adelaide, Australia

2

Abstract: In an effort to identify the factors that are involved in the pathogenesis of Alzheimer’s disease (AD), epidemiological studies have featured prominently in contemporary research. Of those epidemiological factors, accumulating evidence implicates traumatic brain injury (TBI) as a possible predisposing factor in AD development. Exactly how TBI triggers the neurodegenerative cascade of events in AD remains controversial. There has been extensive research directed towards understanding the potential relationship between TBI and AD and the putative inﬂuence that apolipoprotein E (APOE) genotype has on this relationship. The aim of the current paper is to provide a critical summary of the experimental and human studies regarding the association between TBI, AD and APOE genotype. It will be shown that despite signiﬁcant discrepancies in the literature, there still appears to be an increasing trend to support the hypothesis that TBI is a potential risk factor for AD. Furthermore, although it is known that APOE genotype plays an important role in AD, its link to a deleterious outcome following TBI remains inconclusive and ambiguous. Keywords: amyloid precursor protein; traumatic brain injury; apolipoprotein E; Alzheimer’s disease Epidemiology of TBI and AD

(AD) is the most common neurodegenerative disorder of modern societies accounting for 50–60% of all age-related dementia (Andersen et al., 2006). Speciﬁcally, AD afﬂicts 8–10% of the population over the age of 65 and almost 50% over the age of 85 (Mattson, 1997). Over 24 million people worldwide are estimated to be currently suffering from AD, and with the projected rise in the elderly population over the coming decades, this has been estimated to rise to 81 million by 2040 (Miller et al., 2006). The possibility that TBI may predispose a person to developing AD in later life has signiﬁcant social and medical implications, and reinforces the need for preventative efforts and health service planning to cope with the potential large increase in the number of AD patients (Lye and Shores, 2000). It is therefore critical to establish whether any link between TBI and AD does

Traumatic brain injury (TBI) is the leading cause of death and disability in people under the age of 45 years in industrialized countries. Studies in Australia, the United States, France and Spain indicate that the incidence of death from TBI is 20–30 per 100,000 (Finfer and Cohen, 2001) with motor vehicle accidents accounting for the majority of fatal head injuries (Kraus, 1993). Those individuals who survive TBI are often left with permanent neurological deﬁcits, which adversely affect their quality of life, and contributes to the enormous social and economic costs of TBI that are borne by communities. Alzheimer’s disease Corresponding author. Tel.: +61 8 8222 3370; Fax: +61 8

8222 3392; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61021-2

303

304

exist, and whether the mechanisms associated with this link can be exploited to develop an interventional therapy.

The ‘Amyloid Cascade Hypothesis’ Neuropathologically, AD is characterized by the presence of neuroﬁbrillary tangles (NFTs) and neuritic plaques. NFTs in the brain consist primarily of hyperphosphorylated tau. Neuritic amyloid plaques consist primarily of aggregated amyloid-b peptides (Ab), a peptide of 40–42 amino acids, which are surrounded by dystrophic neurites, microglia and reactive astrocytes (Selkoe, 2001). It is now generally accepted that the development of these two pathologies is central to the pathogenesis of AD and the leading ‘amyloid cascade hypothesis’ suggests that it is the accumulation of Ab in the brain which is the primary inﬂuence in AD (Hardy and Selkoe, 2002). Ab has been hypothesized to cause the pathologic and behavioural manifestations of AD, including neuroﬁbrillary tangle formation, neuronal degeneration, synaptic dysfunction and loss as well as impaired memory (Janus et al., 2000). According to this theory, the accompanying NFT formation and neuronal loss are downstream events resulting from an imbalance between Ab production from its precursor Amyloid Precursor Protein (APP). Evidence in support of the hypothesis comes from the knowledge that genetic defects such as mutations in the APP gene, and the presenilin 1 and 2 genes, results in aberrant amyloidogenic processing of APP leading to Ab deposition and rare autosomal dominant forms of familial AD (Selkoe, 1997). Also, active immunization against Ab is associated with reduced numbers of amyloid deposits in human AD brains (Hardy and Selkoe, 2002), while passive immunization with antibodies against Ab results in neutralization of the pathologic Ab resulting in rapid improvement in spatial learning and memory, and restoration of bloodbrain barrier integrity in transgenic mice (Lee et al., 2006). Degenerating neurones in AD display impaired energy production, increased oxidative damage and disruption of calcium homeostasis, and importantly Ab has been shown to be involved

in all of these cell malfunctions (Mattson, 2004). Moreover, Ab causes apoptotic cell death of neuronal cells in culture by the induction of caspases, known instigators of apoptotic cell death (Awasthi et al., 2005). Thus, when Ab is injected into the hippocampal region of rats, an almost complete loss of CA1 neurones is observed (Miguel-Hidalgo and Cacabelos, 1998).

APP processing Neurotoxic Ab peptides are derived from enzymatic processing of the ubiquitously expressed, highly conserved type-1 transmembrane glycoprotein called APP. Understanding the normal structure, molecular function and processing of APP is therefore critical to unravelling the molecular basis of AD. APP is found in all neurons and some glial cells in the central and peripheral nervous systems as well as being expressed in platelets, endothelial cells and ﬁbroblasts. In the brain, where it is constitutively expressed, APP serves a synaptic function and is located predominantly in neuronal cell bodies, concentrated at axosomatic and other synaptic sites (Beyreuther et al., 1993). After translation of APP, which occurs within the endoplasmic reticulum, the protein is transported through the secretory pathway becoming N- and O-glycosylated and tyrosylsulphated while moving through the trans-Golgi network (Tomita et al., 1998). Once mature, APP can be processed by two mutually exclusive complex pathways, either the non-amyloidogenic or the amyloidogenic pathway (Fig. 1). The non-amyloidogenic pathway accounts for the majority of APP processing (Suh and Checler, 2002), and results in a secreted a form of APP (sAPPa) via a-secretase cleavage. The b- and g-secretase pathway is responsible for producing secreted APPb (sAPPb) and the toxic Ab which is found within amyloid plaques in AD (Mills and Reiner, 1999). Thus, APP can result in both a beneﬁcial or detrimental product depending on the way in which it is post-translationally processed within cells. The secretases which play a crucial role in APP processing and cleavage have now been well characterised. The a-secretase includes members of the A disintegrin and metalloprotease (ADAM) family of

305

Fig. 1. Schematically illustrating the cleavage pathways available to amyloid precursor protein and the mutually exclusive nature of the a- and b-secretase pathways.

proteases (Lammich et al., 1999), b-secretase is a type-1 membrane-spanning aspartic protease termed b-site APP-cleavage enzyme 1 (BACE1) (Vassar et al., 1999), and g-secretase comprises a complex of at least four proteins — presenilin 1 or 2, anterior pharynx defective homologue 1 (APH1), presenilin enhancer 2 homologue (PEN2) and nicastrin (reviewed by Wilquet and De Strooper (2004)). Secretases are obvious targets for drug development for the prevention and treatment of AD. Therapies either aim to increase a-secretase activity, thereby increasing non-amyloidogenic cleavage of APP, or inhibit the b- and g-secretase (Dewachter and Van Leuven, 2002; Arbel et al., 2005). g-Secretase inhibitors have also been produced (Dewachter and Van Leuven, 2002) but these proteases have a widespread tissue distribution and their use in humans may be compromised by side effects resulting from blockade of g-secretase cleavage of Notch and other protein substrates (King et al., 2004). In contrast, BACE inhibitors may prove beneﬁcial in reducing Ab production without major side effects. Indeed BACE deﬁcient mice have reduced Ab production and do not exhibit any overt abnormal phenotypes (Roberds et al., 2001). To date, drugs that target speciﬁc sites of this Ab cascade have only been employed in cell culture and animal models of AD, such that their potential in the clinic remains unclear. As indicated above, a-secretase processing not only precludes the formation of Ab (Mattson, 1997)

but it also gives rise to sAPPa, which has been reported to have many neuroprotective and neurotrophic functions within the central nervous system (as reviewed by Mattson et al. (1993)). The discovery that sAPPa has many beneﬁcial functions, and the knowledge that production of sAPPa precludes the formation of the toxic Ab, has led to an increased awareness of the factors that govern its production. These factors are the focus of many investigations that attempt to modulate APP processing toward production of sAPPa, thus slowing or inhibiting Ab production (Racchi et al., 1999). Stimulation of glutamatergic G-protein-coupled receptors linked to the phospholipase C/protein kinase C (PKC) signalling system regulate and increase the release of sAPPa (Lee et al., 1997; Lee and Wurtman, 2000). Recently it has been shown that neuronal adaptor proteins X 11a and X 11b interact with APP processing, possibly by modulating BACE cleavage of APP, and resulting in inhibition of Ab production (Miller et al., 2006). Furthermore, SorLA/LR11, which is a sorting receptor, regulates the intracellular transport and processing of APP in neurons thus inhibiting Ab production (Andersen et al., 2006).

APP and Ab within damaged axons It is feasible that a relationship may exist between TBI and AD since TBI leads to overexpression of

306

APP within neuronal cell bodies (McKenzie et al., 1994; Bramlett et al., 1997; Van Den Heuvel et al., 1999) and accumulation of APP within traumatically injured axons (Blumbergs et al., 1995). Theoretically, the accumulation of APP results in an increase in the substrate that could potentially be processed to form amyloidogenic Ab. It has therefore been hypothesized by a number of groups that the overexpression of APP may exceed the limit of normal processing capacity, resulting in mismetabolization of APP into potentially amyloidogenic fragments (Graham et al., 1995, 1996). In support of this hypothesis, both axonal APP accumulation and long-term accumulation of Ab has been reported in axons following TBI (Iwata et al., 2002) suggesting that damaged axons may provide a key source of Ab following TBI (Uryu et al., 2004). Co-accumulation of Ab with APP in swollen axons and neuronal cell bodies has been found within days following trauma in a pig model of diffuse axonal injury (Smith et al., 1999), and the same group also found widespread axonal Ab accumulation associated with Ab plaques in brain-injured humans with diffuse axonal injury (Smith et al., 2003). It has also been demonstrated that BACE and presenilin-1 co-localize with APP and Ab within the damaged axons and that BACE and presenilin 1 mediate the Ab production from APP within the axonal membrane (Kamal et al., 2001). This was later conﬁrmed by Chen et al. (2004) in a pig model who demonstrated that the Ab accumulation in the swollen axons persisted for 6 months following injury in the subcortical white matter and basal ganglia. They concluded that TBI leading to impaired axonal transport induces a long-term pathological co-accumulation of APP with BACE, presenilin 1 and activated cysteine-dependent aspartate-speciﬁc proteases (caspases), thereby providing a possible mechanism for APP cleavage and production of Ab within axons following TBI. Following neuronal cell injury, activation of caspase-3 (Raghupathi et al., 2000) occurs in parallel with increased APP accumulation and Ab production in several different experimental paradigms (Stone et al., 2002), suggesting a potential association between caspase cleavage of APP, caspase-mediated cell death, increased production

of Ab and subsequent AD. Similarly, excitotoxic neuronal injury and global ischaemia result in accumulation and colocalization of activated caspase-3, caspase-cleaved APP fragments and apoptotic (TUNEL) labelling in affected hippocampal neurons and plaque-associated neurites in AD brains (Gervais et al., 1999). Therefore, the observations in TBI that excitotoxic neuronal injury and global ischaemia resemble changes seen in dying neurons and plaque-associated neurites in AD brains (Masliah et al., 1998), supports the hypothesis that TBI induced caspase activation may potentially result in increased APP proteolysis and may directly or indirectly inﬂuence Ab levels. Evidence for the role of caspase-3 in APP cleavage and Ab production has come from recent studies examining the effects of caspase inhibition following TBI in mice with the human Ab coding sequence ‘knocked in’ to their endogenous APP gene. Caspase inhibitors prevented the elevated level of Ab normally seen in these animals as well as reducing hippocampal cell death, although attenuation of cell death could not be completely attributed to less Ab formation (Abrahamson et al., 2006). Further human autopsy and experimental studies are required to conﬁrm these ﬁndings and to ascertain the relevance of this co-accumulation within neuronal cell bodies and axons and to better elucidate the exact mechanisms involved in APP cleavage and subsequent deposition of Ab plaques.

Alzheimer’s pathology following fatal head injury in humans Preliminary evidence implicating a possible role for TBI in the development of AD came from an early case report documenting early-onset classic AD pathology in a 38-year-old man who had suffered a previous single episode of severe head trauma 16 years ago (Rudelli et al., 1982). Since then, studies of the brains of boxers suffering from dementia pugilistica have also demonstrated AD-like pathology with diffuse Ab plaque deposition (Roberts et al., 1990). We can speculate that such Ab deposition resulted from repeated blows to the head over a long period of time, and that

307

such events may also occur in the brains of headinjured individuals. Histopathological studies of such individuals who died after suffering a single severe TBI demonstrate widespread cerebral Ab deposition in short (Roberts et al., 1991, 1994; Gentleman et al., 1997; Ikonomovic et al., 2004) and long term (Clinton et al., 1991) survivors irrespective of age. In contrast to these studies, Adle-Biassette et al. (1996) were unable to detect any Ab deposits in head injury cases of individuals below the age of 63 years, in a similar age range and survival time to those studied by Roberts et al. (1991). Also, the extensive non-selective autopsy investigation by Braak and Braak (1997) rarely detected Ab plaques in the younger subjects (below the age of 40 years). None of the 58 cases examined in the 31–35 years age range, and only two of the 83 subjects in the 36–40 years age range had some Ab plaques. Furthermore, Ab deposits were detected in only 20 of the 207 people in their ﬁfth decade of life (Braak and Braak, 1997). Therefore, signiﬁcant discrepancies exist in neuropathological studies. There also appears to be several morphological similarities, as well as differences, between the Ab plaques in AD and those found deposited in the brain following TBI. In AD, there are numerous, compact and hard core Ab deposits as well as neuroﬁbrillary tangles and neuropil threads, whereas following TBI there appears to be a higher incidence of diffuse Ab plaques (Clinton et al., 1991; Roberts et al., 1991, 1994; Ikonomovic et al., 2004). Exactly what this morphological difference in plaques means is unknown. One similarity between the deposits seen in TBI and AD is that they are both composed primarily of Ab1–42. These deposits appear much more abundantly than Ab1–40 deposits seen in TBI (Gentleman et al., 1997; Horsburgh et al., 2000a; Ikonomovic et al., 2004) or in AD (Iwatsubo et al., 1994). Another consequence of APP gene mutations is elevated Ab levels in the cerebrospinal ﬂuid (CSF) in AD (Nakamura et al., 1994). Studies have since aimed to determine if CSF levels of Ab are similarly increased following TBI. Raby et al. (1998) reported signiﬁcantly elevated levels of Ab1–42 in the CSF in six patients (age 19–51 years) following severe TBI, an observation later conﬁrmed by

Olsson et al. (2004) who reported elevated levels of Ab1–42 and APP in ventricular CSF of severely injured TBI patients. These groups postulate that the increased levels of Ab1–42 in the CSF are directly related to the elevated Ab levels in the brain (Raby et al., 1998). It has also been speculated that the rise in Ab1–42 in both brain and CSF may be a direct result of the increased APP levels in neuronal cell bodies and axons after TBI (McKenzie et al., 1996). In direct contrast, a recent study of 29 TBI patients found a reduction in Ab1–42 levels in CSF, which appeared to correlate with poor outcome after TBI (Franz et al., 2003). Future studies are required to validate these CSF ﬁndings and to resolve the contradictions. It is important to determine how relevant the CSF concentration of Ab1–42 is as a marker of potential AD, and whether monitoring CSF levels might prove helpful in assessing the degree of neuronal injury resulting from TBI. There is also the potential of identifying those individuals potentially at risk of developing AD later in life (Raby et al., 1998).

Alzheimer’s pathology following experimental head injury Studies using different forms of experimental TBI have given some insight into how brain injury may lead to AD, although results remain somewhat inconclusive. In support of the argument that increased APP levels following TBI may potentiate AD pathology, experimental TBI in rats induced overexpression and accumulation of APP in the cerebral cortex and hippocampus, which subsequently led to neuronal degeneration in the CA3 region of the hippocampus as early as 3 days postinjury (Murakami et al., 1998). Also, a pig model of rotational head injury developed rapid Ab accumulation, manifested in axonal bulbs and diffuse plaques 3 days after trauma (Smith et al., 1999). However, post-traumatic Ab deposition has not been observed in the majority of nontransgenic animal studies where most failed to identify any Ab staining (Pierce et al., 1996, 1998; Masumura et al., 2000; Laurer et al., 2001; Ciallella et al., 2002; Hamberger et al., 2003).

308

Experimental results from APP transgenic mice also remain inconclusive. Transgenic mice with AD-like pathology (platelet-derived-APP; PD-APP), which overexpress human APP by 10-fold and form amyloid plaques by 6 months of age, enable the interaction between TBI and AD to be studied. TBI in these animals resulted in a transient rise in Ab levels, particularly in the hippocampus. This elevation in Ab correlated with a much greater loss of CA3 hippocampal neurones than injured wild type mice, which also had a signiﬁcant loss of hippocampal neurones compared to uninjured controls (Smith et al., 1998). However, TBI did not lead to premature amyloid plaque formation in these transgenic mice, and on long term examination there was a reported reduction in amyloid plaques ipsilateral to the injury (Smith et al., 1998; Nakagawa et al., 1999). Paradoxically, TBI induced a regression of previously existing plaques in aged PD-APP-transgenic mice (Nakagawa et al., 2000). Thus, elevated Ab levels, which are expected to be seen in the APP-transgenic mice, resulted in hippocampal neuronal death but the TBI did not accelerate Ab deposition. Interestingly, in transgenic mice that overexpress human APP by only twofold, Ab deposition did not occur following TBI and similar cognitive and motor deﬁcits to wild-type mice were observed (Murai et al., 1998). In contrast, repetitive TBI in a Tg2576 APP-transgenic mice model did result in greater Ab deposition and an increase in the production of both soluble and insoluble cortical Ab40 and Ab42, which may be accounted for by the higher levels of oxidative stress in repetitive TBI (Uryu et al., 2002). Normally, repetitive TBI in animals does not lead to the formation of neuroﬁbrillary tangles. Yet in a study by Yoshiyama et al. (2005), repetitive TBI in one of 12 mice resulted in the development of neuroﬁbrillary tangles, which correlated with a remarkably poor neurobehavioural score. Consequently the presence of iron deposits, which also occurs in AD, was thought to contribute to the formation of these neuroﬁbrillary tangles (Yoshiyama et al., 2005). The results of TBI studies in transgenic models of AD are also in conﬂict with that seen in human TBI. Firstly, rapid Ab deposition has not been

demonstrated in any of the described transgenic models, unlike human studies (Roberts et al., 1991, 1994; Gentleman et al., 1997; Ikonomovic et al., 2004). Secondly, an increased severity of injury does not result in increased Ab deposition; if anything it actually seems to correlate with reduced Ab deposition or even resolution of already established plaques, as reviewed by Szczygielski et al. (2005). This trend seems to be reversed in humans, since both the risk (Plassman et al., 2000) and the post-traumatic Ab deposition increased with TBI severity (Roberts et al., 1994).

Epidemiological studies The epidemiological studies that have reported on the relationship between TBI and AD are also contradictory, and accordingly, the association still remains inconclusive. Some reports suggest a positive association with AD (French et al., 1985; Mortimer et al., 1985, 1991; Graves et al., 1990; Van Duijn et al., 1992; Rasmusson et al., 1995; Salib and Hillier, 1997) while other studies were unable to conﬁrm head trauma as a risk factor for AD (Katzman et al., 1989; Li et al., 1992; Fratiglioni et al., 1993). With respect to the casecontrol studies, one of the earliest showed that patients with dementia of the Alzheimer type had a signiﬁcantly greater incidence of antecedent head trauma (French et al., 1985). Subsequent pooled re-analysis of some of the case-control studies has shown positive associations between the development of AD and a history of head trauma, particularly amongst males (Mortimer et al., 1991; Fleminger et al., 2003). Speciﬁcally, the metaanalysis of seven case-control studies by Mortimer et al. (1991) allowed for a more powerful statistical investigation of the association between TBI and AD and provided the ﬁrst convincing evidence in support of a strong association between these two pathologies. They achieved a high level of statistical power (0.92) as opposed to the low mean statistical power of the individual studies (0.22). The ﬁndings of the EURODERM re-analysis highlighted the importance of designing studies with adequate statistical power and suggested that many of the early negative ﬁndings stem from the

309

fact that they suffered from insufﬁcient statistical power (see review by Lye and Shores (2000)). This is particularly important when the risk factor under investigation, in these cases TBI, occurs infrequently in the general population. Recent casecontrol studies have reﬁned their designs to avoid the methodological ﬂaws found in earlier studies and have shown signiﬁcant associations between AD and TBI (Graves et al., 1990; Van Duijn et al., 1992; Rasmusson et al., 1995; Salib and Hillier, 1997). An additional strength of the study by Rasmusson et al. (1995) was that it examined the association between AD and TBI of any severity, concluding that even mild head injury may serve as a predisposing factor for some cases of AD. More recently, a detailed systematic review of case-control studies conducted over the past 10 years sought to replicate those ﬁndings, and supported an association between head injury and AD, but only in males (Fleminger et al., 2003). The possible explanation for the gender differences in the risk of AD following TBI may be attributed to the neuroprotective and neuroregenerative effects of the female hormones oestrogen and progesterone (Stein, 2001), or it could be merely the fact that men typically suffer more severe injuries than women. The association between TBI and early onset of dementia was also demonstrated in a recent population-based study of fall-related TBI, concluding that fall-related TBI may hasten, increase or even trigger the onset of AD (Luukinen et al., 2005). As with the case-control studies, there have been conﬂicting reports in cohort studies. For example, no signiﬁcant association was found between head injury and the risk of developing AD in the early work (Katzman et al., 1989; Williams et al., 1991). Also, more recently a history of head trauma with unconsciousness was excluded as a risk factor for AD in both the Rotterdam Study (Mehta et al., 1999) and the large EURODERM study (Launer et al., 1999). The EURODERM study was a pooled analysis of four European populationbased studies of individuals 65 years and older. Plassman et al. (2000) examined the association between early adult head injury, as documented by military hospital records, and dementia in late life. In the study of 548 world war II (US Navy)

brain-injured veterans and 1228 age-matched noninjured controls, follow up after 50 years revealed that moderate to severe head injuries in young men may be associated with increased risk of AD and other dementias in late life. These ﬁndings were supported by the MIRAGE study, which is the largest study to date (Guo et al., 2000). This study analysed 2233 deﬁnite and probable AD patients, and 14,668 ﬁrst-degree relatives and showed that head injury with loss of consciousness signiﬁcantly increased the AD risk. The magnitude of risk was proportional to injury severity and was heightened amongst ﬁrst-degree relative of AD patients (Guo et al., 2000). Several studies have shown that there is a greater risk of developing AD when head injury has occurred in later life (within 10 years of onset of AD) as opposed to head injury occurring earlier in life (beyond 10 years of onset of AD) (Graves et al., 1990; Mortimer et al., 1991; Van Duijn et al., 1992). In contrast, other studies have shown that head trauma occurring mainly in younger childhood is associated with an increased risk of AD in later life (Schoﬁeld et al., 1997; Plassman et al., 2000). After examining the incidence of AD pathology in 58 consecutive patients with residual closed TBI lesions and the frequency of TBI residuals in 57 age-matched autopsy controls, Jellinger et al. (2001) concluded that severe TBI may have some inﬂuence on the development of AD irrespective of age when the TBI occurred. For an excellent detailed review of case-control studies from 1984 to 1997 see Lye and Shores (2000). A related hypothesis concerns the relationship between TBI and AD focussing on the time delay between TBI and the onset of AD. These studies have highlighted the importance of considering more speciﬁc factors such as duration of loss of consciousness, severity of injury and the time of onset of AD symptoms. Cohort studies have suggested that TBI may interact with other risk factors to hasten the onset of AD in persons susceptible to the disease (Sullivan et al., 1987; Gedye et al., 1989; Schoﬁeld et al., 1997; Nemetz et al., 1999). Of 17 AD cases reported, there was a younger mean age of onset for cases with a history of head injury than for those cases without it (Sullivan et al., 1987). In a larger study of 148

310

referral patients with conﬁrmed probable AD, Gedye et al. (1989) found that the mean age of AD onset was signiﬁcantly younger for those with a history of TBI. Also, in a slightly larger community based longitudinal study of aging in 271 participants in North Manhattan, participants who suffered loss of consciousness exceeding 5 min following a head trauma were at a signiﬁcantly increased risk of earlier onset AD (Schoﬁeld et al., 1997). Furthermore, among 1283 TBI cases in Olmsted County, 31 patients developed AD. Given the incidence of AD in the community, this number was not unexpected, however the time between TBI and onset of AD was less than expected suggesting that TBI reduces the time of onset of AD among persons at risk of developing AD (Nemetz et al., 1999). In contrast, Rasmusson et al. (1995) did not ﬁnd any effect of head injury on age of onset of AD in a cohort of 68 AD cases, nor did the MIRAGE study (Guo et al., 2000).

The inﬂuence of APOE genotype on TBI outcome Following reports that Ab is not found in all headinjured brains, it was speculated that it may only be deposited in the brains of people who are genetically susceptible to developing AD, for example, those individuals possessing an apolipoprotein E4 (APOE e4) allele. In 1993, it was demonstrated that the APOE e4 allele was a risk factor for AD (Strittmatter et al., 1993) and the association between APOE e4 with late onset familial and sporadic AD has now been conﬁrmed worldwide (Corder et al., 1993; Saunders et al., 1993; Mayeux et al., 1995; Nicoll et al., 1995; Horsburgh et al., 2000b). APOE protein has been co-localized with the neuropathological hallmarks of AD, including neuroﬁbrillary tangles, Ab plaques and amyloid angiopathy (Namba et al., 1991; Rebeck et al., 1993). It is now known to be a major susceptibility factor associated with approximately 40–50% of sporadic and familial AD compared with 30% of the normal population (Roses, 1996). While it is known that APOE is a key factor involved in lipid transport within the human central nervous system (CNS), the three alleles (2, 3 and 4) have signiﬁcantly varied effects within the

CNS (Nicoll, 1996). The e4 allele has been linked to impaired branching and growth of neurites and it is therefore suggested to exacerbate neurological disturbances (Kerr and Kraus, 1998). Numerous studies have since conﬁrmed this association by demonstrating that Ab deposition following head injury occurs more prominently in those who possess an APOE e4 allele. Also, clinical studies of TBI have shown that possession of APOE e4 is associated with a relatively poor outcome (Sorbi et al., 1995; Teasdale et al., 1997). In the study by Teasdale et al. (1997), 57% of APOE e4 carriers had an unfavourable outcome (deﬁned as dead, vegetative state or severe disability) compared with 27% of non-carriers of APOE e4. The relationship between functional recovery in patients with head injury and APOE e4 genotype was later conﬁrmed in an independent study (Lichtman et al., 2000), with rehabilitation outcome for TBI survivors being adversely affected in the presence of the APOE allele. It was subsequently shown that APOEe4 individuals were 10 times more likely to develop AD after TBI than those who did not posses the allele (Mayeux et al., 1995). A recent case study has demonstrated that possession of APOE e4 is associated with a greater incidence of moderate/ severe contusional injury and severe ischaemic damage in fatal cases of TBI (Smith et al., 2006). These authors postulate that APOE has a role in cerebrovascular and haematological mechanisms. Animal studies, using APOE knockout and transgenic mice, have also provided further information about APOE mechanisms and the response of the brain to injury (as reviewed by Horsburgh et al. (2000b)). Some studies have shown that the APOE molecule may have a direct neurotoxic role (Neve and Robakis, 1998), whereas others have speculated that APOE e4 may directly interact with Ab and impact on the metabolism of APP (Growdon, 1998). Transgenic mice that express human APOE e4 have poorer outcome and greater mortality following TBI than mice expressing human APOE e3 (Sabo et al., 1999). TBI mice expressing APOE e4 also had a decrease in neuroprotective sAPPa within the hippocampus, whereas mice expressing APOE e3 had an increase in sAPPa in the cortex. These studies concluded that the alteration in APP metabolism contributed

311

to the increased mortality and poorer outcome in APOE e4 transgenic mice, and the greater recovery and less cortical damage in APOE e3 transgenic mice (Ezra et al., 2003). In addition, transgenic APOE e4 mice, which also overexpress APP, had accelerated deposition of Ab following injury, suggesting APOE e4 may reduce the clearance of Ab, thereby favouring its deposition (Hartman et al., 2002). In contrast to the studies demonstrating an association between APOE e4 and an unfavourable outcome following TBI, a number of recent studies examining both severe and mild TBI have failed to support such ﬁndings (Guo et al., 2000; Plassman et al., 2000; Jellinger et al., 2001; Chamelian et al., 2004). Therefore there is no conclusive evidence linking APOE genotype with the development of AD following TBI (Jellinger, 2004). Further experimental studies and larger clinicopathological and prospective human studies are warranted to clarify the relationship between TBI, APOE genotype and AD. Additional studies with APOE transgenic mice are also expected to contribute to the greater understanding of the pathophysiological role of APOE in the brain.

Conclusion Although ﬁndings to date have been contradictory, there have been major advances from epidemiological studies and both experimental and human autopsy studies regarding the relationship between TBI and AD, and the inﬂuence that the APOE e4 genotype has on this putative association. Human studies have certainly supported the hypothesis that either repeated mild to moderate TBI, or moderate to severe single TBI with loss of consciousness, are risk factors for the subsequent development of AD. Moreover, those patients who have the APOE e4 are at greater risk of a poorer outcome following TBI. However, the exact mechanism by which APOE genotype inﬂuences outcome and increases risk of AD following TBI is unclear. Experimental and human studies have also found accumulation of Ab and APP within damaged axons following TBI, but the exact implications of this and exactly how this translates

into release of Ab from axons, Ab plaque formation and toxicity remains unclear. Despite some experimental and human studies suggesting that APP overexpression following TBI may lead to Ab formation, there are numerous studies which have failed to demonstrate any such relationship. Accordingly, further investigation is required, particularly with larger prospective cohort studies of TBI survivors and larger autopsy series before an association between TBI and onset of AD can be deﬁnitively supported. References Abrahamson, E.E., Ikonomovic, M.D., Ciallella, J.R., Hope, C.E., Paljug, W.R., Isanski, B.A., Flood, D.G., Clark, R.S. and Dekosky, S.T. (2006) Caspase inhibition therapy abolishes brain trauma-induced increases in Abeta peptide: implications for clinical outcome. Exp. Neurol., 197: 437–450. Adle-Biassette, H., Duyckaerts, C., Wasowicz, M., He, Y., Fornes, P., Foncin, J.F., Lecomte, D. and Hauw, J.J. (1996) BAP deposition and head trauma. Neurobiol. Aging, 17: 415–419. Andersen, O.M., Schmidt, V., Spoelgen, R., Gliemann, J., Behlke, J., Galatis, D., Mckinstry, W.J., Parker, M.W., Masters, C.L., Hyman, B.T., Cappai, R. and Willnow, T.E. (2006) Molecular dissection of the interaction between amyloid precursor protein and its neuronal trafﬁcking receptor SorLA/LR11. Biochemistry, 45: 2618–2628. Arbel, M., Yacoby, I. and Solomon, B. (2005) Inhibition of amyloid precursor protein processing by b-secretase through site-directed antibodies. PNAS, 102: 7718–7723. Awasthi, A., Matsunaga, Y. and Yamada, T. (2005) Amyloidbeta causes apoptosis of neuronal cells via caspase cascade, which can be prevented by amyloid-beta-derived short peptides. Exp. Neurol., 196: 282–289. Beyreuther, K., Pollwein, P., Multhaup, G., Monning, U., Konig, G., Dyrks, T., Schubert, D. and Masters, C.L. (1993) Regulation and expression of the Alzheimer’s B/A4 amyloid protein precursor in health, disease and Down’s Syndrome. Ann. N.Y. Acad. Sci., 695: 91–102. Blumbergs, P.C., Scott, G., Manavis, J., Wainwright, H., Simpson, D.A. and Mclean, A.J. (1995) Topography of axonal injury as deﬁned by amyloid precursor protein and the sector scoring method in mild and severe closed head injury. J. Neurotrauma, 12: 565–572. Braak, H. and Braak, E. (1997) Frequency of stages of Alzheimer-related lesions in different age categories. Neurobiol. Aging, 18: 351–357. Bramlett, H.M., Kraydieh, S., Green, E.J. and Dietrich, W.D. (1997) Temporal and regional patterns of axonal damage following traumatic brain injury: a beta-amyloid precursor protein immunocytochemical study in rats. J. Neuropathol. Exp. Neurol., 56: 1132–1141.

312 Chamelian, L., Reis, M. and Feinstein, A. (2004) Six-month recovery from mild to moderate traumatic brain injury: the role of APOE-4 allele. Brain, 127: 2621–2628. Chen, X.H., Siman, R., Iwata, A., Meaney, D.F., Trojanowski, J.Q. and Smith, D.H. (2004) Long-term accumulation of amyloid-beta, beta-secretase, presenilin-1, and caspase-3 in damaged axons following brain trauma. Am. J. Pathol., 165: 357–371. Ciallella, J.R., Ikonomovic, M.D., Paljug, W.R., Wilbur, Y.I., Dixon, C.E., Kochanek, P.M., Marion, D.W. and Dekosky, S.T. (2002) Changes in expression of amyloid precursor protein and interleukin-1beta after experimental traumatic brain injury in rats. J. Neurotrauma, 19: 1555–1567. Clinton, J., Ambler, M.W. and Roberts, G.W. (1991) Post traumatic Alzheimer’s disease: preponderance of a single plaque type. Neuropathol. Appl. Neurobiol., 17: 69–74. Corder, E.H., Saunders, A.M., Strittmatter, W.J., Schmechal, D.E., Gaskell, P.C., Small, G.W., Roses, A.D., Haines, J.L. and Pericak Vance, M.A. (1993) Gene dosage of apolipoprotein E type allele and the risk of Alzheimer’s disease in late onset families. Science, 261: 921–923. Dewachter, I. and Van Leuven, F. (2002) Secretases as targets for the treatment of Alzheimer’s disease: the prospects. Lancet Neurol., 1: 409–416. Ezra, Y., Oron, L., Moskovich, L., Roses, A.D., Beni, S.M., Shohami, E. and Michaelson, D.M. (2003) Apolipoprotein E4 decreases whereas apolipoprotein E3 increases the level of secreted amyloid precursor protein after closed head injury. Neuroscience, 121: 315–325. Finfer, S.R. and Cohen, J. (2001) Severe traumatic brain injury. Resuscitation, 48: 77–90. Fleminger, S., Oliver, D.L., Lovestone, S., Rabe-Hesketh, S. and Giora, A. (2003) Head injury as a risk factor for Alzheimer’s disease: the evidence 10 years on; a partial replication. J. Neurol. Neurosurg. Psychiatry, 74: 857–862. Franz, G., Beer, R., Kampﬂ, A., Engelhardt, K., Schmutzhard, E., Ulmer, H. and Deisenhammer, F. (2003) Amyloid beta 1–42 and tau in cerebrospinal ﬂuid after severe traumatic brain injury. Neurology, 60: 1457–1461. Fratiglioni, L., Ahlbom, A., Viitanen, M. and Winblad, B. (1993) Risk factors for late-onset alzheimer’s disease: a population-based, case-control study. Ann. Neurol., 33: 258–266. French, L.R., Schuman, L.M., Mortimer, J.A., Hutton, J.T., Boatman, R.A. and Christians, B. (1985) A case-control study of dementia of the Alzheimer type. Am. J. Epidemiol., 121: 414–421. Gedye, A., Beattie, B.L., Tuokko, H., Horton, A. and Korsarek, E. (1989) Severe head injury hastens age of onset of Alzheimer’s disease. J. Am. Geriatr. Soc., 37: 970–973. Gentleman, S.M., Greenberg, B.D., Savage, M.J., Noori, M., Newman, S.J., Roberts, G.W., Grifﬁn, W.S. and Graham, D.I. (1997) AB42 is the predominant form of amyloid-b-protein in the brains of short-term survivors of head injury. Neuroreport, 8: 1519–1522. Gervais, F.G., Xu, D., Robertson, G.S., Vaillancourt, J.P., Zhu, Y., Huang, J., Leblanc, A., Smith, D., Rigby, M., Shearman, M.S., Clarke, E.E., Zheng, H., Van Der Ploeg, L.H.,

Ruffolo, S.C., Thornberry, N.A., Xanthoudakis, S., Zamboni, R.J., Roy, S. and Nicholson, D.W. (1999) Involvement of caspases in proteolytic cleavage of Alzheimer’s amyloidbeta precursor protein and amyloidogenic A beta peptide formation. Cell, 97: 395–406. Graham, D.I., Gentleman, S.M., Lynch, A. and Roberts, G.W. (1995) Distribution of beta-amyloid protein in the brain following severe head injury. Neuropathol. Appl. Neurobiol., 21: 27–34. Graham, D.I., Gentleman, S.M., Nicoll, J.A., Royston, M.C., McKenzie, J.E., Roberts, G.W. and Grifﬁn, W.S. (1996) Altered beta-APP metabolism after head injury and its relationship to the aetiology of Alzheimer’s disease. Acta Neurochir. Suppl. (Wien.), 66: 96–102. Graves, A.B., White, E., Koepsell, T.D., Reiﬂer, B.V., Van Belle, G., Larson, E.B. and Raskind, M. (1990) The association between head trauma and Alzheimer’s disease. Am. J. Epidemiol., 131: 491–501. Growdon, J.H. (1998) Apolipoprotein E and Alzheimer’s disease. Arch. Neurol., 44: 1053–1054. Guo, Z., Cuppels, L.A., Kurz, A., Auerbach, S.H., Volicer, L., Chui, H., Green, R.C., Sadovnick, A.D., Durara, R., Decarli, C., Johnson, K., Go, R.C., Growdon, J.H., Haines, J.L., Kukull, W.A. and Farrer, L.A. (2000) Head injury and risk of AD in the MIRAGE study. Neurology, 54: 1316–1323. Hamberger, A., Huang, Y.L., Zhu, H., Bao, F., Ding, M., Blennow, K., Olsson, A., Hansson, H.A., Viano, D. and Haglid, K.G. (2003) Redistribution of neuroﬁlaments and accumulation of beta-amyloid protein after brain injury by rotational acceleration of the head. J. Neurotrauma, 20: 169–178. Hardy, J.A. and Selkoe, D.J. (2002) The amyloid hypothesis of Alzheimer’s disease; progress and problems on the road to therapeutics. Science, 297: 353–356. Hartman, R.E., Laurer, H., Longhi, L., Bales, K.R., Paul, S.M., McIntosh, T.K. and Holtzman, D.M. (2002) Apolipoprotein E4 inﬂuences amyloid deposition but not cell loss after traumatic brain injury in a mouse model of Alzheimer’s disease. J. Neurosci., 22: 10083–10087. Horsburgh, K., Cole, G.M., Yang, F., Savage, M.J., Greenberg, B.D., Gentleman, S.M., Graham, D.I. and Nicoll, J.A.R. (2000a) b-Amyloid (Ab)42(43), Ab42, Ab40 and apoE immunostaining of plaques in fatal head injury. Neuropath. Appl. Neurobiol., 26: 124–132. Horsburgh, K., Mccarron, M.O., White, F. and Nicoll, J.A. (2000b) The role of apolipoprotein E in Alzheimer’s disease, acute brain injury and cerebrovascular disease: evidence of common mechanisms and utility of animal models. Neurobiol. Aging, 21: 245–255. Ikonomovic, M.D., Uryu, K., Abrahamson, E.E., Ciallella, J.R., Trojanowski, J.Q., Lee, V.M.-Y., Clark, R.S., Marion, D.W., Wisniewski, S.R. and Dekosky, S.T. (2004) Alzheimer’s pathology in human temporal cortex surgically excised after severe brain injury. Exp. Neurol., 190: 192–203. Iwata, A., Chen, X.H., McIntosh, T.K., Browne, K.D. and Smith, D.H. (2002) Long-term accumulation of amyloid-beta

313 in axons following brain trauma without persistent upregulation of amyloid precursor protein genes. J. Neuropathol. Exp. Neurol., 61: 1056–1068. Iwatsubo, T., Odaka, A., Suzuki, N., Mizusawa, H., Nukina, N. and Ihara, Y. (1994) Visualization of AB42 and AB40 in senile plaques with end speciﬁc AB monoclonals: evidence that an initially deposited species is AB42. Neuron, 13: 45–53. Janus, C., Pearson, J., Mclaurin, J., Mathews, P.M., Jiang, Y., Schmidt, S.D., Chishti, M.A., Horne, P., Heslin, D., French, J., Mount, H.T., Nixon, R.A., Mercken, M., Bergeron, C., Fraser, P.E., St George-Hyslop, P. and Westaway, D. (2000) A beta peptide immunization reduces behavioural impairment and plaques in a model of Alzheimer’s disease. Nature, 408: 979–982. Jellinger, K.A. (2004) Head injury and dementia. Curr. Opin. Neurol., 17: 719–723. Jellinger, K.A., Paulus, W., Wrocklage, C. and Litvan, I. (2001) Effects of closed traumatic brain injury and genetic factors on the development of Alzheimer’s disease. Eur. J. Neurol., 8: 707–710. Kamal, A., Almenar-Queralt, A., Leblanc, J.F., Roberts, E.A. and Goldstein, L.S. (2001) Kinesin-mediated axonal transport of a membrane compartment containing b-secretase and presenilin-1 requires APP. Nature, 414: 643–648. Katzman, R., Aronson, M., Fuld, P., Kawas, C., Brown, T., Morgenstern, H., Frishman, W., Gidez, L., Eder, H. and Ooi, W.L. (1989) Development of dementing illness in an 80-year-old volunteer cohort. Ann. Neurol., 25: 317–324. Kerr, M.E. and Kraus, M. (1998) Genetics and the central nervous system: apolipoprotein E and brain injury. AACN Clin. Issues, 9: 524–530. King, G.D., Cherian, K. and Turner, S.R. (2004) X11a impairs g- but not b-cleavage of amyloid precursor protein. J. Neurochem., 88: 971–982. Kraus, J.F. (1993) Epidemiology of head injury. In: Cooper P.R. (Ed.), Head Injury (3rd ed). Williams and Wilkins, Baltimore. Lammich, S., Kojro, E., Postina, R., Gilbert, S., Pfeiffer, R., Jasionowski, M., Haass, C. and Fahrenholz, F. (1999) Constitutive and regulated a-secretase cleavage of Alzheimer’s amyloid precursor protein by a disintegrin metalloprotease. Proc. Natl. Acad. Sci. U.S.A., 96: 3922–3927. Launer, L.J., Andersen, K., Dewey, M.E., Letenneur, L., Ott, A., Amaducci, L.A., Brayne, C., Copeland, J.R., Dartigues, J.F., Kragh-Sorensen, P., Lobo, A., Martinez-Lage, J.M., Stijnen, T. and Hofman, A. (1999) Rates and risk factors for dementia and Alzheimer’s disease: results from EURODEM pooled analyses. EURODEM Incidence Research Group and Work Groups. European Studies of Dementia. Neurology, 52: 78–84. Laurer, H.L., Bareyre, F.M., Lee, V.M., Trojanowski, J.Q., Longhi, L., Hoover, R., Saatman, K.E., Raghupathi, R., Hoshino, S., Grady, M.S. and McIntosh, T.K. (2001) Mild head injury increasing the brain’s vulnerability to a second concussive impact. J. Neurosurg., 95: 859–870. Lee, E.B., Leng, L.Z., Zhang, B., Kwong, L., Trojanowski, J.Q., Abel, T. and Lee, V.M.Y. (2006) Targeting amyloid-b

peptide (Ab) oligomers by passive immunization with a conformation-selective monoclonal antibody improves learning and memory in Ab precurosr protein (APP) transgenic mice. J. Biol. Chem., 281: 4292–4299. Lee, J. and Wurtman, R.J. (2000) Regulation of APP synthesis and secretion by neuroimmunophilin ligands and cyclooxygenase inhibitors. Ann. N.Y. Acad. Sci., 920: 261–268. Lee, R.K.K., Araki, W. and Wurtman, R.J. (1997) Stimulation of amyloid precursor protein synthesis by adrenergic receptors coupled to cAMP formation. Proc. Natl. Acad. Sci. U.S.A., 94: 5422–5426. Li, G., Shen, Y.C., Li, Y.T., Chen, C.H., Zhau, Y.W. and Silverman, J.M. (1992) A case-control study of Alzheimer’s disease in China. Neurology, 42: 1481–1488. Lichtman, S.W., Seliger, G., Tycko, B. and Marder, K. (2000) Apolipoprotein E and functional recovery from brain injury following post acute rehabilitation. Neurology, 55: 1536–1539. Luukinen, H., Viramo, P., Herala, M., Kervinen, K., Kesaniemi, Y.A., Savola, O., Winqvist, S., Jokelainen, J. and Hillbom, M. (2005) Fall-related injuries and the risk of dementia in elderly people: a population-based study. Eur. J. Neurol., 12: 86–92. Lye, T.C. and Shores, A.E. (2000) Traumatic brain injury as a risk factor for Alzheimer’s disease: a review. Neuropsychol. Rev., 10: 115–129. Masliah, E., Mallory, M., Alford, M. and Hansen, L.A. (1998) Caspase dependent DNA fragmentation might be associated with excitotoxicity in Alzheimer’s disease. J. Neuropathol. Exp. Neurol., 57: 1041–1052. Masumura, M., Hata, R., Uramoto, H., Murayama, N., Ohno, T. and Sawada, T. (2000) Altered expression of amyloid precursor proteins after traumatic brain injury in rats: in situ hybridization and immunochemical study. J. Neurotrauma, 17: 123–134. Mattson, M.P. (1997) Cellular actions of b-amyloid precursor protein and it’s soluble and ﬁbrillogenic derivatives. Physiol. Rev., 77: 1081–1132. Mattson, M.P. (2004) Pathways towards and away from Alzheimer’s disease. Nature, 430: 631–639. Mattson, M.P., Cheng, B., Culwell, A.R., Esch, F.S., Lieberburg, I. and Rydel, R.E. (1993) Evidence for excitoprotective and intraneuronal calcium-regulating roles for secreted forms of the beta-amyloid precursor protein. Neuron, 10: 243–254. Mayeux, R., Ottman, R., Maestre, G., Ngai, C., Tang, M.X., Ginsberg, H., Chun, M., Tycko, B. and Shelanski, M. (1995) Synergistic effects of traumatic head injury and apolipoprotein E4 in patients with Alzheimer’s disease. Neurology, 45: 555–557. McKenzie, J.E., Gentleman, S.M., Roberts, G.W., Graham, D.I. and Royston, M.C. (1994) Increased numbers of beta APP-immunoreactive neurons in the entorhinal cortex after head injury. Neuroreport, 6: 161–164. McKenzie, K.J., Mclellan, D.R., Gentleman, S.M., Maxwell, W.L., Gennarelli, T.A. and Graham, D.I. (1996) Is X-APP a marker of axonal damage in short-surviving head injury? Acta Neuropathol. (Berl.), 92: 608–613.

314 Mehta, K.M., Ott, A., Kalmijn, S., Slooter, A.J., Van Duijn, C.M., Hofman, A. and Breteler, M. (1999) Head trauma and risk of dementia and Alzheimer’s disease: the Rotterdam study. Neurology, 53: 1959–1962. Miguel-Hidalgo, J.J. and Cacabelos, R. (1998) Beta-amyloid(1–40)-induced neurodegeneration in the rat hippocampal neurons of the CA1 subﬁeld. Acta Neuropathol. (Berl.), 95: 455–465. Miller, C.C., Mcloughlin, D.M., Lau, K.F., Tennant, M.E. and Rogelj, B. (2006) The X11 proteins, A beta production and Alzheimer’s disease. Trends Neurosci., 29: 280–285. Mills, J. and Reiner, P.B. (1999) Regulation of amyloid precursor protein cleavage. J. Neurochem., 72: 443–460. Mortimer, J.A., French, L.R., Hutton, J.T. and Schuman, L.M. (1985) Head injury as a risk factor for Alzheimer’s disease. Neurology, 35: 264–267. Mortimer, J.A., Van Duijn, C.M., Chandra, V., Fratiglioni, L., Graves, A.B., Heyman, A., Jorm, A.F., Kokmen, E., Kondo, K. and Rocca, W.A. (1991) Head trauma as a risk factor for Alzheimer’s disease: a collaborative re-analysis of casecontrol studies. EURODEM Risk Factors Research Group. Int. J. Epidemiol., 20(Suppl 2): S28–S35. Murai, H., Pierce, J.E.S., Raghupathi, R., Smith, D.H., Saatman, K.E., Trojanowski, J.Q., Lee, V.M.Y., Loring, J.F., Eckman, C.B., Younkin, S.G. and McIntosh, T.K. (1998) Twofold overexpression of human b-amyloid precursor proteins in transgenic mice does not affect the neuromotor, cognitive or neurodegenerative sequelae following experimental brain injury. J. Comp. Neurol., 392: 428–438. Murakami, N., Yamaki, T., Iwamoto, Y., Sakakibara, T., Kobori, N., Fushiki, S. and Ueda, S. (1998) Experimental brain injury induces expression of amyloid precursor protein, which may be related to neuronal loss in the hippocampus. J. Neurotrauma, 15: 993–1003. Nakagawa, Y., Nakamura, M., McIntosh, T.K., Rodriguez, A., Berlin, J.A., Smith, D.H., Saatman, K.E., Raghupathi, R., Clemens, J., Saido, T.C., Schmidt, M.L., Lee, V.M. and Trojanowski, J.Q. (1999) Traumatic brain injury in young, amyloid-beta peptide overexpressing transgenic mice induces marked ipsilateral hippocampal atrophy and diminished Abeta deposition during aging. J. Comp. Neurol., 411: 390–398. Nakagawa, Y., Reed, L., Nakamura, M., McIntosh, T.K., Smith, D.H., Saatman, K.E., Raghupathi, R., Clemens, J., Saido, T.C., Lee, V.M. and Trojanowski, J.Q. (2000) Brain trauma in aged transgenic mice induces regression of established abeta deposits. Exp. Neurol., 163: 244–252. Nakamura, T., Shoji, M., Harigaya, Y., Watanabe, M., Hosoda, K., Cheung, T.T., Shaffer, L.M., Golde, T.E., Younkin, L.H. and Younkin, S.G. (1994) Amyloid B levels in cerebrospinal ﬂuid are elevated in early-onset Alzheimer’s disease. Ann. Neurol., 36: 903–911. Namba, Y., Tomonaga, M., Kawasaki, H., Otomo, E. and Ikeda, K. (1991) Apolipoprotein E immunoreactivity in cerebral amyloid deposits and neuroﬁbrillary tangles in Alzheimer’s disease and kuru plaque amyloid in Creutzfeldt-Jakob disease. Brain Res., 541: 163–166.

Nemetz, P.N., Leibson, C., Naessens, J.M., Beard, M., Kokmen, E., Annegers, J.F. and Kurland, L.T. (1999) Traumatic brain injury and time to onset of Alzheimer’s disease: a population-based study. Am. J. Epidemiol., 149: 32–40. Neve, R.L. and Robakis, N.K. (1998) Alzheimer’s disease: a re-examination of the amyloid hypothesis. Trends Neurosci., 21: 15–19. Nicoll, J.A. (1996) Genetics and head injury. Neuropathol. Appl. Neurobiol., 22: 515–517. Nicoll, J.A.R., Roberts, G.W. and Graham, D.I. (1995) Apolipoprotein E e4 allele is associated with deposition of amyloid b-protein following head injury. Nat. Med., 1: 135–137. Olsson, A., Scsajbok, L., Ost, M., Hoglund, K., Nylen, K., Rosengre, L., Nellgard, B. and Blennow, K. (2004) Marked increase of beta-amyloid (1–42) and amyloid precursor protein in ventricular cerebrospinal ﬂuid after severe traumatic brain injury. J. Neurol., 251: 870–876. Pierce, J.E., Smith, D.H., Trojanowski, J.Q. and McIntosh, T.K. (1998) Enduring cognitive, neurobehavioral and histopathological changes persist for up to one year following severe experimental brain injury in rats. Neuroscience, 87: 359–369. Pierce, J.E., Trojanowski, J.Q., Graham, D.I., Smith, D.H. and McIntosh, T.K. (1996) Immunohistochemical characterization of alterations in the distribution of amyloid precursor proteins and beta-amyloid peptide after experimental brain injury in the rat. J. Neurosci., 16: 1083–1090. Plassman, B.L., Havlik, R.J., Steffens, D.C., Helms, M.J., Newman, T.N., Drosdick, D., Phillips, C., Gau, B.A., WelshBohmer, K.A., Burke, J.R., Guralnik, J.M. and Breitner, J.C.S. (2000) Documented head injury in early adulthood and risk of Alzheimer’s disease and other dementias. Neurology, 55: 1158–1166. Raby, C.A., Morganti-Kossman, M.C., Kossman, T., Stahel, P.F., Watson, M.D., Evans, L.M., Mehta, P.D., Spiegel, K., Kuo, Y., Roher, A.E. and Emmerling, M.R. (1998) Traumatic brain injury increases b-amyloid peptide 1–42 in cerebrospinal ﬂuid. J. Neurochem., 71: 2505–2509. Racchi, M., Solano, C.D., Sironi, M. and Govoni, S. (1999) Activity of a-secretase as the common ﬁnal effector of protein kinase c-dependent and independent modulation of amyloid precursor protein metabolism. J. Neurochem., 72: 2464–2470. Raghupathi, R., Graham, D.I. and McIntosh, T.K. (2000) Apoptosis after traumatic brain injury. J. Neurotrauma, 17: 927–938. Rasmusson, D.X., Brandt, J., Martin, D.B. and Folstein, M.F. (1995) Head injury as a risk factor in Alzheimer’s disease. Brain Inj., 9: 213–219. Rebeck, G.W., Reiter, J.S., Strickland, D.K. and Hyman, B.T. (1993) Apolipoprotein E in sporadic Alzheimer’s disease: allelic variation and receptor interactions. Neuron, 11: 575–580. Roberds, S.L., Anderson, J., Basi, G., Bienkowski, M.J., Branstetter, D.G., Chen, K.S., Freedman, S.B., Frigon, N.L., Games, N., Hu, G., Johnson-Wood, K., Kappenman, K.E., Kawabe, T.T., Kola, I., Kuehn, R., Lee, M., Liu, W., Motter, R., Nichols, N.F., Power, M., Robertson, D.W., Schenk, D., Schoor, M., Shopp, G.M., Shuck, M.E.,

315 Sinha, S., Svensson, K.A., Tatsuno, G., Tintrup, H., Wijsman, J., Wright, S. and Mcconlogue, L. (2001) BACE knockout mice are healthy despite lacking the primary bsecretase activity in brain: implications for Alzheimer’s disease therapeutics. Hum. Mol. Genet., 10: 1317–1324. Roberts, G.W., Allsop, D. and Bruton, C.J. (1990) The occult aftermath of boxing. J. Neurol. Neurosurg. Psychiatry, 53: 373–378. Roberts, G.W., Gentleman, S.M., Lynch, A. and Graham, D.I. (1991) beta A 4 amyloid protein deposition in brain after head trauma. Lancet, 338: 1422–1423. Roberts, G.W., Gentleman, S.M., Lynch, A., Landon, M. and Graham, D.I. (1994) Beta amyloid protein deposition in the brain after severe head injury: implications for the pathogenesis of Alzheimer’s disease. J. Neurol. Neurosurg. Psychiatry, 57: 419–425. Roses, A.D. (1996) Apolipoprotein E alleles as risk factors in Alzheimer’s disease. Ann. Rev. Med., 47: 387–400. Rudelli, R., Strom, J.O., Welch, P.T. and Ambler, M.W. (1982) Post-traumatic premature Alzheimer’s disease. Arch. Neurol., 39: 570–575. Sabo, S.L., Lanier, L.M., Ikin, A.F., Khorkova, O., Sahasrabudhe, S., Greengard, P. and Buxbaum, J.D. (1999) Regulation of X-amyloid secretion by FE65, an amyloid protein precursor-binding protein. J. Biol. Chem., 274: 7952–7957. Salib, E. and Hillier, V. (1997) Head injury and the risk of Alzheimer’s disease: a case control study. Int. J. Geriat. Psychiatry, 12: 363–368. Saunders, A.M., Strittmatter, W.J. and Schmechel, D. (1993) Association of apolioprotein E allele e4 with late-onset familial and sporadic Alzheimer’s disease. Proc. Natl. Acad. Sci. U.S.A., 90: 8098–8102. Schoﬁeld, P.W., Tang, M.X., Marder, K., Bell, K., Dooneief, G., Chun, M., Sano, M., Stern, Y. and Mayeux, R. (1997) Alzheimer’s disease after remote head injury: an incidence study. J. Neurol. Neurosurg. Psychiatry, 62: 119–124. Selkoe, D.J. (1997) Alzheimer’s disease: genotypes, phenotypes, and treatments. Science, 275: 630–631. Selkoe, D.J. (2001) Alzheimer’s disease: genes, proteins, and therapy. Physiol. Rev., 81: 741–766. Smith, C., Graham, D.I., Murray, L.S., Stewart, J. and Nicoll, J.A.R. (2006) Association of APOe4 and cerebrovascular pathology in traumatic brain injury. J. Neurol. Neurosurg. Psychiatry, 77: 363–366. Smith, D.H., Chen, X.H., Iwata, A. and Graham, D.I. (2003) Amyloid-X accumulation in axons after traumatic brain injury in humans. J. Neurosurg., 98: 1072–1077. Smith, D.H., Chen, X.H., Nonaka, M., Trojanowski, J.Q., Lee, V.M., Saatman, K.E., Leoni, M.J., Xu, B.N., Wolf, J.A. and Meaney, D.F. (1999) Accumulation of amyloid beta and tau and the formation of neuroﬁlament inclusions following diffuse brain injury in the pig. J. Neuropathol. Exp. Neurol., 58: 982–992. Smith, D.H., Nakamura, M., McIntosh, T.K., Wang, J., Rodriguez, A., Chen, X.H., Raghupathi, R., Saatman, K.E., Clemens, J., Schmidt, M.L., Lee, V.M. and Trojanowski, J.Q.

(1998) Brain trauma induces massive hippocampal neuron death linked to a surge in beta-amyloid levels in mice overexpressing mutant amyloid precursor protein. Am. J. Pathol., 153: 1005–1010. Sorbi, S., Nacmias, N., Piacentini, S., Repice, A., Latorraca, S., Forleo, P. and Amaducci, L. (1995) ApoE as a prognostic factor for post-traumatic coma. Nat. Med., 1: 852. Stein, D.G. (2001) Brain damage, sex hormones and recovery: a new role for progesterone and estrogen? Trends Neurosci., 24: 386–391. Stone, J.R., Okonkwo, D.O., Singleton, R.H., Mutlu, L.K., Helm, G.A. and Povlishock, J.T. (2002) Caspase-3-mediated cleavage of amyloid precursor protein and formation of amyloid b peptide in traumatic axonal injury. J. Neurotrauma, 19: 601–614. Strittmatter, W.J., Saunders, A.M. and Schmechel, D. (1993) Apolioprotein E high avidity binding to B-amyloid and increased frequency of the E4 allele in late onset familial Alzheimer’s disease. Proc. Natl. Acad. Sci. U.S.A., 90: 1977–1981. Suh, Y.H. and Checler, F. (2002) Amyloid precursor protein, presenilins, and a-synuclein: molecular pathogenesis and pharmacological applications in Alzheimer’s disease. Pharm. Rev., 54: 469–525. Sullivan, P., Petitti, D. and Barbaccia, J. (1987) Head trauma and age of onset of dementia of the Alzheimer type (Letter). JAMA, 257: 2289–2290. Szczygielski, J., Mautes, A., Steudel, W.I., Falkai, P., Bayer, T.A. and Wirths, O. (2005) Traumatic brain injury: cause or risk of Alzheimer’s disease? A review of experimental studies. J. Neural. Transm., 112: 1547–1564. Teasdale, G.M., Nicoll, J.A.R., Murray, G. and Fiddes, M. (1997) Association of apolipoprotein E polymorphism with outcome after head injury. Lancet, 350: 1069–1071. Tomita, S., Kirino, Y. and Suzuki, T. (1998) Cleavage of Alzheimer’s amyloid precursor protein (APP) by secretases occurs after O-glycosylation of APP in the protein secretory pathway. J. Biol. Chem., 273: 6277–6284. Uryu, K., Chen, X.H. and Graham, D.J. (2004) Short-term accumulation of beta-amyloid in axonal pathology following traumatic brain injury in humans. Neurobiol. Aging, 25(Suppl 1): P2–P250. Uryu, K., Laurer, H., McIntosh, T.K., Pratico, D., Martinez, D., Leight, S., Lee, V.M.-Y. and Trojanowski, J.Q. (2002) Repetitive mild brain trauma accelerates Ab deposition, lipid peroxidation, and cognitive impairment in a transgenic mouse model of Alzheimer amyloidosis. J. Neurosci., 22: 446–454. Van Den Heuvel, C., Blumbergs, P.C., Finnie, J.W., Manavis, J., Jones, N.R., Reilly, P.L. and Pereira, R.A. (1999) Upregulation of amyloid precursor protein messenger RNA in response to traumatic brain injury: an ovine head impact model. Exp. Neurol., 159: 441–450. Van Duijn, C.M., Tanja, T.A., Haaxma, R., Schulte, W., Saan, R.J., Lameris, A.J., Antonides-Hendriks, G. and Hofman, A. (1992) Head trauma and the risk of Alzheimer’s disease. Am. J. Epidemiol., 135: 775–782.

316 Vassar, R., Bennett, B.D., Babu-Khan, S., Kahn, S., Mendiaz, E.A., Denis, P., Teplow, D.B., Ross, S., Amarante, P., Loeloff, R., Luo, Y., Fisher, S., Fuller, J., Edenson, S., Lile, J., Jarosinski, M.A., Biere, A.L., Curran, E., Burgess, T., Louis, J., Collins, F., Treanor, J., Rogers, G. and Citron, M. (1999) X-Secretase cleavage of Alzheimer’s amyloid precursor protein by the transmembrane aspartic protease BACE. Science, 286: 735–740. Williams, D.B., Annegers, J.F., Kokmen, E., O’brien, P.C. and Kurland, L. (1991) Brain injury and neurologic sequelae: a cohort study of dementia, parkinsonism, amyotrophic lateral sclerosis. Neurology, 41: 1554–1557.

Wilquet, V. and De Strooper, B. (2004) Amyloid-X precursor protein processing in neurodegeneration. Curr. Opin. Neurobiol., 14: 582–588. Yoshiyama, Y., Uryu, K., Higuchi, M., Longhi, L., Hoover, R., Fujimoto, S., McIntosh, T., Lee, V.M. and Trojanowski, J.Q. (2005) Enhanced neuroﬁbrillary tangle formation, cerebral atrophy, and cognitive deﬁcits induced by repetitive mild brain injury in a transgenic tauopathy mouse model. J. Neurotrauma, 22: 1134–1141.

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 22

The neurotrophic protein S100B: value as a marker of brain damage and possible therapeutic implications Andrea Kleindienst1,, Felicitas Hesse1, M. Ross Bullock2 and Michael Buchfelder1 1 Department of Neurosurgery, Friedrich-Alexander-University, Erlangen-Nuremberg, Germany Department of Neurosurgery, Virginia Commonwealth University Medical Center, Richmond, VA 23298-0631, USA

2

Abstract: We provide a critical analysis of the value of S100B as a marker of brain damage and possible therapeutic implications. The early assessment of the injury severity and the consequent prognosis are of major concern for physicians treating patients suffering from traumatic brain injury (TBI). A reliable indicator to accurately determine the extent of the brain damage has to meet certain requirements: (i) to originate in the central nervous system (CNS) with no contribution from extracerebral sources; (ii) a passive release from damaged neurons and/or glial cells without any stimulated active release; (iii) a lack of speciﬁc effects on neurons and/or glial cells interfering with the initial injury; (iv) an unlimited passage through the blood-brain barrier (BBB). The measurement of putative biochemical markers, such as the S100B protein, has been proposed in this role. Over the past decade, numerous studies have reported a positive correlation of S100B serum levels with a poor outcome following TBI. However, some studies raise doubt whether the serum measurement of S100B is a valid biochemical marker of brain damage. We summarize the speciﬁc properties of S100B and analyze whether they support or counteract the necessary requirements to designate this protein as an indicator of brain damage. Finally, we report recent experimental ﬁndings suggesting a possible therapeutic potential of S100B. Keywords: S100B; TBI; neuroregeneration; brain damage; biomarker; neurotrophic factor role of S100B while in the clinical setting the measurement of S100B became more frequently performed. The S100B protein belongs to a multigenic family of low molecular weight (9–13 kDa) calciumbinding S100 proteins (Donato, 2001; Heizmann et al., 2002). S100B is most abundant in glial cells of the central nervous system (CNS), predominately in astrocytes (Donato, 1986), and constitutes 1–1.5 mg/mg of soluble protein (Matsutani et al., 1985). S100 proteins contain no detectable carbohydrate, lipid, nucleic acid or phosphate, and are dimers consisting of at least two types of subunits of either identical or different amino acid composition. S100 proteins are highly conserved in amino acid composition among vertebrate species

General features and origin of the S100B protein As early as 1965, a neurotrophic factor was puriﬁed from bovine brain for the ﬁrst time (Moore, 1965), and in 1978 this factor was demonstrated to consist of two distinct proteins, S100b and S100a (Isobe et al., 1978). After the identiﬁcation of the chromosomal localization of S100 proteins in 1995, the nomenclature was changed from S100b to S100B and from S100a to S100A1 (Schafer et al., 1995). In the following decade, experimental research was focused on identifying the speciﬁc Corresponding author. Tel.: +49-(0)-9131-85-34577;

Fax: +49-(0)-9131-85-34551; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61022-4

317

318

(Donato, 1986), and comparison of the human and bovine S100A1 and the human and rat S100B subunit reveals an almost complete homology (Isobe et al., 1984). The extracerebral origin of S100B has also been recognized in several studies and this data conﬂicts with the interpretation of S100B serum levels following traumatic brain injury (TBI). The S100B tissue distribution has been assessed using immunocytochemical approaches (Donato, 1991). Quantiﬁcation in rat tissue found a S100B concentration in brain tissue of 3000 ng per milligram of soluble protein, in adipose tissue of 1000 ng, in the skin of 80 ng, in the testes of 40 ng, and in all other tissue of around 1–2 ng (Zimmer and Van Eldik, 1987). Thus, elevated S100B levels following injury may also come from either damaged skeletal muscle (Arcuri et al., 2002) or adipose tissue (Suzuki et al., 1984).

Release mechanism of S100B Experimental ﬁndings demonstrate that S100B is actively secreted from astroglia although the exact mechanism has yet not been identiﬁed. Under physiological circumstances, S100B release can be stimulated by 5 HT 1a agonists in cultured rat astrocytes (Shashoua et al., 1984; Van Eldik and Zimmer, 1987; Whitaker-Azmitia et al., 1990), and is observed within a few minutes after activation of A1 adenosine or mGlu3 metabotropic glutamate receptors and release may last up to 10 h (Ciccarelli et al., 1999). In neuronal plus glial cultures, the level of S100B in the growth medium was measurable (Willoughby et al., 2004). Following experimental TBI, applying a 50 ms strain (stretch) on these cell cultures grown on deformable silicone membranes (Ellis et al., 1995), there was a dramatic rise in S100B 15 s after injury (Willoughby et al., 2004), and thereafter a continued release of S100B until at least 48 h post-injury (Slemmer et al., 2002; Willoughby et al., 2004). The injury-induced S100B release is likely to be the result of at least two possible processes. Firstly, electron microscopic studies of stretch-injured astrocyte cultures show that immediately after stretch-injury structural and membrane integrity

is severely compromised (Ellis et al., 1995) allowing the cell release of preformed S100B. Secondly, since an injury-induced ATP and glutamate release has been shown also, the release of S100B might be in part due to an astrocyte receptor activation by ATP and glutamate (Ciccarelli et al., 1999). The ﬁnding of an increased S100B release at 24 and 48 h post-injury may thus imply that this is a longterm metabolic response to brain damage, which promotes a ‘‘healing process’’ in injured neurons or astrocytes.

Cellular action of S100B Intracellularly, S100B is involved in signal transduction via the inhibition of protein phoshorylation, regulation of enzyme activity and by affecting the calcium homeostasis as well as the regulation of cell morphology by interaction with elements of the cytoplasmatic cytoskeleton (Zimmer and Van Eldik, 1986; Rustandi et al., 1998; Wilder et al., 1998). After release into the extracellular ﬂuid, S100B acts in an autocrine and paracrine manner. In vitro studies demonstrate mitogenic properties of S100B, such as an increased proliferative activity on melanoma cell lines at concentrations below 50 nM and above 5 mM (Klein et al., 1989), as well as on rat C6 glioma cells at 50 pM–1.5 nM (Selinfreund et al., 1991), preferably in its oxidized state (Scotto et al., 1998). Further investigations also conﬁrmed a dose-dependent action of S100B exerting a neuroprotective and neurotrophic inﬂuence at nanomolar concentrations (Rickmann et al., 1995; Li et al., 1998; Huttunen et al., 2000), but at micromolar concentrations, an activation of the inducible nitric oxide (NO) synthase was seen with subsequent NO generation, potentially leading to astrocytic death (Hu et al., 1997; Petrova et al., 2000). Different studies provide evidence for neurotrophic properties of S100B. In cultured mesencephalic rat neurons (Azmitia et al., 1990) and dorsal root ganglia (Van Eldik et al., 1991), S100B has been shown to promote neurite outgrowth (Kligman and Marshak, 1985; WinninghamMajor et al., 1989). In primary rat spinal cord culture, S100B rapidly promoted reassembly

319

and/or stabilization of the cytoskeletal system (Nishi et al., 1997), and prevented apoptosis in a neuroblastoma clonal cell line (Brewton et al., 2001). Most importantly, however, after sciatic nerve section in newborn rats, S100B rescued motor neuron death and preserved neuron diameter (Iwasaki et al., 1997). In addition to its trophic properties, a neuroprotective effect of S100B was found in vitro. S100B decreased neuronal cell death and mitochondrial dysfunction in rat hippocampal neurons after both glucose deprivation, and increased intracellular free calcium (Barger et al., 1995). S100B is, thus, thought to be involved in regulation of energy metabolism by stimulation of the fructose1,6-biphosphate aldolase (Zimmer and Van Eldik, 1986) and phosphoglucomutase (Landar et al., 1996). S100B has also been implicated in cytosolic Ca2+ buffering (Xiong et al., 2000). In cultures of embryonal chick and neonatal rat neurons, S100B was found to protect against glutamate- and staurosporin-induced damage (Ahlemeyer et al., 2000). The above-mentioned neuroprotective and neurotrophic properties of S100B have not been elucidated after TBI.

Passage of S100B through the blood-brain barrier While in cell cultures, the injury-induced S100B release continues to increase up to 48 h (Slemmer et al., 2002; Willoughby et al., 2004), S100B serum levels in patients are highest directly after injury, and become normalized within 24 h in a high percentage of cases, even in those patients with a bad outcome (Jackson et al., 2000). The underlying mechanism of the passage of S100B through the blood-brain barrier (BBB) has not been clariﬁed yet, nor do data exist about cerebral S100B levels and their correlation to serum S100B levels. Since one of the purposes of the BBB is to prevent proteins to enter the brain, there is reasonable doubt whether S100B of cerebral origin may be able to cross the intact BBB in the opposite direction. The principle of spectroscopy is widely applied in chemistry for the analysis of molecules in solution, and MR spectroscopy can be used to identify important molecules in living tissue. Recently, MR

proton spectroscopy has been shown to identify brain metabolites like N-acteylaspartate (NAA), creatine, choline and lactate (Bruhn et al., 1989; Frahm et al., 1989a–c). We demonstrated MR proton spectroscopy to detect different concentrations of aqueous solutions of S100B in vitro, with a strong correlation between the S100B concentration and the area under the curve of the respective S100B MR peak at 4.5 ppm (Kleindienst et al., 2005b). Increased cerebral S100B levels following an intraventricular S100B infusion in normal rats were conﬁrmed by MR proton spectroscopy while the S100B serum concentration did not increase according to the notion that proteins do not cross an intact BBB (Kleindienst et al., 2005b). Following controlled cortical impact injury in the rat, S100B serum levels were elevated up to 48 h post-injury (Rothoerl et al., 2000). In ﬂuid percussion injury in the rat, serum S100B levels increased immediately after injury, peaked at 3 h and returned to normal values at 24 h after injury (Kleindienst et al., 2004). This time proﬁle of serum S100B levels parallels the BBB disruption found after experimental TBI (Povlishock et al., 1978). In contrast, cerebral S100B levels as quantiﬁed by MR proton spectroscopy were raised slightly 3 h after injury, and increased significantly thereafter to more than twice these values by day 5 (Kleindienst et al., 2005b). Thus, no clear relationship exists between the cerebral S100B dynamics and the serum S100B levels.

Clinical studies In 2003, a thorough review of the role of S100B as a marker of brain damage was published summarizing the results of 18 clinical studies in a total of 1085 patients (Rothermundt et al., 2003). In 2004 and 2005, another six studies comprising more than 600 adult patients were performed supporting the correlation of elevated serum levels of S100B with a poor outcome after brain injury (Pelinka et al., 2004; Savola et al., 2004; Vos et al., 2004; Wunderlich et al., 2004; Berger et al., 2005; Chen and Zhu, 2005). The time proﬁle of S100B reported in the serum was variable, and even a

320

delayed increase of S100B serum levels on the 6th day after injury has been reported and speculated to occur due to secondary brain cell damage (Raabe and Seifert, 2000). Measurements of cerebrospinal ﬂuid S100B levels allow a true assessment of the cerebral S100B release following brain insults (Kleine et al., 2003; Petzold et al., 2003; Hayakata et al., 2004; Shore et al., 2004), but require for repetitive measurements either lumbar or ventricular cerebrospinal ﬂuid drainage. The importance of S100B in the acutely injured brain is not well known yet. In view of a substantial body of evidence demonstrating an association between S100B and bad outcome after TBI, it is important to be aware that proof of association is not a proof of causation, in science, and this is especially true of S100B. In contrast to TBI, following electroconvulsive therapy in patients, increased serum S100B levels were associated with an improved cognitive performance (Agelink et al., 2001). Furthermore, strong evidence exists that in traumatized patients S100B serum levels are increased without concomitant brain damage (Ashraf et al., 1999; Anderson et al., 2001a, b; Svenmarker et al., 2002; Kleine et al., 2003; Pelinka et al., 2003; Nygren De Boussard et al., 2004). Finally, the exact function and effects of increased cerebral levels of S100B after TBI are not thoroughly understood, yet.

Effect of S100B on recovery following experimental TBI Besides this experimental evidence for the beneﬁcial effect of S100B on neuronal maintenance, a speciﬁc role of S100B has also been proposed at a higher level in developmental plasticity (Marshak, 1990), and in cell processes thought to be involved in learning and memory, such as long-term potentiation (Fazeli et al., 1990). This has been conﬁrmed by injection of S100B anti-serum into the hemisphere of chicks causing amnesia for a passive avoidance task (O’Dowd et al., 1997). Vice versa, S100B infused into the rat hippocampus, has been shown to facilitate long-term memory for an inhibitory avoidance task (Mello e Souza et al., 2000). Finally, light- and electron microscopic

studies are consistent with the hypothesis that S100B plays a role in lesion-induced collateral sprouting and reactive synaptogenesis (Rickmann et al., 1995; Li et al., 1998; McAdory et al., 1998; Huttunen et al., 2000), and that repair may occur by interaction with growth factors (Gomide and Chadi, 1999). The hippocampus is a region critical for learning and memory. It displays increased susceptibility to injury, and cognitive impairment following injury has been linked to hippocampal dysfunction (Hicks et al., 1993). Recent ﬁndings of neurogenesis within the subgranular zone of the dentate gyrus of the hippocampus in adult mammals provides a possible source to replace lost neurons (Eriksson et al., 1998; Kornack and Rakic, 1999; Gould and Gross, 2002). Indeed, neuronal progenitor cells within the hippocampus present increased cellular proliferation and subsequent neuron formation following experimental TBI (Dash et al., 2001; Chirumamilla et al., 2002). This adult neurogenesis has been found to be actively regulated by hippocampal astrocytes, which are a cellular source for mitogenic/neurotrophic factors and thus promote proliferation of adult neural stem cells and instruct the fate commitment of developing neurons (Song et al., 2002). Consequently, the astrocytic neurotrophic protein S100B is a potential candidate to increase this progenitor cell proliferation and subsequent neuron formation following TBI, thereby enhancing hippocampal network repair and improving cognitive recovery. This hypothesis of a neurogenic effect of S100B has been examined by an intraventricular infusion of S100B following ﬂuid percussion injury in the rat (Kleindienst et al., 2004). Functional recovery was assessed by the Morris water maze 5 weeks postinjury and revealed a significantly improved cognitive performance following intraventricular S100B infusion (Kleindienst et al., 2004). Within the hippocampus, S100B increased the proliferative response significantly by more than 40% on day 5 post-injury (Fig. 1; Kleindienst et al., 2005a). S100B promoted the survival of the injury-induced progenitor cell proliferation as well as the subsequent differentiation into neurons (Fig. 2; Kleindienst et al., 2005a).

321

Fig. 1. Enhancement of progenitor cell proliferation in the hippocampus following TBI and S100B treatment. The photomicrograph shows the distribution of BromodeoxyUridine (BrdU) immunoreactive cells within the ipsilateral dentate gyrus following lateral ﬂuid percussion injury and an intraventricular S100B infusion. BrdU immunoreactive cells on day 5 post-injury are predominately identiﬁed in the area of their origin, the subgranular zone. The arrows point to typical cluster of cells.

Thus, S100B has been demonstrated to increase hippocampal stem/progenitor cell proliferation and neuronal differentiation following TBI, and this enhanced neurogenesis is correlated with an improved cognitive recovery (Kleindienst et al., 2005a). These ﬁndings stress the importance of astrocytic factors for neurogenesis and provide compelling evidence for a therapeutic potential of S100B improving functional recovery following TBI. Further studies are needed to elucidate the beneﬁcial role of S100B on hippocampal network repair, thus offering a potential therapy to augment important innate repair mechanisms of the brain in order to promote memory consolidation in patients with brain insults.

Conclusion Taken together, although the desire for a marker of brain damage is reasonable designating S100B for this role warrants considerable simpliﬁcation, and

interpretation of serum S100B levels following TBI requires thorough knowledge of factors inﬂuencing the measured concentration. Firstly, one has to be aware of a contribution to serum S100B levels originating from extracerebral sources. Secondly, the injury-induced cerebral S100B release may be assumed to be a combination of a passive release by damaged astrocytes and an active release by stimulated astrocytes initiating repair mechanisms, with both release patterns varying over time and in the presence of secondary insults. Thirdly, the passage of cerebral S100B is modulated by the BBB on its way into the extracerebral compartment, and serum S100B levels do not reﬂect the corresponding cerebral S100B release. Finally, a beneﬁcial effect of S100B on neuronal maintenance, neurogenesis and cognitive performance has been demonstrated promoting repair mechanisms and functional recovery following TBI. Supplementing measurements of S100B serum levels, we established MR proton spectroscopy to more accurately reﬂect cerebral S100B levels and

322

Fig. 2. Neuronal differentiation of progenitor cells in the ipsilateral dentate gyrus following TBI and S100B treatment. The confocal photomicrograph shows the staining of the neuronal Marker NeuN (red) and BrdU immunoreactivity (green) in the dentate gyrus at 5 weeks following lateral ﬂuid percussion injury and an intraventricular S100B infusion. The arrows point at progenitor cells demonstrating a co-localization of NeuN and BrdU (orange) thereby indicating neuronal differentiation.

thus provide a clinical method to non-invasively monitor whole brain S100B levels repeatedly in patients with brain insults. The determination of cerebral S100B levels may allow a more accurate estimation of the prognosis in head-injured patients. Moreover, we propose that S100B far from being a negative determinant of outcome, as suggested previously in the human TBI and ischemia literature, may improve neurogenesis and functional recovery following acute brain injury and may be in fact potentially a new treatment option, which might improve outcome of patients with many forms of acute brain damage.

References Agelink, M.W., Andrich, J., Postert, T., Wurzinger, U., Zeit, T., Klotz, P. and Przuntek, H. (2001) Relation between

electroconvulsive therapy, cognitive side effects, neuron speciﬁc enolase, and protein S-100. J. Neurol. Neurosurg. Psychiatry, 71: 394–396. Ahlemeyer, B., Beier, H., Semkova, I., Schaper, C. and Krieglstein, J. (2000) S-100beta protects cultured neurons against glutamate- and staurosporine-induced damage and is involved in the antiapoptotic action of the 5 HT(1A)-receptor agonist, Bay 3702. Brain Res., 858: 121–128. Anderson, R.E., Hansson, L.O., Nilsson, O., Dijlai-Merzoug, R. and Settergren, G. (2001a) High serum S100B levels for trauma patients without head injuries. Neurosurgery, 48: 1255–1258; discussion 1258–1260. Anderson, R.E., Hansson, L.O., Nilsson, O., Liska, J., Settergren, G. and Vaage, J. (2001b) Increase in serum S100A1-B and S100BB during cardiac surgery arises from extracerebral sources. Ann. Thorac. Surg., 71: 1512–1517. Arcuri, C., Giambanco, I., Bianchi, R. and Donato, R. (2002) Annexin V, annexin VI, S100A1 and S100B in developing and adult avian skeletal muscles. Neuroscience, 109: 371–388. Ashraf, S., Bhattacharya, K., Tian, Y. and Watterson, K. (1999) Cytokine and S100B levels in paediatric patients

323 undergoing corrective cardiac surgery with or without total circulatory arrest. Eur. J. Cardiothorac. Surg., 16: 32–37. Azmitia, E.C., Dolan, K. and Whitaker-Azmitia, P.M. (1990) S-100B but not NGF, EGF, insulin or calmodulin is a CNS serotonergic growth factor. Brain Res., 516: 354–356. Barger, S.W., Van Eldik, L.J. and Mattson, M.P. (1995) S100 beta protects hippocampal neurons from damage induced by glucose deprivation. Brain Res., 677: 167–170. Berger, R.P., Adelson, P.D., Pierce, M.C., Dulani, T., Cassidy, L.D. and Kochanek, P.M. (2005) Serum neuron-speciﬁc enolase, S100B, and myelin basic protein concentrations after inﬂicted and noninﬂicted traumatic brain injury in children. J. Neurosurg., 103: 61–68. Brewton, L.S., Haddad, L. and Azmitia, E.C. (2001) Colchicine-induced cytoskeletal collapse and apoptosis in N-18 neuroblastoma cultures is rapidly reversed by applied S-100beta. Brain Res., 912: 9–16. Bruhn, H., Frahm, J., Gyngell, M.L., Merboldt, K.D., Hanicke, W. and Sauter, R. (1989) Cerebral metabolism in man after acute stroke: new observations using localized proton NMR spectroscopy. Magn. Reson. Med., 9: 126–131. Chen, D.Q. and Zhu, L.L. (2005) Dynamic change of serum protein S100b and its clinical significance in patients with traumatic brain injury. Chin. J. Traumatol., 8: 245–248. Chirumamilla, S., Sun, D., Bullock, M.R. and Colello, R.J. (2002) Traumatic brain injury induced cell proliferation in the adult mammalian central nervous system. J. Neurotrauma, 19: 693–703. Ciccarelli, R., Di Iorio, P., Bruno, V., Battaglia, G., D’Alimonte, I., D’Onofrio, M., Nicoletti, F. and Caciagli, F. (1999) Activation of A(1) adenosine or mGlu3 metabotropic glutamate receptors enhances the release of nerve growth factor and S-100beta protein from cultured astrocytes. Glia, 27: 275–281. Dash, P.K., Mach, S.A. and Moore, A.N. (2001) Enhanced neurogenesis in the rodent hippocampus following traumatic brain injury. J. Neurosci. Res., 63: 313–319. Donato, R. (1986) S-100 proteins. Cell Calcium, 7: 123–145. Donato, R. (1991) Perspectives in S-100 protein biology. Review article. Cell Calcium, 12: 713–726. Donato, R. (2001) S100: a multigenic family of calciummodulated proteins of the EF-hand type with intracellular and extracellular functional roles. Int. J. Biochem. Cell Biol., 33: 637–668. Ellis, E.F., McKinney, J.S., Willoughby, K.A., Liang, S. and Povlishock, J.T. (1995) A new model for rapid stretchinduced injury of cells in culture: characterization of the model using astrocytes. J. Neurotrauma, 12: 325–339. Eriksson, P.S., Perﬁlieva, E., Bjork-Eriksson, T., Alborn, A.M., Nordborg, C., Peterson, D.A. and Gage, F.H. (1998) Neurogenesis in the adult human hippocampus. Nat. Med., 4: 1313–1317. Fazeli, M.S., Errington, M.L., Dolphin, A.C. and Bliss, T.V. (1990) Extracellular proteases and S100 protein in long-term potentiation in the dentate gyrus of the anaesthetized rat. Adv. Exp. Med. Biol., 268: 369–375. Frahm, J., Bruhn, H., Gyngell, M.L., Merboldt, K.D., Hanicke, W. and Sauter, R. (1989a) Localized high-resolution proton

NMR spectroscopy using stimulated echoes: initial applications to human brain in vivo. Magn. Reson. Med., 9: 79–93. Frahm, J., Bruhn, H., Gyngell, M.L., Merboldt, K.D., Hanicke, W. and Sauter, R. (1989b) Localized proton NMR spectroscopy in different regions of the human brain in vivo. Relaxation times and concentrations of cerebral metabolites. Magn. Reson. Med., 11: 47–63. Frahm, J., Michaelis, T., Merboldt, K.D., Hanicke, W., Gyngell, M.L., Chien, D. and Bruhn, H. (1989c) Localized NMR spectroscopy in vivo. Progress and problems. NMR Biomed., 2: 188–195. Gomide, V.C. and Chadi, G. (1999) The trophic factors S100beta and basic ﬁbroblast growth factor are increased in the forebrain reactive astrocytes of adult callosotomized rat. Brain Res., 835: 162–174. Gould, E. and Gross, C.G. (2002) Neurogenesis in adult mammals: some progress and problems. J. Neurosci., 22: 619–623. Hayakata, T., Shiozaki, T., Tasaki, O., Ikegawa, H., Inoue, Y., Toshiyuki, F., Hosotubo, H., Kieko, F., Yamashita, T., Tanaka, H., Shimazu, T. and Sugimoto, H. (2004) Changes in CSF S100B and cytokine concentrations in early-phase severe traumatic brain injury. Shock, 22: 102–107. Heizmann, C.W., Fritz, G. and Schafer, B.W. (2002) S100 proteins: structure, functions and pathology. Front. Biosci., 7: d1356–d1368. Hicks, R.R., Smith, D.H., Lowenstein, D.H., Saint Marie, R. and McIntosh, T.K. (1993) Mild experimental brain injury in the rat induces cognitive deficits associated with regional neuronal loss in the hippocampus. J. Neurotrauma, 10: 405–414. Hu, J., Ferreira, A. and Van Eldik, L.J. (1997) S100beta induces neuronal cell death through nitric oxide release from astrocytes. J. Neurochem., 69: 2294–2301. Huttunen, H.J., Kuja-Panula, J., Sorci, G., Agneletti, A.L., Donato, R. and Rauvala, H. (2000) Coregulation of neurite outgrowth and cell survival by amphoterin and S100 proteins through receptor for advanced glycation end products (RAGE) activation. J. Biol. Chem., 275: 40096–40105. Isobe, T., Takahashi, K. and Okuyama, T. (1984) S100a0 (alpha alpha) protein is present in neurons of the central and peripheral nervous system. J. Neurochem., 43: 1494–1496. Isobe, T., Tsugita, A. and Okuyama, T. (1978) The amino acid sequence and the subunit structure of bovine brain S-100 protein (PAP I-b). J. Neurochem., 30: 921–923. Iwasaki, Y., Shiojima, T. and Kinoshita, M. (1997) S100 beta prevents the death of motor neurons in newborn rats after sciatic nerve section. J. Neurol. Sci., 151: 7–12. Jackson, R.G., Samra, G.S., Radcliffe, J., Clark, G.H. and Price, C.P. (2000) The early fall in levels of S-100 beta in traumatic brain injury. Clin. Chem. Lab. Med., 38: 1165–1167. Klein, J.R., Hoon, D.S., Nangauyan, J., Okun, E. and Cochran, A.J. (1989) S-100 protein stimulates cellular proliferation. Cancer Immunol. Immunother., 29: 133–138. Kleindienst, A., Harvey, H.B., Rice, A.C., Muller, C., Hamm, R.J., Gaab, M.R. and Bullock, M.R. (2004) Intraventricular infusion of the neurotrophic protein S100B improves

324 cognitive recovery after ﬂuid percussion injury in the rat. J. Neurotrauma, 21: 541–547. Kleindienst, A., McGinn, M.J., Harvey, H.B., Colello, R.J., Hamm, R.J. and Bullock, M.R. (2005a) Enhanced hippocampal neurogenesis by intraventricular S100B infusion is associated with improved cognitive recovery after traumatic brain injury. J. Neurotrauma, 22: 645–655. Kleindienst, A., Tolias, C.M., Corwin, F.D., Muller, C., Marmarou, A., Fatouros, P. and Bullock, M.R. (2005b) Assessment of cerebral S100B levels by proton magnetic resonance spectroscopy after lateral ﬂuid-percussion injury in the rat. J. Neurosurg., 102: 1115–1121. Kleine, T.O., Benes, L. and Zofel, P. (2003) Studies of the brain speciﬁcity of S100B and neuron-speciﬁc enolase (NSE) in blood serum of acute care patients. Brain Res. Bull., 61: 265–279. Kligman, D. and Marshak, D.R. (1985) Puriﬁcation and characterization of a neurite extension factor from bovine brain. Proc. Natl. Acad. Sci. U.S.A., 82: 7136–7139. Kornack, D.R. and Rakic, P. (1999) Continuation of neurogenesis in the hippocampus of the adult macaque monkey. Proc. Natl. Acad. Sci. U.S.A., 96: 5768–5773. Landar, A., Caddell, G., Chessher, J. and Zimmer, D.B. (1996) Identiﬁcation of an S100A1/S100B target protein: phosphoglucomutase. Cell Calcium, 20: 279–285. Li, Y., Wang, J., Sheng, J.G., Liu, L., Barger, S.W., Jones, R.A., Van Eldik, L.J., Mrak, R.E. and Grifﬁn, W.S. (1998) S100 beta increases levels of beta-amyloid precursor protein and its encoding mRNA in rat neuronal cultures. J. Neurochem., 71: 1421–1428. Marshak, D.R. (1990) S100 beta as a neurotrophic factor. Prog. Brain Res., 86: 169–181. Matsutani, T., Nagayoshi, M., Tamaru, M., Hirata, Y. and Kato, K. (1985) Changes in the levels of neural cell speciﬁc proteins in the developing rat brain. Neurochem. Res., 10: 1155–1172. McAdory, B.S., Van Eldik, L.J. and Norden, J.J. (1998) S100B, a neurotropic protein that modulates neuronal protein phosphorylation, is upregulated during lesion-induced collateral sprouting and reactive synaptogenesis. Brain Res., 813: 211–217. Mello e Souza, T., Rohden, A., Meinhardt, M., Goncalves, C.A. and Quillfeldt, J.A. (2000) S100B infusion into the rat hippocampus facilitates memory for the inhibitory avoidance task but not for the open-ﬁeld habituation. Physiol. Behav., 71: 29–33. Moore, B.W. (1965) A soluble protein characteristic of the nervous system. Biochem. Biophys. Res. Commun., 19: 739–744. Nishi, M., Kawata, M. and Azmitia, E.C. (1997) S100beta promotes the extension of microtubule associated protein2 (MAP2)-immunoreactive neurites retracted after colchicine treatment in rat spinal cord culture. Neurosci. Lett., 229: 212–214. Nygren De Boussard, C., Fredman, P., Lundin, A., Andersson, K., Edman, G. and Borg, J. (2004) S100 in mild traumatic brain injury. Brain Inj., 18: 671–683.

O’Dowd, B.S., Zhao, W.Q., Ng, K.T. and Robinson, S.R. (1997) Chicks injected with antisera to either S-100 alpha or S-100 beta protein develop amnesia for a passive avoidance task. Neurobiol. Learn Mem., 67: 197–206. Pelinka, L.E., Kroepﬂ, A., Leixnering, M., Buchinger, W., Raabe, A. and Redl, H. (2004) GFAP versus S100B in serum after traumatic brain injury: relationship to brain damage and outcome. J. Neurotrauma, 21: 1553–1561. Pelinka, L.E., Szalay, L., Jafarmadar, M., Schmidhammer, R., Redl, H. and Bahrami, S. (2003) Circulating S100B is increased after bilateral femur fracture without brain injury in the rat. Br. J. Anaesth., 91: 595–597. Petrova, T.V., Hu, J. and Van Eldik, L.J. (2000) Modulation of glial activation by astrocyte-derived protein S100B: differential responses of astrocyte and microglial cultures. Brain Res., 853: 74–80. Petzold, A., Keir, G., Lim, D., Smith, M. and Thompson, E.J. (2003) Cerebrospinal ﬂuid (CSF) and serum S100B: release and wash-out pattern. Brain Res. Bull., 61: 281–285. Povlishock, J.T., Becker, D.P., Sullivan, H.G. and Miller, J.G. (1978) Vascular permeability alterations to horseradish peroxidase in experimental brain injury. Brain Res., 153: 223–239. Raabe, A. and Seifert, V. (2000) Protein S-100B as a serum marker of brain damage in severe head injury: preliminary results. Neurosurg. Rev., 23: 136–138. Rickmann, M., Wolff, J.R. and Meyer, D.L. (1995) Expression of S100 protein in the vestibular nuclei during compensation of unilateral labyrinthectomy symptoms. Brain Res., 688: 8–14. Rothermundt, M., Peters, M., Prehn, J.H. and Arolt, V. (2003) S100B in brain damage and neurodegeneration. Microsc. Res. Tech., 60: 614–632. Rothoerl, R.D., Brawanski, A. and Woertgen, C. (2000) S-100B protein serum levels after controlled cortical impact injury in the rat. Acta Neurochir. (Wien.), 142: 199–203. Rustandi, R.R., Drohat, A.C., Baldisseri, D.M., Wilder, P.T. and Weber, D.J. (1998) The Ca(2+)-dependent interaction of S100B(beta beta) with a peptide derived from p53. Biochemistry, 37: 1951–1960. Savola, O., Pyhtinen, J., Leino, T.K., Siitonen, S., Niemela, O. and Hillbom, M. (2004) Effects of head and extracranial injuries on serum protein S100B levels in trauma patients. J. Trauma, 56: 1229–1234; discussion 1234. Schafer, B.W., Wicki, R., Engelkamp, D., Mattei, M.G. and Heizmann, C.W. (1995) Isolation of a YAC clone covering a cluster of nine S100 genes on human chromosome 1q21: rationale for a new nomenclature of the S100 calcium-binding protein family. Genomics, 25: 638–643. Scotto, C., Mely, Y., Ohshima, H., Garin, J., Cochet, C., Chambaz, E. and Baudier, J. (1998) Cysteine oxidation in the mitogenic S100B protein leads to changes in phosphorylation by catalytic CKII-alpha subunit. J. Biol. Chem., 273: 3901–3908. Selinfreund, R.H., Barger, S.W., Pledger, W.J. and Van Eldik, L.J. (1991) Neurotrophic protein S100 beta stimulates glial cell proliferation. Proc. Natl. Acad. Sci. U.S.A., 88: 3554–3558.

325 Shashoua, V.E., Hesse, G.W. and Moore, B.W. (1984) Proteins of the brain extracellular ﬂuid: evidence for release of S-100 protein. J. Neurochem., 42: 1536–1541. Shore, P.M., Thomas, N.J., Clark, R.S., Adelson, P.D., Wisniewski, S.R., Janesko, K.L., Bayir, H., Jackson, E.K. and Kochanek, P.M. (2004) Continuous versus intermittent cerebrospinal ﬂuid drainage after severe traumatic brain injury in children: effect on biochemical markers. J. Neurotrauma, 21: 1113–1122. Slemmer, J.E., Matser, E.J., De Zeeuw, C.I. and Weber, J.T. (2002) Repeated mild injury causes cumulative damage to hippocampal cells. Brain, 125: 2699–2709. Song, H., Stevens, C.F. and Gage, F.H. (2002) Astroglia induce neurogenesis from adult neural stem cells. Nature, 417: 39–44. Suzuki, F., Kato, K. and Nakajima, T. (1984) Hormonal regulation of adipose S-100 protein release. J. Neurochem., 43: 1336–1341. Svenmarker, S., Sandstrom, E., Karlsson, T. and Aberg, T. (2002) Is there an association between release of protein S100B during cardiopulmonary bypass and memory disturbances? Scand. Cardiovasc. J., 36: 117–122. Van Eldik, L.J., Christie-Pope, B., Bolin, L.M., Shooter, E.M. and Whetsell Jr., W.O. (1991) Neurotrophic activity of S-100 beta in cultures of dorsal root ganglia from embryonic chick and fetal rat. Brain Res., 542: 280–285. Van Eldik, L.J. and Zimmer, D.B. (1987) Secretion of S-100 from rat C6 glioma cells. Brain Res., 436: 367–370. Vos, P.E., Lamers, K.J., Hendriks, J.C., van Haaren, M., Beems, T., Zimmerman, C., van Geel, W., de Reus, H., Biert, J. and Verbeek, M.M. (2004) Glial and neuronal proteins in serum predict outcome after severe traumatic brain injury. Neurology, 62: 1303–1310.

Whitaker-Azmitia, P.M., Murphy, R. and Azmitia, E.C. (1990) Stimulation of astroglial 5-HT1A receptors releases the serotonergic growth factor, protein S-100, and alters astroglial morphology. Brain Res., 528: 155–158. Wilder, P.T., Rustandi, R.R., Drohat, A.C. and Weber, D.J. (1998) S100B(betabeta) inhibits the protein kinase C-dependent phosphorylation of a peptide derived from p53 in a Ca2+dependent manner. Protein Sci., 7: 794–798. Willoughby, K.A., Kleindienst, A., Muller, C., Chen, T., Muir, J.K. and Ellis, E.F. (2004) S100B protein is released by in vitro trauma and reduces delayed neuronal injury. J. Neurochem., 91: 1284–1291. Winningham-Major, F., Staecker, J.L., Barger, S.W., Coats, S. and Van Eldik, L.J. (1989) Neurite extension and neuronal survival activities of recombinant S100 beta proteins that differ in the content and position of cysteine residues. J. Cell Biol., 109: 3063–3071. Wunderlich, M.T., Wallesch, C.W. and Goertler, M. (2004) Release of neurobiochemical markers of brain damage is related to the neurovascular status on admission and the site of arterial occlusion in acute ischemic stroke. J. Neurol. Sci., 227: 49–53. Xiong, Z., O’Hanlon, D., Becker, L.E., Roder, J., MacDonald, J.F. and Marks, A. (2000) Enhanced calcium transients in glial cells in neonatal cerebellar cultures derived from S100B null mice. Exp. Cell Res., 257: 281–289. Zimmer, D.B. and Van Eldik, L.J. (1986) Identiﬁcation of a molecular target for the calcium-modulated protein S100. Fructose-1,6-bisphosphate aldolase. J. Biol. Chem., 261: 11424–11428. Zimmer, D.B. and Van Eldik, L.J. (1987) Tissue distribution of rat S100 alpha and S100 beta and S100-binding proteins. Am. J. Physiol., 252: C285–C289.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 23

Cerebellar injury: clinical relevance and potential in traumatic brain injury research Eugene Park1,2, Jinglu Ai1 and Andrew J. Baker1,2, 1 St. Michael’s Hospital, Trauma Research, Toronto, ON, M5B 1W8, Canada University of Toronto, Institute of Medical Sciences, Toronto, ON, M5S 1A, Canada

2

Abstract: A treatment for traumatic brain injury (TBI) remains elusive despite compelling evidence from animal models for a variety of therapeutic targets. Numerous animal models have been developed to address the wide spectrum of mechanisms involved in the progression of secondary injury after TBI. Evidence from well-established models such as the ﬂuid percussion injury (FPI) device, cortical impact model, and the impact acceleration model has demonstrated diffuse pathophysiological mechanisms throughout various brain structures. More specifically, we have recently extended characterization of the FPI model to include pathophysiological changes in the cerebellum following unilateral ﬂuid percussion. Data suggest that the cerebellum is susceptible to selective Purkinje cell loss as well as white matter dysfunction. Despite the cerebellum’s low proﬁle in TBI research, there is evidence to warrant further study of the cerebellum to examine mechanisms of neuronal death and traumatic axonal injury. Furthermore, evidence from clinical literature and basic science suggests that some components of TBI pathophysiology have a basis in cerebellar dysfunction. This review highlights some of the recent ﬁndings in cerebellar trauma and builds an argument for including the cerebellum as a model to assess mechanisms of secondary injury and its potential contribution to the pathology of TBI. Keywords: animal models; cerebellum; electrophysiology; FPI; TAI; TBI injury mechanisms exists in variety of injured axon subpopulations (reviewed by Buki and Povlishock (2006)). These data support a need for in-depth scrutiny of mechanisms encompassing all regions susceptible to traumatic brain injury (TBI) in order to formulate an accurate and thorough description of pathophysiologic processes. The cerebellum’s role in TBI has received relatively limited scrutiny and characterization as other structures including the brain stem, hippocampus, and cerebral cortex. It is becoming increasingly evident that general mechanistic descriptions of cell death and axonal injury are overly simplistic and do not accurately reﬂect the full spectrum of events occurring after TBI. Clinical cases of TBI and parallel research in animal

Introduction Intense scrutiny of neuronal cell death and traumatic axonal injury (TAI) has clariﬁed numerous molecular and cellular events contributing to the progression of secondary injury cascades. As our understanding of molecular events continues to increase, our description of these events have become increasingly complex. For example, all traumatically injured axons were previously thought to undergo a series of pathophysiological events leading to disconnection and bulb formation. However, it has become evident that a spectrum of Corresponding author. Tel.: +1-416-864-5510; Fax: +1-416-

864-5512; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61023-6

327

328

models have conﬁrmed that the mechanisms of injury extend to the cerebellum thus building a case for its relevance in neurotrauma. A thorough characterization of all affected structures throughout the brain, including the cerebellum, would provide a better framework with which to optimize treatments for TBI. Furthermore, the potential for discovering or elucidating mechanisms of injury not yet described would also add to our existing knowledge of secondary injury mechanisms. It may be argued that cerebellar trauma is of limited clinical importance; however, recent studies have begun to elucidate roles in higher order cognitive function not previously appreciated in cerebellar processing (Petrosini et al., 1998; Ramnani, 2006). These higher order functions as well as other neurological processing functions not previously known may account for some of the observed behavioral abnormalities in people afﬂicted by TBI. In the present discussion we propose several lines of evidence to support the role of cerebellar injury as an important component of TBI, and also several unique features that render the cerebellum a potentially useful structure in which to assess mechanisms of secondary injury after trauma.

Clinical evidence of cerebellar trauma Clinical manifestation of cerebellar dysfunction include perturbations to stance and gait, characterized by a loss of equilibrium, a wide-based stance, irregular steps, and lateral veering (Mariotti et al., 2005). Cerebellar deficits can also give rise to tremors (e.g., Parkinsonian), alterations to speech resulting in slowed or slurred vocalization, as well as cerebellar mutism (Gordon, 1996). Recent studies have also indicated an important contribution of cerebellar function to visuo-motor control (Glickstein, 2000). Patients with cerebellar ataxias exhibit ocular motor defects in target ﬁxation and coordinating head movements (Mariotti et al., 2005). Although these deficits are generally not described in the context of trauma, it is not unreasonable to draw inference to traumatic cerebellar injury which could result in similar pathological impairments, particularly since

many of the motor, cognitive, and speech impairments described above are also observed in TBI patients. There are many reviews on diagnoses, mechanisms, outcome, epidemiology, cost analysis, rehabilitation, etc., and speciﬁc case studies in TBI involving cerebral injury. However there is limited clinical literature with direct emphasis on cerebellar trauma. In 1917, Gordon Holmes described observations of World War I soldiers who had sustained gunshot wounds to the cerebellum. He described a drunken-sailor type gait, as well as the postural disturbances commonly associated with cerebellar dysfunction (Holmes, 1917). A recent report of 26 patients who had sustained infratentorial injuries is the only clinical study to date to specifically address the issue of direct trauma to the cerebellum (Nathoo et al., 2002). This particular type of injury, although rare in the civilian population, has been documented predominantly in military literature (Rish et al., 1983; Brandvold et al., 1990). Moreover, the authors of this study note that cerebellar injury had no effect on outcome. It should be noted however, that the ﬁve point Glasgow Outcome Scale (GOS) used in this particular study might not be an adequately sensitive measure of functional and behavioral deficits to detect cerebellar dysfunction. Furthermore, the GOS is primarily a measure of global outcome and does not factor in the consequences of motor, speech, and cognitive impacts on daily living. Despite limited literature on direct cerebellar trauma there is a varied collection of literature in which the cerebellum has been referenced as exhibiting pathological changes including selective cell loss, altered metabolism, and white matter injury after focal and diffuse TBI. For example, Purkinje cell loss has been noted in the brains of boxers and implicated as a significant contributor to ataxias and Parkinsonian tremors (Guterman and Smith, 1987; Unterharnscheidt, 1995a, b). Another report cites hypertrophic olivary and Purkinje cell degeneration following chronic TBI (Anderson and Treip, 1973). Post-traumatic intracerebellar hemorrhage, though an unusual clinical occurrence, has been observed after head trauma and is an important clinical entity with respect to

329

neurological outcome (Bostrom et al., 1992; D’Avella et al., 2001). There have also been several reports of cerebellar atrophy following TBI (Krauss et al., 1995; Soto-Ares et al., 2001; Gale et al., 2005). Furthermore, a subgroup of ataxic TBI patients may have etiologic basis in cerebellar dysfunction after injury (Chester and Reznick, 1987; Mysiw et al., 1990). Crossed cerebellar diaschisis (CCD), a well-documented phenomenon of altered or depressed metabolic ﬂow and activity in the hemisphere of the cerebellum contralateral to the side of cortical lesion or injury, has also been documented after TBI (Alavi et al., 1997). Ipsilateral metabolic changes as well as hypermetabolic changes have also been documented (Shamoto and Chugani, 1997; Niimura et al., 1999). A loss of afferent inputs from the cortico-ponto-cerebellar pathway is believed to result in deactivation of the targets in the cerebellum, although the exact physiological mechanism is not entirely understood (Feeney and Baron, 1986). The occurrence of CCD underscores the importance of synaptic connectivity between the cerebrum and the cerebellum. This would suggest TBI resulting in a loss of cortical synaptic input to the cerebellum that results in an unknown volume of information processing that is not taking place. The relationship between ischemia and cerebellar injury is also of significance as Purkinje cells are highly susceptible to ischemic injury (Bhatia et al., 1995; Welsh et al., 2002). Ischemia is a common component of non-penetrating head injuries and an important contributor to secondary injury mechanisms (Graham et al., 1978, 1989). Examination of 151 cases of fatal head injury revealed that the majority of ischemic damage occurred in the hippocampus (81% of patients); however, there was also a significant proportion of patients with evidence of ischemic injury in the cerebellum (44%; Graham et al., 1978). Furthermore, TBI, including multisystem trauma patients, often present with hypoxichypotensive or hemorrhagic complications (Miller et al., 1978, 1981; The Brain Trauma Foundation, 2000) increasing the likelihood of global brain ischemia and subsequent cerebellar injury.

Cerebellar injury in animal models Purkinje cell vulnerability Several studies in animal models of TBI have documented vulnerability of Purkinje cells following forebrain injury. Both midline and unilateral forebrain FPI have been shown to result in significant and delayed cell death of Purkinje neurons accompanied by the presence of activated microglia (Fukuda et al., 1996; Mautes et al., 1996). Other studies have supported these ﬁndings by demonstrating the presence of FluorojadeB positive cells in the cerebellum, consistent with the morphology of Purkinje neurons (Sato et al., 2001; Hallam et al., 2004). Using double label ﬂuorescent immunohistochemistry, we recently demonstrated selective vulnerability of Purkinje neurons in the posterior regions of the cerebellum following unilateral FPI in the cerebrum (Park et al., 2006b). In this study we demonstrated that the majority of Purkinje neurons were lost in the acute phase of injury (24 h) while delayed cell loss occurred in the middle and anterior regions of the cerebellum at 7 and 14 days post-injury at higher grades of injury severity (Fig. 1). Examination of the coronal plane cerebellar sections following forebrain FPI indicates no evidence of medio-lateral banding (unpublished observations). Others, however, have reported parasagittal banding patterns of Purkinje cell loss using the CCI model of trauma (Weber, J., personal communication), perhaps indicative of variations in injury biomechanics to the cerebellum. Although the mechanisms of Purkinje cell loss following trauma are still unknown, there are several potential explanations to consider with relevance to neurotrauma research. Presynaptic hyperexcitability, differential gene expression, pre- and post-synaptic cell survival, and neuronal–glial interactions are several possible avenues to be explored. The cerebellum offers several unique features to address these mechanisms as contributors to secondary injury. Furthermore, these areas have a broad range of application to the study of TBI in general and are not limited to analysis of cerebellum speciﬁc effects.

330

Fig. 1. Double label immunohistochemistry indicates selective Purkinje cell death in the cerebellum following forebrain ﬂuid percussion trauma: (a) Sham, (b) 1 day post-injury, (c) 14 days post-injury. Scale bar ¼ 50 mm. (d) Quantiﬁcation of surviving Purkinje cells from the posterior cerebellum following four grades of ﬂuid percussion trauma indicates a dose–response effect at 1 day postinjury (red asterisk). A significant decline in Purkinje cell numbers relative to sham animals is observed as early as 1 day post-injury in the 2–2.5 atm injury groups (*). The cartoon (left) indicates the area of ﬂuid percussion trauma (red arrow) and the region of Purkinje cell quantiﬁcation (red circle). (Adapted with permission from Park et al., 2006b.)

Presynaptic hyperexcitability The synaptic architecture of the Purkinje cells, consisting of glutamatergic inputs from climbing and parallel ﬁbers, creates an environment in which the potential for synaptic mediated excitotoxicity is high (for review see Slemmer et al., 2005). Our laboratory has demonstrated that direct cerebellar trauma results in delayed

presynaptic hyperexcitability in parallel ﬁberPurkinje cell synaptic connections (Ai and Baker, 2002). Given that a single Purkinje cell receives input from 200,000 parallel ﬁbers (Fox and Barnard, 1957), the potential for a presynaptic excitotoxic event is high. In addition, a single climbing ﬁber, originating from the inferior olive, forms up to 1500 synaptic connections on the proximal dendrite of a Purkinje cell (Strata and

331

Rossi, 1998). Excitation of this afferent pathway with ibogaine administration results in excitotoxic Purkinje cell death while ablation with 3-AP is protective to Purkinje cells (O’Hearn and Molliver, 1997). The pattern of cell death from climbing ﬁber depolarization results in very distinct medio-lateral banding pattern of cell death likely corresponding to the terminal afferent projections of the climbing ﬁbers. However, there are factors beyond simple afferent targets to consider in this discussion of Purkinje cell vulnerability.

Gene expression Despite a seemingly homogeneous and redundant arrangement of cellular architecture throughout the cerebellum (Voogd and Glickstein, 1998; Ramnani, 2006), there exist subtle differences within cell populations that can give rise to vastly different pathologies. In particular, differential expression of genes involved in metabolism and cell signaling have been demonstrated within subsets of Purkinje neurons with both medio-lateral and anterior–posterior patterns of expression (reviewed by Herrup and Kuemerle (1997) and Sarna and Hawkes (2003)) (Fig. 2). Furthermore patterned gene expression in Purkinje neurons occurs independently of afferent synaptic organization. The effectiveness of neuroprotective strategies targeting these differentially expressed gene products can be readily ascertained through histological examination for the presence or absence of banding patterns (see O’Hearn and Molliver, 1997). There are numerous candidate proteins that are expressed in distinct patterns within the cerebellum that are of interest to neurotrauma research including heat shock proteins, zebrin expression, calcium binding proteins, and amino acid transporters (Herrup and Kuemerle, 1997). There may also be others that have not yet been characterized to date. Also of importance is the maintenance of the compartmentalized expression of these gene products across mammalian species permitting for a degree of consistency in crossspecies comparisons of mechanism and role in TBI.

Pre- and post-synaptic cell survival The cerebellum is a potentially useful structure in which to assess the effects of neuronal cell loss in pre- and post-synaptic targets following TBI. For example, the lurcher mouse phenotype resulting from a mutation of the Grid2 gene for the d2-glutamate receptor (d2-GluR) results in chronic depolarization and loss of Purkinje cells (Zuo et al., 1997). Subsequent to Purkinje neuron death is the loss of granule cells as well as retrograde loss of inferior olive input (Heckroth and Eisenman, 1991; Heckroth, 1992; Zanjani et al., 1998). Anterograde effects manifest as a loss of deep cerebellar nuclei (Heckroth, 1994). Whether these effects are speciﬁc to the mutation of the d2-GluR or are a common response to Purkinje cell loss has not been elucidated but may be a valuable area of research to pursue. The extensive literature describing afferent tracing to the cerebellum as well as its relatively simple synaptic organization makes it a suitable model in which to evaluate retrograde and anterograde fates of injured or dying neurons. These studies could add to our understanding of how focal injuries affect distal targets through retrograde or anterograde signaling dysfunction or cell loss.

Neuronal– glial cell communication The role of neuronal–glial communication following neurotrauma remains a controversial issue with evidence to supporting glial scarring as inhibitor of endogenous repair mechanisms along with compelling evidence to indicate neuroprotective roles as well (reviewed by Sofroniew (2005)). In the cerebellum, Bergmann glia residing in the molecular layer form a complex functional and structural relationship with Purkinje cells (reviewed by Bellamy (2006)). These specialized astrocytes form a sheath around the Purkinje neuron soma and synapses. This physical interaction has functional implications as stimulation of climbing and parallel ﬁbers have been shown to activate inward currents in Bergmann glia (Bergles et al., 1997; Clark and Barbour, 1997; Bellamy and Ogden, 2005). These currents are in part

332

Fig. 2. Cartoon representation of cerebellar compartmentation in the mouse, as revealed by the expression of zebrin II in subsets of Purkinje cells (Adapted with permission from Sillitoe and Hawkes, 2002): anterior, dorsal and posterior views are shown. The cerebellar vermis is divided into 10 lobules (I–X). However, a more fundamental parcellation is into four transverse zones — anterior (AZ), central (CZ), posterior (PZ), and nodular (NZ). The AZ and PZ are striped — all Purkinje cells in the CZ and NZ express zebrin II uniformly (although stripes can be revealed here by using other markers, such as the small heat shock protein, HSP25 (Armstrong et al., 2000)). Complementary views of a cerebellum whole mount immunoperoxidase stained for zebrin II are shown on the right. (Adapted with permission from Sarna and Hawkes, 2003.)

contributed to by glial-expressed alpha-amino-3hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptors, GluR1 and GluR4, as well as glial glutamate transporters, GLAST and GLT-1 (Bellamy, 2006). Although the Bergmann glia

represent a specialized form of astrocyte, they maintain the characteristic reactive astrocyte response to injury following trauma. We have demonstrated the presence of reactive astrocytes in regions of Purkinje cell loss following forebrain

333

Fig. 3. (a) Confocal image of sham tissue expression of GFAP in Bergmann glia astrocytes, in red, in close proximity to Purkinje cells, labeled in green. (b) Following forebrain ﬂuid percussion trauma there is a marked increased in GFAP immunoreactivity in regions of cerebellar injury indicated by the loss of Purkinje cells and increase in GFAP immunoreactivity. Scale bar ¼ 50 mm. (Adapted with permission from Park et al., 2006b.)

trauma as indicated by increased GFAP expression (Fig. 3). The close anatomical coupling and functional relationship between Purkinje cells and Bergman glia presents an excellent opportunity to examine the changes in neuronal and glial communication following TBI. Specifically, the cerebellum is an ideal system in which to perform patch-clamping experiments in brain slices, which permits maintenance of cellular architecture and synaptic connectivity. These features may be helpful in elucidating the changes in communication and synaptic plasticity between glial cells and neurons in pathophysiological states. Cerebellar white matter injury Numerous animal models of TBI have reported evidence of TAI in the cerebellum (Lighthall et al., 1990; Shima and Marmarou, 1991; Foda and Marmarou, 1994; Hoshino et al., 2003). Information regarding the functional consequences of such injuries, however, is limited. We recently characterized functional and structural changes in cerebellar white matter following forebrain FPI and demonstrated significant and persistent deficits and pathological changes in the cerebellum (Park et al., 2006a). Early neuroﬁlament degradation and

parallel accumulation in a subset of axons was observed at 1 day post-injury while persistent calpainmediated degradation of aII-spectrin was observed up to 14 days post-injury. The acute degradation of heavy neuroﬁlament chain (NF200) and prolonged degradation of calpain-mediated aII-spectrin would suggest temporally preferential targeting of calpain substrates, NF200 and aII-spectrin, in the cerebellum. Another possibility is that proteolysis of these substrates occurred in subsets of axons with different modes of axonal injury progression. The observations of NF200 and aII-spectrin degradation are not novel concepts in the study of TAI. However, the results reveal a further level of complexity in TAI with respect to calpain’s temporal substrate speciﬁcity. Compound action potential recordings from cerebellar white matter have also demonstrated the usefulness of electrophysiological techniques in assessment of white matter function after TBI. The results indicate a consistent CAP response within selected regions. Interestingly, recordings from the middle cerebellar lobe produce a primarily fast conducting myelinated response whereas the posterior and anterior cerebellar lobes exhibit a two-peaked response including both fast and slow myelinated and unmyelinated responses, respectively (Fig. 4). This

334

Fig. 4. (a) Compound action potentials were recorded from three regions of cerebellar white matter as indicated by the black-boxed regions. (b) CAP waveforms were dependent on location of recording. Anterior and posterior waveforms consisted of distinct fast and slow components. The middle cerebellar CAP response was primarily a fast myelinated signal. (c) CAP values from cerebellar white matter following forebrain ﬂuid percussion trauma indicated a significant decline in electrophysiological function (*) in the posterior and middle regions of the cerebellum which persisted at 14 days post-injury. (Adapted with permission from Park et al., 2006a.)

presents an opportunity to examine the effects of TBI and potential therapeutic interventions on both myelinated and demyelinated axon populations. A recent study demonstrated varying electrophysiological susceptibilities of these two axonal populations in the corpus callosum (Reeves et al., 2005).

Given the relative lack of electrophysiological data available for TBI white matter injury studies, a comparison of results between the cerebellum and the corpus callosum may reveal whether these susceptibilities are structure speciﬁc or are a general principle of injured axons throughout the brain.

335

The cerebellum as a model of neurotrauma A key question in the discussion of cerebellar injury and its application in neurotrauma research is ultimately whether it is clinically relevant. We would argue ‘yes’ for the following reasons. There is sufﬁcient evidence of clinical manifestation of cerebellar injury following TBI. In addition to the role in motor coordination, the role of the cerebellum in multiple higher order function is becoming more appreciated and hence a recognition of its potential importance if injured. The traumatized cerebellum offers numerous technical opportunities to examine relevant areas of research related to mechanisms of cellular injury, neuronal–glial pathophysiological interactions, TAI, post-injury synaptic plasticity, selective neuronal vulnerability, and mechanisms of diffuse injury remote from the location of initial trauma. From an anatomical perspective, there are notable differences in human and rodent cerebellar placement that is likely to result in differences on the production of injury biomechanics between these two species. However, there are structural similarities in the cerebellum, such as the cellular organization and foliated segments that are maintained throughout all mammalian species (Voogd and Glickstein, 1998). This gyrencephalic property is not maintained in the cerebral cortex between species and represents an area in which the biomechanics of injury production may differ significantly between higher and lower order species. Despite the pros and cons of anatomical similarity between species, the intent of this discussion is to provide a rationale for inclusion of the cerebellum not only for its clinical and functional importance in TBI pathophysiology, but also as a structure in which to examine the mechanistic phenomena of TBI mechanisms in general. We propose that comparison and complement of studies across various injury models will not only validate but also elucidate novel mechanism of secondary injury.

Conclusion Traumatic brain injury continues to be a health care issue of epidemic proportions as a leading

cause of morbidity and mortality in young adults. Despite the prevalence of this silent epidemic, there is little therapeutic beneﬁt that has translated from the neuroprotective strategies developed in animal models of TBI (Bullock et al., 1999; Faden, 2001; Narayan et al., 2002; Tolias and Bullock, 2004). The failure of clinical trials highlights a lack of complete understanding of the complexity of TBI pathophysiology. This includes optimizing therapeutic time windows and clarifying the multiple pathways leading to cell death. Addressing these issues will require multifaceted approaches and balancing the inhibition of secondary injury mechanisms while not adversely affecting normal physiologic function (Faden, 2002; Ikonomidou and Turski, 2002). We believe that this will be best achieved through complementation and comparisons between existing and novel injury paradigms. The cerebellum is one such novel area that remains to be fully appreciated and described in the context of TBI. Abbreviations AMPA CCD FPI GluR GOS NF200 TAI TBI

alpha-amino-3-hydroxy-5methyl-4-isoxazole propionic acid crossed cerebellar diaschisis ﬂuid percussion injury glutamate receptor Glasgow outcome scale heavy neuroﬁlament chain traumatic axonal injury traumatic brain injury

Acknowledgment We thank the Ontario Neurotrauma Foundation for its funding support. References Ai, J. and Baker, A. (2002) Presynaptic hyperexcitability at cerebellar synapses in traumatic injury rat. Neurosci. Lett., 332: 155–158. Alavi, A., Mirot, A., Newberg, A., Alves, W., Gosﬁeld, T., Berlin, J., Reivich, M. and Gennarelli, T. (1997) Fluorine-18FDG evaluation of crossed cerebellar diaschisis in head injury. J. Nucl. Med., 38: 1717–1720.

336 Anderson, J.R. and Treip, C.S. (1973) Hypertrophic olivary degeneration and Purkinje cell degeneration in a case of longstanding head injury. J. Neurol. Neurosurg. Psychiatry, 36: 826–832. Armstrong, C.L., Krueger-Naug, A.M., Currie, R.W. and Hawkes, R. (2000) Constitutive expression of the 25-kDa heat shock protein Hsp25 reveals novel parasagittal bands of purkinje cells in the adult mouse cerebellar cortex. J. Comp. Neurol., 416: 383–397. Bellamy, T.C. (2006) Interactions between Purkinje neurones and Bergmann glia. Cerebellum, 5: 116–126. Bellamy, T.C. and Ogden, D. (2005) Short-term plasticity of Bergmann glial cell extrasynaptic currents during parallel ﬁber stimulation in rat cerebellum. Glia, 52: 325–335. Bergles, D.E., Dzubay, J.A. and Jahr, C.E. (1997) Glutamate transporter currents in bergmann glial cells follow the time course of extrasynaptic glutamate. Proc. Natl. Acad. Sci. U.S.A., 94: 14821–14825. Bhatia, K.P., Brown, P., Gregory, R., Lennox, G.G., Manji, H., Thompson, P.D., Ellison, D.W. and Marsden, C.D. (1995) Progressive myoclonic ataxia associated with coeliac disease. The myoclonus is of cortical origin, but the pathology is in the cerebellum. Brain, 118(Pt 5): 1087–1093. Bostrom, K., Helander, C.G. and Lindgren, S. (1992) Blunt basal head trauma: aspects of unconsciousness. Acta Neurochir. Suppl. (Wien.), 55: 25–28. Brandvold, B., Levi, L., Feinsod, M. and George, E.D. (1990) Penetrating craniocerebral injuries in the Israeli involvement in the Lebanese conﬂict, 1982–1985: analysis of a less aggressive surgical approach. J. Neurosurg., 72: 15–21. Buki, A. and Povlishock, J.T. (2006) All roads lead to disconnection? Traumatic axonal injury revisited. Acta Neurochir. (Wien.), 148: 181–194. Bullock, M.R., Lyeth, B.G. and Muizelaar, J.P. (1999) Current status of neuroprotection trials for traumatic brain injury: lessons from animal models and clinical studies. Neurosurgery, 45: 207–220. Chester, C.S. and Reznick, B.R. (1987) Ataxia after severe head injury: the pathological substrate. Ann. Neurol., 22: 77–79. Clark, B.A. and Barbour, B. (1997) Currents evoked in Bergmann glial cells by parallel ﬁbre stimulation in rat cerebellar slices. J. Physiol., 502(Pt 2): 335–350. D’avella, D., Cacciola, F., Angileri, F.F., Cardali, S., La Rosa, G., Germano, A. and Tomasello, F. (2001) Traumatic intracerebellar hemorrhagic contusions and hematomas. J. Neurosurg. Sci., 45: 29–37. Faden, A.I. (2001) Neuroprotection and traumatic brain injury: the search continues. Arch. Neurol., 58: 1553–1555. Faden, A.I. (2002) Neuroprotection and traumatic brain injury: theoretical option or realistic proposition. Curr. Opin. Neurol., 15: 707–712. Feeney, D.M. and Baron, J.C. (1986) Diaschisis. Stroke, 17: 817–830. Foda, M.A. and Marmarou, A. (1994) A new model of diffuse brain injury in rats. Part II: morphological characterization. J. Neurosurg., 80: 301–313.

Fox, C.A. and Barnard, J.W. (1957) A quantitative study of the Purkinje cell dendritic branchlets and their relationship to afferent ﬁbres. J. Anat., 91: 299–313. Fukuda, K., Aihara, N., Sagar, S.M., Sharp, F.R., Pitts, L.H., Honkaniemi, J. and Noble, L.J. (1996) Purkinje cell vulnerability to mild traumatic brain injury. J. Neurotrauma, 13: 255–266. Gale, S.D., Baxter, L., Roundy, N. and Johnson, S.C. (2005) Traumatic brain injury and grey matter concentration: a preliminary voxel based morphometry study. J. Neurol. Neurosurg. Psychiatry, 76: 984–988. Glickstein, M. (2000) How are visual areas of the brain connected to motor areas for the sensory guidance of movement? Trends Neurosci., 23: 613–617. Gordon, N. (1996) Speech, language, and the cerebellum. Eur. J. Disord. Commun., 31: 359–367. Graham, D.I., Adams, J.H. and Doyle, D. (1978) Ischaemic brain damage in fatal non-missile head injuries. J. Neurol. Sci., 39: 213–234. Graham, D.I., Ford, I., Adams, J.H., Doyle, D., Teasdale, G.M., Lawrence, A.E. and Mclellan, D.R. (1989) Ischaemic brain damage is still common in fatal non-missile head injury. J. Neurol. Neurosurg. Psychiatry, 52: 346–350. Guterman, A. and Smith, R.W. (1987) Neurological sequelae of boxing. Sports Med., 4: 194–210. Hallam, T.M., Floyd, C.L., Folkerts, M.M., Lee, L.L., Gong, Q.Z., Lyeth, B.G., Muizelaar, J.P. and Berman, R.F. (2004) Comparison of behavioral deficits and acute neuronal degeneration in rat lateral ﬂuid percussion and weight-drop brain injury models. J. Neurotrauma, 21: 521–539. Heckroth, J.A. (1992) Development of glutamic acid decarboxylase-immunoreactive elements in the cerebellar cortex of normal and lurcher mutant mice. J. Comp. Neurol., 315: 85–97. Heckroth, J.A. (1994) Quantitative morphological analysis of the cerebellar nuclei in normal and lurcher mutant mice. I: morphology and cell number. J. Comp. Neurol., 343: 173–182. Heckroth, J.A. and Eisenman, L.M. (1991) Olivary morphology and olivocerebellar topography in adult lurcher mutant mice. J. Comp. Neurol., 312: 641–651. Herrup, K. and Kuemerle, B. (1997) The compartmentalization of the cerebellum. Annu. Rev. Neurosci., 20: 61–90. Holmes, G. (1917) The symptoms of acute cerebellar injuries due to gunshot wounds. Brain, 40: 461–535. Hoshino, S., Kobayashi, S., Furukawa, T., Asakura, T. and Teramoto, A. (2003) Multiple immunostaining methods to detect traumatic axonal injury in the rat ﬂuid-percussion brain injury model. Neurol. Med. Chir. (Tokyo), 43: 165–174. Ikonomidou, C. and Turski, L. (2002) Why did NMDA receptor antagonists fail clinical trials for stroke and traumatic brain injury? Lancet Neurol., 1: 383–386. Krauss, J.K., Wakhloo, A.K., Nobbe, F., Trankle, R., Mundinger, F. and Seeger, W. (1995) Lesion of

337 dentatothalamic pathways in severe post-traumatic tremor. Neurol. Res., 17: 409–416. Lighthall, J.W., Goshgarian, H.G. and Pinderski, C.R. (1990) Characterization of axonal injury produced by controlled cortical impact. J. Neurotrauma, 7: 65–76. Mariotti, C., Fancellu, R. and Di Donato, S. (2005) An overview of the patient with ataxia. J. Neurol., 252: 511–518. Mautes, A.E., Fukuda, K. and Noble, L.J. (1996) Cellular response in the cerebellum after midline traumatic brain injury in the rat. Neurosci. Lett., 214: 95–98. Miller, J.D., Butterworth, J.F., Gudeman, S.K., Faulkner, J.E., Choi, S.C., Selhorst, J.B., Harbison, J.W., Lutz, H.A., Young, H.F. and Becker, D.P. (1981) Further experience in the management of severe head injury. J. Neurosurg., 54: 289–299. Miller, J.D., Sweet, R.C., Narayan, R. and Becker, D.P. (1978) Early insults to the injured brain. JAMA, 240: 439–442. Mysiw, W.J., Corrigan, J.D. and Gribble, M.W. (1990) The ataxic subgroup: a discrete outcome after traumatic brain injury. Brain Inj., 4: 247–255. Narayan, R.K., Michel, M.E., Ansell, B., Baethmann, A., Biegon, A., Bracken, M.B., Bullock, M.R., Choi, S.C., Clifton, G.L., Contant, C.F., Coplin, W.M., Dietrich, W.D., Ghajar, J., Grady, S.M., Grossman, R.G., Hall, E.D., Heetderks, W., Hovda, D.A., Jallo, J., Katz, R.L., Knoller, N., Kochanek, P.M., Maas, A.I., Majde, J., Marion, D.W., Marmarou, A., Marshall, L.F., McIntosh, T.K., Miller, E., Mohberg, N., Muizelaar, J.P., Pitts, L.H., Quinn, P., Riesenfeld, G., Robertson, C.S., Strauss, K.I., Teasdale, G., Temkin, N., Tuma, R., Wade, C., Walker, M.D., Weinrich, M., Whyte, J., Wilberger, J., Young, A.B. and Yurkewicz, L. (2002) Clinical trials in head injury. J. Neurotrauma, 19: 503–557. Nathoo, N., Chite, S.H., Edwards, P.J. and van Dellen, J.R. (2002) Civilian infratentorial gunshot injuries: outcome analysis of 26 patients. Surg. Neurol., 58: 225–233. Niimura, K., Chugani, D.C., Muzik, O. and Chugani, H.T. (1999) Cerebellar reorganization following cortical injury in humans: effects of lesion size and age. Neurology, 52: 792–797. O’Hearn, E. and Molliver, M.E. (1997) The olivocerebellar projection mediates ibogaine-induced degeneration of Purkinje cells: a model of indirect, trans-synaptic excitotoxicity. J. Neurosci., 17: 8828–8841. Park, E., Liu, E., Shek, M., Park, A. and Baker, A. (2006a) Heavy neuroﬁlament accumulation and alpha-spectrin degradation accompanies cerebellar white matter functional deficits following forebrain ﬂuid percussion injury. Exp. Neurol., in press, Oct. 26, Epub ahead of print. Park, E., McKnight, S., Ai, J. and Baker, A.J. (2006b) Purkinje cell vulnerability to mild and severe forebrain head trauma. J. Neuropathol. Exp. Neurol., 65: 226–234. Petrosini, L., Leggio, M.G. and Molinari, M. (1998) The cerebellum in the spatial problem solving: a co-star or a guest star? Prog. Neurobiol., 56: 191–210.

Ramnani, N. (2006) The primate cortico-cerebellar system: anatomy and function. Nat. Rev. Neurosci., 7: 511–522. Reeves, T.M., Phillips, L.L. and Povlishock, J.T. (2005) Myelinated and unmyelinated axons of the corpus callosum differ in vulnerability and functional recovery following traumatic brain injury. Exp. Neurol., 196: 126–137. Rish, B.L., Dillon, J.D. and Weiss, G.H. (1983) Mortality following penetrating craniocerebral injuries: an analysis of the deaths in the Vietnam Head Injury Registry population. J. Neurosurg., 59: 775–780. Sarna, J.R. and Hawkes, R. (2003) Patterned Purkinje cell death in the cerebellum. Prog. Neurobiol., 70: 473–507. Sato, M., Chang, E., Igarashi, T. and Noble, L.J. (2001) Neuronal injury and loss after traumatic brain injury: time course and regional variability. Brain Res., 917: 45–54. Shamoto, H. and Chugani, H.T. (1997) Glucose metabolism in the human cerebellum: an analysis of crossed cerebellar diaschisis in children with unilateral cerebral injury. J. Child. Neurol., 12: 407–414. Shima, K. and Marmarou, A. (1991) Evaluation of brain-stem dysfunction following severe ﬂuid-percussion head injury to the cat. J. Neurosurg., 74: 270–277. Sillitoe, R.V. and Hawkes, R. (2002) Whole-mount immunohistochemistry: a high-throughput screen for patterning defects in the mouse cerebellum. J. Histochem. Cytochem., 50: 235–244. Slemmer, J.E., De Zeeuw, C.I. and Weber, J.T. (2005) Don’t get too excited: mechanisms of glutamate-mediated Purkinje cell death. Prog. Brain Res., 148: 367–390. Sofroniew, M.V. (2005) Reactive astrocytes in neural repair and protection. Neuroscientist, 11: 400–407. Soto-Ares, G., Vinchon, M., Delmaire, C., Abecidan, E., Dhellemes, P. and Pruvo, J.P. (2001) Cerebellar atrophy after severe traumatic head injury in children. Child’s Nerv. Syst., 17: 263–269. Strata, P. and Rossi, F. (1998) Plasticity of the olivocerebellar pathway. Trends Neurosci., 21: 407–413. The Brain Trauma Foundation. The American Association of Neurological Surgeons. The Joint Section on Neurotrauma and Critical Care. (2000) Resuscitation of blood pressure and oxygenation. J. Neurotrauma, 17: 471–478. Tolias, C.M. and Bullock, M.R. (2004) Critical appraisal of neuroprotection trials in head injury: what have we learned? NeuroRx, 1: 71–79. Unterharnscheidt, F. (1995a) A neurologist’s reﬂections on boxing. I: impact mechanics in boxing and injuries other than central nervous system damage. Rev. Neurol., 23: 661–674. Unterharnscheidt, F. (1995b) A neurologist’s reﬂections on boxing. V: Concluding remarks. Rev. Neurol., 23: 1027–1032. Voogd, J. and Glickstein, M. (1998) The anatomy of the cerebellum. Trends Neurosci., 21: 370–375. Welsh, J.P., Yuen, G., Placantonakis, D.G., Vu, T.Q., Haiss, F., O’Hearn, E., Molliver, M.E. and Aicher, S.A. (2002) Why do Purkinje cells die so easily after global brain

338 ischemia? Aldolase C, EAAT4, and the cerebellar contribution to posthypoxic myoclonus. Adv. Neurol., 89: 331–359. Zanjani, H.S., Vogel, M.W., Martinou, J.C., Delhaye-Bouchaud, N. and Mariani, J. (1998) Postnatal expression of Hu-bcl-2 gene in Lurcher mutant mice fails to rescue Purkinje

cells but protects inferior olivary neurons from target-related cell death. J. Neurosci., 18: 319–327. Zuo, J., De Jager, P.L., Takahashi, K.A., Jiang, W., Linden, D.J. and Heintz, N. (1997) Neurodegeneration in Lurcher mice caused by mutation in delta2 glutamate receptor gene. Nature, 388: 769–773.

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 24

Sex differences in brain damage and recovery of function: experimental and clinical ﬁndings Donald G. Stein Department of Emergency Medicine, Emory University School of Medicine, Atlanta, GA, USA

Abstract: Until the last decade or so, there was very little systematic examination of sex differences in recovery from brain injury — most of the work was anecdotal or based on very small studies comparing males to females. This chapter reviews some of the physiological, morphological, and functional evidence for sex differences in response to brain injury across the spectrum of development. It also examines more recent data showing that ﬂuctuations in hormonal status during the menstrual and estrous cycle can play a determining role in functional outcome in both normal and brain-injured females, and that these hormonal inﬂuences can be measured at both the cellular and behavioral levels. Keywords: sex differences; brain damage; recovery; hormones; neurosteroids animal studies females were excluded because of concerns that their hormonal cycling would interfere with or inﬂuence drug metabolism and therefore ‘confound’ the interpretation of results. However, there have been a number of highly publicized initiatives calling for the systematic study of gender differences, as well as several edited volumes on sex differences and brain function (McIntosh, 1996; Kimura, 1999; Morrison, 2000), and the number of empirical studies examining sex differences in brain injury has increased substantially in the last few years.

Introduction Over the last 30 years, development of some 40 different compounds, and industry and governmentsupported clinical trials for those compounds, have failed to ﬁnd any pharmacologic agent that safely and effectively led to better functional outcomes after traumatic brain injury (TBI). With the qualiﬁed exception of tissue plasminogen activator (TPA), the search for an effective treatment for ischemic stroke yielded the same disappointing results. Throughout all these trials, no substantive, systematic attention was paid to whether there were any differences in brain injury severity and outcomes between males and females. Many of the clinical trials did not include females and most of the laboratory research has focused closely on male animals. In the human trials, women were often excluded because investigators were worried about possible side effects of new drugs on fecundity or hormonal cycling. Indeed, in both human and

Traumatic brain injury is not a unitary event As most of the articles in this volume will show, TBI is not a simple event in which a sudden impact or penetration of the brain parenchyma kills a certain number of cells. Rather it is a complex cascade of events that continues to unfold over time and involves a relatively large number of genomic, metabolic, cellular and anatomic changes that can take months to years before complete

Corresponding author. Tel: +1-404-712-9704; Fax: +1-404-

727-2388; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61024-8

339

340

stabilization. To promote functional and adaptive central nervous system (CNS) repair in the acute stage of injury, a number of steps are required. First, the initial progression of events causing cytotoxicity, inﬂammation, swelling and shortand long-term degeneration of nerve and glial cells must be reduced. This process of repair engages not just the damaged neurons, but the entire organism’s adaptive capacity. The prevention of further loss of vulnerable or injured nerve cells can play an important role in rebuilding damaged circuitry and in providing a physiological matrix for later, successful rehabilitation training and therapy. Second, once the cascade of cytotoxic events is controlled, endogenous, growth-promoting factors as well as the administration of exogenous, trophic agents may then stimulate the formation of new axons, dendrites and synapses or even enhance neurogenesis in the damaged adult brain (Brinton and Wang, 2006; Kobayashi et al., 2006; Sailor et al., 2006; Shankaran et al., 2006; Shivraj Sohur et al.; 2006, Taupin, 2006). Taken together, all these changes require considerable time and all occur in a complex genomic and proteomic milieu that can be dramatically affected by a number of contextual parameters. We now know that such parameters include the subject’s health, sex and genomic and hormonal status during development and at the time of injury. It is important to note here that there is no empirically based ‘principle’ of neuroscience claiming that unless recovery of function occurs within the ﬁrst few weeks after an injury, there will never be any recovery at all (see Dancause et al. (2005) for review and discussion on this issue). For the purposes of this chapter, one of the key questions to be asked is, ‘‘Are there, in fact, sex differences in the outcome of brain injury, and could they be traced to differences in hormonal status or factors related to steroid synthesis in the brain before, during, or after traumatic damage?’’ It has typically been assumed that sex differences in the cognitive, sensory, or motor performance of males and females are due to the divergences in the anatomical structure and neuronal connectivity of the male and female brain. This is often referred to in the experimental literature as sexual dimorphism. In an excellent recent review, ‘‘Why Sex

Matters for Neuroscience,’’ Larry Cahill (2006) points out that there are ‘copious’ sex inﬂuences on brain function and anatomy that are often disregarded in neuroscience research. He lists ﬁve misconceptions concerning sex inﬂuences on brain: (1) sex differences are small and unreliable (therefore not worth considering?); (2) average differences between the sexes are due to a few extreme cases in a distribution — this was the same argument presented in the early days of work on recovery of function — i.e., extremely rare phenomena are not worthy of serious study; (3) differences within sex are more substantial than between sex and therefore can be dismissed as trivial; (4) any sex differences are likely to be explained by the effects of estrogen, primarily during development; and (5) if no sex differences in behavior exist, it must mean that the neural mechanisms underlying behavior(s) are identical for males and females, although a number of CNS recording and imaging studies now show that this is not the case (Gur and Gur, 2004; Rider and Abdou, 2004; Stein, 2004; Cahill, 2006).

Are there structural differences between the male and female brain that can affect plasticity and outcome of injury? As Cahill has noted, much of the earlier literature on sex differences in vertebrate brain structure emphasized the locus of control for mating, vocalization (in songbirds), reproduction, and aggression and territorial behaviors in the male and female nervous systems (Arnold et al., 2004; Collaer et al., 2004). Following on the rapid developments in non-invasive imaging techniques, a substantial number of recent experiments have studied sex differences in metabolic activity in a variety of brain areas as men and women go about performing sensory, motor and cognitive tasks. This is perhaps the one area of clinical neuroscience showing the most interest in studying whether sexual dimorphisms underlie the differences in male and female behaviors (see Gur and Gur (2004) and Becker (2002) for comprehensive reviews). For the present chapter and volume the goal is to determine whether sexual dimorphisms

341

in the anatomy of the brain could account for any beneﬁcial or detrimental differences in the response of the nervous system to traumatic injuries.

Sex differences in the gross and cellular anatomy of the brain There are now many anatomical and physiological studies describing sex differences in structure/ function relationships in the brain (Kimura, 1992; Shaywitz et al., 1995; Kimura, 1999; Cahill, 2006). Recently, Rabinowicz et al. (2002) examined 86 different areas in the brains of 6 males and 5 females, 12–24 years old. These investigators measured cortical thickness, neuronal and glial packing densities (the volume of cells in a geometrically deﬁned area), and neuronal and astrocyte sizes (the size of the ‘neuropil’). The authors reported that females had more overall neuropil than males, but the males had higher neuronal densities (more synaptic contacts /mm3). The men had smaller but more numerous units, while women had larger cells, especially in the left hemisphere (women are more frequently righthanded and have better language abilities). In another recent report, Goldstein et al. (2001) used MRI to obtain T1-weighted three-dimensional images from 48 normal adults of similar education, age, socioeconomic status, intelligence, and handedness. These investigators found that, in general, women had larger cortical brain volumes than men; however, the frontomedial cortex, the amygdala and the hypothalamus were larger in men. In particular, there was greater sexual dimorphism in those brain regions which, compared to homologous animal studies, show the greatest numbers of sex steroid receptors during development. One area of the cortex that can be affected by sex, age, and hormonal milieu of the subject is the medial frontal cortex, implicated by numerous lesion studies in spatial learning and cognition (Markham and Juraska, 2002). The frontal cortex and its related structures have also received considerable attention in the functional recovery literature in both humans and laboratory animals. Markham et al. (2005) examined both young and aged male and female rats (the latter in various

stages of the estrous cycle) and measured dendritic spine density and arborization (the major site of excitatory activity) in the anterior cingulate cortex to determine whether age, sex, and ovarian hormonal status would change the morphology of these cells. In general, young males had greater spine densities and arborization than females, although both sexes showed an expected decline in the medial frontal cortex with age. However, the aged females showed less age-related drop-off of spine density and arborization than the males. This may be because the aged female rats, unlike human females, continue to produce ovarian hormones, which could prevent some of the cell loss. It is also worth noting that the Juraska laboratory claims that males also show an age-related loss of dendritic spines in the motor cortex while females do not (unpublished meeting abstract). Other investigators have reported sex differences in the long-term synaptic potentiation (LTP) of dentate granule cells of the hippocampus. This activity is thought to subserve hippocampally mediated learning and memory. For example, Maren (1995) recorded LTP in young male and female rats after single-pulse stimulation of the perforant path. In this study, stimulation led to a greater magnitude of LTP in the male rats compared to females and was attributed to the males having higher levels of NMDA receptor activation than the females. In this context, since males have higher levels of LTP activation, and since hippocampus and temporal cortex are particularly susceptible to seizures, it would be interesting to learn whether male or female patients with brain injuries have a higher incidence of post-traumatic epilepsy. There are two additional points to note in this context. First, Goldstein et al. (2001) showed that when relative difference in overall brain size is taken into consideration, while the female hippocampus is larger than that of the male, males have a bigger volume of the CA1 ﬁeld and larger neurons than females. Second, females with very high levels of estrogen during their menstrual cycle are more likely to suffer from catamenial epilepsy; this could be due to the greater sensitivity of the female hippocampal cells to neurotransmitters, glucocorticoids, and NMDA receptor binding (see Herzog et al.

342

(2004) and Reddy (2004) for more discussion on this issue). Although there are currently no direct studies on humans bearing directly on whether women might have a higher incidence of post-traumatic epilepsy, there are some interesting clues. For example, Smith et al. (2002), using transcranial magnetic stimulation to activate motor-evoked potentials by paired-pulse technique in the motor cortex, evaluated the effects of ovarian hormones on cortical excitability in healthy human females (average age of 34 years) during the stages of their menstrual cycle. The study sought to determine whether levels of estrogen and progesterone could affect cortical excitability in a way similar to what glutamate agonists or benzodiazepines might do at the glutamate or GABAA receptor(s), respectively. When estrogen levels were high during the late follicular stage, the motor-evoked potentials were signiﬁcantly higher as well. During the luteal phase of the cycle when progesterone was at its peak, the motor-evoked potentials were signiﬁcantly lower. These data were taken to show that sex hormones could induce either inhibition or activation of cerebrocortical activity as well as blood ﬂow, thus affecting drug metabolism and possibly CNS mechanisms involved in cognition, perception and the CNS response to injury. These data can be taken to suggest that, at least for females, any dynamic brain imaging studies should take into consideration the patient’s hormonal state during testing and evaluation.

Sex differences in the brain may be present in glial as well as neuronal morphology Recently, Conejo et al. (2003) used GFAPimmunoreactive labeling and unbiased stereological counting to determine the numbers of reactive astrocytes in the CA and CA3 areas of the adult rat hippocampus in both males and females (whose measurements were taken during the proestrous phase of their cycle). Males had higher numbers of astrocytes than females in the CA3 area of the hippocampus, but the females had higher numbers of those cells in the CA1 region. It is interesting to note that females in proestrus

are purportedly less susceptible to ischemic damage of the CA1 region (He et al., 2002) than males. Conejo et al. suggest that the neuroprotection may be conferred by the higher number of glial cells secreting neurotrophic factors, an idea which is receiving growing experimental support. The higher numbers of glial cells present could also play a role in more efﬁcient scavenging of cytotoxic agents such as excitatory amino acids and free radicals. Of the various types of glial cells found in the brain, astrocytes are the most active in producing neurosteroids such as pregnenalone, progesterone, dehydroepiandrosterone (DHEA), testosterone, estradiol, and others (Zwain and Yen, 1999). Recently, Cerghet et al. (2006) reported that the density of oligodendrocytes and the expression of myelin proteins in the corpus callosum, fornix and spinal cord are 20–40% greater in male compared to female rats. But in the corpus callosum, the generation of new glia and their apoptotic loss is twice as great in females as in males. The authors took this to mean that the lifespan and turnover of oligodendrocytes in females is shorter, and this substantial sex difference suggests that hormones are playing a critical role in this ‘turnover’ process which would include the synthesis of new myelin. To know whether the amounts of such neurosteroids differ between the sexes or whether the levels of neurosteroids or their receptors vary as a function of the estrous cycle in human females will require more research. A better understanding of the role these factors play in mediating CNS plasticity could help to explain why females might have more advantage in recovery from TBI than males, especially if they have the capacity to generate more myelin following a brain injury (Cerghet et al., 2006). In support of this hypothesis, it is interesting to note that after brain or spinal cord injury, the synthesis of pregnenalone, progesterone, and allopregnanolone increases two to four times in the areas immediately adjacent to the injury sites. Concomitant with this rise there is a signiﬁcant increase in glial hyperplasia and the expression of myelin basic protein and remyelination in the areas surrounding the injury (di Michele et al., 2000). It is also worthwhile to note that the expression of the myelin proteins can

343

be directly affected by treatment with progesterone and some of its metabolites — with progesterone having a direct stimulatory effect (Schumacher et al., 2004; De Nicola et al., 2006; Labombarda et al., 2006; Magnaghi et al., 2006). In the brain itself, McCarthy et al. (2002), using both light and electron microscopy, looked at the effects of the estrous cycle on changes in the activity of the TrkA receptor, a major receptor for neurotrophins in the brain which plays a part in neurite outgrowth in the dendritic ﬁelds of the hippocampus. When estrogen levels were high, there was a 16-fold increase in the number of TrkA-labeled astrocytes, but this number dropped off precipitously when the animals were ovariectomized. After estrogen replacement, there was again a 12-fold increase. At the electron microscope level the authors observed that the TrkAlabeled cells were situated closest to dendrites and to unmyelinated axons. McCarthy and colleagues attributed the beneﬁcial effects to estrogen-induced activity helping the cells play a role in axonal guidance and synaptic regeneration. This notion would appear to support the more recent ﬁndings described above. (For more detailed reviews of the role of hormones and dendritic activity, see McEwen (1998) and Arnold et al. (2004).) Much of contemporary neuroscience involves the use of cell culture preparations, but researchers rarely, if ever, report the sex of the donor cells, probably thinking that the sex origin of such cells would have no consequence for the measures they are making. A more recent study (Zhang et al., 2003) now questions this assumption. Zhang et al. compared cell survival and the levels of phosphokinases ERK1, ERK2, Akt and the ‘survival’ protein Bcl-2 in vitro in embryonic (E19) rat cortical neurons (cortical plate and ventricular zone) taken from male and female donors. To analyze cell survival, the cells were double-labeled with MAP-2 (microtubule-associated protein) and anti-GABA antibodies on day 14 in culture. The phosphokinases were evaluated with Western blots and optical density. Cells taken from females had a highly signiﬁcant increase in survival compared to cells taken from male donors, although there did not appear to be any differences in the number of GABA-ergic neurons. The authors suggest that

the female cells underwent less apoptosis than male cells and that this may have been due to the signiﬁcantly higher levels of ERK1, ERK2, Bcl-2 and Akt, which helped to preserve the stability of mitochondrial membranes and produced less degradation over time. The authors state that, ‘‘these data are the ﬁrst demonstration of sexual dimorphism in neuronal cell signaling,’’ and further, that gender may provide a context in which the same biological stimuli can lead to different outcomes in disease and behavior. For us, the fact that hardly any molecular biological investigators consider sex as a variable in the in vitro pharmacokinetic studies leads us to wonder how valid are many of the reports when it comes to deﬁning the proper parameters for drug design, development and application to the treatment of brain injury and other organic disorders. To emphasize the point that sex differences are important in determining outcome of brain injury, three other studies should be mentioned. First, it is now known that small, sublethal ischemic insults can protect from additional ischemic damage in the adult brain, a phenomenon known as ‘ischemic preconditioning’ (IPC) (Truettner et al., 2002). Truettner and colleagues created bilateral cerebral artery occlusions to cause mild temporary ischemia in rats. They then took tissue samples from the hippocampus, cortex, and striatum to examine the expression of genes thought to be implicated in neuroprotection after injury (e.g., brain-derived neurotrophic factor (BDNF), nerve growth factor (NGF), c-jun, and c-fos), and found signiﬁcant increases. The authors speculated that the increase in neurotrophic factors following the pre-conditioning injury could help with protecting neurons from a second attack. What is particularly interesting about this ﬁnding is that others have shown that the extent of tolerance to hypoxia and ischemia in females varies according to the estrous cycle and relative levels of estrogen or progesterone at the time of the insult. Second, recent investigations (Kasischke et al., 1999; von Arnim et al., 2002) using mouse hippocampal slice preparation recordings show that short-term hypoxia will suppress normal electrical activity of neurons. The return of electrical activity is measured by population spike amplitudes (PSA)

344

and varies according to stage of estrus. In the Kasischke et al. experiment, the best overall recovery of the PSA was found in the males, but for females the worst recovery was observed during proestrus when estrogen was high, and the best recovery was seen during diestrus when progesterone was high. This means that if hormonal status and cycling is not considered when females are included in brain injury studies, the reported ﬁndings must be considered inconclusive or incomplete. As a case in point, in the later study by von Arnim et al. (2002), progesterone receptors were down-regulated by preconditioning lesions (seen when progesterone levels are high and associated with neuroprotection). The authors observe that, ‘‘Not only is net hypoxic tolerance gender dependent but underlying mechanisms are gender specific. Both the net difference in primary and induced hypoxic tolerance and the speciﬁcity of the mechanisms conferring the hypoxic tolerance should be considered in the development of future gender-speciﬁc therapeutic strategies’’ (p. 87 [italics added]).

The inﬂuence of sex differences and hormonal status on cerebral functions in normal and brain-damaged subjects There is now a very substantial literature on the role of sex hormones in the development and locus of reproductive, maternal, social, and aggressive behaviors, to which is now being added a growing number of studies examining sex differences in more complex cognitive and sensory functions. In this area, especially in animal research, lesion experiments combined with behavioral testing have long been used to exemplify dimorphic cognitive responses to brain injury. One of the ﬁrst to use this technique in studying the development of sex differences in cognition and recovery after brain injury was Goldman (1974). Goldman created orbitofrontal cortical lesions in 50-day-old male and female rhesus monkeys and then tested them on an object discrimination learning task. At about two-and-a-half months of age, the males with orbitofrontal cortical lesions were impaired in learning but the females with the same surgery

performed the task as well as intact controls. In another series of projects Goldman created the same injuries in males and females, tested the animals soon after surgery and then again over a year later for their learning abilities. At ﬁrst testing, the very young toddler females performed much better than their male counterparts, but the differences between sexes disappeared by 18 months of age. For Goldman, this suggested that males were initially worse because their orbitofrontal cortex and its expected ‘functions’ matured more rapidly, so the loss of this structure produced the early deﬁcits. We have already seen that even in the embryonic stage of development, female brain cells appear to have higher titers of neurotrophic activity than males and this may confer some potential protection (and better behavioral performance) against neuronal loss following injury during infancy (Zhang et al., 2003). Subsequent studies showed that there are neurosteroid receptors in the developing and adult cortex which can play a role in mediating injury-induced neural plasticity (Clark et al., 1988). In a follow-up primate study, Clark and Goldman-Rakic (1989) gave male and female rhesus monkeys orbitofrontal cortex lesions at about 50 days of age followed by testosterone injections in ﬁve of the females. The injections were given from birth to 46 days of age. These animals were then compared to groups of intact monkeys given testosterone prenatally or at 540 and 675 days after birth. The animals were then tested behaviorally on several cognitive discrimination learning tasks. Interestingly, intact males learned the tasks somewhat better than the intact females but the males with brain injury were more impaired on object reversal learning (the task required being able to give up a previously learned strategy) than the lesion females. When the brain-injured females were given testosterone, the advantage disappeared and the lesion females performed as badly as the injured males, again purportedly demonstrating either the neuroprotective effects of the female sex hormones in a brain injury condition or, quite possibly, the detrimental effects of giving testosterone to females. In adult laboratory rats, sex differences in sparing and recovery of function have also been found

345

after unilateral lesions of the entorhinal cortex, an area of the brain thought to be implicated in working and reference memory processing (Roof et al., 1993). These investigators looking at spatial learning performance in both males and females found that intact animals of both sexes performed about the same on a task which required them to swim to a submerged platform that could be located only by using room cues as a spatial reference. After brain injury, the male rats had much more difﬁculty remembering the location of the platform than did the females with the same damage; the latter did not differ from intact males or females on performance of this task, even though the entorhinal cortex lesions were very large. As we learn more about sex differences in the outcome of brain injury, it is becoming apparent that there are no simple answers. Some studies show females have an advantage while others report that males do better, especially when the lesions are inﬂicted early in life while hormonal parameters are still being developed (Kolb and Cioe, 1996; Forgie and Kolb, 1998). Extremely low birth weight (ELBW) children (o1 kg) also show sex differences in cognitive abilities as early as 2 years of age (Hindmarsh et al., 2000). The authors followed almost 400 ELBW children after discharge and tested them for cognitive abilities at 2 years of age using the Grifﬁths Mental Development Scales (locomotor, personal social, hearing and speech, eye and hand coordination, and ‘practical reasoning’). For the most part, the investigators found that, except for locomotion, the female children were ‘signiﬁcantly superior’ in all areas of testing and regardless of the speciﬁc impairments manifested by the boys and girls at time of testing. These data can be taken to support other ﬁndings showing that, in general, very young females perform better in language and social development and that this difference could be due to exposure to differential gonadal hormones early in life, which then leads to more rapid maturation and enhanced rate of function in cerebral structures of females. Although other explanations may be offered, this may be one of the reasons why males tend to have a higher prevalence of developmental lags and learning disabilities (Arnold, 1996; Berk, 1997).

Raz et al. (1995) examined 58 children (34 boys, 24 girls) 3 years or older who had suffered an intracranial hemorrhage (ICH) and who were of 37 weeks of gestation or less at birth. The children were equated for a variety of factors such as age, race, mother’s IQ, parents’ education, and socioeconomic status. These children were compared on the Revised Wechsler Intelligence Test to high birth risk infants without ICH. The girls signiﬁcantly outperformed the boys when the extent of insult was controlled for. The authors suggest that even after the various controls for lesion size, males seem to be more functionally affected by early brain injury than females because their brains are generally more vulnerable to insult. They state that, ‘‘it is possible that regardless of developmental factors, the male brain may possess structural or physiological properties that could not only affect the extent of cerebral damage, but might also compromise functional recovery’’ (p. 964). A subsequent paper by Donders and Hoffman (2002) also reported that girls with blunt TBI had better verbal outcome scores on the California Verbal Learning Test than boys, even after they were fully equated on a variety of demographic, premorbid measures, including basic intelligence pre- and post-injury, and severity and type of injury. After the injury, the boys could not remember as many words, used fewer adaptive learning strategies and needed more time than girls to work on test problems. Following their review of the literature, Donders and Hoffman suggest that the outcome of sex differences in pediatric TBI could be due to the neuroprotective effects of sex steroids, especially progesterone, although there are few, if any, studies examining the role of hormonal treatments in survivors of pediatric TBI. Although there are reports of females doing better, some studies draw the opposite conclusions. For example, Morrison et al. (2004) found no statistically signiﬁcant differences in outcome between boys and girls with brain injury and there was even some trend toward the females being somewhat worse. Sex differences in the brain-injured adult The inﬂuence of hormonal status (e.g., estrus) on cognitive performance in adult humans has been

346

more controversial than the ﬁndings in children, perhaps because of the ways in which hormonal assays are conducted and performance is measured over the menstrual cycle (see Epting and Overman (1998) and Mumenthaler et al. (2001) for detailed discussions on this issue). Data from laboratory animals may be more precise and consistent because, among other things, it is easier to obtain a variety of carefully timed measures of hormonal ﬂuctuations (see DeGraba and Pettigrew (2000) for a discussion on some of the issues). In rats (Kinsley et al., 1999), females that had a number of litters previously were compared to age-matched females that never had a litter and to foster mothers who never delivered any litters but raised pups that were given to them. The multiparous females and foster mothers, with higher circulating levels of progesterone, performed much better on a complex maze task than did the nulliparous animals. In rhesus monkeys (Lacreuse et al., 2001), spatial recognition memory performance also varies across the menstrual cycle. Performance of the animals was measured on a number of learning tasks and found to be better during the follicular and luteal phases than during the peri-ovulatory phase, when estrogen levels were at their highest. This ﬁnding ﬁts with other data showing that high levels of estradiol can disrupt learning and memory in female rats (Galea et al., 2001). Taken together, these data suggest that either reduced levels of estrogen, or relatively higher levels of progesterone, or possibly both, are beneﬁcial to learning. The role of estrogen and related hormones in mediating recovery after brain injury is becoming an increasingly studied issue, especially in light of the recently failed trials with estrogen in the treatment of stroke and Alzheimer’s disease in women. With respect to CNS trauma and possibly late-life degenerative disorders such as Alzheimer’s, it may also be important to determine ﬁrst whether the extent and severity of the diseases vary in relation to the normal hormonal status of the female. If natural or induced hormonal cycling does play a role in the progression of injury or disease, then the timing of drug therapy to the appropriate stage of estrus will need to be considered in the treatment of adolescent learning and attention disorders.

In an animal model of transient forebrain ischemia, b-estradiol has been shown to exacerbate the loss of neurons in the CA1 pyramidal cell ﬁeld. When rats were ovariectomized the ischemia caused a loss of 32% of pyramidal cells, but with estradiol replacement in intact and ovariectomized rats the mean cell loss was 54% and 49%, respectively (Harukuni et al., 2001). Would administration of progesterone alone or in combination with the estradiol have prevented this loss? Estrogen is a proexcitatory agent and could have caused neuronal loss by increasing excitotoxicity and subsequent inﬂammation in neurons rendered particularly vulnerable by the ischemic event. The hippocampus is also particularly sensitive to seizure activity, which could have triggered the additional loss of the injured neurons. Progesterone, in contrast, has been shown to inhibit seizure activity, especially in women with catemenial epilepsy, and inhibition of progesterone metabolism increases seizure activity (Herzog and Frye, 2003).

Do females generally have better outcomes after brain damage than males? There is considerable controversy in the literature over the issue of whether females recover better from TBI and stroke than males, though anecdotal reports claim an advantage for females. Some clinical case-study reports suggest that women recover better from stroke than men, while others claim the opposite or no differences between the sexes. For example, a paper by Kertesz and McCabe (1977) examined 23 males and 13 females and found no differences in recovery between the sexes, whereas Basso et al. (1982) examined males and females and found long-term aphasia severity to be much more marked in males than in females, even though initial impairments were judged to be the same. Groswasser et al. (1998) looked at the outcome of ‘severe’ brain injuries in males and females (72 females, 262 males) and controlled for age range and severity. Females apparently had better overall responses to rehabilitation therapy than males when return-to-work was used as an outcome measure.

347

Some reports contest the Groswasser and colleagues ﬁndings. Kraus et al. (2000) looked at gender differences in outcome after TBI in two trauma centers and found that mortality was substantially higher in females than in males with the same extent of injury. The females who survived had poorer outcomes compared to males. Supporting this ﬁnding, Farace and Alves (2000) performed a meta-analysis of previously published TBI studies to determine whether women did better than men in TBI outcome on 20 different measures. These investigators found that fatality rates for women with TBI were much higher than for men (because of body size women may suffer more severe injuries; there are also some data to suggest that they are more often hit by motor vehicles than injured while driving them). Poorer outcomes were also reported for women on measures of disability, seizures, coma, etc. Farace and Alves mention that sex differences in drug metabolism and other variables could contribute to the negative ﬁndings. The report makes no mention of attempting to determine the women’s hormonal status before, during or after injury to see if these factors could have contributed to the more detrimental outcomes. Farace and Alves list a number of reasons why females may have poorer outcomes than males, one of them being sex hormones. Recently, Farin et al. (2003) reported that women have signiﬁcantly greater frequencies of brain swelling than men and worse outcomes than men if they are under 50 years of age at the time of their injury. These results could have been due to higher levels of estrogen relative to progesterone in the younger women at the time of injury, but this variable was not studied (i.e., the authors did not mention whether some of the older women were on hormone replacement therapy (HRT) and whether the younger women were on contraceptives, both of which could affect outcomes). It was interesting to note that after 50 years of age, when both estrogen and progesterone are diminished, the amount of brain swelling in older females dropped and was almost equivalent to that of the males. Another retrospective study by Coimbra et al. (2003) also found no evidence of any sex differences in TBI outcomes between males and females on any of their measures of recovery or mortality.

To add to this controversy, Ratcliffe et al. (2003) looked at the relationship between gender and cognitive recovery one year after blunt TBI. They selected 325 records from the TBI Model Systems National Database and used multivariate analyses to evaluate the association between sex and six cognitive measures of attention — working memory, verbal memory, language, visual analytic skills, problem solving, and motor functioning. Females represented 31% of the total number of patients tested. The women showed better performance on ﬁve of the six cognitive outcome measures. The authors point out that pre-injury differences were not a factor in this design because men did not perform better than women after injury on tasks like visual memory, on which in the normal condition they typically outperform females. The discrepancies among the different outcome studies can be due to a number of factors, not the least of which is the way in which patients are selected for studies. For instance, Obler et al. (1995) reported in ‘early’ (i.e., 1970s) studies that lesion site rather than gender was the determining factor in selecting patients and that this could have biased outcome results. We now know that the loci of brain injuries related to aphasia severity and disorders in linguistic abilities differ between men and women. According to Kimura (1999), the incidence and severity of speech disorders is much higher in women if damage is to the anterior cortex in the left hemisphere, while the reverse is true for men. Women also have higher incidence of manual apraxias after lateral frontal cortical injuries, while men show greater deﬁcits if the damage is in the posterior cortex. Could the outcome of these studies be inﬂuenced by the women’s hormonal status at the time of their injury? There is some tantalizing but indirect evidence bearing on this question. Hausmann et al. (2002) used radioimmunoassays taken every three days to evaluate the hormonal status of healthy women volunteers participating in a perceptual asymmetry study using a variety of behavioral tasks that are taken to measure cerebral asymmetries. When the women were high in serum progesterone, there was less task-dependent cerebral asymmetry on some of the behavioral tests, leading to what the

348

investigators called ‘interhemispheric decoupling,’ and this was interpreted to mean that progesterone mediates the extent of cerebral lateralization. Therefore, as noted earlier, when functional MRIs are used to evaluate structure–function/metabolic activities in the brains of women, it may be important to determine the subjects’ speciﬁc hormonal status at the time of testing in order to obtain more reliable information concerning CNS organization and whether there is, indeed, contextual and/or sex related cerebral asymmetry. With respect to a number of earlier studies (and in some current ones), Obler et al. (1995) note that many of the variables that could determine aphasia severity were not published (e.g., gender distribution, time since onset of the injury and testing, speciﬁc locus of the injury, prior health status, hearing status, medication history, etc.), and this lack of information could certainly have contributed to the variability in the ﬁndings. For purposes of the present discussion, especially for female subjects, hormonal status at the time of injury is never mentioned. Moreover, it is highly unlikely that the patients were queried as to whether they were taking any kind of birth control medication or HRT prior to or at the time of their TBI. Such a chronic history of hormone regulation could affect the outcome of the brain injury, if, in fact, hormone levels were playing any role in mediating inﬂammatory processes that accompany TBI (Stein, 2005). In general, and especially when people suffer from a severe brain injury, there is also accompanying severe bodily injury and shock. There is evidence that females, especially when progesterone levels are high relative to estrogen, maintain better immune function activity and response to physiologic shock than males (Wichmann et al., 1996; Zellweger et al., 1997; Kuebler et al., 2003). These better levels of immunodepression may also beneﬁt the response to brain injury. Especially in older women with stroke or injury, whether they are/were on HRT prior to or after the damage is never noted. This could be particularly important if the older women were or were not on HRT at the time of their injury. All these factors could contribute to differences in brain injury outcome, and if there are, as some have in fact

claimed, no differences between the sexes, then this would need to be determined in systematic experiments controlling for at least some of these variables before any ﬁrm conclusions can be drawn. This highlights the validity of doing animal experiments on brain injury outcomes; most of the variables relevant to gender-speciﬁc issues can be controlled better in a laboratory setting, thereby reducing the variability typical of uncontrolled clinical assessments. Clearly, the issues surrounding sex differences in brain injury are complex and more research is needed to tease out whether sex differences can play a role in normal cognitive functions as well as in TBI prognoses. Given the growing literature showing that hormones like progesterone, estrogen or even testosterone can serve as treatments for TBI (see for example Wright et al. (2006)), it is likely that the study of sex differences will begin to receive more attention if for no other reason than to improve the quality and success of clinical trials seeking treatments for the victims of acquired brain injuries. Acknowledgment Thanks to Leslie McCann for editorial assistance. References von Arnim, C.A., Etrich, S.M., Timmler, M. and Riepe, M.W. (2002) Gender-dependent hypoxic tolerance mediated via gender-speciﬁc mechanisms. J. Neurosci. Res., 68: 84–88. Arnold, A.P. (1996) Genetically triggered sexual differentiation of brain and behavior. Horm. Behav., 30: 495–505. Arnold, A.P., Agate, R.J. and Carruth, L.C. (2004) Hormonal and non-hormonal mechanisms of sexual differentiation of the brain. In: Legato M.J. (Ed.), Principles of GenderSpeciﬁc Medicine. Academic Press, New York, pp. 84–95. Basso, A., Capitani, E. and Moraschini, S. (1982) Sex differences in recovery from aphasia. Cortex, 18: 469–475. Becker, J.B. (2002) Behavioral Endocrinology. MIT Press, Cambridge, MA. Berk, L.E. (1997) Child Development. Allyn & Bacon, Boston. Brinton, R.D. and Wang, J.M. (2006) Therapeutic potential of neurogenesis for prevention and recovery from Alzheimer’s disease: allopregnanolone as a proof of concept neurogenic agent. Curr. Alzheimer Res., 3: 185–190. Cahill, L. (2006) Why sex matters for neuroscience. Nat. Rev. Neurosci., 7: 477–484.

349 Cerghet, M., Skoff, R.P., Bessert, D., Zhang, Z., Mullins, C. and Ghandour, M.S. (2006) Proliferation and death of oligodendrocytes and myelin proteins are differentially regulated in male and female rodents. J. Neurosci., 26: 1439–1447. Clark, A.S. and Goldman-Rakic, P.S. (1989) Gonadal hormones inﬂuence the emergence of cortical function in nonhuman primates. Behav. Neurosci., 103: 1287–1295. Clark, A.S., Maclusky, N.J. and Goldman-Rakic, P.S. (1988) Androgen binding and metabolism in the cerebral cortex of the developing rhesus monkey. Endocrinology, 123: 932–940. Coimbra, R., Hoyt, D.B., Potenza, B.M., Fortlage, D., Hollingsworth-Fridlund, P. (2003) Does sexual dimorphism inﬂuence outcome of traumatic brain injury patients? The answer is no! J. Trauma, 54: 689–700 Collaer, M.L., Tory, H.O. and Valkenburgh, M.C. (2004) Do sex steroid hormones contribute to sexual differentiation of the human brain? In: Legato M.J. (Ed.), Principles of Gender-Speciﬁc Medicine, Vol. 1. Academic Press, New York, pp. 71–83. Conejo, N.M., Gonzalez-Pardo, H., Pedraza, C., Navarro, F.F., Vallejo, G. and Arias, J.L. (2003) Evidence for sexual difference in astrocytes of adult rat hippocampus. Neurosci. Lett., 339: 119–122. Dancause, N., Barbay, S., Frost, S.B., Plautz, E.J., Chen, D., Zoubina, E.V., Stowe, A.M. and Nudo, R.J. (2005) Extensive cortical rewiring after brain injury. J. Neurosci., 25: 10167–10179. DeGraba, T.J. and Pettigrew, L.C. (2000) Why do neuroprotective drugs work in animals but not humans? Neurol. Clin., 18: 475–493. De Nicola, A.F., Gonzalez, S.L., Labombarda, F., Deniselle, M.C., Garay, L., Guennoun, R. and Schumacher, M. (2006) Progesterone treatment of spinal cord injury: effects on receptors, neurotrophins, and myelination. J. Mol. Neurosci., 28: 3–15. Donders, J. and Hoffman, N.M. (2002) Gender differences in learning and memory after pediatric traumatic brain injury. Neuropsychology, 16: 491–499. Epting, L.K. and Overman, W.H. (1998) Sex-sensitive tasks in men and women: a search for performance ﬂuctuations across the menstrual cycle. Behav. Neurosci., 112: 1304–1317. Farace, E. and Alves, W.M. (2000) Do women fare worse: a metaanalysis of gender differences in traumatic brain injury outcome. J. Neurosurg., 93: 539–545. Farin, A., Deutsch, R., Biegon, A. and Marshall, L.F. (2003) Sex-related differences in patients with severe head injury: greater susceptibility to brain swelling in female patients 50 years of age and younger. J. Neurosurg., 98: 32–36. Forgie, M.L. and Kolb, B. (1998) Sex differences in the effects of frontal cortex injury: role of differential hormonal experience in early development. Behav. Neurosci., 112: 141–153. Galea, L.A., Wide, J.K., Paine, T.A., Holmes, M.M., Ormerod, B.K. and Floresco, S.B. (2001) High levels of estradiol disrupt conditioned place preference learning, stimulus response learning and reference memory but have limited effects on working memory. Behav. Brain Res., 126: 115–126.

Goldman, P.L. (1974) An alternative to developmental plasticity: heterology of CNS structures in infants and adults. In: Stein D.G., Rosen J.J. and Butters N. (Eds.), Plasticity and Recovery of Function in the Central Nervous System. Academic Press, New York, pp. 149–174. Goldstein, J.M., Seidman, L.J., Horton, N.J., Makris, N., Kennedy, D.N., Caviness Jr., V.S., Faraone, S.V. and Tsuang, M.T. (2001) Normal sexual dimorphism of the adult human brain assessed by in vivo magnetic resonance imaging. Cereb. Cortex, 11: 490–497. Groswasser, Z., Cohen, M. and Keren, O. (1998) Female TBI patients recover better than males. Brain Inj., 12: 805–808. Gur, R.G. and Gur, R.E. (2004) Gender differences in the functional organization of the brain. In: Legato M.J. (Ed.), Principles of Gender-Speciﬁc Medicine, Vol. 1. Academic Press, New York, pp. 63–70. Harukuni, I., Hurn, P.D. and Crain, B.J. (2001) Deleterious effect of beta-estradiol in a rat model of transient forebrain ischemia. Brain Res., 900: 137–142. Hausmann, M., Becker, C., Gather, U. and Gunturkun, O. (2002) Functional cerebral asymmetries during the menstrual cycle: a cross-sectional and longitudinal analysis. Neuropsychologia, 40: 808–816. He, Z., He, Y.J., Day, A.L. and Simpkins, J.W. (2002) Proestrus levels of estradiol during transient global cerebral ischemia improves the histological outcome of the hippocampal CA1 region: perfusion-dependent and -independent mechanisms. J. Neurol. Sci., 193: 79–87. Herzog, A.G. and Frye, C.A. (2003) Seizure exacerbation associated with inhibition of progesterone metabolism. Ann. Neurol., 53: 390–391. Herzog, A.G., Harden, C.L., Liporace, J., Pennell, P., Schomer, D.L., Sperling, M., Fowler, K., Nikolov, B., Shuman, S. and Newman, M. (2004) Frequency of catamenial seizure exacerbation in women with localization-related epilepsy. Ann. Neurol., 56: 431–434. Hindmarsh, G.J., O’Callaghan, M.J., Mohay, H.A. and Rogers, Y.M. (2000) Gender differences in cognitive abilities at 2 years in ELBW infants. Extremely low birth weight. Early Hum. Dev., 60: 115–122. Kasischke, K., Huber, R., Li, H., Timmler, M. and Riepe, M.W. (1999) Primary hypoxic tolerance and chemical preconditioning during estrus cycle in mice. Stroke, 30: 1256–1262. Kertesz, A. and McCabe, P. (1977) Recovery patterns and prognosis in aphasia. Brain, 100(Pt 1): 1–18. Kimura, D. (1992) Sex differences in the brain. Sci. Am., 267: 118–125. Kimura, D. (1999) Sex and Cognition. MIT University Press, Cambridge, MA. Kinsley, C.H., Madonia, L., Gifford, G.W., Tureski, K., Grifﬁn, G.R., Lowry, C., Williams, J., Collins, J., McLearie, H. and Lambert, K.G. (1999) Motherhood improves learning and memory. Nature, 402: 137–138. Kobayashi, T., Ahlenius, H., Thored, P., Kobayashi, R., Kokaia, Z. and Lindvall, O. (2006) Intracerebral infusion of glial cell line-derived neurotrophic factor promotes striatal neurogenesis after stroke in adult rats. Stroke, 37: 2361–2367.

350 Kolb, B. and Cioe, J. (1996) Sex-related differences in cortical function after medial frontal lesions in rats. Behav. Neurosci., 110: 1271–1281. Kraus, J.F., Peek-Asa, C. and McArthur, D. (2000) The independent effect of gender on outcomes following traumatic brain injury: a preliminary investigation. Neurosurg. Focus, 8: e5. Kuebler, J.F., Yokoyama, Y., Janar, D., Toth, B., Rue III, L.W., Bland, K.I., Wang, P. and Chaudry, I.H. (2003) Administration of progesterone after trauma and hemorrhagic shock prevents hepalocellular injury. Arch. Surg., 138: 727–734. Labombarda, F., Gonzalez, S., Gonzalez Deniselle, M.C., Garay, L., Guennoun, R., Schumacher, M. and De Nicola, A.F. (2006) Progesterone increases the expression of myelin basic protein and the number of cells showing NG2 immunostaining in the lesioned spinal cord. J. Neurotrauma, 23: 181–192. Lacreuse, A., Verreault, M. and Herndon, J.G. (2001) Fluctuations in spatial recognition memory across the menstrual cycle in female rhesus monkeys. Psychoneuroendocrinology, 26: 623–639. Magnaghi, V., Veiga, S., Ballabio, M., Gonzalez, L.C., GarciaSegura, L.M. and Melcangi, R.C. (2006) Sex-dimorphic effects of progesterone and its reduced metabolites on gene expression of myelin proteins by rat Schwann cells. J. Peripher. Nerv. Syst., 11: 111–118. Maren, S. (1995) Sexually dimorphic perforant path long-term potentiation (LTP) in urethane-anesthetized rats. Neurosci. Lett., 196: 177–180. Markham, J.A. and Juraska, J.M. (2002) Aging and sex inﬂuence the anatomy of the rat anterior cingulate cortex. Neurobiol. Aging, 23: 579–588. Markham, J.A., McKian, K.P., Stroup, T.S. and Juraska, J.M. (2005) Sexually dimorphic aging of dendritic morphology in CA1 of hippocampus. Hippocampus, 15: 97–103. McCarthy, J.B., Barker-Gibb, A.L., Alves, S.E. and Milner, T.A. (2002) TrkA immunoreactive astrocytes in dendritic ﬁelds of the hippocampal formation across estrous. Glia, 38: 36–44. McEwen, B.S. (1998) Multiple ovarian hormone effects on brain structure and function. J. Gend. Specif. Med., 1: 33–41. McIntosh, T.K. (1996) Neuropathological sequelae of traumatic brain injury: relationship to neurochemical and biomechanical mechanisms. Lab. Invest., 74: 315–342. di Michele, F., Lekieffre, D., Pasini, A., Bernardi, G., Benavides, J. and Romeo, E. (2000) Increased neurosteroids synthesis after brain and spinal cord injury in rats. Neurosci. Lett., 284: 65–68. Morrison, M.F. (2000) Hormones, Gender, and the Aging Brain: The Endocrine Basis of Geriatric Psychiatry. Cambridge University Press, New York, NY. Morrison, W.E., Arbelaez, J.J., Fackler, J.C., de Maio, A. and Paidas, C.N. (2004) Gender and age effects on outcome after pediatric traumatic brain injury. Pediatr. Crit. Care. Med., 5: 145–151. Mumenthaler, M.S., O’Hara, R., Taylor, J.L., Friedman, L. and Yesavage, J.A. (2001) Relationship between variations in

estradiol and progesterone levels across the menstrual cycle and human performance. Psychopharmacology (Berl.), 155: 198–203. Obler, L.K., Goral, M. and Albert, M.L. (1995) Variability in aphasia research: aphasia subject selection in group studies. Brain Lang., 48: 341–350. Rabinowicz, T., Petetot, J.M., Gartside, P.S., Sheyn, D., Sheyn, T. and De, C.M. (2002) Structure of the cerebral cortex in men and women. J. Neuropathol. Exp. Neurol., 61: 46–57. Ratcliffe, J.J., Greenspan, A., Goldstein, F.C., Wright, D.W., Stringer, A., Bell, D., Bushnik, T., Englander, J., Hammond, F., Novack, T. and Whyte, J. (2003) Gender and traumatic brain injury: do the sexes fare differently? Society for Experimental Medicine Abstract. Acad. Emerg. Med., 10: 491. Raz, S., Lauterbach, M.D., Hopkins, T.L., Glogowski, B.K., Porter, C.L., Riggs, W.W. and Sander, C.J. (1995) A female advantage in cognitive recovery from early cerebral insult. Dev. Psychol., 31: 958–966. Reddy, D.S. (2004) Role of neurosteroids in catamenial epilepsy. Epilepsy Res., 62: 99–118. Rider, V.A. and Abdou, N.I. (2004) Sexual dimorphism and the immune system. In: Legato M.J. (Ed.), Principles of GenderSpeciﬁc Medicine, Vol. 2. Academic Press, New York, pp. 1071–1081. Roof, R.L., Duvdevani, R. and Stein, D.G. (1993) Gender inﬂuences outcome of brain injury: progesterone plays a protective role. Brain Res., 607: 333–336. Sailor, K.A., Ming, G.L. and Song, H. (2006) Neurogenesis as a potential therapeutic strategy for neurodegenerative diseases. Expert Opin. Biol. Ther., 6: 879–890. Schumacher, M., Guennoun, R., Robert, F., Carelli, C., Gago, N., Ghoumari, A., Gonzalez Deniselle, M.C., Gonzalez, S.L., Ibanez, C., Labombarda, F., Coirini, H., Baulieu, E.E. and De Nicola, A.F. (2004) Local synthesis and dual actions of progesterone in the nervous system: neuroprotection and myelination. Growth Horm. IGF Res., 14(Suppl A): S18–S33. Shankaran, M., King, C., Lee, J., Busch, R., Wolff, M. and Hellerstein, M.K. (2006) Discovery of novel hippocampal neurogenic agents by using an in vivo stable isotope labeling technique. J. Pharmacol. Exp. Ther., 319: 1172–1181. Shaywitz, B.A., Shaywitz, S.E., Pugh, K.R., Constable, R.T., Skudlarski, P., Fulbright, R.K., Bronen, R.A., Fletcher, J.M., Shankweiler, D.P., Katz, L., et al. (1995) Sex differences in the functional organization of the brain for language [see comments]. Nature, 373: 607–609. Shivraj Sohur, U., Emsley, J.G., Mitchell, B.D. and Macklis, J.D. (2006) Adult neurogenesis and cellular brain repair with neural progenitors, precursors and stem cells. Philos. Trans. R. Soc. Lond. B. Biol. Sci., 361: 1477–1497. Smith, M.J., Adams, L.F., Schmidt, P.J., Rubinow, D.R. and Wassermann, E.M. (2002) Effects of ovarian hormones on human cortical excitability. Ann. Neurol., 51: 599–603. Stein, D.G. (2004) Brain trauma, sex hormones, neuronal survival and recovery of function. In: Legato M. (Ed.), Principles of Gender-Speciﬁc Medicine, Vol. 1. Academic Press, New York, pp. 104–115.

351 Stein, D. G. (2005) The case for progesterone. In: Singh M. and Simpkins J.W. (Eds.), The Future of Estrogen and Hormone Theraphy in Postmenopausal Women: What Basic Science and Clinical Studies Teach Us, Vol. 1052. Annals of the New York Academy of Science, pp. 152–169. Taupin, P. (2006) Adult neurogenesis in mammals. Curr. Opin. Mol. Ther., 8: 345–351. Truettner, J., Busto, R., Zhao, W., Ginsberg, M.D. and PerezPinzon, M.A. (2002) Effect of ischemic preconditioning on the expression of putative neuroprotective genes in the rat brain. Brain Res. Mol. Brain Res., 103: 106–115. Wichmann, M.W., Zellweger, R., Demaso, C.M., Ayala, A. and Chaudry, I.H. (1996) Enhanced immune responses in females, as opposed to decreased responses in males following haemorrhagic shock and resuscitation. Cytokine, 8: 853–863.

Wright, D.W., Kellermann, A.L., Hertzberg, V.S., Clark, P.L., Frankel, M., Goldstein, F.C., Salomone, J.P., Dent, L.L., Harris, O.A., Ander, D.S., Lowery, D.W., Patel, M.M., Denson, D.D., Gordon, A.B., Wald, M.M., Gupta, S., Hoffman, S.W., Stein, D.G. (2006) Progesterone for acute traumatic brain injury. Ann. Emerg. Med. Zellweger, R., Wichmann, M.W., Ayala, A., Stein, S., Demaso, C.M. and Chaudry, I.H. (1997) Females in proestrus state maintain splenic immune functions and tolerate sepsis better than males. Crit. Care Med., 25: 106–110. Zhang, L., Li, P.P., Feng, X., Barker, J.L., Smith, S.V. and Rubinow, D.R. (2003) Sex-related differences in neuronal cell survival and signaling in rats. Neurosci. Lett., 337: 65–68. Zwain, I.H. and Yen, S.S. (1999) Neurosteroidogenesis in astrocytes, oligodendrocytes, and neurons of cerebral cortex of rat brain. Endocrinology, 140: 3843–3852.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 25

Heat acclimation: a unique model of physiologically mediated global preconditioning against traumatic brain injury Na’ama A. Shein1,2, Michal Horowitz2 and Esther Shohami1, 1

Department of Pharmacology, Hebrew University, Jerusalem, Israel 2 Department of Physiology, Hebrew University, Jerusalem, Israel

Abstract: Sub-lethal exposure to practically any harmful stimulus has been shown to induce consequent protection against more severe stress. This preconditioning (PC) effect may be achieved by exposure to different stressors, indicating that the induction of tolerance involves activation of common protective pathways. Chronic exposure to moderate heat (heat acclimation, HA) is a unique PC model, since this global physiological adaptation, as opposed to discrete organ PC, has been shown to induce cross-tolerance against other stressors, including closed head injury (CHI). HA animals show accelerated functional recovery after injury which is accompanied by reduced secondary brain damage. However, the precise mechanisms underlying this phenomenon have not been thoroughly studied until recently. Here we will address the concept of PC, highlighting the unique properties of HA as a model which can be used for the study of endogenous protective pathways triggered by PC procedures. Several molecular mechanisms which are suggested to mediate HA-induced neuroprotection will also be discussed, bringing to light their potential contribution to the development of traumatic brain injury treatment strategies utilizing therapeutic augmentation of endogenous defense mechanisms. Keywords: closed head injury; erythropoietin; heat acclimation; hypoxia inducible factor-1; inﬂammation; neuroprotection; preconditioning which enhance intrinsic defensive ability. Conceivably, the coping ability of pre-exposed organisms will consequently be augmented in the setting of recurring challenges (Fig. 1). Janoff (1964), ﬁrst introduced the terms ‘preconditioning’ (PC) and ‘tolerance’ to describe the phenomenon, setting the stage for numerous later studies which described many different PC models and demonstrated the beneﬁcial effects of this protocol in providing consequent protection. Traditionally, PC protocols were aimed at providing enhanced defensive ability to a speciﬁc organ, with cardiac PC being the most extensively studied model. In the mid 1980s, it was initially

The phenomenon of preconditioning Throughout history, several scholars have approached the notion that harmful substances or stressors may also engulf beneﬁcial effects. These apparently conﬂicting observations were linked by early-time savants such as Hippocrates and Paracelsus to the extent of exposure or the given dose, suggesting that sub-lethal subjection to a noxious insult may bring about the activation of endogenous protective response mechanisms, Corresponding author. Tel.: +972-2-675-7513; Fax: +972-2-

675-8741; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61025-X

353

Preconditioned

Non-preconditioned

354

Basal defensive ability

Overall outcome

Endogenous protection activation Deleterious processes

time Attenuated stress / harmful stimuli exposure

Time interval Hours-weeks

Severe stress/harmful stimuli induction

Outcome measurement

Fig. 1. General model of preconditioning. Pre-exposure to attenuated stress/harmful stimuli enhances endogenous protective mechanisms, leading to improved overall outcome following subsequent more severe insult. Note that the beneﬁcial effect brought about by preconditioning results from augmented defensive ability rather than a direct effect on the extent of deleterious processes.

reported that pre-exposure to an attenuated ischemic insult affords subsequent myocardial protection (Murry et al., 1986). The beneﬁcial effects of myocardial ischemic PC were later shown to include the reduction of infarct size, a decline in the occurrence of cardiac arrhythmias and accelerated functional recovery following ischemia (Shiki and Hearse, 1987; Cleveland et al., 1996). In vitro experiments also indicated the protective potential of myocardial PC (Ikonomidis et al., 1994). Since then, ischemic PC-induced cardio-protection has been reproduced by numerous groups in various species including humans (Vohra and Galinanes, 2006). In addition, myocardial PC has shown to be clinically relevant in the setting of myocardial infarction and in patients receiving repeated cycles of aortic clamping and reperfusion prior to cardiopulmonary bypass (CPB) (Yellon and Dana, 2000). Subsequently, the beneﬁcial effects of ischemic PC on lung preservation in patients undergoing open heart surgery with CBP were also reported (see Luh and Yang, 2006 for review). Similarly, the neuroprotective potential of ischemic/hypoxic PC against ischemic injury within

the central nervous system has been the subject of extensive research. Ischemic PC, as well as PC by exposure to hyperbaric oxygen provided protection in various models of spinal cord injury (Toumpoulis and Anagnostopolus, 2003; Nie et al., 2006). Numerous studies have revealed the profound effect of PC by brief ischemic exposure on the outcome of brain ischemic injury. The beneﬁcial consequences of such procedures have been demonstrated in different in vitro (Reshef et al., 1996; Liu et al., 2000) as well as in vivo models (Sharp et al., 2004; Glantz et al., 2005, and see Dirnagl et al., 2003; Blanco et al., 2006 for review). The substantiation of PC-afforded neuroprotection in animal models consequently instigated a wide-scale research effort aimed at exploring methods of employing ischemic PC in stroke therapy (Schaller, 2005). Interestingly, over the years, a body of evidence accumulated, indicating that the initial deﬁnition of PC as the induction of tolerance toward a speciﬁc stressor by attenuated pre-exposure to the same type of stress might have been too narrow. It has been shown that the effects of PC may be mimicked by exogenous treatments. PC-like effects have been achieved in various models by pharmacological interventions utilizing low doses

355

of otherwise harmful substances. Boeck et al. (2004) have shown that exposure to low doses of NMDA induces protection against the hippocampal neurotoxicity of quinolic acid. Analogous observations were reported while using PC with thrombin (Hua et al., 2003; Cannon et al., 2005). Importantly, Dahl and Balfour (1964) demonstrated that in rats, exposure to short periods of anoxia induces a ‘whole organism’ protective effect, indicating that PC can be applied toward conferring protection to organisms and not only to discrete organs. Taken together, these ﬁndings suggest that PC is a global and well-preserved adaptation mechanism which is aimed at enabling dealing with stress. The fact that this endogenous mechanism can be prompted, not only by exposure to stress but also via administration of exogenous agents, is of great importance since it suggests that PC-like effects may be induced at will in a relevant clinical setting. In addition, since global exposure to external ambient anoxia can cause PC of the entire organism, it is conceivable that activation of the protective mechanisms which underlie PC-induced tolerance could be achieved by exposure to other types of environmental stress.

Long-term heat acclimation: physiologically mediated global preconditioning Heat acclimation (HA) is a conserved adaptive response to exposure to moderately high ambient temperature. Conceptually, the process may be delineated as a transition from an early, transient, ‘‘inefﬁcient’’ state to an ‘‘efﬁcient’’ acclimated homeostatic state (Horowitz, 2003, 2007). Long-term HA is achieved by a 3–4 week exposure to mild environmental heat. The acclimation process is biphasic, consisting of two phases which differ in terms of the controlling mechanisms for heat dissipation. While increased excitability of the autonomic nervous system compensates for impaired cellular performance during short-term acclimation, the features of HA include enhanced cellular function accompanied by decreased excitability (Horowitz and Meiri, 1985; Horowitz, 1994 and for review, see Horowitz, 2002, 2007).

Individual thermoregulatory effectors display a similar pattern of biphasic response during HA. Salivation, which is utilized by several rodent species as the major pathway of evaporative cooling, clearly demonstrates this phenomenon. While saliva production is inadequate for sufﬁcient heat dissipation at the onset of acclimation due to impaired responsiveness of muscarinic receptors, once the new acclimated homeostasis has been achieved, 3 weeks later, glandular efﬁciency is regained due to upregulation and a return of these receptors to pre-acclimation afﬁnity (Kloog et al., 1985; Horowitz, 2002). Similarly, enhanced autonomic sympathetic activity is utilized in order to compensate for the desensitization of cardiac adrenergic receptors during short-term HA, whereas the HA heart maintains the ability to produce adequate contractile force, to enhance work efﬁciency and to enhance pressure in the face of reduced ATP utilization (Levi et al., 1993; Horowitz, 1998, 2002). At the organism level, the HA phenotype has been characterized by reduced metabolic and heart rates, decreased core body temperature, lower temperature thresholds for the activation of heat dissipation mechanisms and an elevated temperature threshold for the development of thermal injury. Evidence now indicates that this acclimated phenotype results from both post-translational modiﬁcations as well as genomic responses. Maloyan et al. (1999) showed that heat acclimated rats display about three times more inducible heat shock protein 72 kDa than do non-acclimated counterparts, thus showing better coping with heat stress without the need for de novo HSP synthesis (which characterizes the heat shock response of non-acclimated animals). Alterations in genomic responses during the course of HA have been shown to include changes in the expression pattern of genes which are involved in anti-apoptosis, antioxidation and heat shock response (Horowitz et al., 2004). Altogether, these responses contribute to an overall expansion of the dynamic thermoregulatory range. Consequently, HA results in enhanced ability of coping with exposure to environmental heat. Hence, in accordance with the terminology introduced by Janoff (1964) HA may be viewed as a long-term physiologically mediated heat-PC.

356

An inseparable outcome of HA is that adjusting to one stressor can, in addition to evolving primary adaptations, add to the amount of adjustment to additional stressors. Such cross-reinforcement raises the possibility of inducing adaptation to a stressor without previous exposure to that particular one. Indeed, the ability of HA to induce ‘‘crosstolerance’’ against a variety of stressors has been reported. HA has been shown to convey improved cardiac mechanical and metabolic performance and reduced injury upon ischemic-reperfusion insult to the heart. Levi et al. (1993) showed that following total ischemia, acclimated hearts displayed enhanced preservation of the cardiac ATP pool and a delayed decline in intracellular pH. HA has also been shown to induce cross-tolerance toward exposure to hyperbaric oxygen. Arieli et al. (2003) reported a delay in the onset of central nervous system oxygen toxicity following HA, demonstrating that the latency to the appearance of evident toxicity was twice as long in heat acclimated rats. Taken together, these ﬁndings indicate a wide scale protective effect which is induced by HA. This observation may be of particular importance in the setting of traumatic brain injury. A neuroprotective effect elicited by HA will potentially have a fundamental effect on functional recovery upon subsequent injury. Additionally, it could provide valuable insight into common protective pathways which are shared by HA and post-injury recovery mechanisms.

HA-induced neuroprotection in closed head injury The effects of HA on the outcome of closed head injury (CHI) have been examined using a wellestablished CHI model (described by Shapira et al., 1988 and Chen et al., 1996). The model reproduces the post-traumatic sequence of events observed in humans and consistently recreates several pathophysiological features such as edema formation, blood-brain barrier disruption and functional dysfunction. Shohami et al. (1994b) demonstrated that when CHI is induced in HA rats, they display faster and

better functional recovery as well as reduced secondary tissue damage. The motor ability of the rats was evaluated according to a neurological severity score (NSS) which is based on the presence of some reﬂexes and the ability to perform motor and behavioral tasks such as beam walking, beam balance and spontaneous locomotion. HA rats showed signiﬁcantly better motor function 48 h following injury as compared with normothermic controls. Additionally, brain edema and blood-brain barrier disruption were reduced in the acclimated group. These ﬁndings were subsequently reinforced in mice (Shein et al., 2005a), demonstrating signiﬁcantly better motor function 24 h post-injury accompanied with a reduction in brain water accumulation in HA mice. These mice also perform better in a memory test when examined by the object recognition test (Ennaceur and Delacour, 1988). When tested 3 days following CHI, HA mice spent most of their exploration time at the new object while normothermic counterparts failed to prefer the ‘novel’ object over the familiar one. Interestingly, it has also been shown that neuroprotection may be achieved in vitro by exposing cell cultures to serum derived from HA animals. Beit Yannai et al. (1998) demonstrated that when exposing cultures of PC12 cells (a sympathetic pheochromocytoma cell line widely used as a neuronal cell model) to 1% serum from HA rats, a signiﬁcant reduction in cell death occurs as compared with cultures exposed to serum taken from normothermic ones. Additionally, the HA serum treated cultures showed increased neuroprotection upon exposure to a free radical generator causing oxidative stress. Taken together, current data establishes that HA elicits a continuum of processes resulting in enhanced neuroprotection. The fact that the beneﬁcial effect can be demonstrated in different species as well as in varying experimental models is in agreement with the observations pointing to the widescale nature of PC-induced tolerance. Although HA-induced neuroprotection was hypothesized to be facilitated via common protective pathways shared by different PC paradigms (Horowitz, 2004), the underlying mechanisms remained poorly deﬁned for some time. However, recent work has

357

provided new insight into several molecular processes which may contribute to the overall functional improvement. Mechanisms of HA-induced neuroprotection The research of HA-induced neuroprotection is generally aimed at identifying molecular effectors and signaling cascades which are involved in the development of the protected phenotype. In the following sections we describe our own studies which examined the effects of HA on pathways which are known to either convey a well-established protective effect or foster deleterious consequences in CHI, in an effort to examine whether these pathways may also play a role in the beneﬁcial outcome induced by HA. These include low molecular weight antioxidants and anti-inﬂammatory capacity, hypoxia inducible factor-1, erythropoietin receptor expression and signaling as well as acute post-CHI inﬂammation and the expression of neurotrophic factors. Effect of HA on the ability to cope with oxidative stress: endogenous and post-CHI low-molecular weight antioxidant profile Reactive oxygen species (ROS) have been extensively studied and proposed as candidates for the elicitation of pathological responses in the pathogenesis of ischemia and trauma (e.g., Chan, 2001; Kontos, 2001). A number of therapeutic approaches, based on intervention by scavenging ROS, have been attempted both in experimental models and in the clinical setting (for review, see Vink and Van Den Heuvel, 2004). Generally, several lines of defense have been developed by living cells in order to cope with oxidative stress. These include preventive and repair mechanisms as well as an antioxidant defense system, consisting of antioxidative enzymes and low molecular weight antioxidants (LMWA). The main direct-acting LMWA found within the brain include the tripeptide glutathione (glu-cys-gly), tocopherols (vitamin E), ascorbic acid (vitamin C), histidine-related compounds, melatonin, uric acid and lipoic acid. Evaluation of individual LMWA

is difﬁcult given the large number of scavengers and the fact that they interact with each other, such that increase of one may lead to a decrease of the other. However, this limitation can be overcome by the cyclic voltammetry technique. Cyclic voltammetry was traditionally used in chemistry to study electron transfer between molecules and was shown initially by Kissinger et al. (1973) to be applicable for evaluating antioxidant levels in vivo. A new method was then developed, enabling the measurement of overall antioxidant activity based on the determination of the total reducing power of the biological tissue or ﬂuid (Kohen et al., 1999; Kohen and Nyska, 2002). In our model of CHI, we have previously shown the activation of the arachidonic acid metabolic cascade which produces free radicals (Shohami et al., 1987, 1989) as well as enhanced cytokine production (Shohami et al., 1994a), a typical inﬂammatory response involving ROS. In addition, we have demonstrated that a marked decrease in LMWA concentrations occurs within minutes following injury, indicating their immediate utilization in an effort to overcome acute oxidative stress. The detrimental role of ROS in early-CHI pathology was also supported by the observation that ROS-neutralizing compounds are protective when given soon after CHI (Beit-Yannai et al., 1996). We have examined the LMWA in the brains of HA rats, in an effort to correlate the observed protection with the ability of the brain to neutralize ROS. A cyclic voltammetry study performed on the water-soluble fraction of LMWA showed that the chemical nature of LMWA is not altered by HA. Interestingly, the concentrations of the reducing equivalents were signiﬁcantly lower in the HA group, i.e. the basal LMWA levels were found to be lower in HA rats (Beit Yannai et al., 1997). However, while the relative changes in anodic currents during the post-CHI period revealed a decrease from basal (sham) levels in the normothermic animals, this did not occur in the HA group which sustained levels higher than those measured in sham controls, for up to a week following trauma. In light of these results we suggested that similarly to observations in the ischemic heat

358

acclimated heart (Horowitz et al., 2004), in the setting of CHI long-term exposure to high ambient temperature provides a more efﬁcient mechanism of dealing with ROS displayed by better ability of adjusting to the subsequent acute stress.

Involvement of hypoxia inducible factor-1 and erythropoietin signaling Oxygen deprivation is known to evoke a sequele of adaptive responses. These responses are aimed at compensating for decreased aerobic ATP production and are regulated by the transcriptional activator hypoxia inducible factor-1 (HIF-1). HIF-1 regulates the expression of over 70 known target genes, leading to increased tissue oxygen delivery and ATP production (Semenza, 2004). The functional factor is a heterodimer, consisting of a constitutively expressed b subunit and an inducible a subunit, which undergo dimerization by means of basic helix-loop-helix PAS domains. Interestingly, the accumulation of HIF-1a and the transcriptional activation of HIF-1 have recently been shown to also be triggered by non-oxygen-dependent pathways including ROS, cytokines and several growth factors and hormones (Chandel et al., 2000; Haddad, 2002). The fact that many HIF-1 regulated genes are related to oxygen and energy homeostasis, taken together with its importance in environmental adaptive processes led to the hypothesis that this factor may also be involved in the metabolic responses during HA. Furthermore, preliminary data from Maloyan et al. (2001) and Bromberg and Horowitz (2004) indicated that the levels of HIF-1a are increased following HA. However, the role of HIF-1 in HA remained unclear and it was not known until 2 years later whether the increase in the levels of this factor are of functional importance or only an epiphenomenon. The question of the involvement of HIF-1 in HA was assessed by Treinin et al. (2003). Their study elegantly demonstrated that HIF-1 is essential to the development of HA, using a ‘hif-1 loss of function’ mutant strain of the C. elegans nematode. As opposed to wild type nematodes, which demonstrated an acclimated phenotype following

an appropriate acclimation protocol, the loss of function strain was unable to acclimate to heat. The contribution of HIF-1 targeted pathways to HA-induced cardiac cross-tolerance has also recently been established. Maloyan et al. (2005) reported an increase in HIF-1a protein levels and HIF-1 DNA binding activity as well as enhanced target gene mRNA expression in the hearts of HA rats. HIF-1 regulates the expression of erythropoietin (Epo). Epo, traditionally recognized as the main erythropoietic cytokine, has been shown to be extensively expressed within the brain (Buemi et al., 2002). The cytokine as well as its speciﬁc receptor (EpoR) are expressed by neuronal, glial and brain capillary endothelial cells and are upregulated by ischemic as well as metabolic stress (Bernaudin et al., 1999, 2000). Importantly, Epo has been shown to engulf neuroprotection in a large variety of in vitro and in vivo models of brain injury as well as in human stroke patients (Morishita et al., 1997; Agnello et al., 2002; Catania et al., 2002; Ehrenreich et al., 2002; Grasso et al., 2002). We have previously established the neuroprotective effect of Epo in our model of CHI (Yatsiv et al., 2005). When treated with recombinant human Epo (rhEpo) 1 and 24 h following CHI, mice displayed lower motor deﬁcits and accelerated restoration of cognitive function. In addition, a reduction in the number of apoptotic neurons as well as caspase-3 expression was found in the rhEpo-treated group. This was also accompanied by better preservation of axons at the trauma area and reduced activation of glial cells in the injured hemisphere. In light of this, we hypothesized that HIF-1 and in turn, Epo signaling, may also play a part in the beneﬁts brought about by HA. We proceeded to initially examine the basal and post-injury levels of HIF-1a in HA mice and normothermic counterparts. The levels of HIF-1a were found to be much higher following HA and this upregulation was sustained 4 h following injury (Shein et al., 2005a). Subsequently, the expression of Epo and EpoR were evaluated. An increase in both basal and post-injury receptor levels was found, but surprisingly this was not accompanied by upregulation of Epo itself. This was unexpected, given the large

359

number of studies which highlighted the association of HIF-1 upregulation and increased expression of Epo. However, several other reports have demonstrated that the behavior of HIF-1 target genes varies among different genes and is also greatly dependent on the experimental model used (Jones and Bergeron, 2001). In addition, HIF-1 activation which is not followed by increased Epo expression has previously been described in the presence of pro-inﬂammatory cytokines (Hellwig-Burgel et al., 1999). The levels of EpoR are known to be upregulated by ischemia and hypoxia (Siren et al., 2001; Genc et al., 2004). The established involvement of ischemia in CHI pathophysiology may therefore explain the post-injury upregulation of this receptor. Since HA has been shown to mediate enhanced heat tolerance as well as cross-tolerance via increased gene responsiveness (Horowitz et al., 2004), we propose that it is likely that the marked increase in EpoR in HA mice following injury is due to an additive effect of both injury-induced ischemia and increased responsiveness resulting from HA. Since the levels of EpoR were consistently higher following HA only, as well as in HA mice subjected to CHI, we calculated the overall receptor/ligand Epo ratio as a measure of overall Epo signaling ability. This ratio was signiﬁcantly higher in the HA group and led us to further examine downstream intracellular signaling which is mediated by this receptor. We chose to focus on the extent of phosphorylation of Akt as an indicator of intracellular antiapoptotic Epo/EpoR signaling. Following EpoR activation, Akt is phophorylated via the phosphatidylinositol-3 kinase pathway. When phosphorylated, Akt subsequently acts as a kinase and phosphorylates pro-apoptotic mediators including Bad, caspase-9, the forkhead transcription factor and glycogen synthase kinase 3b. Each of the cellular systems targeted by Akt is inactivated by phosphorylation, resulting in a blockade of apoptotic cell death. Additionally, Akt phosphorylation has been associated with neuroprotection and has been suggested as one of the molecular pathways mediating Epo-induced neuroprotection (Digicaylioglu et al., 2001; Chong et al., 2002; Kilic et al., 2006).

When we examined the levels of total and phosphorylated Akt in HA mice and compared them with the levels found in non-acclimated controls, we observed an increase in post-injury phosphorylated Akt in the HA group (Shein et al., 2005b). We may therefore conclude that HA has both constitutive and dynamic effects; namely establishing higher basal levels and rapidly elevated post-injury levels of both HIF-1a and EpoR and that this in turn leads to increased post-CHI enhancement of Akt phosphorylation. Effect of HA anti-inflammatory capacity and acute post-CHI inflammation The process of acute inﬂammation, initiated following CHI has been shown to involve the upregulation of pro-inﬂammatory cytokines (Shohami et al., 1994a). These cytokines, namely tumor necrosis factor alpha (TNF-a) and interleukin-1b (IL-1b), when present during the initial time period following CHI, have been implicated of involvement in blood-brain barrier disruption and associated with subsequent edema formation (Shohami et al., 1997; Stahel et al., 2000). Knoblach and Faden (1998) have previously demonstrated that the post-injury expression of these deleterious effectors may be attenuated by pre-injury treatment with the antiinﬂammatory cytokine interleukin-10 (IL-10). We therefore conceived that elevated pre-injury levels of anti-inﬂammatory mediators may lead to decreased post-CHI acute inﬂammation in HA animals. Our current data indicates that HA mice do indeed display both an increase in basal, pre-injury anti-inﬂammatory capacity as well as a subsequent decrease in post-injury expression of pro-inﬂammatory cytokine mRNA (Shein et al., 2006). Immunohistochemical ﬁndings now indicate that HA also enhances the presence of brain-derived neurotrophic factor-positive microglia (Shein et al., 2006). Summary Current data suggests that the neuroprotective effect of HA is likely to be mediated by a network of pathways which work in concert to yield overall

360

functional improvement. In light of the ﬁndings to date, it can be suggested that HA leads to both the potentiation of endogenous protective pathways as well as to the attenuation of detrimental postinjury processes. Although some changes are constitutive and observed following the acclimation period itself while others are dynamic and only apparent after injury, they all eventually confer modiﬁcations of the post-injury state (Fig. 2). The fact that the effects of HA can be demonstrated individually for different molecular mediators substantiates the importance of this physiological Higher levels in HA

adaptive mechanism as a model for enabling better deﬁnition of the role of discrete endogenous neuroprotective pathways in the development of tolerance. HA is a unique model in terms of providing insight into physiologically induced neuroprotection. Unraveling the physiological mechanisms of tolerance development could have profound implications for treatment in the pathophysiological setting of brain trauma, since future intervention strategies may be aimed at enhancing endogenous protective ability in a physiologicalmimetic manner.

Lower levels in HA

Abbreviations HIF-1α Pre-CHI

EpoR IL-10/IL-4

LMWA TNF-α HIF-1α

CHI CPB Epo EpoR HA HIF-1 IL-1b IL-10 LMWA

IL-1β

Post-CHI EpoR p-Akt

PC ROS TNFa

closed head injury cardiopulmonary bypass erythropoietin erythropoietin receptor long-term heat acclimation hypoxia inducible factor-1 interleukin-1beta interleukin-10 low molecular weight antioxidants preconditioning reactive oxygen species tumor necrosis factor alpha.

Note, factors which are higher in HA (either pre or post CHI) are associated with neuroprotection. Factors which are lower in HA after CHI are associated with damage Fig. 2. A schematic overview summarizing current experimental data for mechanisms suggested as being involved in HAinduced neuroprotection in closed head injury (CHI). Following long-term heat acclimation (HA), the pre-injury levels of hypoxia inducible factor-1alpha (HIF-1a), erythropoietin receptor (EpoR) and anti-inﬂammatory cytokines are increased. Higher levels of HIF-1a and EpoR are sustained in HA animals following injury as well and are complemented by an increase in post-CHI low molecular weight antioxidants (LMWA) and phosphorylated Akt (p-Akt) as well as a reduction in early postinjury levels of pro-inﬂammatory cytokines. Abbreviations: CHI, closed head injury; HA, long-term heat acclimation; LMWA, low molecular weight antioxidants; IL-10, interleukin10; IL-4, interleukin 4; TNF-a, tumor necrosis factor alpha; IL-1b, interleukin-1beta; HIF-1a, hypoxia inducible factor-1alpha; EpoR erythropoietin receptor; p-Akt, phosphorylated Akt.

References Agnello, D., Bigini, P., Villa, P., Mennini, T., Cerami, A., Brines, M.L. and Ghezzi, P. (2002) Erythropoietin exerts anti-inﬂammatory effect on the CNS in a model of experimental autoimmune encephalomyelitis. Brain Res., 952: 128–134. Arieli, Y., Eynan, M., Gancz, H., Arieli, R. and Kashi, Y. (2003) Heat acclimation prolongs the time to central nervous system oxygen toxicity in the rat: possible involvement of HSP72. Brain Res., 962: 15–20. Beit-Yannai, E., Kohen, R., Horowitz, M., Trembovler, V. and Shohami, E. (1997) Changes in biological reducing activity in rat brain following closed head injury: a cyclic voltammetry study in normal and acclimated rats. J. Cereb. Blood Flow Metab., 17: 273–279.

361 Beit-Yannai, E., Trembovler, V., Horowitz, M., Lazarovici, P., Kohen, R. and Shohami, E. (1998) Neuroprotection against oxidative stress by serum from heat acclimated rats. Neurosci. Lett., 254: 89–92. Beit-Yannai, E., Zhang, R., Trembovler, V., Samuni, A. and Shohami, E. (1996) Cerbroprotective effect of stable nitroxide radicals in closed head injury in the rat. Brain Res., 717: 22–28. Bernadin, M., Bellail, A., Marti, H.H., Yvon, A., Vivien, D., Duchatelle, I., Mackenzie, E.T. and Petit, E. (2000) Neurons and astrocytes express Epo mRNA: oxygen-sensing mechanisms that involve the redox-state of the brain. Glia, 30: 271–278. Bernaudin, M., Marti, H.H., Roussel, S., Divoux, D., Nouvelot, A., MacKenzie, E.T. and Petit, E. (1999) A potential role for erythropoietin in focal permanent cerebral ischemia in mice. J. Cereb. Blood Flow Metab., 19: 643–651. Blanco, M., Lizasoain, I., Sobrino, T., Vivancos, J. and Castillo, J. (2006) Ischemic preconditioning: a novel target for neuroprotective therapy. Cerebrovasc. Dis., 21(Suppl 2): 38–47. Boeck, C.R., Ganzella, M., Lottermann, A. and Vendite, D. (2004) NMDA preconditioning protects against seizures and hippocampal neurotoxicity induced by quinolinic acid in mice. Epilepsia, 45: 745–750. Bromberg, Z. and Horowitz, M. (2004) Mus musculus: a model for studying the dynamics of heat acclimation and cross tolerance between heat acclimation and oxygen deprivation. Proceedings of the 1st Integrated meeting on Physiology and Pharmacology of Thermoregulation (PPTR 2004), October 2004, Rhodes, Greece, p. 198. Buemi, M., Cavallaro, E., Floccari, F., Sturiale, A., Aloisi, C., Trimarchi, M., Grasso, G., Corica, F. and Frisina, N. (2002) Erythropoietin and the brain: from neurodevelopment to neuroprotection. Clin. Sci., 103: 275–282. Cannon, J.R., Keep, R.F., Hua, Y., Richardson, R.J., Schallert, T. and Xi, G. (2005) Thrombin preconditioning provides protection in a 6-hydroxydopamine Parkinson’s disease model. Neurosci. Lett., 373: 189–194. Catania, M.A., Marciano, M.C., Parisi, A., Sturiale, A., Buemi, M., Grasso, G., Squadrito, F., Caputi, A.P. and Calapai, G. (2002) Erythropoietin prevents cognition impairment induced by transient brain ischemia in gerbils. Eur. J. Pharmacol., 437: 147–150. Chan, P.H. (2001) Reactive oxygen radicals in signaling and damage in the ischemic brain. J. Cereb. Blood Flow Metab., 21: 2–14. Chandel, N.S., McClintock, D.S., Feliciano, C.E., Wood, T.M., Melendez, J.A., Rodriguez, A.M. and Schumacker, P.T. (2000) Reactive oxygen species generated at mitochondrial complex III stabilize hypoxia inducible factor-1a during hypoxia: a mechanism of oxygen sensing. J. Biol. Chem., 275: 25130–25138. Chen, Y., Constantini, S., Trembovler, V., Weinstock, M. and Shohami, E. (1996) An experimental model of closed head injury in mice: pathophysiology, histopathology, and cognitive deﬁcits. J. Neurotrauma, 13: 557–568.

Chong, Z.Z., Kang, J.Q. and Maiese, K. (2002) Hematopoietic factor erythropoietin fosters neuroprotection through novel signal transduction cascades. J. Cereb. Blood Flow Metab., 22: 503–514. Cleveland Jr., J.C., Wollmering, M.M., Meidrum, D.R., Rowland, R.T., Rehring, T.F., Sheridan, B.C., Harken, A.H. and Banerjee, A. (1996) Ischemic preconditioning in human and rat ventricle. Am. J. Physiol., 271: H178–H179. Dahl, N.A. and Balfour, W.M. (1964) Prolonged anoxic survival due to anoxia pre-exposure: brain ATP, lactate, and pyruvate. Am. J. Physiol., 207: 452–456. Digicaylioglu, M., Bichet, S. and Marti, H.H. (2001) Erythropoietin-mediated neuroprotection involves cross-talk between Jak-2 and NF-kappaB signaling cascades. Nature, 412: 641–647. Dirnagl, U., Simon, R.P. and Hallenbeck, J.M. (2003) Ischemic tolerance and endogenous neuroprotection. Trends Neurosci., 26: 248–254. Ehrenreich, H., Hasselblatt, M., Dembowski, C., Cepek, L., Lewczuk, P., Stiefel, M., Rustenbeck, H.H., Breiter, N., Jacob, S., Knerlich, F., Bohn, M., Poser, W., Ruther, E., Kochen, M., Gefeller, O., Gleiter, C., Wessel, T.C., De Ryck, M., Itri, L., Prange, H., Cerami, A., Brines, M. and Siren, A.L. (2002) Erythropoietin therapy for acute stroke is both safe and beneﬁcial. Mol. Med., 8: 495–505. Ennaceur, A. and Delacour, J. (1988) A new one-trial test for neurobiological studies of memory in rats. 1: behavioral data. Behav. Brain Res., 31: 47–59. Genc, S., Koroglu, F. and Genc, K. (2004) Erythropoietin and the nervous system. Brain Res., 1000: 19–31. Glantz, L., Avramovich, A., Trembovler, V., Gurvitz, V., Kohen, R., Eidelman, L.A. and Shohami, E. (2005) Ischemic preconditioning increases antioxidants in the brain and peripheral organs after cerebral ischemia. Exp. Neurol., 192: 117–124. Grasso, G., Buemi, M., Alafaci, C., Sfacteria, A., Passalacqua, M., Sturiale, A., Calapai, G., De Vico, G., Piedimonte, G., Salpietro, F.M. and Tomasello, F. (2002) Beneﬁcial effects of systemic administration of recombinant human erythropoietin in rabbits subjected to subarachnoid hemorrhage. Proc. Natl. Acad. Sci. U.S.A., 99: 5627–5631. Haddad, J.J. (2002) Recombinant human interleukin (IL)-1bmediated regulation of hypoxia inducible factor 1a (HIF-1a) stabilization, nuclear translocation and activation requires an antioxidant/reactive oxygen species (ROS)-sensitive mechanism. Eur. Cytokine Netw., 13: 250–260. Hellwig-Burgel, T., Rutkowski, K., Metzen, E., Fandrey, J. and Jelkmann, W. (1999) Interleukin-1beta and tumor necrosis factor-alpha stimulate DNA binding of hypoxia-inducible factor-1. Blood, 94: 1561–1567. Horowitz, M. (1994) Heat stress and heat acclimation: the cellular response-modiﬁer of autonomic control. Integrative and cellular aspects of autonomic function. In: Pleschka K. and Gerstberger R. (Eds.), Temperature and Osmoregulation. Libbey Eurotext, Paris, pp. 87–95. Horowitz, M. (1998) Do cellular heat acclimation responses modulate central thermoregulatory activity? News Physiol. Sci., 13: 218–225.

362 Horowitz, M. (2002) From molecular and cellular to integrative heat defense during exposure to chronic heat. Comp. Biochem. Physiol., 131: 475–483. Horowitz, M. (2003) Matching the heart to heat-induced circulatory load: heat-acclimatory responses. NIPS: News Physiol. Sci., 18: 215–221. Horowitz, M. (2007) Heat acclimation and cross-tolerance against novel stressors: genomic-physiological linkage. In: Sharma H.S. (Ed.), Progress in Brain Research: The Neurobiology of Hyperthermia, Elsevier, Amsterdam, The Netherlands (in press). Horowitz, M., Eli-Berchoer, L., Wapinski, I., Friedman, N. and Kodesh, E. (2004) Stress-related genomic responses during the course of heat acclimation and its association with ischemic-reperfusion cross-tolerance. J. Appl. Physiol., 97(4): 1496–1507. Horowitz, M. and Meiri, U. (1985) Altered responsiveness to parasympathetic activation of submaxillary salivary gland in the heat acclimated rat. Comp. Biochem. Physiol., 80A: 57–60. Hua, Y., Keep, R.F., Hoff, J.T. and Xi, G. (2003) Thrombin preconditioning attenuates brain edema induced by erythrocytes and iron. J. Cereb. Blood Flow Metab., 23: 1448–1454. Ikonomidis, J.S., Tumiati, L.C., Weisel, R.D., Mickel, D.A.G. and Li, R.K. (1994) Preconditioning human ventricular cardiomyocytes with brief periods of simulated ischemia. Cardiovasc. Res., 28: 623–634. Janoff, A. (1964) Alterations in lysosomes (intracellular enzymes) during shock; effects of preconditioning (tolerance) and protective drugs. Int. Anesthesiol. Clin., 2: 251–259. Jones, N.M. and Bergeron, M. (2001) Hypoxic preconditioning induces changes in HIF-1 target genes in neonatal rat brain. J. Cereb. Blood Flow Metab., 21: 1105–1114. Kilic, E., Kilic, U., Wang, Y., Bassetti, C.L., Marti, H.H. and Hermann, D.M. (2006) The phosphatidylinositol-3-kinase/ Akt pathway mediates VEGF’s neuroprotective activity and induces blood brain barrier permeability after focal cerebral ischemia. FASEB J., 20(8): 1185–1187. Kissinger, P.T., Jonathan, B.H. and Adams, R.W. (1973) Voltammetry in brain tissue: a new neurophysiological measurement. Brain Res., 55: 209–213. Kloog, Y., Horowitz, M., Meiri, U., Galron, R. and Avron, A. (1985) Regulation of submaxillary gland muscarinic receptors during heat acclimation. Biochim. Biophys. Acta, 845: 428–435. Kohen, R., Beit Yannai, E., Berry, E.M. and Tirosh, O. (1999) Overall low molecular antioxidant activity of biological ﬂuids and tissues by cyclic voltammetry. Methods Enzymol., 300: 285–296. Kohen, R. and Nyska, A. (2002) Oxidation of biological systems: oxidative stress phenomena, antioxidants, redox reactions, and methods for their quantiﬁcation. Toxicol. Pathol., 30: 620–650. Kontos, H.A. (2001) Oxygen radicals in cerebral ischemia. Stroke, 11: 2712–2716. Knoblach, S.M. and Faden, A.I. (1998) Interleukin-10 improves outcome and alters proinﬂammatory cytokine

expression after experimental traumatic brain injury. Exp. Neurol., 153: 143–151. Levi, E., Vivi, A., Hasin, Y., Tassini, M., Navon, G. and Horowitz, M. (1993) Heat acclimation improves cardiac mechanics and metabolic performance during ischemia and reperfusion. J. Appl. Physiol., 75: 833–839. Liu, J., Ginis, I., Spatz, M. and Hallenbeck, J.M. (2000) Hypoxic preconditioning protects cultured neurons against hypoxic stress via TNF-alpha and ceramide. Am. J. Physiol. Cell Physiol., 278: C144–C153. Luh, S. and Yang, P. (2006) Organ preconditioning: the past, current status and related lung studies. J. Zhejiang Univ. Sci. B, 7: 331–341. Maloyan, A., Eli-Berchoer, L., Semenza, G.L., Gerstenblith, G., Stern, M.D. and Horowitz, M. (2005) HIF-1alphatargeted pathways are activated by heat acclimation and contribute to acclimation-ischemic cross-tolerance in the heart. Physiol. Genomics, 23(1): 79–88. Maloyan, A., Palmon, A. and Horowitz, M. (1999) Heat acclimation increases the basal HSP72 level and alters its production dynamics during heat stress. Am. J. Physiol., 276: R1506–R1515. Maloyan, A., Semenza, G.L., Gerstenblith, G., Stern, M. and Horowitz, M. (2001) Heat-acclimation-ischemia cross tolerance: does HIF-1 play a role? J. Mol. Cell Cardiol., 33: A72. Morishita, E., Masuda, S., Nagao, M., Yasuda, Y. and Sasaki, R. (1997) Erythropoietin receptor is expressed in rat hippocampal and cerebral cortical neurons, and erythropoietin prevents in vitro glutamate-induced neuronal death. Neuroscience, 76: 105–116. Murry, C.E., Jennings, R.B. and Reimer, K.A. (1986) Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium. Circulation, 74: 1124–1136. Nie, H., Xiong, L., Lao, N., Chen, S., Xu, N. and Zhu, Z. (2006) Hyperbaric oxygen preconditioning induces tolerance against spinal cord ischemia by upregulation of antioxidant enzymes in rabbits. J. Cereb. Blood Flow Metab., 26: 666–674. Reshef, A., Sperling, O. and Zoref-Shani, E. (1996) Preconditioning of primary rat neuronal cultures against ischemic injury: characterization of the ‘‘time window of protection.’’ Brain Res., 741: 252–257. Schaller, B. (2005) Ischemic preconditioning as induction of ischemic tolerance after transient ischemic attacks in human brain: its clinical relevance. Neurosci. Lett., 377: 206–211. Semenza, G.L. (2004) Hydroxylation of HIF-1: oxygen sensing in the molecular level. Physiology, 19: 176–182. Shapira, Y., Shohami, E., Sidi, A., Soffer, D., Freeman, S. and Cotev, S. (1988) Experimental closed head injury in rats: mechanical, pathophysiologic, and neurologic properties. Crit. Care Med., 16: 258–265. Sharp, F.R., Ran, R., Lu, A., Tang, Y., Strauss, K.I., Glass, T., Ardizzone, T. and Bernaudin, M. (2004) Hypoxic preconditioning protects against ischemic brain injury. NeuroRx, 1: 26–35. Shein, N.A., Doron, H., Alexandrovich, A.G., Horowitz, M. and Shohami, E. (2006) Heat acclimation increases pre-injury anti-inﬂammatory capacity and leads to attenuation of

363 pro-inﬂammatory mRNA expression after closed head injury in mice, 5th Forum of European Neuroscience, Vienna, Austria. Shein, N.A., Horowitz, M., Alexandrovich, A.G., Tsenter, J. and Shohami, E. (2005a) Heat acclimation increases hypoxiainducible factor 1alpha and erythropoietin receptor expression: implication for neuroprotection after closed head injury in mice. J. Cereb. Blood Flow Metab., 25: 1456–1465. Shein, N.A., Horowitz, M., Alexandrovich, A.G., Tsenter, J. and Shohami, E. (2005b) Akt-1 phosphorylation is increased and tumor necrosis factor is decreased after closed head injury in heat acclimated mice. J. Neurotrauma, 22(10): 1227. Shiki, K. and Hearse, D.J. (1987) Preconditioning of ischemic myocardium: reperfusion induced arrhythmia. Am. J. Physiol., 253: H1470–H1476. Shohami, E., Gallily, R., Mechoulam, R., Bass, R. and Ben-Hur, T. (1997) Cytokine production in the brain following closed head injury: dexanabinol (HU-211) is a novel TNF-alpha inhibitor and an effective neuroprotectant. J. Neuroimmunol., 72: 169–177. Shohami, E., Novikov, M., Bass, R., Yamin, A. and Gallily, R. (1994a) Closed head injury triggers early production of TNF alpha and IL-6 by brain tissue. J. Cereb. Blood Flow Metab., 14(4): 615–619. Shohami, E., Novikov, M. and Horowitz, M. (1994b) Long term exposure to heat reduces edema formation after closed head injury in the rat. Acta Neurochir. Suppl. (Wien.), 60: 443–445. Shohami, E., Shapira, Y., Sidi, A. and Cotev, S. (1987) Head injury induces increased prostaglandin synthesis in rat brain. J. Cereb. Blood Flow Metab., 7: 58–63. Shohami, E., Shapira, Y., Yadid, G., Reisfeld, N. and Yedgar, S. (1989) Brain phospholipase A2 is activated after experimental closed head injury in the rat. J. Neurochem., 53: 1541–1546.

Siren, A.L., Knerlich, F., Poser, W., Gleiter, C.H., Bruck, W. and Ehrenreich, H. (2001) Erythropoietin and erythropoietin receptor in human ischemic/hypoxic brain. Acta Neuoropathol. (Berl.), 101: 271–276. Stahel, P.F., Shohami, E., Younis, F.M., Kariya, K., Otto, V.I., Lenzlinger, P.M., Grosjean, M.B., Eugster, H.P., Trentz, O., Kossmann, T. and Morganti-Kossmann, M.C. (2000) Experimental closed head injury: analysis of neurological outcome, blood-brain barrier dysfunction, intracranial neutrophil inﬁltration, and neuronal cell death in mice deﬁcient in genes for pro-inﬂammatory cytokines. J. Cereb. Blood Flow Metab., 20(2): 369–380. Toumpoulis, I.K. and Anagnostopoulos, C.E. (2003) Early ischemic preconditioning for spinal cord protection. Ann. Thorac. Surg., 76: 1340–1341. Treinin, M., Shliar, J., Jiang, H., Powell-Coffman, J.A., Bromberg, Z. and Horowitz, M. (2003) HIF-1 is required for heat acclimation in the nematode C. elegans. Physiol. Genomics, 14: 17–24. Vink, R. and Van Den Heuvel, C. (2004) Recent advances in the development of multifactorial therapies for the treatment of traumatic brain injury. Expert Opin. Investig. Drugs, 13(10): 1263–1274. Vohra, H.A. and Galinanes, M. (2006) Myocardial preconditioning against ischemia-reperfusion apoptosis and necrosis in man. J. Surg. Res., 134(1): 138–144. Yatsiv, I., Grigoriadis, N., Simeonidou, C., Stahel, P.F., Schmidt, O.I., Alexandrovitch, A.G., Tsenter, J. and Shohami, E. (2005) Erythropoietin is neuroprotective, improves functional recovery, and reduces neuronal apoptosis and inﬂammation in a rodent model of experimental closed head injury. FASEB J., 19: 1701–1703. Yellon, D.M. and Dana, A. (2000) The preconditioning phenomenon: a tool for the scientist or a clinical reality. Circ. Res., 87: 543–550.

This page intentionally left blank

SECTION VI

The Future of Neurotrauma: Developing Novel Treatment Strategies

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 26

In vivo tracking of stem cells in brain and spinal cord injury Eva Sykova and Pavla Jendelova Institute of Experimental Medicine ASCR, EU Centre of Excellence, Prague, Czech Republic; Center for Cell Therapy and Tissue Repair, Charles University, Second Medical Faculty, Prague, Czech Republic; Department of Neuroscience, Charles University, Second Medical Faculty, Prague, Czech Republic

Abstract: Cellular magnetic resonance (MR) imaging is a rapidly growing ﬁeld that aims to visualize and track cells in living organisms. Superparamagnetic iron oxide (SPIO) nanoparticles offer a sufﬁcient signal for T2 weighted MR images. We followed the fate of embryonic stem cells (ESCs) and bone marrow mesenchymal stem cells (MSCs) labeled with iron oxide nanoparticles (Endorems) and human CD34+ cells labeled with magnetic MicroBeads (Miltenyi) in rats with a cortical or spinal cord lesion, models of stroke and spinal cord injury (SCI), respectively. Cells were either grafted intracerebrally, contralaterally to a cortical photochemical lesion, or injected intravenously. During the ﬁrst post-transplantation week, grafted MSCs or ESCs migrated to the lesion site in the cortex as well as in the spinal cord and were visible in the lesion on MR images as a hypointensive signal, persisting for more than 30 days. In rats with an SCI, we found an increase in functional recovery after the implantation of MSCs or a freshly prepared mononuclear fraction of bone marrow cells (BMCs) or after an injection of granulocyte colony stimulating factor (G-CSF). Morphometric measurements in the center of the lesions showed an increase in white matter volume in cell-treated animals. Prussian blue staining conﬁrmed a large number of iron-positive cells, and the lesions were considerably smaller than in control animals. Additionally, we implanted hydrogels based on poly-hydroxypropylmethacrylamide (HPMA) seeded with nanoparticle-labeled MSCs into hemisected rat spinal cords. Hydrogels seeded with MSCs were visible on MR images as hypointense areas, and subsequent Prussian blue histological staining conﬁrmed positively stained cells within the hydrogels. To obtain better results with cell labeling, new polycation-bound iron oxide superparamagnetic nanoparticles (PC-SPIO) were developed. In comparison with Endorem, PC-SPIO demonstrated a more efﬁcient intracellular uptake into MSCs, with no decrease in cell viability. Our studies demonstrate that magnetic resonance imaging (MRI) of grafted adult as well as ESCs labeled with iron oxide nanoparticles is a useful method for evaluating cellular migration toward a lesion site. Keywords: cell transplantation; iron oxide nanoparticles; magnetic resonance; MicroBeads; photochemical lesion; scaffold; spinal cord injury disorders of the central nervous system (CNS; Park et al., 2002; McKay, 2004; Newman et al., 2004; Roitberg, 2004; Emsley et al., 2005; Fairless and Barnett, 2005; Pluchino et al., 2005; Zhao et al., 2005; Uccelli et al., 2006). Crucial to the future success of cell transplantation in the clinical setting is the ability of transplanted cells to migrate

Introduction Stem cells and progenitor cells are being explored in regenerative medicine for cell therapy in Corresponding author. Tel.: +420-241062230;

Fax:+420-241062782; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61026-1

367

368

from the site of transplantation to the lesioned area and to survive, differentiate, and/or produce growth factors and cytokines for the prolonged periods of time necessary for the patient to beneﬁt from their regenerative properties. Visualizing transplanted cells in vivo is essential for preclinical studies in rodents, and the magnetic tracking of cells appears to be a valuable tool for such studies. Magnetic resonance (MR) imaging may also serve to study cell migration to lesions, its timecourse, and how long the cells persist in the target region. Such information could help to elucidate the time window during which the transplantation of therapeutic cells may be clinically effective, the number of cells needed, and the optimal method of their administration. For magnetic resonance imaging (MRI) detection, cells can be labeled with MR contrast agents in order to make them visible in vivo (Bulte et al., 2002; Jendelova et al., 2003, 2004; Sykova and Jendelova, 2005, 2006; Sykova et al., 2006).

Superparamagnetic cell labels Superparamagnetic iron oxide (SPIO) nanoparticles were introduced as contrast agents shortly after the use of gadolinium-chelates (Mendonca Dias and Lauterbur, 1986; Renshaw et al., 1986; Pachernik et al., 2005). They are currently preferred as contrast agents primarily due to the following properties: (a) they provide the greatest signal contrast change, in particular on T2 and T2* weighted images; (b) they are composed of biodegradable iron; (c) their surface coating makes them soluble and stable and allows for the chemical linkage of functional groups and ligands; and (d) they can be easily detected by both light and electron microscopy (EM; Jendelova et al., 2003; Bulte and Kraitchman, 2004). Iron oxide nanoparticle stabilization in order to prevent aggregation is most commonly accomplished by a surface coating of dextran. Dextrancoated SPIO nanoparticles include the products Feridexs and Endorems, the ultrasmall superparamagnetic iron oxide (USPIO) nanoparticles Combidexs and Sinerems, monocrystalline iron

oxide nanoparticles, and cross-linked iron oxide nanoparticles. Another approach for stabilizing iron oxide nanoparticles is the use of carboxylated polyamidoamine dendrimers (Bulte et al., 2001). These highly soluble nanocomposites of iron oxides and dendrimers are stable under a wide range of temperatures and pH and have an overall size of 20–30 nm. Iron oxide nanoparticles also require an appropriate outer surface layer that induces the internalization of the particles into the cytoplasm, so that they can be nonspecifically taken up by a variety of cultured mammalian cells, regardless of cell origin or animal species. In magnetodendrimers, the highly charged carboxylated dendrimers bind to multiple sites on the cell membrane, inducing membrane bending followed by endocytosis (Zhang and Smith, 2000). In the case of anionic magnetic nanoparticles, the negative surface charge induces an uptake three orders of magnitude greater than that of conventional dextran-coated SPIO nanoparticles (Billotey et al., 2003). A disadvantage of these particles, as well as of the magnetodendrimers, is that they have not yet been commercially developed. Another method is based on the mixing of a commercially available (U)SPIO formulation, Feridexs (Frank et al., 2003) or Sinerems, and a commercially available transfection agent, for example poly-L-lysine, Lipofectamin, or FugeneTM (Frank et al., 2003). We have developed new polycation-bound iron oxide superparamagnetic nanoparticles (PC-SPIO) that can be used for intracellular labeling (Horak et al., 2006).

Cell labeling with dextran-coated SPIO nanoparticles MR tracking of stem and progenitor cells in the lesioned CNS has been performed by several groups that utilized different methods of administration. The ﬁrst studies of imaging cell transplants labeled by superparamagnetic contrast agents in rat brains were reported in 1992 (Norman et al., 1992; Hawrylak et al., 1993). Rats received grafts of fetal rat tissue prepared as cell suspensions and

369

labeled by incubation with reconstituted Sendai viral envelopes containing iron oxide particles. When magnetically labeled neurospheres were transplanted into the ventricles of experimental allergic encephalomyelitis (EAE) rats at the peak of their disease, the migration of glial precursors into white matter structures was observed on MR images (Bulte et al., 2003). Ferromagnetic-labeled neural progenitor cells were also transplanted into the cisterna magna of rats that underwent experimental MCAO (middle cerebral artery occlusion). MR images showed the migration of cells throughout the ventricular system toward the ischemic brain parenchyma (Zhang et al., 2003). Olfactory ensheathing cells (OECs) labeled with magnetodendrimers were implanted into the transected rat spinal cord, and their distribution was followed in vivo using MR imaging (Lee et al., 2004). In our experiments (Jendelova et al., 2003, 2004), we have shown that a suitable contrast agent for mesenchymal stem cells (MSCs), embryonic stem cells (ESCs), and OECs is a commercially available contrast agent based on dextran-coated SPIO nanoparticles (Fig. 1A), Endorems (Guerbet, France), which has also been approved as a blood pool agent for human use. The contrast agent Endorem can be easily incorporated by endocytosis, and all the cells survive and further divide in vitro. Therefore, Endorem uptake does not need to be facilitated by a transfection agent, which can damage large numbers of cells (Arbab et al., 2004). Bone marrow-derived MSCs are adult stem cells that reside within the bone marrow compartment. Recent data have presented evidence for their multilineage differentiation potential (Pittenger et al., 1999). More recently, MSCs have been reported to have the ability to stimulate or participate in the regeneration of diverse tissues and organs, including the liver, myocardium, endothelium, and CNS (Takahashi et al., 1999; Orlic et al., 2001a, b; Jiang et al., 2002; Toma et al., 2002). In addition, they can be genetically modiﬁed, and, due to their migratory properties, they can serve as a carrier for drug delivery in tumor therapies (Anderson et al., 2005). In cell therapy, bone marrow cells (BMCs) have some advantages over

other sources of cells: (a) they are relatively easy to isolate; (b) they can be used in autologous transplantation protocols; and (c) they have already been approved for the treatment of hematopoietic diseases. In our experiments, cell cultures of human MSCs or rat MSCs were incubated 2–3 days in media containing Endorem. Nanoparticles were detected by staining for iron (Fig. 1B). Transmission electron microscopy conﬁrmed the presence of iron oxide particles inside the cells, observed as large membrane-bound clusters scattered within the cell cytoplasm (Fig. 1C). On the day that nanoparticles were withdrawn, the efﬁciency of MSC labeling (i.e., how many cells of the total number of analyzed cells were labeled) was 50–70%.

Tracking mesenchymal stem cells in an experimental model of stroke Endorem-labeled MSCs were co-labeled with bromdeoxyuridine (BrdU) and grafted into rats with a cortical photochemical lesion (Jendelova et al., 2003). Rats were examined weekly for a period of 3–7 weeks post-transplantation using a 4.7 T Bruker spectrometer. Single sagittal, coronal, and transversal images were obtained by a fast gradient echo sequence for localizing subsequent T2 weighted transversal images measured by a standard turbospin echo sequence. The lesion was clearly visible on MR images 2 h after lesioning as a hyperintense signal and remained visible during the entire measurement period. No recognizable hypointense signal in the lesion was detected during the ﬁrst 2 days after implantation. A decrease in the MR signal was found only at the injection site in animals with cells injected contralaterally to the lesion. One week after grafting, we observed a hypointense signal in the lesion, which intensiﬁed during the second and third weeks (Fig. 1D). Histology conﬁrmed that a large number of Prussian blue-positive cells had entered the lesion. No hypointense signal was found in other brain regions. The hypointense signal occurred only in damaged areas populated with MSCs, and its intensity corresponded to Prussian

370

Fig. 1. (A) Scheme of an iron oxide nanoparticle. The contrast agent Endorem consists of a superparamagnetic Fe3O4 core coated by a dextran shell. (B) Rat mesenchymal stem cells (MSCs) in culture labeled with superparamagnetic nanoparticles (blue dots). The cell nuclei are counterstained with hematoxylin. (C) Transmission electron microphotograph of a cluster of iron nanoparticles surrounded by a cell membrane. (D) An implant of Endorem-BrdU co-labeled MSCs (in the hemisphere contralateral to the lesion, arrow) and the lesion (arrowhead) itself are both hypointense in an MR image taken 2 weeks after implantation. (E), (F) A hypointense signal (black arrowhead) was observed in the lesion 6 days after the intravenous (i.v.) injection of Endorem-BrdU co-labeled MSCs, becoming more hypointense and persisting for 47 days (F). (G, H) Massive invasion of rat MSCs (Prussian blue staining counterstained with hematoxylin) into a photochemical lesion 7 weeks after i.v. injection into a rat with a photochemical lesion. (I) Serial section stained for BrdU, 7 weeks after i.v. injection. (Adapted with permission from Jendelova et al. (2003).)

blue or BrdU staining. Only a few (less than 3%) of the MSCs that migrated into the lesion expressed the neuronal marker NeuN when tested 28 days post-implantation. No GFAP-positive cells were found in the lesion. After the intravenous injection of MSCs, we found a similar hypointense MR signal in the lesion site. The signal was observed 6 days after cell infusion and persisted for 7 weeks (Figs. 1E, F). Prussian blue and anti-BrdU staining conﬁrmed the presence of iron oxide-BrdU co-labeled cells in

the lesion, which densely populated the borders of the lesion (Figs. 1G–I). Only a few cells weakly stained for Prussian blue were found in photochemical lesions without any implanted cells. The staining represents iron, which most likely originated in hemorrhages and iron degradation products released from ironcontainingproteins (such as hemoglobin, ferritin, and hemosiderin) and phagocytized by microglia/ macrophages. We did not observe any BrdUpositive cells in the brains of nongrafted animals.

371

Tracking embryonic stem cells in an experimental model of stroke ESCs are pluripotent cell lines with a capacity for self-renewal and broad differentiation plasticity. They are derived from embryos and can be propagated as a homogeneous, uncommitted cell population for an almost unlimited period of time without losing their pluripotency and their stable karyotype. Mouse ES cell-derived glial precursors, transplanted into rats with myelin disease, interacted with the host neurons to produce myelin in the brain and spinal cord (Brustle et al., 1999). Retinoic acid-treated embryoid bodies from mouse ESCs, when transplanted into rat spinal cord 9 days after traumatic injury, differentiated into astrocytes, oligodendrocytes, and neurons and promoted recovery (McDonald et al., 1999). Bjorklund et al. (2002) reported that undifferentiated mouse ESCs can become dopamine-producing neurons in the brain in a rat model of Parkinson’s disease and can lead to partial functional recovery. We used mouse ESCs transfected with the pEGFP-C1 vector (Pachernik et al., 2005) and labeled with SPIO nanoparticles (Jendelova et al., 2004). Since undifferentiated ESCs may form embryonic tumors when grafted into a host animal, we induced neural differentiation by culturing enhanced green ﬂuorescent protein (eGFP) ESCs in serum containing DMEM (Dulbecco’s modiﬁed Eagle medium)/F12 without leukemia inhibitory factor (LIF) for 2 days and then transferred the cells into serum-free media supplemented with insulin, transferrin, selenium, and ﬁbronectin for further culture (Pachernik et al., 2005). Feeder-free eGFP ESCs were labeled with Endorem (112.4 mg/ml) during three passages (Fig. 2A). Cell counting of Prussian blue stained cells in suspension revealed that the labeling efﬁciency was 80%. Cells were transplanted intracerebrally or intravenously on the 8th day of differentiation into adult Wistar rats with a cortical photochemical lesion 1 week after lesioning (Jendelova et al., 2004), and were detected by staining for iron and by GFP ﬂuorescence. When we implanted the ESCs 7 days post-lesion, we found a massive migration of Endorem-labeled GFP-positive cells into the lesion site regardless of the route of administration (Fig. 2D). We observed

very similar MR images to those obtained after the implantation of MSCs. In rats with a photochemical lesion and contralaterally injected cells, the cell implants were visible as a hypointense area at the injection site. Two weeks after grafting, a hypointense signal was also observed in the corpus callosum and in the lesion (Fig. 2B). At the same time, histology showed that a large number of Prussian blue-positive cells had entered the lesion. Many labeled cells were also detected in the corpus callosum, suggesting a migration from the contralateral hemisphere toward the lesion (Fig. 2C). When the ESCs were injected intravenously into lesioned rats, we found a hypointense MR signal only at the site of the lesion. However, the extent of the differentiation of eGFP ESCs labeled with nanoparticles, and found in the lesion, was much greater than with grafted MSCs. Of all the eGFP ESCs containing nanoparticles found in the lesion, 70% of them proved to be astrocytes, very few (less than 1%) were oligodendrocytes, and 5% of nanoparticle-labeled eGFP ESCs had differentiated into neurons (Jendelova et al., 2004). Electron microscopy revealed mature astrocytes, neurons, and oligodendrocytes in the lesion site containing iron oxide nanoparticles, providing strong supporting evidence that these cells differentiated in the lesion from implanted embryonic cells (Figs. 2E, F).

Tracking mesenchymal stem cells in a rat model of spinal cord injury Evaluating the effect of different BMC populations on morphological and functional recovery after spinal cord injury (SCI) is an important aspect of our preclinical research. In rats with a balloon-induced spinal cord compression lesion, we studied the effect of an intravenous injection of Endorem-labeled nonhematopoietic MSCs (Urdzikova et al., 2006). The results obtained were compared with those following the implantation of a freshly isolated mononuclear fraction of bone marrow containing stromal cells, hematopoietic and nonhematopoietic stem and precursor cells, and lymphocytes (BMCs). Furthermore, we studied the effect of granulocyte colony stimulating factor (G-CSF) mobilization of endogenous BMCs containing mainly

372

Fig. 2. (A) Trypsinized nanoparticle-labeled suspensions of ESCs after the third passage. (B) The cell implant (in the hemisphere contralateral to the lesion) and the lesion are hypointense in MR images 2 weeks after implantation. A hypointensive signal is also found in the corpus callosum. (C) Dense Prussian blue staining of the injection site in the contralateral hemisphere, the corpus callosum, and the photochemical lesion, 4 weeks after grafting. (D) Invasion of GFP-labeled cells showing GFP-positive ESCs in the lesion 4 weeks after the i.v. injection of eGFP ESCs (serial section to the slice shown in Fig. 2C). (E) Astrocyte from the brain tissue of a rat with nanoparticle-labeled eGFP ESCs implanted contralaterally to the lesion; the nanoparticles are visible in small dense clusters in the cell cytoplasm (arrowhead). (F) Neuron with nanoparticles marginated to the nuclear membrane. (Adapted with permission from Jendelova et al. (2004).)

hematopoietic stem cells, along with progenitor cells and lymphocytes (Sasaki et al., 2001; Akiyama et al., 2002). MSCs labeled with Endorem were injected intravenously into the femoral vein 1 week after lesioning. MR images were taken ex vivo 4 weeks

after implantation using a standard whole body resonator. Functional status was assessed weekly for 5 weeks after spinal cord lesioning, using the Basso-Beattie-Bresnehan (BBB) locomotor rating score and the plantar test. Our data indicated that lesioned animals with grafted MSCs had higher

373

locomotor scores as indicated by their BBB scores and showed better responses in sensitivity testing using the plantar test than did control animals. Particularly, the plantar test showed the recovery of sensitivity in the hind limbs. On MR images we observed the lesion as an inhomogeneity in the tissue texture with a hyperintense signal only in the area of the SCI (Fig. 3A). Images of longitudinal spinal cord sections from animals grafted with nanoparticle-labeled MSCs showed the lesion as a dark hypointense area (Fig. 3B). Prussian blue

staining conﬁrmed a large number of positive cells present in the lesion site (Fig. 3C). On serial sections, Prussian blue-positive cells corresponded to cells labeled with PKH 26, as seen in (Figs. 3D, E). To exclude the possibility that free nanoparticles were taken up by macrophages, we performed immunohistochemical and Prussian blue staining to co-localize iron in activated microglia/ macrophages. Although we found strong positive staining for macrophages in the lesions using ED1 antibody, Prussian blue staining did not,

Fig. 3. (A) Longitudinal MR image of a spinal cord lesion 5 weeks after induction. The formation of the lesion cavity is visible as a strong hyperintensive signal (arrow). (B) Longitudinal MR image 4 weeks after MSC grafting. The lesion with nanoparticle-labeled cells is visible as a dark hypointensive area (arrow). (C) Prussian blue staining of a lesion populated with nanoparticle-labeled MSCs (blue dots). (D) Magniﬁed region from Fig. C. (E) Intravenously injected MSCs were also detectable in the spinal cord lesion using the membrane ﬂorescent dye PKH 26. Serial section to D. (F) Immunostaining with ED-1 antibody revealed a number of macrophages in the lesion (white arrows). The majority of Prussian blue-positive cells did not co-localize with ED-1 staining; however, a few macrophages were Prussian blue-positive (black arrowheads). (Adapted with permission from Urdzikova et al. (2006).)

374

for the most part, co-localize with the ED1 staining (Fig. 3F). Morphometric measurements of the spared white and gray matter were performed in the center of the lesions. The spared cross-sectional area of the white matter, as well as the volume of spared white matter, was significantly greater in MSCtreated animals. The spared cross-sectional area of the gray matter was also significantly larger in MSC-treated animals. In further studies, lesioned animals grafted with BMCs or injected with G-

SCF also had higher scores in BBB testing than did control animals and also showed a faster recovery of sensitivity in their hind limbs using the plantar test. However, the functional improvement was more pronounced in MSC-treated rats (Figs. 4A, B). Morphological measurements of the spared cross-sectional area of the white matter showed a statistically significant increase in groups treated with BMCs or G-SCF, compared to controls, cranially to the lesion center; a statistically significant increase in the volume of spared white

Fig. 4. (A) Behavioral open ﬁeld BBB motor scores of MSC, BMC, and G-CSF treated rats were significantly higher than those of saline-injected (control) animals, 14, 21, 28, and 35 days after SCI (po0.05). The scores of the MSC, BMC, and G-CSF treated animals were not significantly different from one another at any time point. (B) Time course of the animals’ response to radiant heat measured with the plantar test in treated and saline-injected rats. In all treated rats, the latency times decreased as their recovery progressed. The most pronounced effect was seen in MSC treated rats. The latency time in the saline-injected (control) rats did not change during the 35 days survival period. Data are averaged between right and left hind limbs and expressed as mean7SEM. *po0.05 compared to control group. (Adapted with permission from Urdzikova et al. (2006).)

375

Fig. 5. (A) The total volume of the white matter in 11 mm long segments of the spinal lesion in MSCs, BMCs, and G-CSF-treated and saline-injected animals. (B) The total volume of gray matter in 11 mm long segments of the spinal lesion in MSCs, BMCs, and G-CSFtreated and saline-injected animals. No statistically significant differences were observed between treated and saline-injected animals. All data are presented as means7SEM. *po0.05 compared to saline-injected (control) rats. (Adapted with permission from Urdzikova et al. (2006).)

matter was observed in the BMC-treated group. The spared white matter volume in the G-CSFtreated rats was also increased, but the increase did not reach statistical significance (Fig. 5).

Stem cell labeling with polycation-bound nanoparticles In all the above-described experiments, dextrancoated SPIO nanoparticles (the contrast agent Endorem) were used as an intracellular label. However, the efﬁciency of the cell labeling was maximally 70%. In addition, on the day of particle withdrawal a decrease in cell viability (using the WST1 colorimetric assay) was observed. We

therefore developed new polycation-bound SPIO nanoparticles (Fig. 6A; Horak et al., 2006). We compared the inﬂuence of both types of nanoparticles on cell viability and labeling efﬁciency in rat and human MSCs. The PC-SPIO nanoparticle suspension was used at a much lower iron concentration per milliliter of culture media (15.4 mg/ml), and the cells were incubated with PC-SPIO for 3 days. The results were compared with Endorem labeling (Tables 1 and 2). The measurements were performed on the day of withdrawal of the nanoparticles (day 3) and 4 days later (day 7). Labeling cells with PC-SPIO nanoparticles was more efﬁcient than labeling with Endorem, i.e., more cells from the total number of analyzed cells were labeled with PC-SPIO nanoparticles than

376

Fig. 6. (A) Schematic illustration of a polycation-bound superparamagnetic iron oxide nanoparticle. (B) Transmission electron photomicrograph of MSCs labeled with PC-SPIO showing clusters of iron nanoparticles (arrows) in the cell cytoplasm, which are not surrounded by a cell membrane. (C) Prussian blue staining of MSCs in culture labeled with PC-SPIO. (D, E) Axial and coronal (E) MR images of a rat brain with 1000 cells labeled with PC-SPIO implanted to the left hemisphere and 1000 Endorem-labeled cells implanted to the right hemisphere. MR images were taken 3 days after implantation. Table 1. The percentage of cells labeled with nanoparticles after 3 days incubation

Rat MSCs Human MSCs

Endorem-labeled cells (%)

PC-SPIO-labeled cells (%)

59.3 65.2

92.2 87.5

with Endorem (Table 1; Fig. 6C). In Table 2 the cell viability of labeled cells is expressed as a percentage of the number of viable unlabeled cells (taken as the control value and deﬁned as 100%).

In Endorem-labeled human MSCs, on day 3 the cell viability dropped to 50%; however, within 4 days it recovered to 87%. This transient decrease in cell viability might be due to the relatively high concentration of iron in the culture media (112.4 g/ml), to which some cell cultures might be sensitive. The average amount of iron present in rat MSCs was determined by spectrophotometry after mineralization of iron-labeled cell suspensions. In PCSPIO-labeled cells, even though the concentration of iron in the culture media was 10 times lower (15.4 mg/ml culture media), the average amount of

377 Table 2. Cell viability (in %) after incubation with nanoparticles (day 3) and 4 days after withdrawal (day 7)

Rat MSCs Human MSCs

Endorem-labeled cells (%)

PC-SPIO-labeled cells (%)

Day 3

Day 7

Day 3

Day 7

68.9 49.0

72.9 87.0

92.8 97.2

89.7 98.9

iron was 38 pg/cell, while in Endorem-labeled cells only 17 pg/cell. Transmission electron photomicrographs conﬁrmed that more PC-SPIO nanoparticles were taken up by the rat MSCs than were Endorem nanoparticles (Fig. 6B). Moreover, Endorem was observed as membrane-bound clusters within the cell cytoplasm, indicating an endocytotic process of nanoparticle uptake. The higher cellular uptake of PC-SPIO nanoparticles is possible due to the interaction of the surface coating with the negatively charged cell surface and subsequent endosomolytic uptake. Nanoparticles are transported in endosomes and ﬁnally fused with lysosomes, a process during which the vesicle membranes disappear. To check the sensitivity of the MRI technique and to mimic signal behavior in CNS tissue, suspensions of unlabeled cells and cells labeled with PC-SPIO were imaged in vitro. MR images of 1.7% gelatin phantoms containing iron-labeled MSCs were obtained using a 4.7 T Bruker spectrometer equipped with a standard resonator coil. Even the sample containing the lowest concentration (200 cells/ml, which corresponds on average to 2 cells per image voxel) provided visible contrast compared to a control phantom containing the same number of unlabeled cells. A similar set of experiments was performed in earlier work (Jendelova et al., 2003), in which MR images of gelatin phantoms showed a hypointense signal at concentrations above 625 labeled cells/ml. For in vivo imaging, rats were examined 3 days post-transplantation in an MR imager. Figure 6 shows that PC-SPIO labeled cells are also clearly recognizable in vivo. The iron oxide-labeled cell implants are visible as a hypointense area at the injection site. Cells labeled by PC-SPIO provide

stronger contrast change in the signal than do Endorem-labeled cells (Figs. 6D, E). Labeling with PC-SPIO thus enables the detection of a lower number of cells in the tissue.

Cell labeling with magnetic MicroBeads The disadvantage of an intracellular label is that it can affect cell metabolism and subsequently cell viability. In addition, these labels are rather nonspeciﬁc; they can be loaded by virtually any cell present in the medium. The possibility of labeling only selected cell types would therefore be very useful. In addition, cells labeled and separated by means of immunomagnetic selection would not require in vitro culturing, since the label is attached during separation. This would also allow the immediate clinical use of labeled cells. The ﬁrst experiments using contrast agents bound to an antibody that can specifically bind to a single cell type were performed by Bulte et al. (1992). They described experiments with human lymphocytes labeled with biotinylated anti-lymphocyte-directed monoclonal antibodies (mabs), to which streptavidin and subsequently biotinylated dextran-magnetite particles were coupled. We tested a new type of speciﬁc cell labeling using commercially available cell isolation kits for the magnetic separation of CD34+ cells. CD34+ cells are known as hematopoietic progenitor cells. The cells were separated by means of immunomagnetic selection with anti-CD34 antibodies. For sorting, a SPIO core coated with a polysaccharide that is linked to an antibody is bound to the respective cell (Fig. 7A). The size of the label is comparable to commonly used superparamagnetic MR contrast agents; thus it can provide sufﬁcient contrast on MR images. Human CD34+ cells from peripheral blood were selected by CliniMACS CD34 Selection Technology (Miltenyi). On EM images, we determined that after cryopreservation, the nanoparticles remained bound to the cell surface, and we observed several iron labels attached to the cell surface (Fig. 7B). Puriﬁed CD34+ cells were implanted intracerebrally into rats with a cortical

378

Fig. 7. (A) Scheme of MicroBeads. A superparamagnetic iron oxide core coated with a polysaccharide is linked to an antibody, which in turn is bound to a cell via an antigen–antibody complex. (B) Transmission electron microphotograph of MicroBeads binding to the cell surface. (C) Human cells (positive staining for human nuclei) found in the lesion 4 weeks after the grafting of CD34+ cells. (D) Staining with anti-CD34 revealed CD34+ cells in the subventricular zone. (Adapted with permission from Jendelova et al. (2005).)

photochemical lesion, contralaterally to the lesion (Jendelova et al., 2005; Sykova and Jendelova, 2006). The average iron content per cell, determined by spectrometry, was 0.275 pg. This value is lower by two orders of magnitude than in the case of cell labeling using Endorem or PC-SPIO, which enter the cell (17 or 38 pg, respectively); nevertheless, it still provides sufﬁcient MR contrast (Jendelova et al., 2005). The cells were detected as a hypointense spot on T2 weighted images 24 h after grafting. The

hypointensity of the implant slightly decreased during the ﬁrst week and then remained without significant changes for the entire measurement period (4 weeks). During the second week, we observed a weak hypointense signal in the lesion that persisted for the next 2 weeks. Prussian blue and anti-human nuclei staining (Fig. 7C) conﬁrmed the presence of magnetically labeled human cells in the corpus callosum and in the lesion. CD34+ cells were detected not only in the corpus callosum and in the lesion, but also in the subventricular

379

zone (Fig. 7D). Finally, human DNA was detected in the lesioned brain tissue using polymerase chain reaction (PCR), conﬁrming the presence of human cells.

therefore serve as an alternative to the conventional grafting of dissociated cells, beneﬁting from advances in surface chemistry and the cell–cell or cell–matrix interactions that occur during development or regeneration.

Labeling of stem cells in polymer hydrogels as stem cell carriers

Discussion and conclusion

In the case of large lesions, cells alone are not able to repair the injury. It is necessary to bridge the gap left by the lost cell population in order to provide support for tissue restoration, reduce the glial scar, and create a permissive environment for cellular ingrowth and for the diffusion of neuroactive molecules. To bridge such a lesion, we have used biocompatible polymer hydrogels based on poly-hydroxypropylmethacrylamide (HPMA) in combination with stem cell grafting. Before transplantation into a lesion, the hydrogels were seeded in vitro with MSCs. In this case the hydrogels form an inert environment, allowing for the free diffusion of intrinsic growth factors, in which the cells start to differentiate and migrate. The inert and biocompatible environment of the hydrogels also provides an adequate standard background for MR imaging of the cells (Sykova and Jendelova, 2005; Sykova et al., 2006). We employed these cell-polymer constructs in order to facilitate the regeneration of injured spinal cord (Lesny et al., 2006). The right half of a spinal cord segment was removed by hemisection, and a block of HPMA hydrogel seeded with Endorem-labeled MSCs was inserted (Fig. 8A; Sykova and Jendelova, 2005). Six weeks after implantation, the hydrogel had formed a continuous bridge between the hemisected spinal segments, reestablishing the anatomical continuity of the tissue. The hydrogel was visible on MR images as a hypointense area (Fig. 8B), and Prussian blue staining conﬁrmed positively stained cells within the hydrogel (Fig. 8C). Immunostaining for blood vessels (Reca-1, Abcam, UK) conﬁrmed neovascularization of the hydrogel implant (Fig. 8D). Staining for neuroﬁlaments (NF160, Sigma, St. Louis, USA) showed axonal ingrowth into the hydrogel (Fig. 8E; Sykova and Jendelova, 2005; Sykova et al., 2006). Hydrogels seeded with stem cells may

For clinical studies it is obvious that traditional histopathological methods for cell detection are not sufﬁcient to inform us about the migration and fate of the grafted cells in the host tissue. Because of its high spatial resolution, MR imaging is suitable for imaging the distribution of magnetically labeled cells. Labeling methods that use the internalization of iron oxide nanoparticles have some limitations. To label cells with commercially available contrast agents, such as Endorem, a relatively high concentration of iron in the culture media is necessary (Jendelova et al., 2003, 2004). This may cause a transient drop in cell viability, which can be dependent on the type of labeled cell. For improved uptake of magnetic nanoparticles, the surface of the contrast agent needs to be optimized, so that it can induce the internalization of the particles into the cytoplasm. One approach is to use internalizing mabs. The mouse anti-Tfr mab OX-26 induces the internalization of Tft upon binding. Nanoparticles have been conjugated to OX-26 to deliver them to cells by receptor-mediated endocytosis (Bulte et al., 1999). A disadvantage of the use of internalizing mabs, however, is that they are species-speciﬁc, and a newly synthesized antibody is required when performing studies in a different animal. In addition, for eventual clinical use, there will be regulatory issues regarding the use of a xenogeneic (i.e., mouse) protein. Recently, a combination of dextran-coated SPIO nanoparticles with a transfection agent has been used (Kalish et al., 2003). When the complexes are added to a cell culture, the transfection agent effectively transports the nanoparticles into the cells through electrostatic interactions. However, each combination of transfection agent and iron oxide nanoparticle has to be carefully titrated and optimized for different cell cultures, since lower concentrations may result in insufﬁcient cellular uptake, whereas higher concentrations may

380

Fig. 8. (A) Endorem-labeled cells seeded into a hydrogel. (B) On MR images 6 weeks after implantation, the hydrogel was visible as a hypointensive area. (C) Prussian blue staining conﬁrmed the presence of Endorem-labeled cells in the hydrogel. (D) Neovascularization of the implant (Reca-1). (E) Ingrowth of NF160-positive axons into the gel.

381

induce the precipitation of complexes or may be toxic to the cells. In addition, although it has been reported that a poly-L-lysin-Feridex labeling complex does not appear to affect the viability or proliferation of human MSCs, it was found that their differentiation into chondrocytes was markedly inhibited (Kostura et al., 2004), while adipogenic and osteogenic differentiation were not affected. PCSPIO combine a low concentration of iron in the cell culture media with the high efﬁciency of transfection agents and thus have a broad potential for in vivo studies. However, for any intracellular label, detailed studies examining the possible biological side effects, which may vary among different cell types, are necessary. On the other hand, particles that do not internalize do not cause any metabolic alternations and do not have any inﬂuence on cell viability or cell proliferation (Jendelova et al., 2005). Their disadvantage is that particles that do not internalize and thus stay attached to the outer cell membrane are more likely to interfere with cell surface interactions (including cell homing into tissues), may detach easily from the membrane, or may be transferred to other cells. MR tracking may serve in experimental models to study how certain lesions target cell migration, at what speed the cells migrate, and for how long they persist in the target organ. High contrast effects on MR images are easily detected within an experimental time frame of 1–2 h per animal, which is ideal for short and repetitive in vivo MRI. In lesioned tissue, hemorrhage products give rise to Prussian blue-positive deposits that are difﬁcult to distinguish from iron-containing nanoparticles (Urdzikova et al., 2006). The hemorrhage degradation products may also be partially localized to macrophages because macrophages constitute the major cellular pathway for the redistribution of iron in mammals. Furthermore, hemorrhage contributes to T2 weighted hypointensity, thus interfering with the detection of labeled cells and complicating the interpretation of MR images. Therefore, it is important to determine whether cell labels remain co-localized with cell transplants, especially under pathological conditions. With proper attention to the limitations described above, labeling cells with superparamagnetic

agents would enable us to follow the migration of such cells when transplanted into humans, establish the optimal number of transplanted cells, deﬁne therapeutic windows, and monitor cell growth and possible side effects (malignancies). Currently, the described immunolabeling of speciﬁc cell types with clinically approved MicroBeads may help to elucidate the fate of implanted stem cells and evaluate the effect of cell therapy in patients with various diseases of the brain or SCI. Abbreviations BBB BMCs BrdU CNS DMEM eGFP EM ESCs G-CSF HPMA LIF Mab MCAO MR MRI MSCs OECs PCR PC-SPIO SCI SPIO USPIO

Basso-Beattie-Bresnehan bone marrow cells bromdeoxyuridine central nervous system Dulbecco’s modiﬁed Eagle medium enhanced green ﬂuorescent protein electron microscopy embryonic stem cells granulocyte colony stimulating factor hydroxypropylmethacrylamide leukemia inhibitory factor monoclonal antibody middle cerebral artery occlusion magnetic resonance magnetic resonance imaging mesenchymal stem cells olfactory ensheathing cells polymerase chain reaction polycation-bound iron oxide superparamagnetic nanoparticles spinal cord injury superparamagnetic iron oxide ultrasmall superparamagnetic iron oxide

Acknowledgment This work was supported by grants from the Academy of Sciences of the Czech Republic AV0Z50390512, the Ministry of Education, Youth, and Sports of the Czech Republic

382

1M0021620803, the National Grant Agency of the Czech Republic GACR 309/06/1594, and the ECFP6 project DiMI: LSHB-CT-2005-512146.

References Akiyama, Y., Radtke, C., Honmou, O. and Kocsis, J.D. (2002) Remyelination of the spinal cord following intravenous delivery of bone marrow cells. Glia, 39(3): 229–236. Anderson, S.A., Glod, J., Arbab, A.S., Noel, M., Ashari, P., Fine, H.A. and Frank, J.A. (2005) Noninvasive MR imaging of magnetically labeled stem cells to directly identify neovasculature in a glioma model. Blood, 105(1): 420–425. Arbab, A.S., Yocum, G.T., Wilson, L.B., Parwana, A., Jordan, E.K., Kalish, H. and Frank, J.A. (2004) Comparison of transfection agents in forming complexes with ferumoxides, cell labeling efﬁciency, and cellular viability. Mol. Imaging, 3(1): 24–32. Billotey, C., Wilhelm, C., Devaud, M., Bacri, J.C., Bittoun, J. and Gazeau, F. (2003) Cell internalization of anionic maghemite nanoparticles: quantitative effect on magnetic resonance imaging. Magn. Reson. Med., 49(4): 646–654. Bjorklund, L.M., Sanchez-Pernaute, R., Chung, S., Andersson, T., Chen, I.Y., McNaught, K.S., Brownell, A.L., Jenkins, B.G., Wahlestedt, C., Kim, K.S. and Isacson, O. (2002) Embryonic stem cells develop into functional dopaminergic neurons after transplantation in a Parkinson rat model. Proc. Natl. Acad. Sci. U.S.A., 99(4): 2344–2349. Brustle, O., Jones, K.N., Learish, R.D., Karram, K., Choudhary, K., Wiestler, O.D., Duncan, I.D. and McKay, R.D. (1999) Embryonic stem cell-derived glial precursors: a source of myelinating transplants. Science, 285(5428): 754–756. Bulte, J.W., Ben-Hur, T., Miller, B.R., Mizrachi-Kol, R., Einstein, O., Reinhartz, E., Zywicke, H.A., Douglas, T. and Frank, J.A. (2003) MR microscopy of magnetically labeled neurospheres transplanted into the Lewis EAE rat brain. Magn. Reson. Med., 50(1): 201–205. Bulte, J.W., Douglas, T., Witwer, B., Zhang, S.C., Strable, E., Lewis, B.K., Zywicke, H., Miller, B., van Gelderen, P., Moskowitz, B.M., Duncan, I.D. and Frank, J.A. (2001) Magnetodendrimers allow endosomal magnetic labeling and in vivo tracking of stem cells. Nat. Biotechnol., 19: 1141–1147. Bulte, J.W., Hoekstra, Y., Kamman, R.L., Magin, R.L., Webb, A.G., Briggs, R.W., Go, K.G., Hulstaert, C.E., Miltenyi, S., The, T.H., et al. (1992) Speciﬁc MR imaging of human lymphocytes by monoclonal antibody-guided dextranmagnetite particles. Magn. Reson. Med., 25(1): 148–157. Bulte, J.W. and Kraitchman, D.L. (2004) Iron oxide MR contrast agents for molecular and cellular imaging. NMR Biomed., 17(7): 484–499. Bulte, J.W., Zhang, S.C., van Gelderen, P., Herynek, V., Jordan, E.K., Duncan, I.D. and Frank, J.A. (1999) Neurotransplantation of magnetically labeled oligodendrocyte progenitors: magnetic resonance tracking of cell migration

and myelination. Proc. Natl. Acad. Sci. U.S.A., 96: 15256–15261. Bulte, J.W., Zhang, S.C., van Gelderen, P., Herynek, V., Jordan, E.K., Janssen, C.H., Duncan, I.D. and Frank, J.A. (2002) Magnetically labeled glial cells as cellular MR contrast agents. Acad. Radiol., 9(Suppl 1): S148–S150. Emsley, J.G., Mitchell, B.D., Kempermann, G. and Macklis, J.D. (2005) Adult neurogenesis and repair of the adult CNS with neural progenitors, precursors, and stem cells. Prog. Neurobiol., 75(5): 321–341. Fairless, R. and Barnett, S.C. (2005) Olfactory ensheathing cells: their role in central nervous system repair. Int. J. Biochem. Cell. Biol., 37(4): 693–699. Frank, J.A., Miller, B.R., Arbab, A.S., Zywicke, H.A., Jordan, E.K., Lewis, B.K., Bryant Jr., L.H. and Bulte, J.W. (2003) Clinically applicable labeling of mammalian and stem cells by combining superparamagnetic iron oxides and transfection agents. Radiology, 228(2): 480–487. Hawrylak, N., Ghosh, P., Broadus, J., Schlueter, C., Greenough, W.T. and Lauterbur, P.C. (1993) Nuclear magnetic resonance (NMR) imaging of iron oxide-labeled neural transplants. Exp. Neurol., 121: 81–92. Horak, D., Sykova, E., Babic, M., Jendelova, P. and Hajek, M. (2006) Superparamagnetic iron oxide nanoparticles with modiﬁed surface and their use as probes for stem cell labeling. Patent application no: PV 1006–120, Czech Republic. Jendelova, P., Herynek, V., De Croos, J., Glogarova, K., Andersson, B., Hajek, M. and Sykova, E. (2003) Imaging the fate of implanted bone marrow stromal cells labeled with superparamagnetic nanoparticles. Magn. Reson. Med., 50: 767–776. Jendelova, P., Herynek, V., Urdzikova, L., Glogarova, K., Kroupova, J., Bryja, V., Andersson, B., Burian, M., Hajek, M. and Sykova, E. (2004) MR tracking of transplanted bone marrow and embryonic stem cells labeled by iron oxide nanoparticles in rat brain and spinal cord. J. Neurosci. Res., 76(2): 232–243. Jendelova, P., Herynek, V., Urdzikova, L., Glogarova, K., Rahmatova, S., Fales, I., Andersson, B., Prochazka, P., Zamecnik, J., Eckschlager, T., Kobylka, P., Hajek, M. and Sykova, E. (2005) MR tracking of human CD34+ progenitor cells separated by means of immunomagnetic selection and transplanted into injured rat brain. Cell Transplant., 14: 173–182. Jiang, Y., Jahagirdar, B.N., Reinhardt, R.L., Schwartz, R.E., Keene, C.D., Ortiz-Gonzalez, X.R., Reyes, M., Lenvik, T., Lund, T., Blackstad, M., Du, J., Aldrich, S., Lisberg, A., Low, W.C., Largaespada, D.A. and Verfaillie, C.M. (2002) Pluripotency of mesenchymal stem cells derived from adult marrow. Nature, 418(6893): 41–49. Kalish, H., Arbab, A.S., Miller, B.R., Lewis, B.K., Zywicke, H.A., Bulte, J.W., Bryant Jr., L.H. and Frank, J.A. (2003) Combination of transfection agents and magnetic resonance contrast agents for cellular imaging: relationship between relaxivities, electrostatic forces, and chemical composition. Magn. Reson. Med., 50(2): 275–282. Kostura, L., Kraitchman, D.L., Mackay, A.M., Pittenger, M.F. and Bulte, J.W. (2004) Feridex labeling of mesenchymal stem

383 cells inhibits chondrogenesis but not adipogenesis or osteogenesis. NMR Biomed., 17(7): 513–517. Lee, I.H., Bulte, J.W., Schweinhardt, P., Douglas, T., Trifunovski, A., Hofstetter, C., Olson, L. and Spenger, C. (2004) In vivo magnetic resonance tracking of olfactory ensheathing glia grafted into the rat spinal cord. Exp. Neurol., 187(2): 509–516. Lesny, P., Pradny, M., Jendelova, P., Michalek, J., Vacik, J. and Sykova, E. (2006) Macroporous hydrogels based on 2-hydroxyethyl methacrylate. Part 4: growth of rat bone marrow stromal cells in three-dimensional hydrogels with positive and negative surface charges and in polyelectrolyte complexes. J. Mater. Sci. Mater. Med., 17: 829–833. McDonald, J.W., Liu, X.Z., Qu, Y., Liu, S., Mickey, S.K., Turetsky, D., Gottlieb, D.I. and Choi, D.W. (1999) Transplanted embryonic stem cells survive, differentiate and promote recovery in injured rat spinal cord. Nat. Med., 5(12): 1410–1412. McKay, R.D. (2004) Stem cell biology and neurodegenerative disease. Philos. Trans. R. Soc. Lond. B. Biol. Sci., 359(1445): 851–856. Mendonca Dias, M.H. and Lauterbur, P.C. (1986) Ferromagnetic particles as contrast agents for magnetic resonance imaging of liver and spleen. Magn. Reson. Med., 3(2): 328–330. Newman, M.B., Davis, C.D., Borlongan, C.V., Emerich, D. and Sanberg, P.R. (2004) Transplantation of human umbilical cord blood cells in the repair of CNS diseases. Expert Opin. Biol. Ther., 4(2): 121–130. Norman, A.B., Thomas, S.R., Pratt, R.G., Lu, S.Y. and Norgren, R.B. (1992) Magnetic resonance imaging of neural transplants in rat brain using a superparamagnetic contrast agent. Brain Res., 594: 279–283. Orlic, D., Kajstura, J., Chimenti, S., Bodine, D.M., Leri, A. and Anversa, P. (2001a) Transplanted adult bone marrow cells repair myocardial infarcts in mice. Ann. N.Y. Acad. Sci., 938: 221–230. Orlic, D., Kajstura, J., Chimenti, S., Jakoniuk, I., Anderson, S.M., Li, B., Pickel, J., McKay, R., Nadal-Ginard, B., Bodine, D.M., Leri, A. and Anversa, P. (2001b) Bone marrow cells regenerate infarcted myocardium. Nature, 410(6829): 701–705. Pachernik, J., Bryja, V., Esner, M., Kubala, L., Dvorak, P. and Hampl, A. (2005) Neural differentiation of pluripotent mouse embryonal carcinoma cells by retinoic acid: inhibitory effect of serum. Physiol. Res., 54(1): 115–122. Park, K.I., Lachyankar, M., Nissim, S. and Snyder, E.Y. (2002) Neural stem cells for CNS repair: state of the art and future directions. Adv. Exp. Med. Biol., 506(Pt B): 1291–1296. Pittenger, M.F., Mackay, A.M., Beck, S.C., Jaiswal, R.K., Douglas, R., Mosca, J.D., Moorman, M.A., Simonetti, D.W., Craig, S. and Marshak, D.R. (1999) Multilineage potential of adult human mesenchymal stem cells. Science, 284(5411): 143–147.

Pluchino, S., Zanotti, L., Deleidi, M. and Martino, G. (2005) Neural stem cells and their use as therapeutic tool in neurological disorders. Brain Res. Brain Res. Rev., 48(2): 211–219. Renshaw, P.F., Owen, C.S., McLaughlin, A.C., Frey, T.G. and Leigh Jr., J.S. (1986) Ferromagnetic contrast agents: a new approach. Magn. Reson. Med., 3(2): 217–225. Roitberg, B. (2004) Transplantation for stroke. Neurol. Res., 26(3): 256–264. Sasaki, M., Honmou, O., Akiyama, Y., Uede, T., Hashi, K. and Kocsis, J.D. (2001) Transplantation of an acutely isolated bone marrow fraction repairs demyelinated adult rat spinal cord axons. Glia, 35: 26–34. Sykova, E. and Jendelova, P. (2005) Magnetic resonance tracking of implanted adult and embryonic stem cells in injured brain and spinal cord. Ann. N.Y. Acad. Sci., 1049: 146–160. Sykova, E. and Jendelova, P. (2006) Magnetic resonance tracking of transplanted stem cells in rat brain and spinal cord. Neurodegener. Dis., 3: 62–67. Sykova, E., Jendelova, P., Urdzikova, L., Lesny, P. and Hejcl, A. (2006) Bone marrow stem cells and polymer hydrogels: two strategies for spinal cord injury repair. Cell. Mol. Neurobiol., 26(7–8): 1111–1127. Takahashi, T., Kalka, C., Masuda, H., Chen, D., Silver, M., Kearney, M., Magner, M., Isner, J.M. and Asahara, T. (1999) Ischemia- and cytokine-induced mobilization of bone marrow-derived endothelial progenitor cells for neovascularization. Nat. Med., 5(4): 434–438. Toma, C., Pittenger, M.F., Cahill, K.S., Byrne, B.J. and Kessler, P.D. (2002) Human mesenchymal stem cells differentiate to a cardiomyocyte phenotype in the adult murine heart. Circulation, 105(1): 93–98. Uccelli, A., Zappia, E., Benvenuto, F., Frassoni, F. and Mancardi, G. (2006) Stem cells in inﬂammatory demyelinating disorders: a dual role for immunosuppression and neuroprotection. Expert Opin. Biol. Ther., 6(1): 17–22. Urdzikova, L., Jendelova, P., Glogarova, K., Burian, M., Hajek, M. and Sykova, E. (2006) Transplantation of bone marrow stem cells as well as mobilization by granulocytecolony stimulating factor promote recovery after spinal cord injury in rat. J. Neurotrauma, 23: 1379–1391. Zhang, Z.G., Jiang, Q., Zhang, R., Zhang, L., Wang, L., Arniego, P., Ho, K.L. and Chopp, M. (2003) Magnetic resonance imaging and neurosphere therapy of stroke in rat. Ann. Neurol., 53(2): 259–263. Zhang, Z.Y. and Smith, B.D. (2000) High-generation polycationic dendrimers are unusually effective at disrupting anionic vesicles: membrane bending model. Bioconjug. Chem., 11(6): 805–814. Zhao, C., Fancy, S.P., Magy, L., Urwin, J.E. and Franklin, R.J. (2005) Stem cells, progenitors and myelin repair. J. Anat., 207(3): 251–258.

This page intentionally left blank

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 27

Intrathecal drug delivery strategy is safe and efﬁcacious for localized delivery to the spinal cord Molly S. Shoichet1,2,3,, Charles H. Tator4, Peter Poon1,2, Catherine Kang1,2 and M. Douglas Baumann1,2 1

Department of Chemical Engineering and Applied Chemistry, University of Toronto, Donnelly Center for Cellular and Biomolecular Research, Toronto, ON, M5S 3E1, Canada 2 Institute of Biomaterials and Biomedical Engineering, 164 College St., Toronto, ON, M5S 3G9, Canada 3 Department of Chemistry, University of Toronto, 80 St. George Street, Toronto, ON, M5S 3H6, Canada 4 Krembil Neuroscience Centre, Toronto Western Research Institute, and Department of Surgery, University of Toronto, 399 Bathurst Street, Toronto, ON, M5 T 2S8, Canada

Abstract: Neuroprotective and neuroregenerative strategies for spinal cord injury repair are limited in part by poor delivery techniques. A novel drug delivery system is being developed in our laboratory that can provide localized release of therapeutically relevant molecules from an injectable hydrogel. Design criteria were established for the hydrogel to be — injectable, fast-gelling, biocompatible, biodegradable and able to release biologically active therapeutics when injected into the intrathecal space that surrounds the spinal cord. This novel way of localized drug delivery to the spinal cord was tested ﬁrst with a collagen gel and then with a new hydrogel blend of hyaluronan and methylcellulose (HAMC). The underlying principle that this novel methodology is both safe and able to provide localized delivery was proven with a fast gelling collagen solution. Using a recombinant human epidermal growth factor, rhEGF, dispersed in collagen, we demonstrated localized release to the injured spinal cord. We extended this technology to other fast-gelling systems and found that HAMC was injectable due to the shear thinning property of hyaluronan (HA), biocompatible and had some therapeutic beneﬁt when injected into the intrathecal space using a compression injury model in rats. Keywords: spinal cord injury; localized drug delivery; hydrogels; hyaluronan; methyl cellulose; collagen; epidermal growth factor therapeutic strategies that have been investigated hold great promise (Gorio et al., 2002), yet systemic delivery may cause side effects, and many promising therapeutic proteins degrade when delivered systemically. Moreover, many therapeutic molecules are unable to cross the blood-spinal cord barrier (BSCB). These difﬁculties suggest that local delivery strategies are required. Two intrathecal techniques have been used to test localized delivery: (1) bolus injection, the effects of which are short-lived because the therapeutic

Introduction Of the several therapeutic strategies investigated for spinal cord injury repair (Pearse et al., 2004; Tsai et al., 2004; Ramer et al., 2005), only systemic delivery of methylprednisolone (MP) is used clinically; however, results from its clinical trial have been openly criticized (Hurlbert, 2001). Other Corresponding author. Tel.: +1-416-978-1460; Fax: +1-416-

978-4317; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61027-3

385

386

agent is washed away by the cerebrospinal ﬂuid (CSF) ﬂow (Terada et al., 2001; Yaksh et al., 2004); and (2) mini-pump delivery, which is invasive and can lead to complications due to infection and/or blockage of the catheter (Jones and Tuszynski, 2001). A third extradural delivery strategy is being investigated clinically for the delivery of rho-kinase inhibitors; however, in this methodology where the therapeutic molecule is applied in a ﬁbrin glue to the external surface of the dura, the molecule must diffuse across the dura, the intrathecal space and then the spinal cord, resulting in only a very small fraction of the molecule reaching the target tissue (Dergham et al., 2002; McKerracher and Higuchi, 2006).

Injectable delivery strategy Given the limitations of current clinical strategies, we sought to develop an improved localized delivery strategy. As with any localized strategy, the therapeutic beneﬁts should include decreased systemic toxicity and decreased dose applied. In the proposed model, we sought to develop a delivery strategy that would localize the therapeutic molecules at the site of injection in the intrathecal space (or subarachnoid space) to localize delivery to the injured spinal cord. To this end, our goal was to develop a biocompatible, biodegradable, injectable, fast gelling polymer within which therapeutic molecules could be dispersed and released. We aimed to inject the drug delivery system via a 30 gauge needle into the intrathecal cavity and have it gel quickly (less than 1 min) in order to localize release at this site (see Fig. 1 for conceptual understanding of new strategy). In the ﬁrst set of studies we aimed to demonstrate safety because this was a completely new and novel approach toward drug delivery to the spinal cord. Several questions arose in the development of this system, such as: What would happen if we injected a polymeric hydrogel into the intrathecal space? Would this strategy cause any harm to either non-injured or spinal cord injured rats? Given that the volume of CSF in the adult rat is 250 ml, we chose to investigate the injection of 20 ml, which represents less than 10% of the total

CSF volume, a volume which we expected to be tolerated based on previous studies where liquids of a similar volume had been injected (Rieselbach et al., 1962). All animal procedures were performed in accordance with the Guide to the Care and Use of Experimental Animals (Canadian Council on Animal Care) and protocols were approved by the Animal Care Committee of the Research Institute of the University Health Network. Importantly, injection into the intrathecal space of both non-injured and spinal cord injured rats was tolerated and had no deleterious affects relative to the injection of an artiﬁcial CSF (aCSF). The safety of the delivery strategy methodology was proven with a fast gelling collagen solution (Jimenez Hamann et al., 2003) that was injected into adult rats either non-injured or after acute spinal cord compression injury. The compression injury was achieved at thoracic level 2 (T2) using a modiﬁed aneurysm clip with a closing force ranging from 35 to 56 g, corresponding to injuries ranging from moderate to severe (Rivlin and Tator, 1977, 1978). Safety was assessed by histology, immunohistochemistry and open-ﬁeld motor function using the Basso, Beattie and Bresnahan (BBB) scoring scale (Basso et al., 1996). This is a 21-point scale that ranks no locomotion at 0 and normal gait at 21. Neither healthy nor injured spinal cord tissue was affected by this injection strategy involving collagen gel. Similarly, locomotor function was unaffected.

Localized release of bioactive factors Having demonstrated the safety of this new intrathecal delivery strategy (Jimenez Hamann et al., 2003), our next goal was to demonstrate localized delivery and efﬁcacy. Based on previous research which demonstrated that mini-pump delivery into the subarachnoid space of EGF and FGF2 stimulated endogenous stem cell proliferation (Kojima and Tator, 2000), we chose to study the effects of their delivery by the injectable collagen system. Using recombinant human (rh) EGF and rhFGF2, we ﬁrst investigated the bioactivity and release proﬁle of each factor from the collagen gel in vitro. Released EGF was bioactive for 14 days and

387

Fig. 1. Injectable delivery strategy to the intrathecal space of the spinal cord. A ﬁne, 30 gauge needle pierces the dura of the spinal cord and is used to deliver a fast gelling, biodegradable, biocompatible hydrogel for localized delivery of therapeutic molecules to the injured spinal cord. (Adapted with permission, copyright Michael Corrin (2005).)

FGF2 for 56 days. Conﬁdent that our system allowed release of bioactive factors, we investigated the release of these factors in vivo (in the presence of heparin as a stabilizer). Using a similar animal model — 35 g clip compression injury at T2 — we compared the delivery of rhEGF/rhFGF2 vs. aCSF. We had determined that there was no cross-reactivity between: (1) the rhEGF and the endogenous EGF in the rat; and (2) the rhFGF2 and endogenous FGF2, so that we could follow the depth of penetration into the tissue by immunohistochemistry. We determined that rhEGF penetrated deeply into the injured spinal cord tissue as evident at 30 min and 6 h after injection of the collagen gel. At 1 and 7 days, it was difﬁcult to detect the rhEGF in the injured spinal cord due to either a fast release from the collagen gel or the limited sensitivity of the immunohistochemical method. Interestingly, in non-injured controls, rhEGF penetrated the spinal cord tissue, but the penetration was superﬁcial and not deep. The greater depth of penetration observed in the injured tissue likely reﬂects its greater permeability. Figure 2 summarizes the depth of rhEGF penetration observed in the injured and non-injured

spinal cord. Unlike rhEGF, rhFGF2 did not penetrate the injured (or non-injured) spinal cord tissue beyond the pia. It is not clear why EGF penetrated the tissue and FGF2 did not; however, this may reﬂect the greater molecular weight of FGF2 relative to EGF or its charge. Importantly, release of EGF/FGF2 from the collagen did stimulate the ependymal cells that line the central canal, which is believed to be the source of endogenous stem cells (Martens et al., 2002). There was also evidence of reduced cavitation around the site of injury in those animals injected with EGF/FGF2 relative to aCSF; however, there was no improvement in function. Stimulation of endogenous stem cells was also observed in studies where EGF/FGF2 was delivered by injection into the ventricles in the brain (Martens et al., 2002). The novel injectable drug delivery system was shown to be safe and to localize release of these agents (Jimenez Hamann et al., 2005). This provided a new paradigm for drug delivery to the spinal cord that we have continued to investigate. However, cellular build-up in the intrathecal space was observed when EGF/FGF2 was incorporated into the collagen gel, likely attracting ﬁbroblasts

388

Fig. 2. Injection of rhEGF via the fast-gelling collagen gel to the (A) uninjured and (B) injured spinal cord demonstrates localized delivery. Using immunohistochemical labeling against rhEGF, we observed a superﬁcial penetration of rhEGF into the uninjured spinal cord (A) and deep penetration into the injured spinal cord (B) as demarcated by the dotted lines. Penetration of rhEGF is observed after 30 min and 6 h of injection, but not after 1 and 7 days. Figure is at 6 h after injection. (Adapted with permission from Jimenez Hamann et al. (2005).)

and other cell types. To overcome this limitation of the delivery vehicle, it was redesigned to be nonadhesive to cells while still meeting the other design criteria of biocompatibility, biodegradability, injectability, and fast gelation. New fast-gelling blend shows therapeutic promise Of the many available chemical and physical gels available, none met all of our design criteria, requiring us to develop a new gel system. While physical gels are not as stable or robust as chemical gels, we believed that they could be sufﬁciently stable in the intrathecal space. Both methylcellulose and hyaluronan met all of our design criteria, except fast gelation. Methylcellulose (MC) is an inverse gelling polymer that forms a weak gel at 371C in water (Li et al., 2001), but does not gel fast enough for our injectable drug delivery system. Hyaluronan (HA) does not gel on its own. However, a blend of HA and MC (HAMC) gel quickly and in fact form a gel at room temperature, likely due to a ‘‘salting out’’ effect that HA has on MC (Xu et al., 2004). In addition to promoting faster gelation of MC, HA is attractive because it is non-immunogenic, biocompatible (Vercruysse and Prestwich, 1998) and known to promote wound healing by reducing inﬂammation and minimizing tissue adhesion and scar formation (Balazs, 1991). HA is also shear

thinning — that is it ﬂows in the direction of stress (Weiss, 2000) — and it is this unique property that allows HAMC to meet the design criteria of injectable and fast gelling. Prior to initiating the in vivo study in rats, HAMC was evaluated in vitro for degradation and cell adhesion. HAMC was found to degrade (or erode) within 14 days and was non-adhesive to cells. The lack of adhesion of 3T3 ﬁbroblasts to HAMC gels demonstrated the non-adhesive property of HAMC in vitro. Thus this new fast-gelling, injectable HAMC was investigated as a drug delivery vehicle to the injured spinal cord. The ﬁrst test was one of safety. Since this was a new blend, we had to ﬁrst demonstrate that it would not cause any deleterious effects in the rat animal model. Thus the biocompatibility of HAMC within the intrathecal space was examined in vivo in both uninjured and spinal cord injured rat animal models relative to control animals receiving aCSF injections. In this set of studies HAMC was injected at T2 after a moderate clip compression injury using a 35 g clip, similar to the in vivo design with the injectable collagen system. Injection of HAMC in spinal cord injured and uninjured rats were compared to a control injection of aCSF. The only difference other than the choice of material was the injection volume, where 10 ml was used instead of the 20 ml of collagen injected.

389

Fig. 3. Representative histology sections of each experimental group stained with Luxol Fast Blue and counterstained with Hematoxylin and Eosin. Sections showing injection site in dura of: (A) uninjured rat injected with aCSF (arrow points to unsealed dura that was punctured with needle); and (B) uninjured rat injected with HAMC (arrow points to resealed dura). (Scale bar ¼ 200 mm.) (Adapted with permission from Gupta et al. (2006).)

It is important to realize that this new technique requires that the dura be punctured, which clinically can result in serious post-dural puncture headaches. Interestingly, in the spinal cord injured animals that received HAMC, the dura was observed to re-seal whereas no re-sealing was observed in animals that received aCSF (Fig. 3). This key result is critical to the long-term application of this material clinically. After 28 days there was no evidence of scar formation, arachnoiditis, syringomyelia or of HAMC itself in the intrathecal space. The beneﬁts of HAMC were observed at the tissue level and the functional level. Both cavitation volume and inﬂammatory response were less in those animals injected with HAMC relative to aCSF. For example, the cavitation volume in animals injected with aCSF was 52.1720.1 mm3 whereas for HAMC the cavitation volume was 36.676.0 mm3. While not statistically different (at 95% conﬁdence), the trend of reduced cavitation for animals injected with HAMC is important and demonstrates that injection of HAMC is safe and affected neither the normal spinal cord nor the severity of injury. The inﬂammatory response was determined by quantiﬁcation of ED-1 labeled macrophages/microglia based on pixel intensity. For animals injected with aCSF, the area occupied was approximately 3.7971.39 105 mm2

vs. 2.4270.35 105 mm2 for animals injected with HAMC. Thus, the inﬂammatory response at 28 days after injection, was significantly reduced in those animals injected with HAMC relative to aCSF controls (po0.05, n ¼ 6). This was an unexpected, yet beneﬁcial effect of HAMC delivery, potentially due to the beneﬁcial anti-inﬂammatory effects attributed to HA in other tissues (Dougados, 2000; Sheehan et al., 2003; Cabrera et al., 2004; Roth et al., 2005). The re-sealing of the dura provides another advantage over other delivery techniques, such as bolus and intrathecal mini-pump where the dura may not seal at the punctured site, resulting in the potential for CSF leakage and/or infection. The beneﬁt or detriment of the inﬂammatory response is widely debated in the neuroscience community (Schwartz, 2003). For example, Michael Schwartz is investigating the therapeutic beneﬁt of macrophage delivery (Rapalino et al., 1998; Knoller et al., 2005). In the case of HAMC delivery, the reduced inﬂammatory response observed, judged by ED-1 immunohistochemistry, was thought to be beneﬁcial and there was also improved locomotor function as described below. To gain a greater perspective on the safety of HAMC delivery, the locomotor function of animals was scored by the BBB and grid walk

390

A

21 18

BBB Score

15 12 9 6 3 0 0

5

10

15

20

25

30

20

25

30

Days after surgery

B

18 16

# of Foot Falls

14 12 10 8 6 4 2 0 0

5

10

15 Days after Surgery

Fig. 4. (A) BBB scores over 1 month to measure the effect of HAMC on animal function: (&) HAMC uninjured, (W) aCSF uninjured, (’) HAMC injured, and (m) aCSF injured. The two uninjured groups scored normal (21 points) at all times; (B) % foot falls of total steps per grid walk run for (&) HAMC uninjured, (W) aCSF uninjured, (’) HAMC injured, and (m) aCSF injured. The two uninjured groups had virtually zero footfalls at all times. Data is shown as mean7standard deviation (uninjured animals: n ¼ 4; injured animals: n ¼ 8). The asterisk (*) at day 7 in (A) indicates significant difference (p ¼ 0.035). (Adapted with permission from Gupta et al. (2006).)

analysis. As shown in Fig. 4A, uninjured animals injected with HAMC or aCSF showed no difference in locomotor function, demonstrating safety of the new injectable HAMC material and conﬁrming safety of the methodology. Importantly, injection of HAMC in spinal cord injured animals showed a general trend of improved

locomotor function relative to injection of aCSF, and a statistically significant difference was observed at day 7 (p ¼ 0.035). This suggests a mild neuroprotective effect of HAMC (Rabchevsky et al., 1999) possibly due to the decreased inﬂammatory response observed by immunohistochemistry.

391

To conﬁrm the functional beneﬁt observed, grid walk analysis was completed following the methodology described by Metz et al. (2000), where previously trained animals were scored on the number of foot falls that occurred when walking across a horizontal ladder. As shown in Fig. 4B, uninjured animals had nearly perfect scores of zero foot falls throughout a run for both HAMCinjected and aCSF-injected animals, thereby corroborating the normal BBB scores shown (cf. Fig. 4A). It is important to note that uninjured animals often miss 1–2 steps on the horizontal ladder (Metz et al., 2000; Hains et al., 2004), likely due to their carelessness when crossing the grid quickly. For injured animals, the HAMC-injected animals had fewer foot falls than aCSF-injected animals, but the difference between HAMC-injected and aCSFinjected animals was not statistically different. Conclusions and future outlook We have demonstrated a new and safe way to achieve localized delivery to the injured spinal cord. First with collagen and then with HAMC, we demonstrated that the underlying methodology is safe, having neither a negative impact on the tissue nor locomotor function. Moreover, there may be a small therapeutic beneﬁt of HAMC alone, indicating that this material may be useful as a delivery and for its inherent ‘‘wound healing’’ properties. Lastly, we found that this delivery strategy resulted in localized release of growth factors into the injured spinal cord, thereby proving the hypothesis that this strategy could achieve local release into the injured tissue. In on-going studies we continue to examine the therapeutic beneﬁt of the release of neuroprotective and neuroregenerative factors to limit degeneration and promote regeneration, respectively, following traumatic injury to the spinal cord. Localized release from hydrogel systems continues to be explored for its potential therapeutic beneﬁt in spinal cord injury. Acknowledgments We are grateful to Maria Jimenez-Hamann and Dimpy Gupta for initial studies. We thank the

Canadian Institute of Health Research, the Canada Foundation for Innovation and the Ontario Innovation Trust and the Ontario Branch of the Canadian Paraplegic Association for funding. References Balazs, E.A. (1991) Cosmetic and pharmaceutical applications of polymers. In: Gebelein C.G. (Ed.), Medical Applications of Hyaluronan and its Derivatives. Plenum Press, New York, pp. 293–310. Basso, D.M., Beattie, M.S. and Bresnahan, J.C. (1996) Graded histological and locomotor outcomes after spinal cord contusion using the NYU weight-drop device versus transection. Exp. Neurol., 139: 244–256. Cabrera, P.V., Blanco, G., Alaniz, L., Greczanik, S., Garcia, M., Alvarez, E. and Hajos, S.E. (2004) CD44 and hyaluronic acid regulate in vivo iNOS expression and metalloproteinase activity in murine air-pouch inﬂammation. Inflamm. Res., 53: 556–566. Dergham, P., Ellezam, B., Essagian, C., Avedissian, H., Lubell, W.D. and McKerracher, L. (2002) Rho signaling pathway targeted to promote spinal cord repair. J. Neurosci., 22: 6570–6577. Dougados, M. (2000) Sodium hyaluronate therapy in osteoarthritis: arguments for a potential beneﬁcial structural effect. Semin. Arthritis Rheum., 30: 19–25. Gorio, A., Gokmen, N., Erbayraktar, S., Yilmaz, O., Madaschi, L., Cichetti, C., Di Giulio, A.M., Vardar, E., Cerami, A. and Brines, M. (2002) Recombinant human erythropoietin counteracts secondary injury and markedly enhances neurological recovery from experimental spinal cord trauma. Proc. Natl. Acad. Sci. U.S.A., 99: 9450–9455. Gupta, D., Tator, C.H. and Shoichet, M.S. (2006) Fast-gelling injectable blend of hyaluronan and methylcellulose for intrathecal, localized delivery to the injured spinal cord. Biomaterials, 27: 2370–2379. Hains, B.C., Saab, C.Y., Lo, A.C. and Waxman, S.G. (2004) Sodium channel blockade with phenytoin protects spinal cord axons, enhances axonal conduction, and improves functional motor recovery after contusion SCI. Exp. Neurol., 188: 365–377. Hurlbert, R.J. (2001) The role of steroids in acute spinal cord injury: an evidence-based analysis. Spine, 26: S39–S46. Jimenez Hamann, M.C., Tator, C.H. and Shoichet, M.S. (2005) Injectable intrathecal delivery system for localized administration of EGF and FGF-2 to the injured rat spinal cord. Exp. Neurol., 194: 106–119. Jimenez Hamann, M.C., Tsai, E.C., Tator, C.H. and Shoichet, M.S. (2003) Novel intrathecal delivery system for treatment of spinal cord injury. Exp. Neurol., 182: 300–309. Jones, L.L. and Tuszynski, M.H. (2001) Chronic intrathecal infusions after spinal cord injury cause scarring and compression. Microsc. Res. Tech., 54: 317–324. Knoller, N., Auerbach, G., Fulga, V., Zelig, G., Attias, J., Bakimer, R., Marder, J.B., Yoles, E., Belkin, M., Schwartz, M.

392 and Hadani, M. (2005) Clinical experience using incubated autologous macrophages as a treatment for complete spinal cord injury: phase I study results. J. Neurosurg. Spine, 3: 173–181. Kojima, A. and Tator, C.H. (2000) Epidermal growth factor and ﬁbroblast growth factor 2 cause proliferation of ependymal precursor cells in the adult rat spinal cord in vivo. J. Neuropathol. Exp. Neurol., 59: 687–697. Li, L., Thangamathesvaran, P.M., Yue, C.Y., Tam, K.C., Hu, X. and Lam, Y.C. (2001) Gel network structure of methylcellulose in water. Langmuir, 17: 8062–8068. Martens, D.J., Seaberg, R.M. and van der Kooy, D. (2002) In vivo infusions of exogenous growth factors into the fourth ventricle of the adult mouse brain increase the proliferation of neural progenitors around the fourth ventricle and the central canal of the spinal cord. Eur. J. Neurosci., 16: 1045–1057. McKerracher, L. and Higuchi, H. (2006) Targeting Rho to stimulate repair after spinal cord injury. J. Neurotrauma, 23: 309–317. Metz, G.A., Merkler, D., Dietz, V., Schwab, M.E. and Fouad, K. (2000) Efﬁcient testing of motor function in spinal cord injured rats. Brain Res., 883: 165–177. Pearse, D.D., Pereira, F.C., Marcillo, A.E., Bates, M.L., Berrocal, Y.A., Filbin, M.T. and Bunge, M.B. (2004) cAMP and Schwann cells promote axonal growth and functional recovery after spinal cord injury. Nat. Med., 10: 610–616. Rabchevsky, A.G., Fugaccia, I., Fletcher-Turner, A., Blades, D.A., Mattson, M.P. and Scheff, S.W. (1999) Basic ﬁbroblast growth factor (bFGF) enhances tissue sparing and functional recovery following moderate spinal cord injury. J. Neurotrauma, 16: 817–830. Ramer, L.M., Ramer, M.S. and Steeves, J.D. (2005) Setting the stage for functional repair of spinal cord injuries: a cast of thousands. Spinal Cord, 43: 134–161. Rapalino, O., Lazarov-Spiegler, O., Agranov, E., Velan, G.J., Yoles, E., Fraidakis, M., Solomon, A., Gepstein, R., Katz, A., Belkin, M., Hadani, M. and Schwartz, M. (1998) Implantation of stimulated homologous macrophages results in partial recovery of paraplegic rats. Nat. Med., 4: 814–821. Rieselbach, R.E., Di Chiro, G., Freireich, E.J. and Rall, D.P. (1962) Subarachnoid distribution of drugs after lumbar injection. N. Engl. J. Med., 267: 1273–1278.

Rivlin, A.S. and Tator, C.H. (1977) Objective clinical assessment of motor function after experimental spinal cord injury in the rat. J. Neurosurg., 47: 577–581. Rivlin, A.S. and Tator, C.H. (1978) Effect of duration of acute spinal cord compression in a new acute cord injury model in the rat. Surg. Neurol., 10: 38–43. Roth, A., Mollenhauer, J., Wagner, A., Fuhrmann, R., Straub, A., Venbrocks, R.A., Petrow, P., Brauer, R., Schubert, H., Ozegowski, J., Peschel, G., Muller, P.J. and Kinne, R.W. (2005) Intra-articular injections of high-molecular-weight hyaluronic acid have biphasic effects on joint inﬂammation and destruction in rat antigen-induced arthritis. Arthritis Res. Ther., 7: R677–R686. Schwartz, M. (2003) Macrophages and microglia in central nervous system injury: are they helpful or harmful? J. Cereb. Blood Flow Metab., 23, 385–394. Sheehan, K.M., DeLott, L.B., Day, S.M. and DeHeer, D.H. (2003) Hyalgan has a dose-dependent differential effect on macrophage proliferation and cell death. J. Orthop. Res., 21: 744–751. Terada, H., Kazui, T., Takinami, M., Yamashita, K., Washiyama, N. and Muhammad, B.A. (2001) J. Thorac. Cardiovasc. Surg., 122: 979–985. Tsai, E.C., Dalton, P.D., Shoichet, M.S. and Tator, C.H. (2004) Synthetic hydrogel guidance channels facilitate regeneration of adult rat brainstem motor axons after complete spinal cord transection. J. Neurotrauma, 21: 789–804. Vercruysse, K.P. and Prestwich, G.D. (1998) Hyaluronate derivatives in drug delivery. Crit. Rev. Ther. Drug Carrier Syst., 15: 513–555. Weiss, C. (2000) New frontiers in medical sciences: redeﬁning Hyaluronan. In: Weigel, G. A. a. P. H. (Eds.), Why Viscoelasticity is Important for the Medical Uses of Hyaluronan and Hylans. Excerpta Medica, pp. 89–103. Xu, Y., Wang, C., Tam, K.C. and Li, L. (2004) Salt-assisted and salt-suppressed sol-gel transitions of methylcellulose in water. Langmuir, 20: 646–652. Yaksh, T.L., Horais, K.A., Tozier, N., Rathbun, M., Richter, P., Rossi, S., Grafe, M., Tong, C., Meschter, C., Cline, J.M. and Eisenach, J. (2004) Toxicol. Sci., 80: 322–334.

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 28

Decompression craniectomy after traumatic brain injury: recent experimental results Nikolaus Plesnila Laboratory of Experimental Neurosurgery, Department of Neurosurgery and Institute for Surgical Research, University of Munich Medical Center, GroX hadern, Marchioninistr 15, 81377 Munich, Germany

Abstract: Among the secondary events occurring after traumatic brain injury (TBI) pathologically increased intracranial pressure (ICP) correlates most closely with poor outcome. In addition to infusion of hypertonic solutions, e.g. mannitol, and other medical measures, decompression of the brain by surgical removal of a portion of the cranium (craniectomy) has been used for many decades as an intuitive strategy for the treatment of post-traumatic ICP increase. The lack of evidence-based clinical and controversial experimental data, however, resulted in decompressive craniectomy to be recommended by most national and international guidelines only as a third tier therapy for the treatment of pathologically elevated ICP. Ongoing clinical trials on the use of decompressive craniectomy after TBI may clarify many aspects of the clinical application of this technique, however, some important pathophysiological issues, e.g. the timing of decompression craniectomy, its effect on brain edema formation, and its role for secondary brain damage, are still widely discussed and can only be addressed in experimental settings. The aim of the current review was therefore to summarize and discuss recent experimental data dealing with the use of decompression craniectomy following TBI. The present results suggest that surgical decompression effectively prevents secondary brain damage when performed early enough. Although caution should be taken when transferring conclusions drawn from experimental settings to the clinical situation, the current literature suggests that the timing of decompression may be of utmost importance in order to exploit the full neuroprotective potential of craniectomy following TBI. Keywords: traumatic brain injury; decompression; craniectomy; intracranial pressure; mice; review

Introduction

History of decompression craniectomy

Despite signiﬁcant improvements in the management of traumatic brain injury (TBI) within the past two decades, brain edema formation resulting in refractory intracranial hypertension remains the leading cause of unfavorable outcome and death among affected patients (Murray et al., 1999).

Opening of the cranial cavity is the most intuitive treatment option for increased intracranial pressure (ICP). Accordingly, as soon as increased ICP was considered an important component of the pathophysiology of TBI, i.e. at the beginning of the last century, surgical decompression was used for the treatment of patients suffering from the sequels of severe TBI. The ﬁrst report on the use of decompression craniectomy following TBI was published by the legendary Swiss surgeon and Noble laureate for medicine Theodor Emil Kocher (1901).

Corresponding author. Tel.: +49-89-2180-76-535; Fax: +49-

89-2180-76-532; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61028-5

393

394

The intense collaboration of Kocher with Harvey Cushing resulted — among others — in the use of decompression craniectomy also for the treatment of other cerebral disorders, e.g., brain tumors and vascular malformations (Cushing, 1905).

Decompression craniectomy after TBI in man Kocher’s initial positive experience with decompression craniectomy was followed by a plethora of controversial clinical ﬁndings on the use of surgical decompression following brain trauma during the past 100 years (Diemath, 1966; Clark et al., 1968; Kerr, 1968; Kjellberg and Prieto, 1971; Ransohoff and Benjamin, 1971; Ogawa et al., 1974; Venes and Collins, 1975; Pereira et al., 1977; Makino and Yamaura, 1979; Yamaura et al., 1979; Gerl and Tavan, 1980). These early reports helped to determine the most effective surgical technique for decompression craniectomy in TBI patients. Another important ﬁnding was that decompression craniectomy, most effectively together with opening of the dura mater, lowers pathologically elevated ICP, a fact conﬁrmed also by more recent studies (Dam et al., 1996; Polin et al., 1997; Yoo et al., 1999; Berger et al., 2002; Schneider et al., 2002; Ruf et al., 2003; Aarabi et al., 2006; Josan and Sgouros, 2006; Sahuquillo and Arikan, 2006). However, despite the positive effect of decompression craniectomy on otherwise uncontrollable intracranial hypertension, the effect of surgical decompression on mortality and overall functional outcome following TBI remained controversial. For example, a recent review of 400 patients with severe TBI reported that craniectomised patients may have beneﬁted from the procedure (Meier and Grawe, 2003). Another study matching TBI patients with craniectomy with non-craniectomised controls from the Traumatic Coma Data Bank reported good recovery or moderate disability in 37% of decompressed patients as compared with only 16% in the historic control group (Polin et al., 1997). Although a number of other studies support these ﬁndings (Kunze et al., 1998; Guerra et al., 1999; De Luca et al., 2000; Kontopoulos et al., 2002; Schneider et al., 2002; Ziai et al., 2003), the general problem of all these investigations is the lack of a prospective,

randomized, and controlled study design and in some cases an insufﬁcient number of observations. In additions to these limitations, there are also a number of publications reporting no beneﬁt from decompressive craniectomy (Clark et al., 1968; Munch et al., 2000; Soukiasian et al., 2002; Messing-Junger et al., 2003). For example, a retrospective study comparing craniectomised patients from one center with a matched control population form the Traumatic Coma Data Bank reported no differences between these two groups regarding ICP, therapy intensity level, and overall mortality (Munch et al., 2000), a ﬁnding also corroborated by others (Soukiasian et al., 2002). Accordingly, despite a report of a randomized trial demonstrating beneﬁcial use of decompression craniectomy in children (Taylor et al., 2001) and clear beneﬁts of surgical decompression in stroke patients (Schwab et al., 1998; Fraser and Hartl, 2005), the current literature does not provide evidence-based proof for the general use of decompression surgery following TBI in adults. As a result current American and European TBI guidelines still recommend decompression craniectomy only as a ‘‘second tier’’ therapeutic option for adult patients with TBI (Maas et al., 1997; The Brain Trauma Foundation. The American Association of Neurological Surgeons. The Joint Section on Neurotrauma and Critical Care, 2000).

Decompression craniectomy after experimental TBI Because heterogeneous patient populations and mechanisms of injury may have hampered the investigation of decompression craniectomy in the clinical environment, many researchers tried to evaluate the therapeutic potential of decompression craniectomy under more homogenous and controlled conditions, i.e. in experimental models of TBI. Most surprisingly also experimental data on the therapeutic use of decompression craniectomy after brain trauma were not univocal (Moody et al., 1968; Cooper et al., 1979; Gaab et al., 1979; Burkert and Plaumann, 1989; Rinaldi et al., 1990). Although most authors agree with the clinical ﬁnding that craniotomy lowers pathologically elevated ICP (Moody et al., 1968; Gaab et al.,

395

1979; Rinaldi et al., 1990), an increase in brain edema formation and enhanced histopathological damage were observed after craniectomy in experimental animals (Moody et al., 1968; Gaab et al., 1979; Rinaldi et al., 1990). Although it is difﬁcult to retrospectively evaluate why experimental studies show such heterogeneous results, the consideration of some pathophysiological concepts not available at the time when these experiments were performed may deepen our understanding on the use of decompression craniectomy after TBI.

Pathophysiological considerations Experimental and clinical ﬁndings demonstrate that the pathophysiology of TBI consists of two components: one is determined by the immediate mechanical damage to the brain by the traumatic insult and the other is determined by a plethora of secondary events on the functional, neurochemical, and molecular level, which ultimately result in additional loss of initially viable brain tissue. While the ﬁrst, or primary injury component occurs within a few milliseconds during the mechanical impact and

32

can only be avoided by prophylactic measures, secondary injury is a dynamic process which develops gradually over several hours (maybe days) and offers therefore the possibility for therapeutic interventions (Murray et al., 1999; Hartl and Ougorets, 2004; Nortje and Menon, 2004). In case of experimental traumatic contusions, as observed after the controlled cortical impact (CCI) brain injury model, these two components can easily be distinguished by standard histopathological techniques, e.g. Nissl staining. Determination of the contusion volume 15 min after injury reveals a value of 20 mm3, while 24 h after TBI the contusion volume plateaus at 32 mm3. This indicates that the primary contused tissue, which is necrotic already minutes after TBI, i.e. it contains only pyknotic nuclei and does not show any perfusion, increases by 60% (Zweckberger et al., 2006). Determination of the contusion volume at intermediate time points, i.e. 2, 6, and 12 h after TBI, shows a fast and linear expansion of contused tissue volume within the ﬁrst 12 h after trauma with only limited additional tissue loss from 12 to 24 h (Fig. 1). Since blood-brain barrier leakage, brain edema formation, and ICP increase show a very similar early time course

: p<0.05 vs 15min; n=7

Contusion volume (mm3)

30 28

+63%

26

+52%

24 +32%

22 20 18 16 0 0

2

6

12

24

Time after trauma (h) Fig. 1. Secondary expansion of the volume of a traumatic cortical contusion after brain injury. As compared with the primary damage measured immediately after the traumatic impact (0 h; controlled cortical impact in mice) secondary processes result in delayed and progressive expansion of the contusion by more than 60% of its initial size during the ﬁrst 24 h after experimental TBI. Adapted with permission from Zweckberger et al. (2006).

396

(Zweckberger et al., 2003, 2006) it seem to be clear that secondary brain damage and its sequel, i.e. contusion expansion, is a linearly ongoing process which starts immediately after TBI and levels off approximately 12 h after injury. The consequence of this ﬁnding is, that the later a therapeutic intervention starts the less effective it will be. In the case of craniectomy this may indicate that delayed initiation of decompression, i.e. at a time point when brain edema and ICP already developed, may not have any beneﬁcial potential. In the contrary, under circumstances of massive brain edema formation and high ICP craniectomy, which among others lowers vascular–interstitial pressure gradients, may also have adverse effects, e.g. an increased risk of internal or external herniation by enhanced vasogenic edema formation with subsequent cerebral ischemia or detrimental brain stem compression, respectively. Evaluating the protocols of the early experimental studies investigating decompression craniectomy following TBI (Moody et al., 1968; Cooper et al., 1979; Gaab et al., 1979; Burkert and Plaumann, 1989; Rinaldi et al., 1990) it becomes clear that, most logically, a clinically relevant treatment scenario was chosen. In parallel to the situation in patients, animals were decompressed several hours after TBI when pathologically elevated ICP already developed. In view of our current understanding of the gradual and irreversible development of secondary brain damage, the point in time of decompression in these studies may well have been chosen too late, i.e. at a time when the vicious circle of secondary brain damage already started and sequels became irreversible. Accordingly, the experimental data available up to the beginning of the new millennium did not necessarily allow to draw deﬁnitive conclusions on the therapeutic potential of decompression craniectomy after TBI.

Experimental results on early decompression after TBI Based on the above-mentioned pathophysiological understanding of the gradual and irreversible progression of parenchymal brain damage following TBI, we designed a series of experiments in a

kinetic experimental TBI model, CCI, investigating the maximal treatment potential of decompression craniectomy. This was achieved by comparing one group of animals which was allowed to develop post-traumatic ICP, while in a second group of animals pathological ICP increase was completely prevented by a large parietal craniectomy (Fig. 2). Under these circumstances the contusion volume of the not craniectomised animals increased by almost 60% within 24 h as compared with the primary damage inﬂicted by the mechanical impact immediately (15 min) after trauma (Fig. 3). In craniectomised animals the secondary contusion expansion was completely blunted thereby demonstrating that decompression craniectomy completely and effectively prevents the secondary loss of brain tissue following TBI. Since two studies from the late 70 s using cold lesion-induced brain injury reported up to sevenfold increase of brain edema volume 8 h following decompressive craniectomy in dogs (Cooper et al., 1979; Gaab et al., 1979), we performed wet–dry ratio measurements for the determination of brain edema in our kinetic TBI model. Brain water content in mice increased dramatically by almost 3% (from 78 to 81%) following CCI. In craniectomised animals surgical decompression prevented brain edema formation by more than 50%. Together with the lack of secondary contusion expansion these ﬁndings indicate that approximately 50% of post-traumatic brain edema formation following CCI are primary, i.e. caused by the initial mechanical impact, while only 50% are secondary and are therefore amenable to therapeutic interventions. Finally, we were interested to evaluate the therapeutic window of decompression craniectomy. Control animals were traumatized but not craniectomised, while treated animals were craniectomised immediately after TBI or with a delay of 1, 3, or 8 h. Contusion volume in not craniectomised mice was over 30 mm3 24 h after TBI indicating progressive secondary brain damage, while animals craniectomised immediately after trauma had contusion volumes of 20 mm3, i.e. comparable with primary contusions. If craniectomy was delayed by 1 or 3 h a signiﬁcant proportion of secondary brain damage was prevented, while decompression 8 h after TBI showed a marginal but statistically not signiﬁcant

397

30

p<0.001 vs. control p<0.001 vs. 0 h (baseline)

Mean +/- SD; n=3

ICP (mmHg)

25 20

Control

15 10 5

Craniectomy

0 0 1 2 3

6

12

24

Time after trauma (h) Fig. 2. Intracranial pressure of craniectomised and not craniectomised mice after experimental TBI (CCI). While intracranial pressure increases signiﬁcantly in not decompressed animals (control; open circles; n ¼ 6), no signiﬁcant ICP increase is observed in animals receiving a large parietal craniectomy immediately after TBI (closed circles; n ¼ 6). Adapted with permission from Zweckberger et al. (2003).

45

Contusion volume (mm3)

40

Mean +/- SD; n=7 p<0.05 vs. Control 24h

Closed skull Craniectomized

35 30 25 20 15 10 5 0

15 min

24 h

24 h

Fig. 3. Effect of decompression craniectomy on the secondary expansion of a traumatic cortical contusion. In not craniectomised mice (gray bars) signiﬁcant secondary expansion of a traumatic contusion (streaked area) is observed (compare also with Fig. 1). In contrast, no secondary lesion expansion is present in craniectomised mice indicating that craniectomy completely prevents secondary brain damage. Adapted with permission from Zweckberger et al. (2006).

effect. These results clearly indicate that the pathophysiology of secondary lesion expansion follows a two step paradigm: a slow starting phase where treatment is effective which is ultimately followed by a self amplifying ﬁnal phase when even the most

effective treatment options, e.g. craniectomy, are not or only marginally effective. Taken together, the recent literature indicates that the timing of craniectomy is of major signiﬁcance for its therapeutic beneﬁt: performed in the

398

presence of massive brain swelling decompression may be detrimental, but when performed early enough craniectomy may completely prevent secondary brain damage (including secondary brain edema formation). Although we are aware of the caution which should be taken when extrapolating experimental data to the human situation one has also to take into consideration that pure physical phenomena like pressure changes in a closed cavity may not be as different between species as compared with other parameters, e.g. gene expression or the initiation of intracellular signal transduction cascades. Accordingly, our results suggest that early decompressive craniectomy may have a clinical potential and should clearly be favored in comparison with delayed surgical decompression, which may even be detrimental. In view of the present data we conclude that timing seems to be of utmost importance and that decompression has to be performed as early as possible in order to exert its therapeutic potential. Abbreviations CCI ICP TBI

controlled cortical impact intracranial pressure traumatic brain injury

Acknowledgments The excellent work of Klaus Zweckberger who performed most experiments investigating the role of decompression craniectomy on secondary brain damage after TBI is highly appreciated. Special thanks also to Dr. Alexander Baethmann, who was always a source of inspiration and help during this project. References Aarabi, B., Hesdorffer, D.C., Ahn, E.S., Aresco, C., Scalea, T.M. and Eisenberg, H.M. (2006) Outcome following decompressive craniectomy for malignant swelling due to severe head injury. J. Neurosurg., 104: 469–479.

Berger, S., Schwarz, M. and Huth, R. (2002) Hypertonic saline solution and decompressive craniectomy for treatment of intracranial hypertension in pediatric severe traumatic brain injury. J. Trauma, 53: 558–563. Burkert, W. and Plaumann, H. (1989) The value of large pressure-relieving trepanation in treatment of refractory brain edema. Animal experiment studies, initial clinical results. Zentralbl. Neurochir., 50: 106–108. Clark, K., Nash, T.M. and Hutchison, G.C. (1968) The failure of circumferential craniotomy in acute traumatic cerebral swelling. J. Neurosurg., 29: 367–371. Cooper, P.R., Hagler, H., Clark, W.K. and Barnett, P. (1979) Enhancement of experimental cerebral edema after decompressive craniectomy: implications for the management of severe head injuries. Neurosurgery, 4: 296–300. Cushing, H. (1905) The establishment of cerebral hernia as a decompressive measure for inaccessible brain tumor: with the description of intramuscular methods of making the bone defect in temporal and occipital regions. Surg. Gynecol. Obstet., 1: 297–314. Dam, H.P., Sizun, J., Person, H. and Besson, G. (1996) The place of decompressive surgery in the treatment of uncontrollable post-traumatic intracranial hypertension in children. Childs Nerv. Syst., 12: 270–275. De Luca, G.P., Volpin, L., Fornezza, U., Cervellini, P., Zanusso, M., Casentini, L., Curri, D., Piacentino, M., Bozzato, G. and Colombo, F. (2000) The role of decompressive craniectomy in the treatment of uncontrollable posttraumatic intracranial hypertension. Acta Neurochir. Suppl., 76: 401–404. Diemath, H.E. (1966) The value of bitemporal relief surgery in severe closed skull brain trauma. Wien. Med. Wochenschr., 116: 1043–1044. Fraser, J.F. and Hartl, R. (2005) Decompressive craniectomy as a therapeutic option in the treatment of hemispheric stroke. Curr. Atheroscler. Rep., 7: 296–304. Gaab, M., Knoblich, O.E., Fuhrmeister, U., Pﬂughaupt, K.W. and Dietrich, K. (1979) Comparison of the effects of surgical decompression and resection of local edema in the therapy of experimental brain trauma. Investigation of ICP, EEG and cerebral metabolism in cats. Childs Brain, 5: 484–498. Gerl, A. and Tavan, S. (1980) Bilateral craniectomy in the treatment of severe traumatic brain edema. Zentralbl. Neurochir., 41: 125–138. Guerra, W.K., Gaab, M.R., Dietz, H., Mueller, J.U., Piek, J. and Fritsch, M.J. (1999) Surgical decompression for traumatic brain swelling: indications and results. J. Neurosurg., 90: 187–196. Hartl, R. and Ougorets, I. (2004) Critical care of neurotrauma. Curr. Neurol. Neurosci. Rep., 4: 481–488. Josan, V.A. and Sgouros, S. (2006) Early decompressive craniectomy may be effective in the treatment of refractory intracranial hypertension after traumatic brain injury. Childs Nerv. Syst., 22: 1268–1274.

399 Kerr, F.W. (1968) Radical decompression and dural grafting in severe cerebral edema. Mayo Clin. Proc., 43: 852–864. Kjellberg, R.N. and Prieto Jr., A. (1971) Bifrontal decompressive craniotomy for massive cerebral edema. J. Neurosurg., 34: 488–493. Kocher, T. (1901) Die Therapie des Hirndruckes. In: Ho¨lder A. (Ed.), Hirnerschu¨tterung, Hirndruck und chirurgische Eingriffe bei Hirnkrankheiten. Ho¨lder Verlag, Wien, pp. 262–266. Kontopoulos, V., Foroglou, N., Patsalas, J., Magras, J., Foroglou, G., Yiannakou-Pephtoulidou, M., Soﬁanos, E., Anastassiou, H. and Tsaoussi, G. (2002) Decompressive craniectomy for the management of patients with refractory hypertension: should it be reconsidered? Acta Neurochir. (Wien.), 144: 791–796. Kunze, E., Meixensberger, J., Janka, M., Sorensen, N. and Roosen, K. (1998) Decompressive craniectomy in patients with uncontrollable intracranial hypertension. Acta Neurochir. Suppl. (Wien.), 71: 16–18. Maas, A.I., Dearden, M., Teasdale, G.M., Braakman, R., Cohadon, F., Iannotti, F., Karimi, A., Lapierre, F., Murray, G., Ohman, J., Persson, L., Servadei, F., Stocchetti, N. and Unterberg, A. (1997) EBIC-guidelines for management of severe head injury in adults. European Brain Injury Consortium. Acta Neurochir. (Wien.), 139: 286–294. Makino, H. and Yamaura, A. (1979) Assessment of outcome following large decompressive craniectomy in management of serious cerebral contusion. A review of 207 cases. Acta Neurochir. Suppl. (Wien.), 28: 193–194. Meier, U. and Grawe, A. (2003) The importance of decompressive craniectomy for the management of severe head injuries. Acta Neurochir. Suppl., 86: 367–371. Messing-Junger, A.M., Marzog, J., Wobker, G., Sabel, M. and Bock, W.J. (2003) Decompressive craniectomy in severe brain injury. Zentralbl. Neurochir., 64: 171–177. Moody, R.A., Ruamsuke, S. and Mullan, S.F. (1968) An evaluation of decompression in experimental head injury. J. Neurosurg., 29: 586–590. Munch, E., Horn, P., Schurer, L., Piepgras, A., Paul, T. and Schmiedek, P. (2000) Management of severe traumatic brain injury by decompressive craniectomy. Neurosurgery, 47: 315–322. Murray, G.D., Teasdale, G.M., Braakman, R., Cohadon, F., Dearden, M., Iannotti, F., Karimi, A., Lapierre, F., Maas, A., Ohman, J., Persson, L., Servadei, F., Stocchetti, N., Trojanowski, T. and Unterberg, A. (1999) The European Brain Injury Consortium survey of head injuries. Acta Neurochir. (Wien.), 141: 223–236. Nortje, J. and Menon, D.K. (2004) Traumatic brain injury: physiology, mechanisms, and outcome. Curr. Opin. Neurol., 17: 711–718. Ogawa, M., Minami, T., Katsurada, K. and Sugimoto, T. (1974) Evaluation of external cranial decompression for traumatic acute brain swelling. Med. J. Osaka Univ., 25: 73–78.

Pereira, W.C., Neves, V.J. and Rodrigues, Y. (1977) Bifrontal decompressive craniotomy as the treatment for severe cerebral edema. Arq. Neuropsiquiatr., 35: 99–111. Polin, R.S., Shaffrey, M.E., Bogaev, C.A., Tisdale, N., Germanson, T., Bocchicchio, B. and Jane, J.A. (1997) Decompressive bifrontal craniectomy in the treatment of severe refractory posttraumatic cerebral edema. Neurosurgery, 41: 84–92. Ransohoff, J. and Benjamin, V. (1971) Hemicraniectomy in the treatment of acute subdural haematoma. J. Neurol. Neurosurg. Psychiatry, 34: 106. Rinaldi, A., Mangiola, A., Anile, C., Maira, G., Amante, P. and Ferraresi, A. (1990) Hemodynamic effects of decompressive craniectomy in cold induced brain oedema. Acta Neurochir. Suppl. (Wien.), 51: 394–396. Ruf, B., Heckmann, M., Schroth, I., Hugens-Penzel, M., Reiss, I., Borkhardt, A., Gortner, L. and Jodicke, A. (2003) Early decompressive craniectomy and duraplasty for refractory intracranial hypertension in children: results of a pilot study. Crit. Care, 7: R133–R138. Sahuquillo, J. and Arikan, F. (2006) Decompressive craniectomy for the treatment of refractory high intracranial pressure in traumatic brain injury. Cochrane Database Syst. Rev., CD003983. Schneider, G.H., Bardt, T., Lanksch, W.R. and Unterberg, A. (2002) Decompressive craniectomy following traumatic brain injury: ICP, CPP and neurological outcome. Acta Neurochir. Suppl., 81: 77–79. Schwab, S., Steiner, T., Aschoff, A., Schwarz, S., Steiner, H.H., Jansen, O. and Hacke, W. (1998) Early hemicraniectomy in patients with complete middle cerebral artery infarction. Stroke, 29: 1888–1893. Soukiasian, H.J., Hui, T., Avital, I., Eby, J., Thompson, R., Kleisli, T., Margulies, D.R. and Cunneen, S. (2002) Decompressive craniectomy in trauma patients with severe brain injury. Am. Surg., 68: 1066–1071. Taylor, A., Butt, W., Rosenfeld, J., Shann, F., Ditchﬁeld, M., Lewis, E., Klug, G., Wallace, D., Henning, R. and Tibballs, J. (2001) A randomized trial of very early decompressive craniectomy in children with traumatic brain injury and sustained intracranial hypertension. Childs Nerv. Syst., 17: 154–162. The Brain Trauma Foundation. The American Association of Neurological Surgeons. The Joint Section on Neurotrauma and Critical Care. (2000) Critical pathway for the treatment of established intracranial hypertension. J. Neurotrauma, 17: 537–538. Venes, J.L. and Collins, W.F. (1975) Bifrontal decompressive craniectomy in the management of head trauma. J. Neurosurg., 42: 429–433. Yamaura, A., Uemura, K. and Makino, H. (1979) Large decompressive craniectomy in management of severe cerebral contusion. A review of 207 cases. Neurol. Med. Chir. (Tokyo), 19: 717–728. Yoo, D.S., Kim, D.S., Cho, K.S., Huh, P.W., Park, C.K. and Kang, J.K. (1999) Ventricular pressure monitoring during

400 bilateral decompression with dural expansion. J. Neurosurg., 91: 953–959. Ziai, W.C., Port, J.D., Cowan, J.A., Garonzik, I.M., Bhardwaj, A. and Rigamonti, D. (2003) Decompressive craniectomy for intractable cerebral edema: experience of a single center. J. Neurosurg. Anesthesiol., 15: 25–32. Zweckberger, K., Eros, C., Zimmermann, R., Kim, S.W., Engel, D. and Plesnila, N. (2006) Effect of early and delayed

decompressive craniectomy on secondary brain damage after controlled cortical impact in mice. J. Neurotrauma, 23: 1083–1093. Zweckberger, K., Stoffel, M., Baethmann, A. and Plesnila, N. (2003) Effect of decompression craniotomy on increase of contusion volume and functional outcome after controlled cortical impact in mice. J. Neurotrauma, 20: 1307–1314.

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 29

Novel neuroproteomic approaches to studying traumatic brain injury Andrew K. Ottens1,2,, Firas H. Kobeissy1,2, Brian F. Fuller1, Ming Chen Liu2, Monika W. Oli3, Ronald L. Hayes2,3 and Kevin K.W. Wang1,2,3 1

Department of Psychiatry, Center for Neuroproteomics and Biomarkers Research at the McKnight Brain Institute of the University of Florida, PO Box 100256, Gainesville, FL 32610, USA 2 Department of Neuroscience, Center for Traumatic Brain Injury Studies at the McKnight Brain Institute of the University of Florida, PO Box 100244, Gainesville, FL 32610, USA 3 Banyan Biomarkers, Inc., 12085 Research Dr., Alachua, FL 32615, USA

Abstract: Neuroproteomics entails wide-scope study of the nervous system proteome in both its content and dynamics. The ﬁeld employs high-end analytical mass spectrometry and novel high-throughput antibody approaches to characterize as many proteins as possible. The most common application has been differential analysis to identify a limited set of highly dynamic proteins associated with injury, disease, or other altered states of the nervous system. Traumatic brain injury (TBI) is an important neurological condition where neuroproteomics has revolutionized the characterization of protein dynamics, leading to a greater understanding of post-injury biochemistry. Further, proteins of altered abundance or posttranslational modiﬁcations identiﬁed by neuroproteomic studies are candidate biochemical markers of TBI. This chapter explores the use of neuroproteomics in the study of TBI and the validation of identiﬁed putative biomarkers for subsequent clinical translation into novel injury diagnostics. Keywords: proteomics; neuroproteomics; brain injury; neurotrauma; TBI Introduction to TBI

is estimated that between 1.6 and 3.8 million sports-related TBI incidents actually occur in the US as compared with the 300,000 reported. TBI costs the American economy an estimated $60 billion per year in medical expenses and lost productivity (Finkelstein et al., 2006). All told, TBI represents a major public health problem with significant societal and economic consequences (Sosin et al., 1995; Wang et al., 2004; Yi and Hazell, 2006). Despite this, there remains no approved therapy for TBI (Narayan et al., 2002; Wieloch and Nikolich, 2006). TBI results from a number of etiologies (Fig. 1) primarily affecting male (78.8% of incidents) adolescents and young adults (31.7% of incidents are among 15- to 24-year olds). In fact, TBI is

Traumatic brain injury (TBI) is neurotrauma caused by a mechanical force applied to the head (Wang et al., 2004). There are approximately 1.4–2 million TBI incidents annually in the United States, resulting in 1.1 million emergency department visits, 235,000 hospitalizations, 90,000 left with long-term disabilities, and 50,000 deaths (Langlois et al., 2006b; Ragnarsson, 2006). Yet the true burden of TBI may not be reﬂected in these numbers as many TBI incidents go unreported or are not classiﬁed as TBI. For example it Corresponding author. Tel.: +1-352-392-8060; Fax: +1-352-

392-2579; E-mail: [email protected]ﬂ.edu DOI: 10.1016/S0079-6123(06)61029-7

401

402

Fig. 1. Breakdown of reported TBI etiologies in the United States between 1995 and 2001 (Langlois et al., 2006a).

considered the leading cause of death and disability in children and young adults. A sharp rise in TBI incidence among, typically young, American combat casualties recently added to this equation (Warden, 2006). Of Iraq and Afghanistan conﬂict casualties evacuated to the Walter Reed Army Medical Center as of January 2006, 28% (1700 cases) had TBI. An unfortunate beneﬁt to increased incidents of military head trauma has been a recent boost in funded research for development of new neurotrauma diagnostics and therapies. This is an important opportunity as military medicine has historically driven medical advancements in trauma, while progress in head injury treatment has been slow to proceed (Narayan et al., 2002). A major limitation is that TBI continues to be difﬁcult to assess clinically, even with modern imaging techniques such as magnetic resonance imaging (MRI) and computer tomography (CT), which are limited by sensitivity, speciﬁcity and availability (Ottens et al., 2006). These diagnostics often do not provide reliable prognoses, particularly for mild TBI patients that account for 80% of those who suffer lifelong impairments (Alexander, 1995). There is clearly a need for more deﬁned TBI diagnostics for patient management and prognosis. Likewise, these tools are needed as quantiﬁable outcome measures for development of new therapies, a recognized deﬁciency of past TBI therapeutics studies (Denslow et al., 2003).

To this end, the present strategy is to develop TBI diagnostics ahead of new therapy trials. For this purpose, it is important to have accurate models of TBI that mimic human sequelae and outcomes (Narayan et al., 2002; Cernak, 2005). Several experimental TBI models have been introduced to investigate the pathobiology and molecular dynamics of brain injury. Generally, a mechanical force is used to impact or torque the brain. The primary injury is comprised of contusion and hemorrhage, characterized by rapid cell death followed by diffuse axonal injury. Consequently, brain function is rapidly disrupted at the site of injury and in distal regions interconnected through white matter tracts, leading to likely deficits in higher cognitive and vital sensory-motor functions (Wieloch and Nikolich, 2006). Secondary injury occurs in proximal brain regions marked by massive edema, intracranial hemorrhages through blood-brain barrier (BBB) disruption, and further cell death and axonal injury. Those biochemical events are mediated by neurotoxic increases in excitatory amino acids (EAAs), ischemic factors (free radicals/oxidative damage), proteolysis, and an inﬂammatory response mediated by microglia and astroglia activation and proliferation (McIntosh et al., 1998; Wieloch and Nikolich, 2006). A major causal factor for secondary degeneration is the disruption of intracellular calcium homeostasis following EAA glutamate overﬂow into extra-synaptic regions

403

(glutamate excitotoxicity), and the additional glutamate release induced by a calcium inﬂux. These biochemical events can lead to prolonged cell membrane depolarization, ATP depletion and subsequent ionic imbalance, manifested clinical as an increase in intracranial pressure (ICP) (Kupina et al., 2003; Yi and Hazell, 2006). Furthermore, secondary insults continue for days, potentially worsening a patient’s condition with unpredicted results (coma or death). Several studies have demonstrated that activated cysteine proteases are major intracellular effectors in neurotrauma pathology. These proteases act on a number of substrates including cytoskeletal, cytosolic, synaptic and membrane proteins, which upon proteolysis yield signature fragments indicative of neuronal cell death dynamics (Farkas et al., 2005). Necrotic oncosis is mediated by calpains activated by the inrush of calcium (Kupina et al., 2003). Apoptosis is activated via more involved caspase pathways during secondary injury (Wennersten et al., 2003). Intrinsic and extrinsic apoptotic effectors invoke caspase-8 or -9 activation that activates the executioner caspases (primarily caspase-3) (Rink et al., 1995; Ottens et al., 2006). Other molecular events occur tangential to cell death. Shifts in pro-survival and pro-death signals (Rink et al., 1995; Shimamura et al., 2005) can result in cells that survive but are susceptible to degeneration or later death. Another feature of TBI is the marked long-term accumulation of proteins in damaged cells. Some of these proteins, such as synuclein (Syn), amyloid-b (Ab), and neuroﬁlament (NF), are major pathological components of neurodegenerative diseases such as Alzheimer’s and Parkinson’s disease (Smith et al., 2003). A detailed understanding of molecular dynamics following TBI can be of high value to elucidate injury mechanism for diagnostics and treatments (Shimamura et al., 2005). Recently, novel neuroproteomic approaches have been utilized to expand our understanding of TBI molecular mechanisms. Neuroproteomics is a new ﬁeld that provides a wide-scope, more global analysis of protein dynamics to investigate differential alteration in protein levels, modiﬁcations and functions post neurotrauma (Jenkins et al., 2002; Denslow et al.,

2003; Kobeissy et al., 2006). A number of papers have been published recently discussing the application of neuroproteomics in the ﬁeld of neurotrauma (Buonocore et al., 1999; Jenkins et al., 2002; Satchell et al., 2003; Conti et al., 2004; Haskins et al., 2005; Kobeissy et al., 2006; Kochanek et al., 2006). One major application of neuroproteomics lies in its ability to identify biomarkers that are of great value in providing insights into injury severity, patient management, and outcome. In this chapter we will explore the application of neuroproteomics to TBI, discussing the relevant preclinical animal models, the mass spectrometry and antibody based neuroproteomic methods employed, and reviewing validation and translation of neuroproteomic data into viable clinical biomarkers of TBI.

Modeling TBI in animals Traumatic brain injury (TBI) has been modeled in animals to replicate clinical trauma in a controlled experimental setting. A variety of animal models are available, and careful selection is required to determine which will work best with the scientiﬁc question at hand. A mechanical force is utilized to inﬂict injury, which must be controlled, reproducible and quantiﬁable. Invariably, the intensity of the mechanical force should dictate the severity of injury. The injury should also be comparable to the human injury, modeled with a direct correlation between injury and outcome. The following section will cover the basic mechanisms, advantages and limitations of current TBI animal models. A summary of the models can be found in Fig. 2, which is based on a classiﬁcation scheme from Ommaya (1995; see also Ommaya et al., 2002) with mechanical paradigms added by Cernak (2005) in his recent review. Static and dynamic brain trauma differs from each other based on amplitude, duration, velocity and acceleration. Static brain trauma is studied utilizing models that have a deﬁned amplitude and duration. A simple example of a static injury model would be to crush a region of the brain using a mechanical force (e.g., a vice) for a deﬁned period of time. If velocity and acceleration were

404

Fig. 2. Schematic representation of in vivo experimental models of traumatic brain injury.

factored into the static model then it would become dynamic. Dynamic brain injury is the result of a mechanical force with a well-deﬁned amplitude, duration, velocity, and acceleration. Dynamic brain injury can be sub-classiﬁed into indirect and direct injuries. Indirect dynamic brain injury occurs from blast, explosive phenomena in which the entire body is affected by the kinetic energy released. Recent studies have noted that exposure to blast waves can cause peripheral blast trauma to the brain without inﬂicting external cranial trauma (Saljo et al., 2000; Cernak et al., 2001). Animal models consist of exposing a test subject, often a rodent, to a blast wave with an over-pressurization range of 154–340 kPa. Importantly, models of rodent blast exposure have produced similar outcomes to studies of human blast injury patients (Cernak et al., 1999, 2000). Direct dynamic brain injury can be further classiﬁed into acceleration and impact models. Acceleration brain injury models study the effects of brain movement within the skull without direct contact with the brain tissue. Rapid head rotation is most commonly used to simulate the effects of forcing the brain into the side of the skull. In this

case special attention has been paid to the relationship between brain mass and acceleration in order to regulate the inertial effects that determine the extent of damage. These models have utilized miniature swine, rabbits, rats and primates as test subjects. Unconstrained head motion is standard in acceleration models; however, it has been noted that unconstrained head motion can lead to significant variability in the outcome. Non-human primates and miniature swine acceleration models have been observed to produce injuries that are strikingly similar to those found in human patients (Povlishock et al., 1994). Often the result is profuse axial injury that causes the test subject to go into a prolonged coma. Rotational acceleration models are disadvantaged by the excessive cost involved in the equipment and use of large animals, as well as the lack of adequate outcome measures of torque induced brain injury. Impact dynamic head injury models reproduce high energy concussive or penetrating injuries that are either open or closed headed. Concussive closed head injury models are utilized in order to replicate human concussive and diffuse brain injury. There have been two controlled concussion (CC) models put forth that utilize cats and rats as

405

test subjects (Tornheim and McLaurin, 1981; Tornheim et al., 1990; Goldman et al., 1991). In both, a force is applied to either the coronal structure (cats) or the skull midline (rat). The cat model has been found to be suitable for the study of cerebral contusion and resulting edema. The downside to the cat model is that the inﬂicted injury usually results in hemorrhage, which limits the ability of the researcher to use this model for the study of molecular or cellular mechanisms related exclusively to the impact. The rat model has been a solid model for the study of mild to moderate injury depending on the amount of force used. It also does not cause skull fracture or significant hemorrhage, and is comparable to similar injuries found in human TBI. These factors, as well as the low cost in using rats, make this model a prime method for the study of both TBI mechanisms and therapy. Open head contusion injuries apply a mechanical force directly to the brain through a hole in the skull (removed by craniotomy). An example is the ﬂuid percussion injury (FPI) model where a ﬂuid pressure pulse is delivered to the revealed dura to cause brain deformation (Erb and Povlishock, 1988; Faden et al., 1989, 2003; McIntosh et al., 1994; Conti et al., 1998; Albensi et al., 2000). FPI is a popular model for use with many different species (rats, mice, cats, pigs, rabbits, dogs, and sheep). The model has been adapted to apply either a central force to the vertex at the midline between bregma and lambda, or a lateral force over the left parietal bone between bregma and lambda to produce different desired outcomes. Craniotomy position is critical for impairment determination and reproducibility. While both central and lateral FPI have comparable pathobiology, it has been observed that central FPI causes death due to autonomic dysfunctions at hypothalamic levels. Thus, lateral FPI can reproduce an injury with clinical characteristics of contusion without skull fracture, but suffers from increased morbidity due to disproportionate brainstem injury and limited control of injury magnitude. Controlled cortical impact (CCI) is a more recent open-head concussive model that applies a mechanical force to the intact dura by means of a pneumatic piston (Dixon et al., 1991; Gennarelli,

1994; Povlishock et al., 1994), which allows for ﬁner control of the mechanical force than with FPI. The use of pneumatics however practically limits this model for use with rodents such as ferrets, rats and mice. CCI is also considered to deliver a more focused injury than FPI. The model successfully replicates the clinical pathobiology of brain injury with skull deformation and cortical compression. CCI is widely used for the study of molecular and cellular mechanisms related to TBI and is a useful tool in the development of novel therapies. The weight drop model is another widely utilized direct impact model, simulating an impact acceleration injury by dropping a brass weight onto a ﬁxed target up against the surgically exposed brain of a rat or mouse. In this model there is a direct correlation between the severity of injury and the weight size and height from which it is dropped (Marmarou et al., 1994; Piper et al., 1996; De Mulder et al., 2000). Historically, this model has suffered from a few technical limitations. Although the weigh is dropped though a Plexiglas tube, there is still a small amount of lateral motion. There is also the likelihood of a second impact as the weight rebounds, since it is dropped freely. More recent improvements to the weight drop model have utilized a laser guided, air-driven, high-velocity impactor, which is more controlled and reproducible then earlier setups (Cernak et al., 2004). The model is able to replicate the biochemical and neurological aspects of TBI in rats and mice as observed in clinical cases. High-velocity penetrating injury models have been designed in order to study the effects of projectile injury to the brain. Initially models were designed for use in cats and sheep (Carey et al., 1989, 1990; Finnie, 1993; Carey, 1995). In each case, a small projectile (relative to the animal’s size) was used to inﬂict a head wound on the anesthetized subject. Each of the models produced sever tissue damage as well as brain displacement. While these models are both useful in the histopathological features of projectile wounds, they are not practical for molecular studies due to the use of large animal test subjects. Williams et al. (2005) recently developed a rat penetrating ballistic brain injury (PBBI) model that mimics aspects

406

of a high energy bullet wound as demonstrated by histopathological and physiological studies. Through the insertion of a rapidly inﬂatable probe, they are able to reproduce the permanent injury track left by a bullet, as well as the temporal cavity generated by the dissipating energy from the penetrating projectile. In all, there are a number of preclinical animal models that adequately reproduce facets of clinical TBI, though the pathobiology of each model will differ somewhat in mirroring the reality of human TBI incidents. In the end, these models are useful for generating preclinical tissue samples in a reproducible, controlled manner for use in neuroproteomic studies reviewed in the following two sections.

Neuroproteomic analysis with mass spectrometry in TBI The introduction of peptide analysis by mass spectrometry in combination with bioinformatics for data processing has revolutionized the ﬁeld of proteomics. Figure 3 illustrates how the ﬁeld of proteomics (and genomics) has been driven by technology. By definition proteomics entails wide-scale protein analysis made possible by high-throughput methods and instrumentation to approach characterization of the complete proteome, or more feasibly a distinct subset such as the neuroproteome, the protein content of the central nervous system (CNS). What is also apparent in Fig. 3 is that further, significant technological advancements are poised to occur in proteomics, which will continue to revolutionize the ﬁeld, bringing us closer to complete neuroproteome characterization. Though mass spectrometers have been used to characterize macro biomolecules (e.g., polypeptides) since the early 1990s (Kaufmann, 1995; Nguyen et al., 1995), it has only been in the last few years that instrument manufacturers have begun to cater to proteomic applications. Since, the proteomics market has swelled with a discernable rise in products and software and correlated publications. Despite this, investigators remain limited in their ability to characterize a given proteome.

Available technology has become quite good for analyzing relative abundance change (differential analysis) among moderately or highly expressed proteins; however, extremes in dynamic range (1010 orders) confound complete proteome coverage. Further, proteins encoded only by some 30,000 genes transcribe to multiple isoforms and can be post-translationally modiﬁed by some 400 different processes (Morrison et al., 2002; Patton, 2002) producing a large array of functionally distinct biomolecules. Though progress has been made in characterizing the more common of these processes, there remains no adequate means to capture the fully functionalized state of proteins, let alone in a wide-scale, high-throughput paradigm. To complicate matters further, any given proteome is highly dynamic, able to shift in its protein compliment in a matter of minutes to hours. As yet proteomic experiments have been limited to temporal snap shots due to resource constraints. In light of the challenges, progress has been made in characterizing neuroproteome dynamics initiated by TBI — the primary motivators being to identify biochemical markers for development of diagnostics and potential therapeutic avenues. Prominent protein dynamics are more easily translated into clinical diagnostics; therefore, the capabilities of current proteomic technologies have coalesced well with the monitoring and prognostic medical needs in neurotrauma (Denslow et al., 2003; Wang et al., 2005). A general platform has emerged for performing mass spectrometry based differential proteomics for TBI research (Fig. 4), though the details for each step do vary. The ﬁrst important consideration is, what adjustments must be made to the proteomics experiment to suit the sample type (e.g., neuronal culture, brain tissue, cerebrospinal ﬂuid (CSF)). Confounding factors introduced by the mechanism of insult and sample processing must be controlled, generally by including matched naı¨ ve or sham controls. Biological variability also will increase with sample heterogeneity (Molloy et al., 2003). Ideally, experiments should be processed multiple times with separate samples, though this is often too costly; thus, subsequent validation of individual protein dynamics is often a more manageable scenario.

407

Fig. 3. Timeline and publication history of proteomics vs. genomics. PubMed entries (searched on Oct. 4, 2006) for genomic studies (search phrase ‘‘genome or genomics’’) and proteomic studies (search phrase ‘‘proteome or proteomics or 2D-PAGE’’) are normalized against the number of all entries per publication year. The graph is overlaid by key technological advancements that have shaped each ﬁeld.

After extraction, proteins must be resolved from one another. The most common approach to proteome separation in general and in TBI research (Buonocore et al., 1999; Jenkins et al., 2002; Satchell et al., 2003; Conti et al., 2004; Kochanek et al., 2006) has been two-dimensional polyacrylamide gel electrophoresis (2D-PAGE). Newer multidimensional separation approaches have been introduced recently, overcoming some limitation of 2D-PAGE (Ottens et al., 2006). For example, we have opted to use ion-exchange liquid chromatography in tandem with gel electrophoresis, in an approached called CAX-PAGE, to enhance differential comparison without having to mix samples and to extend the mass range (Ottens et al., 2005; Kobeissy et al., 2006). However, 2D-PAGE remains the most common approach, having been established over 30 years (Beranova-Giorgianni, 2003). 2D-PAGE involves resolving proteins by their isoelectric point (pI) ﬁrst and then by relative molecular mass (Mr). The peak capacity for each dimension (between 50 and 100) is then multiplied to provide separation of

roughly 5000 proteins (gel spots in Fig. 4). The neuroproteome contains more than this number of proteins though, and does not resolve evenly across the separation space; thus, it is important to have additional dimensions of separation, as afforded by peptide chromatography and mass spectrometry. Mass spectrometry provides accurate and precise mass-to-charge (m/z) information that can be aligned against an established, species speciﬁc, protein database(s) that has been processed in silico into sequence speciﬁc peptides. A general caveat to mass spectrometry is that proteins must be fragmented into polypeptides (generally from 7 to 20 amino acids) in order to permit accurate m/z assignment, though recent progress has been made to alleviate this limitation (Kelleher, 2004; Bogdanov and Smith, 2005). Nevertheless, mass spectrometry provides a degree of selectivity as yet unrivalled for the identiﬁcation of proteins on a large scale (technologies such as Edmund degradation or immunological assays are also selective, but on a more limited scale). Time-of-ﬂight mass

408

Fig. 4. Neuroproteomic analysis scheme using mass spectrometry. The general platform for neuroproteomic research is outlined in six steps: (1) extract proteins from the biological sample and matched control; (2) resolve proteins via multidimensional separations; (3) isolate target (e.g., differential) proteins and digest into tryptic fragments; (4) resolve peptides by chromatography (used with tandem mass spectrometry); (5) analyze with mass spectrometry; and (6) search mass spectra against protein database to identify target proteins.

spectrometers have often been used to identify proteins by precise and accurate measurement of peptide m/z values alone (no tandem mass spectra sequence information). In this approach, called

peptide mass mapping (Aebersold and Mann, 2003), multiple peptide m/z values (at least 4 or more) combined with the Mr and pI of the originating protein (from 2D-PAGE) is selective

409

enough to provide identiﬁcation. However, it is often difﬁcult to attain accurate measurement of 4 or more peptides per protein, particularly for those of lower abundance. Conti et al. (2004) encountered this limitation with lower abundant differential proteins displayed on 2D-PAGE separation of control and TBI CSF. Further, peptide mass mapping will not work well for proteins that have been modiﬁed and therefore shifted in Mr or pI value. The added selectivity afforded by tandem mass spectrometry alleviates these concerns where only the sequence of the originating protein is needed to provide an accurate match to data of a single peptide (condoned when the data are of excellent spectral quality, though data for two or more peptides per protein are generally advised). Tandem mass spectrometry coupled with peptide chromatography is more resource intensive, and therefore is often reserved for identiﬁcation when peptide mass mapping fails, as used by Kochanek et al. (2006) for differential analysis of control and TBI tissue. Tandem mass spectrometry provides multidimensional data via the use of gas-phase molecular fragmentation — traditionally through collision induced dissociation (Abersold and Mann, 2003), though more recent techniques such as electron capture dissociation (Cooper et al., 2005) and electron transfer dissociation (Good and Coon, 2006) have shown some improved functionality. The intact peptide (parent ion) m/z value is analyzed in the ﬁrst stage of mass spectral analysis. In the second stage, the peptide ion is broken up at integral points between amino acids into sequence speciﬁc fragment ions (daughter ions), which are mass analyzed to produce a tandem mass spectrum (Fig. 5). Fragmentation efﬁciency is not uniform between all amino acids of a peptide; however, enough information is generally attained to correlate the data with a database derived peptide sequence having a matched parent ion m/z value. Bioinformatic software is used to automatically process tandem mass spectra against a protein database(s). A score and/or a statistical measure (XC and P (pep), respectively in Fig. 5) of how well the data matches with the peptide sequence is provided. The originating protein(s) is then listed with all associated peptides. As illustrated in Fig. 5, excised gel bands or spots likely contain multiple

proteins, even following multidimensional protein separation. It therefore must be noted that it can still be difﬁcult to associate a single protein with a given differential gel spot or band even after tandem mass spectrometry due to protein comigration. To help ascertain the differential nature of a protein, peptide tandem mass spectrometry data can also be quantitative. The number of peptides per protein is a useful relative quantitative measure of differential abundance (Sanders et al., 2002; Peng et al., 2004; Ottens et al., 2005; Kobeissy et al., 2006), e.g., six MAP2 peptides for naı¨ ve indicates greater abundance than one peptide for TBI (Fig. 5). This information can be used for data reduction in eliminating unlikely candidate proteins, those with no apparent change or opposite change in quantity, and to conﬁrm the differential nature of the ﬁnal protein list. In all, mass spectrometry combined with protein and peptide separations is an effective platform for identifying protein dynamics following TBI (Satchell et al., 2003; Conti et al., 2004; Haskins et al., 2005; Kobeissy et al., 2006; Kochanak et al., 2006). To date, a number of candidate biochemical markers of TBI have been identiﬁed in a handful of neuroproteomic studies, which are in the processes of validation and clinical translation. Characterization of global protein dynamics and pathways following neurotrauma is more elusive, and will require more reﬁned and exhaustive proteomic experiments (further technological advancement will ultimately be required). Greater protein and/or peptide separation, improved sensitivity, and more reﬁned m/z assignment of larger polypeptides will drive the neuroproteomics ﬁeld forward, which will further the understanding of the protein chemistry of brain injury and avenues for new treatments.

TBI neuroproteomics by antibody arrays and highthroughput blotting Alternative to the mass spectrometry-based proteomic approach, several antibody-based approaches are also emerging through the availability of various antibody array or protein array platforms (Zyomyx protein Biochips, BD PowerBlot

410

Fig. 5. Bioinformatics of mass spectrometry data. Illustrated is an example of the bioinformatic output of processed tandem mass spectrometry data of one differential target from our recent TBI neuroproteomics study (Kobeissy et al., 2006). Corresponding naı¨ ve and TBI gel bands at position 23A (ﬁrst differential band in the 23 protein fraction) were excised, digested and analyzed by reversed phase liquid chromatography–tandem mass spectrometry. Peptide tandem mass spectra (peptide probability, po0.001) indicated the presence of the protein MAP2. Protein coverage was lower in the TBI sample than in naı¨ ve control (1 and 6 peptides, respectively), correlating this protein with the less intense TBI gel band.

and BD Clontech antibody microarrays 500) (Moody et al., 2001; Graslund et al., 2002; James, 2002; Kusnezow and Hoheisel, 2002). The following is a brief description of several of these technologies as applicable to TBI research, primarily biomarker development.

Antibody microarray Antibody microarrays are modeled after DNA microarrays. A collection of highly speciﬁc primary antibodies (to different protein antigens) is printed on a standard-size glass slide (e.g., BDClontech microarray 500) (Gu et al., 2006). Biological samples (brain tissue lysate, CSF or serum/ plasma) from control and treatment groups are

labeled with green and red ﬂuorescent Cyanine dye (Cy-3 and Cy-5) pairs. An equal amount of each is mixed and applied to the slide. After washing, the ﬂuorescence intensity for Cy-3 and Cy-5 are then respectively measured. The bound Cy-3 to Cy-5 ratio will then represent a relative differential quantitative measure of each antigen. The method was pioneered with a cytokine panel due to the multitude of cytokine antibodies available. Using this kit, pro-inﬂammatory cytokines such as IL-1b and TNF-alpha were found to be elevated in CSF following TBI (Shiozaki et al., 2005), a general indication for the involvement of CNS inﬂammation following neurotrauma. A drawback of this method is the technical challenge of even sample labeling. As an alternative, it is also possible to construct a ‘‘sandwich format’’ antibody array in a

411

partitioned (e.g., 96-well) plate. In this case, two highly speciﬁc primary antibodies are required to detect each protein antigen. While the coating step of the ﬁrst set of antibodies (capture antibody) is the same as above, the proteins in the biological samples are not labeled or mixed. The second set of detecting antibodies is applied to the plate, forming an antibody–antigen–antibody ‘‘sandwich’’. Finally, the abundance of the antigen is quantiﬁed by measuring the ampliﬁable detection antibody in each well (Huang, 2001).

High-throughput immunoblotting Another novel technique called high-throughput immunoblotting (HTPI) has found some recent utility for neuroproteomics research (Malakhov et al., 2002; Yoo et al., 2002; Lorenz et al., 2003). This proteomic approach involves multiplexing traditional Western blot analysis. Using our own recent TBI study as an example (Liu et al., 2006), two pooled hippocampal lysate samples, naı¨ ve and CCI (200–1000 mg), were prepared and subjected to two sets of ﬁve SDS-polyacrylamide gels (13 10 cm). Proteins in the gel were transferred to a blotting membrane, which was then clamped to a blotting manifold that isolates 40 longitudinal channels. A cocktail of between 4 and 6 antibodies was added to each channel and allowed to hybridize for 1 h at 371C. After washing, a secondary antibody conjugated to Alexa680 ﬂuorescent dye was added, then quantiﬁed by an Odyssey Infrared Imaging system (LI-COR). Each of the antibodies per channel is well characterized and resolved from one another by relative molecular mass, thus achieving coverage of 1000 proteins over ﬁve blots. As HTPI is still a Western blot in principle, it is useful for detecting protein modiﬁcations that alter gel mobility (e.g., proteolytic breakdown products). Using HTPI, we identiﬁed more than 40 decreased proteins following CCI relative to naı¨ ve, many of which were seemingly subjected to calpain and/or caspase proteolysis. Another positive feature is the ease for hit validation, as the same antibodies used for HTPI can also be used for subsequent conventional Western blot conﬁrmation. In this manner, we successfully

conﬁrmed several novel TBI-linked proteolytic substrates (Fig. 6), including betaII-spectrin, striatin, synaptotagmin-1, synaptojanin-1 and NSF (N-ethylmaleimide sensitive fusion protein) (Liu et al., 2006). In summary, antibody-based neuroproteomic techniques offer several attractive features. First, instrumentation requirements for antibody-based platforms are generally far less formidable in cost and operation than those for mass spectrometry based approaches. Second, since the target antigen is known for each antibody probe, initial protein identiﬁcation is rapid and straightforward, and as a result the bioinformatic component of this approach is far less labor-intensive. Third, validation is again relatively simple, since an antibody to the target is already available. One does need to be cautious about antibody cross reactivity to unrelated protein(s) that could give rise to false positive identiﬁcation. One major challenge of this approach is the lack of available antibodies for all 30,000 (approx.) proteins in the average mammalian proteome, let alone the multitude of posttranslational forms, making this approach less than exhaustive (Kusnezow and Hoheisal, 2002). Across-species reactivity is another issue, i.e., can an antibody raised against a human protein antigen cross detect the rat protein homologue? Furthermore, the antibodies used are not standardized in their afﬁnity to target antigens; however, the Human Proteome Organization (HUPO) is taking steps to standardize speciﬁc antibody sets for the whole human proteome (Haab et al., 2005).

Validation of proteomic data for TBI biomarker development In recent years, the term proteomics is often mentioned together with biomarker discovery, as proteomic studies have the capability of identifying unique and unobvious protein biomarkers from tissue or bioﬂuids obtained from human diseased patients or from animal models of disease and disorders. Biomarkers of acute brain injury have also attracted a lot of recent research interest (Wang et al., 2006). First, it is commonly recognized that better patient monitoring and

412

Fig. 6. Example of TBI differential TBI proteome as analyzed by HTPI. HTPI Template A is shown for naı¨ ve rat hippocampus (upper panel) and the controlled cortical impact injury counterpart (lower panel). Molecular mass markers (lane 40) are indicated on the right. Thirteen proteins in template A were reduced in average intensity (abundance) after TBI (solid boxes in upper panel). In addition, several breakdown products were observed (dotted boxes in lower panel). Two proteins found to be up-regulated in TBI were CASK (A3) and Psme3 (A29) (in dotted box). Adapted with permission from Liu et al. (2006).

management often leads to better clinical outcome. Thus, a simple point-of-care (POC) CSF- or blood-based diagnostic test would be extremely useful, and significantly less expensive than any brain image scanner (such as computed tomography). Second, numerous failures in clinical drug trials for the treatment of ischemic strokes and TBI over the last 10–15 years point to the difﬁculty of translating preclinical successes to clinical successes. One remedy of the situation is the development of robust quantitative biomarkers that are

predictive of disease progression and treatment response (Lesko and Atkinson, 2001; Pineda et al., 2004; Wang et al., 2006). In practice, proteins identiﬁed from differential neuroproteomic studies of brain injury (human or animal models) can only be considered potential biomarkers for brain injury, particularly if the goal is to develop a CSF- or blood-based brain injury biomarker test. This is even truer of candidate proteins identiﬁed from differential tissue lysate analysis, not in bioﬂuids.

413

Here are several criteria for an ideal biomarker of brain injury: i. Levels of the biomarker in a relevant bioﬂuid (CSF and/or blood) correlate with disease severity, e.g., Glasgow Coma Score (GCS) for TBI. ii. Levels of the biomarker correlate with clinical outcome, e.g., as measured by the Extended Glasgow Outcome Scale (GOS-E). iii. Biomarker signals are not confounded by other non-brain injury neurological conditions. iv. Biomarker levels are responsive to therapeutic interventions. To date, the most common validation technology for neuroproteomics data to biomarker assay relies heavily on antibody assays, since antibodies (both polyclonal and monoclonal) have the ability to selectively identify and quantify a single protein (antigen) of interest in a rapid, high-throughput manner. Due to the maturity of immunological methods, antibodies-based diagnostic assays have been reﬁned almost to an art form, namely, enzyme-linked immunosorbant assays (ELISA). There are two forms of ELISAs: direct ELISA utilizes a single detection antibody; or sandwich ELISA (swELISA) that requires the formation of an immobilized capture antibody to protein antigen to detection antibody complex. For biomarker-based diagnostic, swELISA is indeed the method of choice, since it provides antigen enrichment that significantly improves signal ﬁdelity. The work ﬂow for developing biomarker swELISA assays is as follows. Once a putative biomarker is identiﬁed and conﬁrmed, one must obtain a pair of compatible and high afﬁnity antibodies. Antibodies from commercial sources are explored ﬁrst; however, custom-made antibodies are often needed. Custom antibodies are generally raised against a peptide epitope-carrier protein complex, a recombinant protein or a fusion protein (e.g., GST fusion protein) as the antigen, which is injected into host animals for an immune response (rabbit, chicken, goat, and mouse). Care must be taken that the antibodies thus raised will only recognize the protein isoform or posttranslationally modiﬁed form of interest (e.g.,

phosphorylated form or breakdown products, such as c-tau or alphaII-spectrin breakdown products (SBDPs) (Pineda et al., 2004). Afterwards, antigen afﬁnity puriﬁcation is highly recommended. Finally, some form of the antigen (puriﬁed native protein or recombinant protein) must be available for the development and testing of the antibody, as well as for use to generate a standard curve for absolute quantiﬁcation. What follows are iterations of trial and error to ﬁnd the best antibody pair (capture and detection), and the optimal assaying conditions to give the best signal to noise ratio as well as the lowest detection limit. The term ELISA implies that the detection antibody is coupled to an enzyme-linked substrate to product system for signal ampliﬁcation (e.g., alkaline phosphatase or horse radish peroxidase). An ideal biomarker ELISA assay should strive to achieve high (490%) selectivity (i.e., the ability to detect the presence of injury) and high (490%) speciﬁcity (i.e., the ability to detect the absence of injury). A major challenge of brain injury biomarker assay development is the availability of high quality antibodies. On this front, there have been significant developments in capturing and detecting agents beyond monoclonal antibodies, such as aptamers (Bunka and Stockley, 2006) and phagedisplayed single chain antibody fragments (Tomizaki et al., 2005). Another major challenge is the detection limits, since it is conceivable that the best brain injury marker might be very low in abundance. ELISA sensitivity can be improved with an enzymatic signal-enhancement method known as tyramide signal ampliﬁcation (TSA). It is also apparent that technological advances in ELISA sensitivity improvement will greatly beneﬁt this ﬁeld. Animal models are commonly used for preclinical biomarker validation, as they confer a number of advantages. One important advantage of preclinical animal models is the ability to prioritize the potential utility of biomarkers by use of relatively simple Western blot studies prior to commitment of substantial time and resources to the development of more sensitive quantitative approaches such as ELISA. Employing a CCI model of TBI, we capitalized on the ability to simultaneously characterize the appearance of biomarkers in

414

injured tissues as well as cerebrospinal ﬂuid (Pike et al., 2001, 2003). This often overlooked advantage allows the rigorous characterization of relationships between biomarker generation in injured brain tissue and appearance in other compartments such as CSF and ultimately in blood. Validations with preclinical animal models make possible comparisons of distribution biokinetics and the temporal proﬁle of biomarker appearance in different, clinically relevant compartments such as CSF and blood, and their relationship to the time of injury. These values can include the half-life and time to maximal concentrations of biomarkers. A clear understanding of the biokinetics of biomarkers is an important component and accurate interpretation of the clinical relevance of changes seen in biomarker levels. Apart from straightforward comparisons of appearance and temporal proﬁle of biomarkers, preclinical animal models provide important information on the potential relationship of biomarker levels to injury magnitudes, lesion volume and outcome. For example, our laboratory (Ringger et al., 2004) has conﬁrmed that a biochemical marker of calpain proteolysis, the cytoskeletal protein alphaII-spectrin, when measured in CSF reliably correlated with the magnitude of CCI injury, lesion volume and neurological outcome.

Translation to clinical diagnostics The primary challenge facing translation of biomarkers to practical clinical diagnostics resides in the approaches to conﬁrm the clinical utility of putative biomarkers of TBI identiﬁed and validated with preclinical animal models. Ideally, the conﬁrmation platform should integrate preclinical validation with clinical assessments of the potential applicability of the candidate biomarkers. A direct comparison of biomarker generation between preclinical models and biomarker data from human clinical studies allows investigators to gain considerable insight into the validity (or challenges to the validity) of the employed preclinical animal models. For example, we have shown (unpublished data) similar proﬁles of alphaII-spectrin degradation by calpain in both CCI and severe human

TBI. Thus, CCI reproduces important features of calpain mediated pathology seen in human TBI. The development of a clinical research platform for TBI poses considerable challenges. An important limitation of clinical models, as typically implemented, is the failure to incorporate systematic studies of the role of secondary insults and age, significant determinants of injury response in human TBI. Mild TBI and the frequent absence of objective evidence of injury by computerized tomographic studies or MRI are especially challenging. Neurological diagnosis of varying magnitudes of mild or moderate TBI is unreliable, since the GCS scale was developed primarily to assess more severely injured comatose patients. Outcome measures are equally controversial and ambiguous. However, greater progress has been made in establishing a clinical research platform for severe TBI. Our program has recently completed the ﬁrst extensive examination of the potential clinical utility of biochemical markers of calpain and caspase3 proteolysis in severe (GCS ¼ 4 8) TBI patients (Pineda, 2006, in press). This research incorporated important elements of a clinical research platform for biomarkers of TBI. Such a platform should be able to test hypotheses about the potential clinical utility and validity of biomarkers assessed, alone or in combination. Generally validation effects should focus on whether biomarkers are reliably associated with the magnitude of brain injury, occurrence of secondary insults and outcome. The platform should allow rigorous determination of whether appropriate markers can provide information critical for diagnosis, triage and management, as well as data on possible injury mechanisms and cellular localization of injury. In the research of Pineda (2006, in press), investigators were able to conﬁrm that a protein biomarker of calpain proteolysis (degradation of alphaII-spectrin) measured in CSF was reliably associated with the magnitude of dichotomized GCS on admission (GCS 3–6 vs. 6–8), while a biomarker of caspase-3 alphaII-spectrin proteolysis showed no such association. The biomarker of calpain proteolysis was also reliably associated with secondary insult, (preadmission hypotension) while a marker of caspase-3 proteolysis showed no significant relationship. The biomarker of calpain proteolysis, but not caspase-3

415

proteolysis was also differentially associated with outcome. Significantly higher level of a biomarker of calpain proteolysis was detected in patients within the ﬁrst 24 h following injury who subsequently had a worse outcome (assessed by the dichotomized GCS 6 months after injury). Importantly, calpain proteolysis is primarily associated with oncotic necrosis while caspase-3 proteolysis is more prominently associated with apoptosis. Thus, these data collectively suggest that oncotic necrosis may more importantly contribute to pathology in severe TBI than apoptosis. This hypothesis is consistent with the temporal proﬁle of accumulation of these biomarkers in CSF of severe TBI patients for the ﬁrst 5 days following injury. Markers of calpain proteolysis were increased more than caspse-3 proteolysis, although caspase-3 proteolysis tended to be more sustained. Our laboratory has recently begun an important series of studies that will significantly enhance our ability to validate relationships between changes in biomarkers and secondary insults following TBI. An important limitation in such studies has been the absence of the ability to continuously record physiological data that would indicate the presence of a secondary insult (e.g., reduced ICP, hypotension, compromised tissue oxygenation). In fact, Hemphill et al. (2005) has recently documented that studies conducted by collecting data of medical records, frequently missed secondary insults documented by continuous recording of physiological parameters. Thus, our program has recently initiated the systematic comparison of biomarker changes assessed from both CSF and blood with physiological variables recorded continuously from severe TBI patient. The ultimate goal of these studies is to determine if changes in biomarker levels can, in fact, predict onset of secondary insults as well as document their occurrence. If so, rapid assessment of biomarkers could provide critical information guiding management of severely injured TBI patients in the intensive care unit.

Closing remarks In conclusion, neuroproteomics has revolutionized the exploration of protein dynamics following TBI.

In the near-term, whether by mass spectrometry or antibody-based arrays, neuroproteomics serves as a high-throughput means to discover potentially novel biochemical markers of TBI. Significant advances have also been made in developing strategies for preclinical and clinical validation of biomarkers of TBI. While often limited by the failure to incorporate the role of secondary insults, age and gender, preclinical animal models in combination with neuroproteomic techniques have proven invaluable for rapidly assessing the potential utility of the relatively large numbers of biomarkers generated. While significant challenges remain in implementing preclinical platforms from mild and moderate TBI, the establishment of effective platforms to validate biomarkers in severe TBI promises to deﬁne clinical useful subsets of biomarkers of TBI within the next few years. Looking forward, neuroproteomic technology will rapidly evolve to more effectively and more efﬁciently characterize the neuroproteome as altered by TBI, much as has been seen in the genomic revolution. This impending opportunity will confer unparalleled detail on the biochemical processes occurring following TBI, and will inspire novel therapies trials, to be evaluated with the biomarkers presently in development.

Disclaimer KKW and RLH own stock, receive royalties from and are executive ofﬁcers of Banyan Biomarkers Inc. and as such may beneﬁt ﬁnancially as a result of the outcomes of this research or work reported in this publication.

References Aebersold, R. and Mann, M. (2003) Mass spectrometry-based proteomics. Nature, 422: 198–207. Albensi, B.C., Knoblach, S.M., Chew, B.G., O’Reilly, M.P., Faden, A.I. and Pekar, J.J. (2000) Diffusion and high resolution MRI of traumatic brain injury in rats: time course and correlation with histology. Exp. Neurol., 162: 61–72. Alexander, M.P. (1995) Mild traumatic brain injury: pathophysiology, natural history, and clinical management. Neurology, 45: 1253–1260.

416 Beranova-Giorgianni, S. (2003) Proteome analysis by two-dimensional gel electrophoresis and mass spectrometry: strengths and limitations. Trac-Trend Anal. Chem., 5: 273–281. Bogdanov, B. and Smith, R.D. (2005) Proteomics by FTICR mass spectrometry: top down and bottom up. Mass Spectrom. Rev., 24: 168–200. Bunka, D.H. and Stockley, P.G. (2006) Aptamers come of age — at last. Nat. Rev. Microbiol., 4: 588–596. Buonocore, G., Liberatori, S., Bini, L., Mishra, O.P., DelivoriaPapadopoulos, M., Pallini, V. and Bracci, R. (1999) Hypoxic response of synaptosomal proteins in term guinea pig fetuses. J. Neurochem., 73: 2139–2148. Carey, M.E. (1995) Experimental missile wounding of the brain. Neurosurg. Clin. N. Am., 6: 629–642. Carey, M.E., Sarna, G.S. and Farrell, J.B. (1990) Brain edema after an experimental missile wound. Adv. Neurol., 52: 301–305. Carey, M.E., Sarna, G.S., Farrell, J.B. and Happel, L.T. (1989) Experimental missile wound to the brain. J. Neurosurg., 71: 754–764. Cernak, I. (2005) Animal models of head trauma. NeuroRx, 2: 410–422. Cernak, I., Savic, J., Ignjatovic, D. and Jevtic, M. (1999) Blast injury from explosive munitions. J. Trauma, 47: 96–104. Cernak, I., Savic, V.J., Kotur, J., Prokic, V., Veljovic, M. and Grbovic, D. (2000) Characterization of plasma magnesium concentration and oxidative stress following graded traumatic brain injury in humans. J. Neurotrauma, 17: 53–68. Cernak, I., Vink, R., Zapple, D.N., Cruz, M.I., Ahmed, F., Chang, T., Fricke, S.T. and Faden, A.I. (2004) The pathobiology of moderate diffuse traumatic brain injury as identiﬁed using a new experimental model of injury in rats. Neurobiol. Dis., 17: 29–43. Cernak, I., Wang, Z., Jiang, J., Bian, X. and Savic, J. (2001) Ultrastructural and functional characteristics of blast injuryinduced neurotrauma. J. Trauma, 50: 695–706. Conti, A.C., Raghupathi, R., Trojanowski, J.Q. and McIntosh, T.K. (1998) Experimental brain injury induces regionally distinct apoptosis during the acute and delayed posttraumatic period. J. Neurosci., 18: 5663–5672. Conti, A., Sanchez-Ruiz, Y., Bachi, A., Beretta, L., Grandi, E., Beltramo, M. and Alessio, M. (2004) Proteome study of human cerebrospinal ﬂuid following traumatic brain injury indicates ﬁbrinogen degradation products as traumaassociated markers. J. Neurotrauma, 21: 854–863. Cooper, H.J., Hakansson, K. and Marshall, A.G. (2005) The role of electron capture dissociation in biomolecular analysis. Mass Spectrom. Rev., 24: 201–222. De Mulder, G., Van Rossem, K., Van Reempts, J., Borgers, M. and Verlooy, J. (2000) Validation of a closed head injury model for use in long-term studies. Acta Neurochir. Suppl., 76: 409–413. Denslow, N.D., Michel, M.E., Temple, M.D., Hsu, C.Y., Saatman, K. and Hayes, R.L. (2003) Application of proteomics technology to the ﬁeld of neurotrauma. J. Neurotrauma, 20: 401–407.

Dixon, C.E., Clifton, G.L., Lighthall, J.W., Yaghmai, A.A. and Hayes, R.L. (1991) A controlled cortical impact model of traumatic brain injury in the rat. J. Neurosci. Methods, 39: 253–262. Erb, D.E. and Povlishock, J.T. (1988) Axonal damage in severe traumatic brain injury: an experimental study in cat. Acta Neuropathol., 76: 347–358. Faden, A.I., Demediuk, P., Panter, S.S. and Vink, R. (1989) The role of excitatory amino acids and NMDA receptors in traumatic brain injury. Science, 244: 798–800. Faden, A.I., Knoblach, S.M., Cernak, I., Fan, L., Vink, R., Araldi, G.L., Fricke, S.T., Roth, B.L. and Kozikowski, A.P. (2003) Novel diketopiperazine enhances motor and cognitive recovery after traumatic brain injury in rats and shows neuroprotection in vitro and in vivo. J. Cereb. Blood Flow Metab., 23: 342–354. Farkas, O., Polgar, B., Szekeres-Bartho, J., Doczi, T., Povlishock, J.T. and Buki, A. (2005) Spectrin breakdown products in the cerebrospinal ﬂuid in severe head injury — preliminary observations. Acta Neurochir., 147: 855–861. Finnie, J.W. (1993) Pathology of experimental traumatic craniocerebral missile injury. J. Comp. Pathol., 108: 93–101. Finkelstein, E., Corso, P. and Miller, T. (2006) The Incidence and Burden of Injuries in the United States. Oxford University Press, New York. Gennarelli, T.A. (1994) Animate models of human head injury. J. Neurotrauma, 11: 357–368. Goldman, H., Hodgson, V., Morehead, M., Hazlett, J. and Murphy, S. (1991) Cerebrovascular changes in a rat model of moderate closed-head injury. J. Neurotrauma, 8: 129–144. Good, D.M. and Coon, J.J. (2006) Advancing proteomics with ion/ion chemistry. Biotechniques, 40: 783–789. Graslund, S., Falk, R., Brundell, E., Hooq, C. and Stahl, S. (2002) A high-stringency proteomics concept aimed for generation of antibodies speciﬁc for cDNA encoded proteins. Biotechnol. Appl. Biochem., 35: 75–82. Gu, Q., Sivanandam, T.M. and Kim, C.A. (2006) Signal stability of Cy3 and Cy5 on antibody microarrays. Proteome Sci., 4: 21. Haab, B.B., Geierstanger, B.H., Michailidis, G., Vitzthum, F., Forrester, S., Okon, R., Saviranta, P., Brinker, A., Sorette, M., Perlee, L., Suresh, S., Drwal, G., Adkins, J.N. and Omenn, G.S. (2005) Immunoassay and antibody microarray analysis of the HUPO Plasma Proteome Project reference specimens: systematic variation between sample types and calibration of mass spectrometry data. Proteomics, 5: 3278–3291. Haskins, W.E., Kobeissy, F.H., Wolper, R.A., Ottens, A.K., Kitlen, J.W., McClung, S.H., O’Steen, B.E., Chow, M.M., Pineda, J.A., Denslow, N.D., Hayes, R.L. and Wang, K.K.W. (2005) Rapid discovery of putative protein biomarkers of traumatic brain injury by SDS-PAGE-capillary liquid chromatography-tandem mass spectrometry. J. Neurotrauma, 22: 629–644. Hemphill III, J.C., Barton, C.W., Marabito, D. and Manley, G.T. (2005) Inﬂuence of data resolution and interpolation method on assessment of secondary brain insults in neurocritical care. Physiol. Meas., 26: 373–386.

417 Huang, R.P. (2001) Simultaneous detection of multiple proteins with an array-based enzyme-linked immunosorbent assay (ELISA) and enhanced chemiluminescence (ECL). Clin. Chem. Lab. Med., 39: 209–214. James, P. (2002) Chips for proteomics: a new tool or just hype? Biotechniques, Dec(Suppl): 4–13. Jenkins, L.W., Peters, G.W., Dixon, C.E., Zhang, X., Clark, R.S.B., Skinner, J.C., Marion, D.W., Adelson, P.D. and Kochanek, P.M. (2002) Conventional and functional proteomics using large format two-dimensional gel electrophoresis 24 hours after controlled cortical impact in postnatal day 17 rats. J. Neurotrauma, 19: 715–740. Kaufmann, R. (1995) Matrix-assisted laser desorption ionization (MALDI) mass spectrometry: a novel analytical tool in molecular biology and biotechnology. J. Biotechnol., 41: 155–175. Kelleher, N.L. (2004) Top-down proteomics. Anal. Chem., 76: 197A–203A. Kobeissy, F.H., Ottens, A.K., Zhang, Z., Liu, M.C., Denslow, N.D., Dave, J.R., Tortella, F.C., Hayes, R.L. and Wang, K.K. (2006) Novel differential neuroproteomics analysis of traumatic brain injury in rats. Mol. Cell. Proteomics, 5: 1887–1898. Kochanek, A.R., Kline, A.E., Gao, W.M., Chadha, M., Lai, Y., Clark, R.S.B., Dixon, C.E. and Jenkins, L.W. (2006) Gelbased hippocampal proteomic analysis 2 weeks following traumatic brain injury to immature rats using controlled cortical impact. Dev. Neurosci., 28: 410–419. Kupina, N.C., Detloff, M.R., Bobrowski, W.F., Snyder, B.J. and Hall, E.D. (2003) Cytoskeletal protein degradation and neurodegeneration evolves differently in males and females following experimental head injury. Exp. Neurol., 180: 55–73. Kusnezow, W. and Hoheisel, J.D. (2002) Antibody microarrays: promises and problems, 33: S14–S23 Langlois, J.A., Rutland-Brown, W. and Thomas, K.E. (2006a) Traumatic Brain Injury in the United States: Emergency Department Visits, Hospitalizations, and Deaths. Center for Disease Control and Prevention, National Center for Injury Prevention and Control, Atlanta, GA. Langlois, J.A., Rutland-Brown, W. and Wald, M.M. (2006b) The epidemiology and impact of traumatic brain injury: a brief overview. J. Head Trauma Rehabil., 21: 375–378. Lesko, L.J. and Atkinson Jr., A.J. (2001) Use of biomarkers and surrogate endpoints in drug development and regulatory decision making: criteria, validation, strategies. Annu. Rev. Pharmacol. Toxicol., 41: 347–366. Liu, M.C., Akle, V., Zheng, W.R., Dave, J.R., Tortella, F.C., Hayes, R.L. and Wang, K.K.W. (2006) Comparing calpainand caspase-3-degradation patterns in traumatic brain injury by differential proteome analysis. Biochem. J., 394: 715–725. Lorenz, P., Ruschpler, P., Koczan, D., Stiehl, P. and Thiesen, H.J. (2003) From transcriptome to proteome, differentially expressed proteins identiﬁed in synovial tissue of patients suffering from rheumatoid arthritis and osteoarthritis by an initial screen with a panel of 791 antibodies. Proteomics, 3: 991–1002. Malakhov, M.P., Kim, K.I., Malakhova, O.A., Jacobs, B.S., Borden, E.C. and Zhang, D.E. (2002) High-throughput

immunoblotting. Ubiquitiin-like protein ISG15 modiﬁes key regulators of signal transduction. J. Biol. Chem., 278: 16608–16613. Marmarou, A., Foda, M.A., van den Brink, W., Campbell, J., Kita, H. and Demetriadou, K. (1994) A new model of diffuse brain injury in rats. Part I: Pathophysiology and biomechanics. J. Neurosurg., 80: 291–300. McIntosh, T.K., Juhler, M. and Wieloch, T. (1998) Novel pharmacologic strategies in the treatment of experimental traumatic brain injury. J. Neurotrauma, 15: 731–769. McIntosh, T.K., Yu, T. and Gennarelli, T.A. (1994) Alterations in regional brain catecholamine concentrations after experimental brain injury in the rat. J. Neurochem., 63: 1426–1433. Molloy, M.P., Brzenzinski, E.E., Hang, J., McDowell, M.T. and VanBogelen, R.A. (2003) Overcoming technical variation and biological variation in quantitative proteomics. Proteomics, 3: 1912–1919. Moody, M.D., Van Arsdell, S.W., Murphy, K.P., Orencole, S.F. and Burns, C. (2001) Array-based ELISAs for highthroughput analysis of human cytokines. Biotechniques, 31: 186–194. Morrison, R.S., Kinoshita, Y., Johnson, M.D., Uo, T., Ho, J.T., McBee, J.K., Conrads, T.P. and Veenstra, T.D. (2002) Proteomic analysis in the neurosciences. Mol. Cell. Proteomics, 1: 553–560. Narayan, R.K., Michel, M.E., Ansell, B., Baethmann, A., Biegon, A., Bracken, M.B., Bullock, M.R., Choi, S.C., Clifton, G.L., Contant, C.F., Coplin, W.M., Dietrich, W.D., Ghajar, J., Grady, S.M., Grossman, R.G., Hall, E.D., Heetderks, W., Hovda, D.A., Jallo, J., Katz, R.L., Knoller, N., Kochanek, P.M., Maas, A.I., Majde, J., Marion, D.W., Marmarou, A., Marshall, L.F., McIntosh, T.K., Miller, E., Mohberg, N., Muizelaar, J.P., Pitts, L.H., Quinn, P., Riesenfeld, G., Robertson, C.S., Strauss, K.I., Teasdale, G., Temkin, N., Tuma, R., Wade, C., Walker, M.D., Weinrich, M., Whyte, J., Wilberger, J., Young, A.B. and Yurkewicz, L. (2002) Clinical trials in head injury. J. Neurotrauma, 19: 503–557. Nguyen, D.N., Becker, G.W. and Riggin, R.M. (1995) Protein mass spectrometry: applications to analytical biotechnology. J. Chromatogr. A, 705: 21–45. Ommaya, A.K. (1995) Head injury mechanisms and the concept of preventive management: a review and critical synthesis. J. Neurotrauma, 12: 527–546. Ommaya, A.K., Goldsmith, W. and Thibault, L. (2002) Biomechanics and neuropathology of adult and paediatric head injury. Br. J. Neurosurg., 16: 220–242. Ottens, A.K., Kobeissy, F.H., Haskins, W.E., Golden, E.C., Zhang, Z.Q., Chen, S., Hayes, R.L., Wang, K.K.W. and Denslow, N.D. (2006) Neuroproteomics in neurotrauma. Mass Spectrom. Rev., 25: 380–408. Ottens, A.K., Kobeissy, F.H., Wolper, R.A., Haskins, W.E., Hayes, R.L., Denslow, N.D. and Wang, K.K.W. (2005) A multidimensional differential proteomic platform using dual phase ion-exchange chromatography — polyacrylamide gel electrophoresis/reverse phase liquid chromatography tandem mass spectrometry. Anal. Chem., 77: 4836–4845.

418 Patton, W.F. (2002) Detection technologies in proteome analysis. J. Chromatogr. B, 771: 3–31. Peng, J., Kim, M.J., Cheng, D., Duong, D.M., Gygi, S.P. and Sheng, M. (2004) Semiquantitative proteomics analysis of rat forebrain postsynaptic density fractions by mass spectrometry. J. Biol. Chem., 279: 21003–21011. Pike, B.R., Flint, J., Dave, J.R., Lu, X.C., Wang, K.K., Tortella, F.C. and Hayes, R.L. (2003) Accumulation of calpain and caspase-3 proteolytic fragments of brain-derived alphaII-spectrin in CSF after middle cerebral artery occlusion in rats. J. Cereb. Blood Flow Metab., 24: 98–106. Pike, B.R., Flint, J., Dutta, S., Johnson, E., Wang, K.K.W. and Hayes, R.L. (2001) Accumulation of non-erythroid alphaIIspectrin and calpain-cleaved alphaII-spectrin breakdown products in cerebrospinal ﬂuid after TBI in rats. J. Neurochem., 78: 1297–1306. Pineda, J.A. (2006) Accumulation of alphaII-spectrin breakdown products in CSF of severely injured traumatic brain injury patients. J. Neurotrauma, in press. Pineda, J.A., Wang, K.K. and Hayes, R.L. (2004) Biomarkers of proteolytic damage following traumatic brain injury. Brain Pathol., 14: 202–209. Piper, I.R., Thomson, D. and Miller, J.D. (1996) Monitoring weight drop velocity and foam stiffness as an aid to quality control of a rodent model of impact acceleration neurotrauma. J. Neurosci. Methods, 69: 171–174. Povlishock, J.T., Hayes, R.L., Michel, M.E. and McIntosh, T.K. (1994) Workshop on animal models of traumatic brain injury. J. Neurotrauma, 11: 723–732. Ragnarsson, K.T. (2006) Traumatic brain injury research since the 1998 NIH consensus conference. J. Head Trauma Rehabil., 21: 379–387. Ringger, N.C., O’Steen, B.E., Brabham, J.G., Siler, X., Pineda, J., Wang, K.K.W. and Hayes, R.L. (2004) A novel marker for traumatic brain injury: CSF alphaII-spectrin breakdown product levels. J. Neurotrauma, 21: 1443–1456. Rink, A., Fung, K.M., Trojanowski, J.Q., Lee, V.M., Neugebauer, E. and McIntosh, T.K. (1995) Evidence of apoptotic cell death after experimental traumatic brain injury in the rat. Am. J. Pathol., 147: 1575–1583. Saljo, A., Bao, F., Haglid, K.G. and Hansson, H.A. (2000) Blast exposure causes redistribution of phosphorylated neuroﬁlament subunits in neurons of the adult rat brain. J. Neurotrauma, 17: 719–726. Sanders, S.L., Jefﬁngs, J., Canutescu, A., Link, A.J. and Well, P.A. (2002) Proteomics of the eukaryotic transcription machinery: identiﬁcation of proteins associated with components of yeast TFIID by multidimensional mass spectrometry. Mol. Cell. Biol., 22: 4723–4738. Satchell, M.A., Zhang, X., Kochanek, P.M., Dixon, C.E., Jenkins, L.W., Melick, J., Szabo, C. and Clark, R.S.B. (2003) A dual role for poly-ADP-ribosylation in spatial memory acquisition after traumatic brain injury in mice involving NAD+ depletion and ribosylation of 14-3-3g. J. Neurochem., 85: 697–708. Shimamura, M., Garcia, J.M., Prough, D.S., Dewitt, D.S., Uchida, T., Shah, S.A., Avila, M.A. and Hellmich, H.L.

(2005) Analysis of long-term gene expression in neurons of the hippocampal subﬁelds following traumatic brain injury in rats. J. Neurosci., 131: 87–97. Shiozaki, T., Hayakata, T., Tasaki, O., Hosotubo, H., Fuijita, K., Mouri, T., Tajima, G., Kajino, K., Nakae, H., Tanaka, H., Shimazu, T. and Sugimoto, H. (2005) Cerebrospinal ﬂuid concentrations of anti-inﬂammatory mediators in early-phase severe traumatic brain injury. Shock, 23: 406–410. Smith, D.H., Uryu, K., Saatman, K.E., Trojanowski, J.Q. and McIntosh, T.K. (2003) Protein accumulation in traumatic brain injury. Neuromolecular Med., 4: 59–72. Sosin, D.M., Sniezek, J.E. and Waxweiler, R.J. (1995) Trends in death associated with traumatic brain injury, 1979 through 1992. Success and failure. JAMA, 273: 1778–1780. Tomizaki, K.Y., Usui, K. and Mihara, H. (2005) Proteindetecting microarrays: current accomplishments and requirements. Chembiochem, 6: 782–799. Tornheim, P.A., McDermott, F. and Shiguma, M. (1990) Effect of experimental blunt head injury on acute regional cerebral blood ﬂow and edema. Adv. Neurol., 52: 377–384. Tornheim, P.A. and McLaurin, R.L. (1981) Acute changes in regional brain water content following experimental closed head injury. J. Neurosurg., 55: 407–413. Wang, K.K.W., Larner, S.F., Robinson, G. and Hayes, R.L. (2006) Neuroprotection targets after traumatic brain injury. Curr. Opin. Neurol., 19: 514–519. Wang, K.K.W., Ottens, A., Haskins, W.E., Liu, M.C., Denslow, N., Chen, S.S. and Hayes, R.L. (2004) Neuroproteomic studies of traumatic brain injury. In: Neuhold L.A. (Ed.), Human Brain Proteome — International Review of Neurobiololgy, Vol. 61. Elsevier, San Diego, pp. 215–240. Wang, K.K.W., Ottens, A.K., Liu, M.C., Lewis, S.B., Meegan, C., Oli, M.W., Tortella, F.C. and Hayes, R.L. (2005) Proteomic identiﬁcation of biomarkers of traumatic brain injury. Expert Rev. Proteomics, 2: 603–614. Warden, D. (2006) Military TBI during the Iraq and Afghanistan wars. J. Head Trauma Rehabil., 21: 398–402. Wennersten, A., Holmin, S. and Mathiesen, T. (2003) Characterization of Bax and Bcl-2 in apoptosis after experimental traumatic brain injury in the rat. Acta Neuropathol., 105: 281–288. Wieloch, T. and Nikolich, K. (2006) Mechanisms of neural plasticity following brain injury. Curr. Opin. Neurobiol., 16: 258–264. Williams, A.J., Hartings, J.A., Lu, X.C., Rolli, M.L., Dave, J.R. and Tortella, F.C. (2005) Characterization of a new rat model of penetrating ballistic brain injury. J. Neurotrauma, 22: 313–331. Yi, J.H. and Hazell, A.S. (2006) Excitotoxic mechanisms and the role of astrocytic glutamate transporters in traumatic brain injury. Neurochem. Int., 48: 394–403. Yoo, G.H., Piechocki, M.P., Ensley, J.F., Nguyen, T., Oliver, J., Meng, H., Kewson, D., Shibuya, T.Y., Lonardo, F. and Tainsky, M.A. (2002) Docetaxel induced gene expression patterns in head and neck squamous cell carcinoma using cDNA microarray and PowerBlot. Clin. Cancer Res., 8: 3910–3921.

Weber & Maas (Eds.) Progress in Brain Research, Vol. 161 ISSN 0079-6123 Copyright r 2007 Elsevier B.V. All rights reserved

CHAPTER 30

Remyelination of the injured spinal cord Masanori Sasaki1,2, Bingcang Li1,2,3, Karen L. Lankford1,2, Christine Radtke1,2,4 and Jeffery D. Kocsis1,2, 1

Department of Neurology and Center for Neuroscience and Regeneration Research, Yale University School of Medicine, New Haven, CT 06510, USA 2 Rehabilitation Research Center, Veterans Affairs Connecticut Healthcare System, West Haven, CT 06516, USA 3 Research Institute of Surgery, Daping Hospital, Third Military Medical University, Chongqing, P.R. China 4 Department of Plastic, Hand and Reconstructive Surgery, Medical School Hannover, Hannover, Germany

Abstract: Contusive spinal cord injury (SCI) can result in necrosis of the spinal cord, but often long white matter tracts outside of the central necrotic core are demyelinated. One experimental strategy to improve functional outcome following SCI is to transplant myelin-forming cells to remyelinate these axons and improve conduction. This review focuses on transplantation studies using olfactory ensheathing cell (OEC) to improve functional outcome in experimental models of SCI and demyelination. The biology of the OEC, and recent experimental research and clinical studies using OECs as a potential cell therapy candidate are discussed. Keywords: spinal cord injury; remyelination; olfactory ensheathing cells

Cellular transplantation of appropriate cells into experimental models of SCI can promote axonal regeneration, provide neuroprotective effects by secretion of neurotrophins and remyelinate axons. One cell of particular interest as a cell therapy candidate to both encourage axonal remyelination and regeneration is a specialized glial cell, the olfactory ensheathing cell (OEC). Adult olfactory receptor neurons continually undergo turnover from an endogenous progenitor pool, and their nascent axons grow through the olfactory nerves and cross the PNS–CNS interface, where they form new synaptic connections in the olfactory bulb (Graziadei et al., 1978). OECs associate with olfactory receptor neurons from their peripheral origin to their central projection in the outer nerve layer of the olfactory bulb (Doucette, 1991). This putative support role of OECs in axonal growth within the adult CNS has spawned extensive research to study the potential of OEC transplants encouraging axonal regeneration and functional recovery in SCI

Introduction The clinical pathophysiology of contusive spinal cord injury (SCI) is complex. A combination of hemorrhage, ischemia and/or edema develops resulting in necrosis with tissue loss (Schwab et al., 2006, review). Additionally long tracts of the injured spinal cord may survive, but become demyelinated. Although regeneration of spinal cord axons is an ultimate objective, the presence in many patients with non-penetrating SCI of a population of surviving axons that do not conduct impulse due to demyelination (Keirstead, 2005; Kocsis and Sasaki, 2005, reviews), suggests a cell-based approach for remyelination as a potential strategy for inducing recovery of function in SCI. One approach to this goal capitalizes on recent progress in the transplantation of cells into the injured CNS. Corresponding author. Tel.: +1-(203)-937-3802;

Fax: +1-(203)-937-3801; E-mail: [email protected] DOI: 10.1016/S0079-6123(06)61030-3

419

420

models (Li et al., 1997, 1998; Ramon-Cueto et al., 1998, 2000; Imaizumi et al., 2000a, b). Transplantation of OECs after SCI is associated with functional improvement even when transplantation is delayed by weeks (Lu et al., 2002; Keyvan-Fouladi et al., 2003). While the precise mechanisms of the functional recovery after OEC transplantation are not fully understood, several mechanisms including elongative axonal regeneration, axonal sparing, sprouting and plasticity associated with novel polysynaptic pathways, recruitment of endogenous SCs and remyelination have been proposed (Raisman, 2001; Bareyre et al., 2004; Sasaki et al., 2004; Keyvan-Fouladi et al., 2002). In animals with SCI and OEC transplants, myelinated axons spanning the lesion site display a characteristic peripheral pattern of myelination similar to that of Schwann cell (SC) myelination (Franklin et al., 1996; Li et al., 1997, 1998; Imaizumi et al., 1998, 2000b; Sasaki et al., 2004). Remyelination by transplantation may be of particular importance because contusive SCI often results in loss of function from demyelination induced by spinal cord trauma. OECs can form myelin when transplanted into demyelinated spinal cord (Franklin et al., 1996; Imaizumi et al., 1998; Sasaki et al., 2004). Moreover, human OECs can form myelin in the immunosuprressed rat (Barnett et al., 2000; Kato et al., 2000) and OECs transplanted into the non-human primate can remyelinate spinal cord axons (Radtke et al., 2004).

Remyelination in demyelinated spinal cord Three types of SCI models are commonly used in experimental studies in rodents: transection, compression and contusion (Rosenzweig and McDonald, 2004, review). However, because of the complex pathophysiological conditions of SCI in these models, the study of remyelination is more complex. Aside from the heterogeneity of the lesion, endogenous remyelination by SCs occurs in these model systems thus making distinction of endogenous remyelination from that of transplanted cells more complex. A model used by several groups including ours to study remyelination

by transplanted cells is the X-irradiation/ethidium bromide model (X-EB). Ethidium bromide (EB), a nucleic acid chelator, kills cells in the target zone (dorsal white matter) and focal X-irradiation to blocks mitosis and kills oligodendrocyte progenitors (Blakemore and Crang, 1985). In this model system, endogenous repair in the center of the lesion is delayed by 6–8 weeks thus allowing a stable time window to assess transplanted cells (Honmou et al., 1996). Generally our analyses are carried out within 3 weeks after X-EB lesion induction, which is well within the time at which persistent demyelination is conﬁrmed in a large number of control studies (Blakemore and Crang, 1985; Franklin et al., 1996; Honmou et al., 1996; Kato et al., 2000). The lesion induced by this procedure is characterized by virtually complete loss of endogenous glial elements (astrocytes and oligodendrocytes) with preservation of axons (Fig. 1A). The higher power light micrograph of the lesion shows compact ﬁelds of demyelinated axons in close apposition separated by areas of myelin debris and macrophages (Figs. 1A2, A3). The lesion occupies nearly the entire dorso-ventral extent of the dorsal columns for 5–7 mm longitudinally. A large body of work supports the proposal that transplantation of OECs into various SCI and demyelination models can promote axonal regeneration, remyelination and functional recovery (Franklin et al., 1996; Li et al., 1998; Ramon-Cueto et al., 1998; Imaizumi et al., 2000a, b; RamonCueto et al., 2000; Lu et al., 2002; Keyvan-Fouladi et al., 2003; Plant et al., 2003). Yet, there is an important controversy as to whether the transplanted OECs associate with axons and form peripheral myelin, as opposed to recruiting endogenous SCs that form myelin (Takami et al., 2002; Boyd et al., 2004). OECs can express a number of trophic factors, transcription factors and extracellular matrix molecules (Ramon-Cueto et al., 1998; Chuah and West, 2002; Au and Roskams, 2003; Ramer et al., 2004), which could facilitate endogenous SC cell invasion, angiogenesis and activation of progenitor cells to facilitate repair. To address this issue, we used OECs from GFP-expressing rats. After 3 weeks of transplantation into X-EB model, methylene blue/Azure II semithin plastic sections demonstrated extensive

Fig. 1. Light micrographs of transverse sections of the dorsal spinal cord stained with methylene blue/azure II showing the dorsal funiculus in demyelinated (A) and OEC transplanted (B) rats. Examination at higher magniﬁcation show demyelinated (A2, A3) and remyelinated (B2, B3) axons in the dorsal columns. Note the extensive myelination after transplantation of OEC (B). Many myelinforming cells were similar to peripheral myelin-forming cells, characterized by large nuclear and cytoplasmic regions (B3). A2, B2 and A3, B3 were prepared from the boxed area in A1, B1 and A2, B2 respectively. Immuno-EM for GFP shows numerous GFP+ cells remyelinating the demyelinated axons (C, D). Counterstaining for these sections was minimal, and dark staining shows electron dense immunoperoxidase reaction product. Note the electron dense reaction product in the cytoplasm and nuclei of most cells forming myelin (C, D). One cell associated with myelin in this ﬁeld displays distinct reaction product, whereas an adjacent cell does not have electron dense reaction product in its cytoplasm (D1). D2 is an enlargement of the box in D1. Note the dense reaction product in the cytoplasm of the myelin-forming cell on the right and the basement membrane surrounding the cell. Scale bars ¼ 700 mm (A1, B1); 70 mm (A2, B2); 1 mm (A3, B3); 5 mm (C); 2 mm (D1); 0.4 (D2). C and D are modiﬁed with permission from Sasaki et al. (2006a).

422

remyelination in the demyelinated axons (Fig. 1B). To fully establish that the OECs derived from GFP rats were indeed responsible for the remyelination, immuno-electron microscopy (EM) for GFP was performed. Intense reaction product was observed within the cytoplasm and nuclei of cell proﬁles surrounding myelinated axons (Fig. 1C). To establish a ﬁrst approximation of the relative contribution of donor cell to host cell remyelination, quantitative analysis was performed on GFP immuno-EM experiments of two transplanted animals. A total of 136 axons were counted in the lesions, and of these 120 were GFP+ cells. Of the GFP+ cells, four (3.3%) showed no direct contact with axons. The myelin-forming status of ﬁve GFP+ cells (4.2%) could not be determined because of poor membrane preservation. Of the remaining 111 GFP+ cells, 90 cells (81%; 90 of 111) showed a distinct myelin structure. The remaining 21 GFP+ cells (19%; 21 of 111) appeared to be in varying stages of loose wrapping or ensheathing the axons but did not show distinct compact myelin. Within this same region, 16 myelinated axon proﬁles were detected, which were associated with myelin-forming cells that did not contain DAB reaction product. An example of a non-GFP myelinated axon adjacent to a GFP+ myelinating cell is shown in Figs. 1, D1 and D2. Note the presence of a basement membrane surrounding both myelinated axons. In frozen sections, OECs from donor GFP rats (GFP-OEC) exhibited a robust distribution within the X-EB lesioned spinal cord dorsal columns. The GFP-OECs were easily distinguished at this time point by their green ﬂuorescence and were distributed longitudinally along the dorsal columns for 8 mm (Fig. 2A). GFP-OECs were mostly conﬁned to the lesion zone of the dorsal funiculus. GFAP immunostaining indicated a near absence of astrocytes within the lesion site, but intense GFAP staining was observed at the outer boundary of the lesion zone (Figs. 2B, C). In contrast to GFAP immunolabeling, immunostaining for P0, a speciﬁc marker of peripheral myelin (Greenﬁeld et al., 1973), was primarily localized within the lesion and transplantation site (Figs. 2D, E). Ringlet-like P0 immunostaining was associated with GFP-OECs and was surrounded by GFP+-OEC cytoplasm, indicating the transplanted

cells within the dorsal column lesion formed that peripheral-type myelin. Immunostaining for P0 and NF in coronal section reveals a central NF+ axonal core surrounded by P0-identiﬁed myelin, which is wrapped by cytoplasm of a GFP-OEC (Fig. 2E), indicating that transplanted OECs remyelinate the demyelinated axons. Moreover, we also transplanted highly puriﬁed OECs isolated from transgenic pigs expressing the alpha1, 2 fucosyltransferase gene (H-transferase or HT) into a demyelinated lesion of the African green monkey spinal cord (Radtke et al., 2004). Four weeks post-transplantation, robust remyelination was found in 62.5% of the lesion sites, whereas there was virtually no remyelination in the non-transplanted controls. This was the ﬁrst demonstration that xenotransplantation of characterized OECs into the primate spinal cord results in remyelination. This together with the immunohistochemical demonstration of the grafted cells within the lesioned area conﬁrmed that remyelination was indeed achieved by OECs. Nodal reconstruction of remyelinated spinal cord axons The restoration of rapid and secure impulse conduction after demyelination is dependent on the acquisition of myelin sheaths and the clustering of speciﬁc molecules within discrete domains of the myelinated axon membrane. In myelinated axons, voltage-gated sodium (Nav) channels are aggregated in high density at nodes of Ranvier, whereas Shaker-type potassium (Kv1) channels are separated from nodal Nav channels by septate-like paranodal junctions (Peles and Salzer, 2000; Rasband and Trimmer, 2001; Girault and Peles, 2002). Of the seven Nav channel isoforms expressed in nervous tissue (Goldin et al., 2000), Nav1.6 is the predominant one at mature nodes in both the PNS and CNS (Caldwell et al., 2000; Boiko et al., 2001) following a transition from Nav1.2 (Boiko et al., 2001; Kaplan et al., 2001; Jenkins and Bennett, 2002; Rios et al., 2003). The channel clustering (Vabnick et al., 1997; Rasband et al., 1999) and the transition from Nav1.2 to Nav1.6 (Boiko et al., 2001; Rios et al., 2003) are dependent on interaction of the axon with myelinating cells (Kaplan

423

Fig. 2. Sagittal frozen sections through the lesion site demonstrate the distribution of transplanted GFP-OEC. Transplanted cells are primarily conﬁned to the lesion site. Some cells migrated into the deep white matter. The dashed line demarcates lesion edge (A). Coronal frozen sections in the lesion show the presence of GFP-OEC within a lesion site. Transplanted cells survived primarily in the dorsal funiculus. There was little GFAP staining within the lesion zone (B). GFAP-positive cells were present at the peripheral margin of the lesion. These results indicate that few astrocytes are present in the transplant region, and that there is a preponderance of GFPOEC in the lesion zone (B, C). P0-immunostaining (red) of the frozen coronal section reveals that most axons remyelinated by the transplanted OECs are surrounded by peripheral type of myelin. Red-P0 rings are associated with green cellular elements, indicating that transplanted OECs remyelinate the demyelinated axons (D). Expansion of a cell indicated by an arrow (D inset). P0 and neuroﬁlament (NF) staining at lesion boundary (dashed line) and transplant zone (E). Higher magniﬁcation showing neuroﬁlamentdeﬁned axon cores surrounded by P0 myelin rings enwrapped by GFP-OECs (E2). B–E are coronal. Scale bars: 1 mm (A); 400 mm (B); 30 mm (C); 10 mm (D); 20 mm (A, inset); 10 mm (B, inset); 30 mm (E1); 10 mm (E2); 2.5 mm (E2, inset). Modiﬁed with permission from Sasaki et al. (2006a).

424

Fig. 3. Nodal structure of remyelinated axons (A, B) at 3 weeks after OEC transplantation. Sagittal section showing a ﬁeld of myelinated axons interspersed among donor OECs. Note the node structure in the center of the ﬁeld (A). High-power electron micrograph showing node and paranodal loops (B1). B2 is an enlargement of the boxed area in B1. Scale bars ¼ 5 mm (A); 1 mm (B1); 0.5 mm (B2). Modiﬁed with permission from Sasaki et al. (2006a).

et al., 1997; Eshed et al., 2005). Remyelinated axons display inappropriately short internodal lengths (Gledhill and McDonald, 1977; Weiner et al., 1980; Blakemore and Murray, 1981; Hildebrand et al., 1985), indicating that new nodes are formed. Despite their location at formerly internodal sites, remyelinated PNS axons have been shown to display high densities of Nav channels at nodes (Novakovic et al., 1996, 1998) and Kv1 aggregations within juxtaparanodal domains (Rasband et al., 1998). The expression and organization of speciﬁc isoforms of Nav and Kv1 channels in remyelinated CNS axons have not been examined. Recently we reported that EM analysis of spinal cords performed at 3 weeks after GFP-OEC transplantation demonstrated distinct nodes of Ranvier (Fig. 3). Large cytoplasmic and nuclear compartments were present in cells associated with the myelin proﬁles (Fig. 3A). Longitudinal sections of the dorsal columns revealed well formed nodal and paranodal regions; paranodal loops from adjacent myelin-forming cells were readily recognized ﬂanking nodes (Figs. 3B1, B2). Also, in immunohistochemical analysis we observed Nav1.6 staining at most nodes, whereas detectable Nav1.2 immunostaining was not apparent at nodes (Fig. 4). In dorsal columns 3 weeks after transplantation, virtually all nodes bounded by GFP-OEC myelin sheaths exhibited Nav1.6 staining (Figs. 4A–D); similar to control spinal cord axons, Nav1.2 immunolabeling was not observed at any nodes (Figs. 4E–H). The Nav1.6 labeling was localized to the nodal domain and was not observed in paranodal or juxtaparanodal regions or

beneath the myelin sheath in remyelinated axons, suggesting that the transplanted GFP-OECs are competent to contribute to the speciﬁc clustering of Nav channels at nodes. As an additional determinant of the ability of axons myelinated by GFP-OECs to support the asymmetric organization of ion channels within remyelinated nodal regions, we examined the distribution of Kv1.2 in the juxtaparanodal region. Kv1.2, as well as Kv1.1, form heteromultimers with Kv1.4 and Kvb2 (Wang et al., 1993; Rasband et al., 1998) and have been shown previously to be aggregated in juxtaparanodal regions of most spinal cord axons (Rasband et al., 1999; Rasband and Trimmer, 2001). At 3 weeks (Figs. 4I–L), Kv1.2 is aggregated within juxtaparanodal regions of the remyelinated axons, with some nodes exhibiting incursion of Kv1.2 channels into adjacent paranodal regions. Kv1.2 was not observed within nodal areas in these axons.

Transected spinal cord Distribution and myelin formation by transplanted GFP-OECs within the spinal cord Spinal cord injuries without OEC transplants can show limited SC-like myelination, presumably from invasion of the injury site from endogenous SCs (Brook et al., 1998; Imaizumi et al., 2000a, b; Namiki et al., 2000; Takami et al., 2002) or possibly from precursor cells. The degree to which OECs can integrate into injury sites and survive and

425

Fig. 4. Nav1.2 and Nav1.6 at GFP-OEC nodes in remyelinated dorsal columns. At 3 weeks after transplantation, Nav1.6 clustering is displayed at most Caspr-delimited (A–D) nodes formed by GFP-OECs at 3 weeks. In contrast, Caspr-delimited nodes formed by GFPOECs do not exhibit Nav1.2 immunostaining (E–H). Merged images of A–C and E–G are shown in D and H, respectively. Juxtaparanodal Kv1.2 immunolabeling at 3 weeks after GFP-OEC transplantation dorsal columns. Paranodes display Caspr staining (I) that is ﬂanked by Kv1.2 aggregations within juxtaparanodal regions. Merged images of I–K are shown in L, respectively. Scale bars ¼ 10 mm. Modiﬁed with permission from Sasaki et al. (2006a).

whether they form myelin or facilitate endogenous myelin repair mechanisms was controversial. OECs in culture are diverse and exhibit characteristics of astrocytes, SCs and oligodendrocytes. Moreover, they express a number of trophic factors, transcription factors and extracellular matrix molecules

(Ramon-Cueto and Avila, 1998; Chuah and West, 2002; Au and Roskams, 2003) that could facilitate endogenous cell invasion, as well as angiogenesis and activation of progenitor cells. A recent study was unable to ﬁnd evidence of myelination in the compressed spinal cord by the OECs isolated from

426

Fig. 5. Montage image of sagittal frozen section showing distribution of GFP-OECs within and beyond the transection site (A). Semithin plastic sections stained with methylene blue/Azure II through the transection site 5 weeks after transplantation of OECs. Low-power micrograph showing completeness of the transection through the entire dorsal funiculus and beyond (B). EM from the same lesion showing myelinated axons surrounded by a cellular element forming a tunnel (C). EM of anti-GFP immunoperoxidase staining of OEC transplant (D). Reaction product can be seen in cytoplasmic regions of the myelin-forming cell but not in the myelin. A crosscut section of axon showing cytoplastic reaction product. Enlargement of the boxed area is shown in D2. Note the presence of extracellular ﬁbrous elements in D1. Scale bars ¼ 250 mm (A); 0.75 mm (B); 5 mm (C); 1 mm (D1); 0.25 mm (D2). Modiﬁed with permission from Sasaki et al. (2004).

embryonic day 18 rat, infected with a LacZexpressing retrovirus, and suggests that OEC transplantation may facilitate endogenous SC invasion into the lesion site (Boyd et al., 2004). We transplanted genetically labeled GFP-OECs into a dorsal funiculus transection model in the

rat. Five weeks after OEC transplantation, the cells survived and distributed extensively within the transection site and more limitedly beyond the lesion site (Fig. 5A1). High-power micrographs (Fig. 5A2) of AnkyrinG and Caspr staining demonstrate nodal and paranodal regions,

427

respectively, associated with some transplanted GFP-OECs, suggesting forming a new node of Ranvier. Methylene blue/Azure II semithin plastic sections through this region of the lesion 5 weeks after transplantation revealed greater structural detail (Fig. 5B). As reported initially by Li et al. (1997), small groups of myelinated axons were often surrounded by a non-myelinating cell forming tube-like structures around the myelinated axons. This organization is not observed after endogenous repair or after transplantation of SCs (Imaizumi et al., 2000b). These proﬁles were observed in both dorsal and ventral regions of the lesion zone (Fig. 5B). EM further indicates that the myelinated axons in the lesion zone are surrounded by cytoplasmic extensions of cells forming tunnels (Fig. 5C). These clusters of myelinated axons were variable from animal to animal and conﬁned mostly to and near the lesion zone. Rings of P0 labeling were observed in some regions of the lesion and were often associated with GFP cellular elements. EM examination of anti-GFP immunoperoxidase-reacted sections revealed that many detectable GFP+ cells were in direct contact with host axons. Reaction product was clearly evident in the cytoplasm of many cells that formed welldeﬁned multi-laminate structures characteristic of

myelin (Fig. 5D). In a longitudinal section of a myelinated axon, intense reaction product can be seen in the cytoplasm of the myelin-forming cell. Large cytoplasmic and nuclear regions as well as the presence of basal lamina and extracellular ﬁbrils (Fig. 5D) indicate a peripheral pattern of myelination. Taken together, we conclude that transplanted GFP-OECs integrate into the injury site of a dorsal funiculus transection, distribute and associate longitudinally with axons spanning the lesion site and do form myelin.

Improved hindlimb locomotor function All OEC-transplanted animals in this model system exhibited a gradual improvement in hindlimb locomotor function during the 5-week recovery period (Fig. 6). The sham injection group recovered to a Basso, Beattie and Bresnahan (BBB) locomotor rating scale (Basso et al., 1995) of 9 and did not display weight-bearing plantar stepping. However, the OEC transplant group recovered to 18 on the BBB score and displayed consistent weight-bearing plantar stepping. Statistical analysis indicated that the open-ﬁeld locomotor scores at

Fig. 6. Open-ﬁeld locomotor scores for OEC transplant (n ¼ 20) and sham injection (n ¼ 6) groups tested 1 week before and for 5 weeks after transplantation. Modiﬁed with permission from Sasaki et al. (2004).

428

3, 4 and 5 weeks after injury and OEC transplantation was signiﬁcantly higher than sham injection.

Neuroprotection of dorsal corticospinal tract A recent study demonstrates that primary motor cortex (M1) pyramidal neurons undergo apoptotic cell death after axotomizing SCI (Hains et al., 2003). Moreover, earlier work has shown that corticospinal neurons become atrophic after spinal cord transection (McBride et al., 1989). To explore the possibility that OECs are neuroprotective for injured corticospinal tract (CST) neurons, we transplanted OECs into the dorsal transected spinal cord (T9) and examined M1 to assess apoptosis and neuronal loss at 1 and 4 weeks post-transplantation. Triple labeling with Hoechst, Fluorogold (FG), and terminal deoxynucleotidyltransferase-mediated deoxyuridine triphosphate–rhodamine nick end-labeling (TUNEL) identiﬁed putatively axotomized pyramidal cells in the SCI group, with and without spinal OEC transplantation. The neuronal subset undergoing apoptosis was readily identiﬁed in 1week tissue. Within these cells, nuclear morphologies of both apoptotic and surrounding normal neurons and glia could be identiﬁed (Figs. 7A, B). Hoechst staining was observed in all cells and uniformly ﬁlled the spherical nuclear compartment. In a proportion of cells undergoing apoptosis, Hoechst staining revealed abnormal morphology, including the formation of chromatin condensates within a less-distinct nucleus (Hains et al., 2003). Merging of Hoechst, FG and TUNEL staining (Figs. 7A, B) showed that irregular nuclear fragments identiﬁed by Hoechst staining overlap securely with FG- and TUNEL-positive cells, permitting the conﬁdent designation of an apoptotic status. The nuclear morphology of TUNELpositive neurons was identical to that seen in our earlier paper (Hains et al., 2003) (Fig. 7A, insets). Apoptotic (FG- and TUNEL-positive cells) pyramidal neurons were observed both after SCI+ FG+DMEM (Fig. 7A, arrowheads) and in the SCI+FG+OEC group (Fig. 7B, arrowheads), but the number of apoptotic neurons was signiﬁcantly reduced in the SCI+FG+OEC group (Fig. 7C). At 1 week after injury, the number of apoptotic

Fig. 7. Hoechst 33342, Fluorogold (FG), and TUNEL triple labeling of corticospinal neurons 1 week after injury. Hoechst staining of non-TUNEL-positive (arrows with tails) and TUNEL-positive (arrowheads) neurons with corresponding FG-backﬁlling are shown. In SCI+FG+OEC animals (B), fewer TUNEL-positive FG-backﬁlled neurons are observed compared with SCI+FG+DMEM (A). Insets in A show two TUNEL-positive neurons exhibiting nuclear compartmentalization and formation of nucleosomes, hallmarks of apoptosis. Quantiﬁcation of neurons that are both TUNEL- and FGpositive (C) reveals that OEC transplantation signiﬁcantly (*Po0.05) reduces apoptotic cell death at 1 week. No evidence of death was observed at any other time-point. Scale bars ¼ 125 mm in A, B; 20 mm in inset in A. Modiﬁed with permission from Sasaki et al. (2006b).

neurons was signiﬁcantly (Po0.05) reduced by 45% (16.476.40 vs. 36.175.69) in the OEC transplant group compared with the SCI+FG+ DMEM group (Fig. 7C). At 4 weeks, both the SCI+FG+DMEM and SCI+FG+OEC groups

429

showed apoptotic activity comparable with sham controls (0.9870.81 and 0.7570.56) (Fig. 7C). We described the supraspinal effects of OEC transplantation on apoptosis and cell survival of CST neurons within the M1 cortex after transection of their axons in the spinal cord. Our results indicate that apoptosis of primary motor cortical neurons is reduced and that cortical neuronal density is increased after OEC transplantation. Enhanced levels of brain-derived neurotrophic factor (BDNF) were observed in the OEC transplanted lesion. Thus, transplantation of OECs into injured spinal cord has a neuroprotective effect on corticospinal neurons. The relative contribution of this effect to the observed functional improvement after OEC transplantation is uncertain, but this data indicates that OEC transplantation results in a larger pool of surviving corticospinal neurons. Thus, OEC transplantation into the injured spinal cord has distant neuroprotective effects on descending cortical projection neurons as well (Sasaki et al., 2006b).

OEC into spinal cord contusion injury We also evaluated the in vivo fate of OECs transplanted into the contused rat spinal cord with the weight-drop device developed at New York University (NYU impactor) (Li et al., 2004). Transplanted GFP-OECs were distributed widely within transplanted spinal cords. In sagittal sections, GFP-OECs were widely distributed along the length of the lesion (Fig. 8A). Immuno-EM using an anti-GFP antibody revealed that the identiﬁed OECs made direct and extensive contact with host axons (Figs. 8B, C). Transplanted GFP-OECs produced multi-laminate structures surrounding axons in a one-to-one relationship with large amounts of cytoplasm and basal lamina characteristic of peripheral myelin. The typical relationships of transplanted SCs and OECs with host axons were quite different. While the overwhelming majority of GFP positive cells in both transplant conditions were in direct contact with host axons, GFP positive SCs

Fig. 8. Distribution of transplanted OECs into a contused rat spinal cord (A). Sagittal section of a segment of the spinal cord 8 mm rostral and caudal to the injured area showed transplanted OECs concentrated near the center and distributed along the longitudinal axes of the spinal cord at 3 weeks after transplantation (A), arrows indicated transplantation points. A2 and A3 are enlargements of boxed areas in A1. Immuno EM for GFP revealed that many detectable GFP+ cells were in direct contact with host axons (Arrows). Reaction product was clearly evident in the cytoplasm of many cells that formed well-deﬁned multi-laminate structures characteristic of myelin at 4 weeks after transplantation (8B, C). Scale bars ¼ 1 mm (A1); 100 mm (A2, A3); 5 mm (B); 1 mm (C).

430

cells were observed almost exclusively in one-toone associations with host axons, with a high proportion of those contacts associated with myelination, while GFP positive OECs typically wrapped or engulfed several axons with a small percentage of contacts associated with myelination. Clinical perspectives Clinical investigations in SCI using OEC transplantation are in progress. A recent report tested the feasibility and safety of transplantation of autologous OECs into the injured spinal cord in human paraplegia in a single blind, Phase I clinical trial (Feron et al., 2005). In this study, 1 year after cell implantation, there were no medical, surgical or other complications to indicate that the procedure is unsafe. They conclude that transplantation of autologous OECs into the injured spinal cord is feasible and is safe up to 1 year post-implantation. Long-term safety studies to compare neurological, functional and psychosocial outcomes are underway. Furthermore, fetal brain tissue has been transplanted into the lesions of more than 400 patients with SCI in China (Senior, 2002; Huang et al., 2003). Dobkin et al. (2006) compared available reports from China with the experiences and objective ﬁndings of patients who received preoperative and post-operative assessments before and up to 1 year after receiving cellular implants. They concluded that the phenotype and the fate of the transplanted cells, described as OECs in the China study, are unknown. Perioperative morbidity and lack of functional beneﬁt were identiﬁed as the most serious clinical shortcomings. The procedures observed did not attempt to meet international standards for either a safety or efﬁcacy trial. In the absence of a valid clinical trial protocol, Dobkin et al. (2006) suggest that physicians should not recommend this procedure to patients.

strategy as an approach that may elicit some degree of functional recovery, raising the possibility that this approach might be useful in the treatment of SCI. Experimental work indicates that cell transplantation approaches can facilitate axonal regeneration, remyelination, neuroprotection and possible neovascularization. Importantly, axons remyelinated by transplanted OECs form appropriate nodal sodium channel and conduction is improved. Several clinical studies are ongoing using cell therapy approaches for SCI (Senior, 2002; Dobkin et al., 2006). Future translational studies are highly expected to bridge the gap between basic and clinical research in OEC transplantation for SCI. Abbreviations BBB

BDNF CNS CST DMEM EB EM FG K v1 M1 Nav NYU OEC PNS SC SCI TUNEL

Basso, Bresnahan and Beattie (1995; i.e., and their rating scale for locomotion) brain-derived neurotrophic factor central nervous system corticospinal tract Dulbecco’s modiﬁed Eagle’s medium ethidium bromide electron microscopy Fluorogold Shaker-type potassium primary motor cortex voltage-gated sodium New York University olfactory ensheathing cells peripheral nervous system Schwann cells spinal cord injury terminal deoxynucleotidyltransferase (TdT)-mediated deoxyuridine triphosphate (dUTP)-rhodamine nick end-labeling

Concluding remarks Acknowledgments Remyelination of the injured spinal cord is one of the key elements for the functional recovery in SCI. Extensive experimental rodent models of SCI describe the feasibility of OEC transplantation

This work was supported in part by the Department of Veterans Affairs, the NIH, and the National Multiple Sclerosis Society. The Center

431

for Neuroscience and Regeneration Research is a collaboration of the Paralyzed Veterans of America and the United Spinal Association with Yale University. We thank Heather Mallozzi and Margaret Borelli for excellent technical assistance.

References Au, E. and Roskams, A.J. (2003) Olfactory ensheathing cells of the lamina propria in vivo and in vitro. Glia, 41: 224–236. Bareyre, F.M., Kerschensteiner, M., Raineteau, O., Mettenleiter, T.C., Weinmann, O. and Schwab, M.E. (2004) The injured spinal cord spontaneously forms a new intraspinal circuit in adult rats. Nat. Neurosci., 10: 269–277. Barnett, S.C., Alexander, C.L., Iwashita, Y., Gilson, J.M., Crowther, J., Clark, L., Dunn, L.T., Papanastassiou, V., Kennedy, P.G. and Franklin, R.J. (2000) Identiﬁcation of a human olfactory ensheathing cell that can effect transplantmediated remyelination of demyelinated CNS axons. Brain, 123: 1581–1588. Basso, D.M., Beattie, M.S. and Bresnahan, J.C. (1995) A sensitive and reliable locomotor rating scale for open ﬁeld testing in rats. J. Neurotrauma, 12: 1–21. Blakemore, W.F. and Crang, A.J. (1985) The use of cultured autologous Schwann cells to remyelinate areas of persistent demyelination in the central nervous system. J. Neurol. Sci., 70: 207–223. Blakemore, W.F. and Murray, J.A. (1981) Quantitative examination of internodal length of remyelinated nerve ﬁbres in the central nervous system. J. Neurol. Sci., 49: 273–284. Boiko, T., Rasband, M.N., Levinson, S.R., Caldwell, J.H., Mandel, G., Timmer, J.S. and Matthews, G. (2001) Compact myelin dictates the differential targeting of two sodium channel isoforms in the same axon. Neuron, 30: 91–104. Boyd, J.G., Lee, J., Skihar, V., Doucette, R. and Kawaja, M.D. (2004) LacZ-expressing olfactory ensheathing cells do not associate with myelinated axons after implantation into the compressed spinal cord. Proc. Natl. Acad. Sci. U.S.A., 17: 2162–2166. Brook, G.A., Plate, D., Franzen, R., Martin, D., Moonen, G., Schoenen, J., Schmitt, A.B., Noth, J. and Nacimiento, W. (1998) Spontaneous longitudinally orientated axonal regeneration is associated with the Schwann cell framework within the lesion site following spinal cord compression injury of the rat. J. Neurosci. Res., 53: 51–65. Caldwell, J.H., Schaller, K.L., Lasher, R.S., Peles, E. and Levinson, S.R. (2000) Sodium channel Na (v) 1.6 is localized at nodes of Ranvier, dendrites, and synapses. Proc. Natl. Acad. Sci. U.S.A., 97: 5616–5620. Chuah, M.I. and West, A.K. (2002) Cellular and molecular biology of ensheathing cells. Microsc. Res. Tech., 58: 216–227. Dobkin, B.H., Curt, A. and Guest, J. (2006) Cellular transplants in China: observational study from the largest human experiment in chronic spinal cord injury. Neurorehabil. Neural Repair, 20: 5–13.

Doucette, R. (1991) PNS-CNS transitional zone of the ﬁrst cranial nerve. J. Comp. Neurol., 312: 451–466. Eshed, Y., Feinberg, K., Poliak, S., Sabanay, H., Sarig-Nadir, O., Spiegel, I., Bermingham, J.R. and Peles, E. (2005) Gliomedin mediates Schwann cell–axon interaction and the molecular assembly of the nodes of Ranvier. Neuron, 47: 215–229. Feron, F., Perry, C., Cochrane, J., Licina, P., Nowitzke, A., Urquhart, S., Geraghty, T. and Mackay-Sim, A. (2005) Autologous olfactory ensheathing cell transplantation in human spinal cord injury. Brain, 128: 2951–2960. Franklin, R.J., Gilson, J.M., Franceschini, I.A. and Barnett, S.C. (1996) Schwann cell-like myelination following transplantation of an olfactory bulb-ensheathing cell line into areas of demyelination in the adult CNS. Glia, 17: 217–224. Girault, J.A. and Peles, E. (2002) Development of nodes of Ranvier. Curr. Opin. Neurobiol., 12: 476–485. Gledhill, R.F. and McDonald, W.I. (1977) Morphological characteristics of central demyelination and remyelination: a single-ﬁber study. Ann. Neurol., 1: 552–560. Goldin, A.L., Barchi, R.L., Caldwell, J.H., Hofmann, F., Howe, J.R., Hunter, J.C., Kallen, R.G., Mandel, G., Meisler, M.H., Netter, Y.B., Noda, M., Tamkun, M.M., Waxman, S.G., Wood, J.N. and Catterall, W.A. (2000) Nomenclature of voltage-dated sodium channels. Neuron, 28: 365–368. Graziadei, P.P., Levine, R.R. and Graziadei, G.A. (1978) Regeneration of olfactory axons and synapse formation in the forebrain after bulbectomy in neonatal mice. Proc. Natl. Acad. Sci. U.S.A., 75: 5230–5234. Greenﬁeld, S., Brostoff, S., Eylar, E.H. and Morell, P. (1973) Protein composition of myelin of the peripheral nervous system. J. Neurochem., 20: 1207–1216. Hains, B.C., Black, J.A. and Waxman, S.G. (2003) Primary cortical motor neurons undergo apoptosis following axotomizing spinal cord injury. J. Comp. Neurol., 462: 328–341. Hildebrand, C., Kocsis, J.D., Berglund, S. and Waxman, S.G. (1985) Myelin sheath remodeling in regenerated rat sciatic nerve. Brain Res., 358: 163–170. Honmou, O., Felts, P.A., Waxman, S.G. and Kocsis, J.D. (1996) Restoration of normal conduction properties in demyelinated spinal cord axons in the adult rat by transplantation of exogenous Schwann cells. J. Neurosci., 16: 3199–3208. Huang, H., Chen, L., Wang, H., Xiu, B., Wang, R., Zhang, J., Zhang, F., Gu, Z., Li, Y., Song, Y., Hao, W., Pang, S. and Sun, J. (2003) Inﬂuence of patients’ age on functional recovery after transplantation of olfactory ensheathing cells into injured spinal cord injury. Chin. Med. J. (Engl.), 116: 1488–1491. Imaizumi, T., Lankford, K.L., Waxman, S.G., Greer, C.A. and Kocsis, J.D. (1998) Transplanted olfactory ensheathing cells remyelinate and enhance axonal conduction in the demyelinated dorsal columns of the rat spinal cord. J. Neurosci., 18: 6176–6185. Imaizumi, T., Lankford, K.L., Burton, W.V., Fodor, W.L. and Kocsis, J.D. (2000a) Xenotransplantation of transgenic pig olfactory ensheathing cells promotes axonal regeneration in rat spinal cord. Nat. Biotechnol., 18: 949–953.

432 Imaizumi, T., Lankford, K.L. and Kocsis, J.D. (2000b) Transplantation of olfactory ensheathing cells or Schwann cells restores rapid and secure conduction across the transected spinal cord. Brain Res., 854: 70–78. Jenkins, S.M. and Bennett, V. (2002) Developing nodes of Ranvier are deﬁned by ankyrin-G clustering and are independent of paranodal axoglial adhesion. Proc. Natl. Acad. Sci. U.S.A., 99: 2303–2308. Kaplan, M.R., Cho, M.H., Ullian, E.M., Isom, L.L., Levinson, S.R. and Barres, B.A. (2001) Differential control of clustering of the sodium channels Na(v)1.2 and Na(v)1.6 at developing CNS nodes of Ranvier. Neuron, 30: 105–109. Kaplan, M.R., Meyer-Franke, A., Lambert, S., Bennett, V., Duncan, I.D., Levinson, S.R. and Barres, B.A. (1997) Induction of sodium channel clustering by oligodendrocytes. Nature, 386: 724–728. Kato, T., Honmou, O., Uede, T., Hashi, K. and Kocsis, J.D. (2000) Transplantation of human olfactory ensheathing cells elicits remyelination of demyelinated rat spinal cord. Glia, 30: 209–218. Keirstead, H.S. (2005) Stem cells for the treatment of myelin loss. Trends Neurosci., 28: 677–683. Keyvan-Fouladi, N., Li, Y. and Raisman, G. (2002) How do transplanted olfactory ensheathing cells restore function? Brain Res. Brain Res. Rev., 40: 325–327. Keyvan-Fouladi, N., Raisman, G. and Li, Y. (2003) Functional repair of the corticospinal tract by delayed transplantation of olfactory ensheathing cells in adult rats. J. Neurosci., 23: 9428–9434. Kocsis, J.D. and Sasaki, M. (2005) Transplantation of peripheral myelin-forming cells to repair demyelinated axons. In: Waxman S.G. (Ed.), Multiple Sclerosis as a Neuronal Disease. Elsevier, San Diego, pp. 421–433. Li, B., Lankford, K.L. and Kocsis, J.D. (2004) Distribution and migration of GFP-expressing mouse Schwann cells and olfactory ensheathing cells transplanted into contused spinal cord of the rat. Abstract of 34th Annual meeting, Society for Neuroscience. Li, Y., Field, P.M. and Raisman, G. (1997) Repair of adult rat corticospinal tract by transplants of olfactory ensheathing cells. Science, 277: 2000–2002. Li, Y., Field, P.M. and Raisman, G. (1998) Regeneration of adult rat corticospinal axons induced by transplanted olfactory ensheathing cells. J. Neurosci., 18: 10514–10524. Lu, J., Feron, F., Mackay-Sim, A. and Waite, P.M. (2002) Olfactory ensheathing cells promote locomotor recovery after delayed transplantation into transected spinal cord. Brain, 125: 14–21. McBride, R.L., Feringa, E.R., Garver, M.K. and Williams Jr., J.K. (1989) Prelabeled red nucleus and sensorimotor cortex neurons of the rat survive 10 and 20 weeks after spinal cord transection. J. Neuropathol. Exp. Neurol., 48: 568–576. Namiki, J., Kojima, A. and Tator, C.H. (2000) Effect of brain-derived neurotrophic factor, nerve growth factor, and neurotrophin-3 on functional recovery and regeneration after spinal cord injury in adult rats. J. Neurotrauma, 17: 1219–1231.

Novakovic, S.D., Deerinck, T.J., Levinson, S.R., Shrager, P. and Ellisman, M.N. (1996) Cluster of axonal Na+ channels adjacent to remyelinating Schwann cells. J. Neurocytol., 25: 403–412. Novakovic, S.D., Levinson, S.R., Schachner, M. and Shrager, P. (1998) Disruption and reorganization of sodium channels in experimental allergic neuritis. Muscle Nerve, 21: 1019–1032. Peles, E. and Salzer, J.L. (2000) Molecular domains of myelinated axons. Curr. Opin. Neurobiol., 10: 558–565. Plant, G.W., Christensen, C.L., Oudega, M. and Bunge, M.B. (2003) Delayed transplantation of olfactory ensheathing glia promotes sparing/regeneration of supraspinal axons in the contused adult rat spinal cord. J. Neurotrauma, 20: 1–16. Radtke, C., Akiyama, Y., Brokaw, J., Lankford, K.L., Wewetzer, K., Fodor, W.L. and Kocsis, J.D. (2004) Remyelination of the nonhuman primate spinal cord by transplantation of H-transferase transgenic adult pig olfactory ensheathing cells. FASEB J., 18: 335–337. Raisman, G. (2001) Olfactory ensheathing cells – another miracle cure for spinal cord injury? Nat. Rev. Neurosci., 2: 369–375. Ramer, L.M., Au, E., Richter, M.W., Liu, J., Tetzlaff, W. and Roskams, A.J. (2004) Peripheral olfactory ensheathing cells reduce scar and cavity formation and promote regeneration after spinal cord injury. J. Comp. Neurol., 473: 1–15. Ramon-Cueto, A. and Avila, J. (1998) Olfactory ensheathing glia: properties and function. Brain Res. Bull., 46: 175–187. Ramon-Cueto, A., Cordero, M.I., Santos-Benito, F.F. and Avila, J. (2000) Functional recovery of paraplegic rats and motor axon regeneration in their spinal cords by olfactory ensheathing glia. Neuron, 25: 425–435. Ramon-Cueto, A., Plant, G.W., Avila, J. and Bunge, M.B. (1998) Long-distance axonal regeneration in the transected adult rat spinal cord is promoted by olfactory ensheathing glia transplants. J. Neurosci., 18: 3803–3815. Rasband, M.N., Peles, E., Trimmer, J.S., Levinson, S.R., Lux, S.E. and Shrager, P. (1999) Dependence of nodal sodium channel clustering on paranodal axoglial contact in the developing CNS. J. Neurosci., 19: 7516–7528. Rasband, M.N. and Trimmer, J.S. (2001) Developmental clustering of ion channels at and near the node of Ranvier. Dev. Biol., 236: 5–16. Rasband, M.N., Trimmer, J.S., Schwarz, T.L., Levinson, S.R., Ellisman, M.N., Schachner, M. and Schrager, P. (1998) Potassium channel distribution, clustering, and function in remyelinating rat axons. J. Neurosci., 18: 36–47. Rios, J.C., Rubin, M., St Martin, M., Downey, R.T., Einheber, S., Rosenbluth, J., Levinson, S.R., Bhat, M. and Salzer, J.L. (2003) Paranodal interactions regulate expression of sodium channel subtypes and provide a diffusion barrier for the node of Ranvier. J. Neurosci., 23: 7001–7011. Rosenzweig, E.S. and McDonald, J.W. (2004) Rodent models for treatment of spinal cord injury: research trends and progress toward useful repair. Curr. Opin. Neurol., 17(2): 121–131. Sasaki, M., Black, J.A., Lankford, K.L., Tokuno, H.A., Waxman, S.G. and Kocsis, J.D. (2006a) Molecular

433 reconstruction of nodes of Ranvier after remyelination by transplanted olfactory ensheathing cells in the demyelinated spinal cord. J. Neurosci., 26: 1803–1812. Sasaki, M., Hains, B.C., Lankford, K.L., Waxman, S.G. and Kocsis, J.D. (2006b) Protection of corticospinal tract neurons after dorsal spinal cord transection and engraftment of olfactory ensheathing cells. Glia, 53: 352–359. Sasaki, M., Lankford, K.L., Zemedkun, M. and Kocsis, J.D. (2004) Identiﬁed olfactory ensheathing cells transplanted into the transected dorsal funiculus bridge the lesion and form myelin. J. Neurosci., 24: 8485–8493. Schwab, J.M., Brechtel, K., Mueller, C.A., Failli, V., Kaps, H.P., Tuli, S.K. and Schluesener, H.J. (2006) Experimental strategies to promote spinal cord regeneration: an integrative perspective. Prog. Neurobiol., 78: 91–116. Senior, K. (2002) Olfactory ensheathing cells to be used in spinal-cord repair trial. Lancet Neurol., 1: 269.

Takami, T., Oudega, M., Bates, M.L., Wood, P.M., Kleitman, N. and Bunge, M.B. (2002) Schwann cell but not olfactory ensheathing glia transplants improve hindlimb locomotor performance in the moderately contused adult rat thoracic spinal cord. J. Neurosci., 22: 6670–6681. Vabnick, I., Messing, A., Chiu, S.Y., Levinson, S.R., Schachner, M., Roder, J., Li, C., Novakovic, S. and Shrager, P. (1997) Sodium channel distribution in axons of hypomyelinated and MAG null mutant mice. J. Neurosci. Res., 50: 321–336. Wang, H., Kunkel, D.D., Martin, T.M., Schwartzkroin, P.A. and Tempel, B.L. (1993) Heteromultimeric K+ channels in terminal and juxtaparanodal regions of neurons. Nature, 365: 75–79. Weiner, L.P., Waxman, S.G., Stohlman, S.A. and Kwan, A. (1980) Remyelination following viral-inducted demyelination: ferric ion-ferrocyanide staining of nodes of Ranvier within the CNS. Ann. Neurol., 8: 580–583.

This page intentionally left blank

Subject Index

4-aminopyridine (4-AP) 220, 221 a-CaMKII 153 aII-spectrin 333, 413, 414 acceleration input 29 acceleration/deceleration 68, 158, 264, 269–271, 284 acetylcholine 62, 274 acidosis 61, 65, 66, 113, 120, 121 acute spinal cord injury study 217, 219, 222 acute subdural hematoma (ASDH) 118, 239, 266, 293, 294, 296, 298, 300 (see also subdural hematoma) Akt 69, 343, 359, 360 Alzheimer’s disease 81, 253, 255, 303–311, 346 (see also neurodegenerative disease) AMPA receptor 68, 70, 155, 159–162, 164, 279, 332 amyloid precursor protein (APP) 48, 86–89, 130, 255, 272–274, 303–308, 310, 311 amyloid-b peptides 304 antibody microarray 410, 411 aphasia 346–348 apnea 281, 295, 298, 299, 300 APOE genotype 303, 310, 311 apolipoprotein E 255, 303 apolipoprotein E4 310 apoptosis inducing factor (AIF) 84, 87 apoptosis 35, 43–45, 47, 67, 81–88, 90, 121, 127, 128, 218, 223, 224, 283, 319, 343, 355, 403, 415, 428, 429 (see apoptotic) apoptotic 34, 35, 43–48, 67, 81–87, 126, 128, 131, 132, 135, 223, 300, 304, 342, 358, 359, 403, 428 apparent diffusion co-efﬁcient (ADC) mapping 236 aquaporins 185 arachidonic acid 175, 357 arterio venous difference of oxygen (AVDO2) 112, 119, 120, 208, 210

astrocyte swelling 66, 189 astrocytes 21, 31, 34, 61–70, 72, 73, 83, 88, 89, 98, 103, 117, 121, 129, 158, 176, 187–190, 196, 227, 228, 317, 318, 320, 331–333, 342, 343, 371, 420, 422, 425 astrocytic cascades 34 reactive astrocytes 304 astroglia 61, 175, 282, 318, 402 (see also glial cells) athletes 253–258, 263, 280, 282 ATP 33, 34, 62–65, 69, 70, 72, 85, 87, 111, 113–115, 117, 121, 122, 164, 279, 280, 318, 355, 356, 358, 403 autophagy 82, 84, 85 (see also cell death) autoregulation 112, 113, 131, 210–212, 281, 282 metabolic 112, 119 pressure 112, 119 viscosity 112, 119 axolemma 52, 175, 271, 285 axonal damage 49, 90, 144, 267, 269, 271, 275, 278 axotomy 43, 48, 263 delayed 273, 274 primary 285 secondary 263, 271, 273, 274, 277–279, 285 BACE inhibitors 305 Bad 47, 359 basal ganglia 99, 295–297, 306 Bax 47, 84, 86 Bcl-2 47, 69, 70, 72, 84, 86, 343 Bergmann glia 331–333 big black brain 293–295, 297–300 biochemical markers 130, 282, 317, 401, 406, 409, 414, 415 biomarkers 30, 35, 317, 401, 403, 411–415 biomechanical mechanisms 28 biomechanics 13, 18, 23–25, 27, 329, 335 acceleration 17 deceleration 17 deformations 14 435

436

force 14 shear strain 15 strains 14 stress 14 blood brain barrier (BBB) 88, 97–99, 102, 104, 115, 117, 126, 136, 143, 144, 158, 186, 188, 190, 237, 246, 282, 317, 319, 321, 356, 359, 395, 402 blunt trauma 158, 268, 270 bone marrow stem cells (BMSCs) 225 bone marrow stromal cells 217, 226, 227 brain derived neurotrophic factor (BDNF) 62, 229, 343, 359, 429 brain edema 28, 86, 187, 235, 356, 393, 395, 396, 398 brain tissue oxygen tension 113, 207–209 brain water 185–189, 191, 192, 356, 396 bromdeoxyuridine (BrdU) 321, 322, 369, 370 burst ﬁring 195, 196, 200, 201 calcineurin 53, 88, 280 calcium 31–34, 46, 47, 52, 53, 55, 62–64, 66, 68–70, 82, 87, 98, 114, 121, 135, 151, 153, 158, 218, 274, 275, 279, 280, 304, 318, 319, 331, 402, 403 calcium ATPase 273, 279 calcium dysregulation 46, 47, 52, 53 calcium homeostasis 122, 279, 304, 318, 402 calcium oscillations 62 calcium signaling 62 calcium waves 62 intracellular Ca2+ 46, 68 intracellular stores 62 calpain 45–47, 49, 52, 53, 81, 82, 85–88, 91, 98, 131, 135, 218, 279, 280, 333, 403, 411, 414 calpain inhibitors 87, 88, 91 calpain proteolysis 414, 415 caspases 45–47, 81, 84, 85, 122, 218, 304, 306, 403 caspase-3 33, 47, 67, 83–85, 87, 89, 132, 306, 358, 403, 414, 415 caspase-12 83, 84 caspase-3 proteolysis 414, 415 caspase-inhibitor 85 causes of neurotrauma 3–6 accidents 4 falls 4 ﬁrearm injuries 5

gunshot wounds 4 trafﬁc injuries 4 cell culture 63, 85, 175, 256, 305, 318, 319, 343, 356, 369, 376, 379, 381 cell death (CD) 13, 20, 21, 23, 24, 33–35, 43–48, 66, 67, 72, 81–87, 114, 122, 125, 126, 128, 131, 132, 135, 144, 147, 153, 164, 218, 255, 257, 263, 279, 304, 306, 319, 327, 329–331, 335, 356, 359, 402, 403, 428 (see also apoptosis, necrosis, autophagy) cell membrane damage 19, 20 cellular 43 mechanisms 62, 147, 156, 157, 259, 405 morphology 61 response 13, 16, 18, 19, 21, 22, 43, 132, 177, 284 cerebellar injury 327 cerebellum 44, 62, 65, 175, 256, 257, 327–329, 331, 333, 335 cerebral blood ﬂow (CBF) 111, 112, 186, 208, 236, 263, 278, 281, 296 cerebral blood volume (CBV) 112, 113 cerebral contusion 27, 66, 235, 266, 283, 405 cerebral cortex 48, 83, 130, 135, 145, 146, 165, 257, 307, 327, 335 cerebral edema 102, 120, 185–188, 190, 246, 279 (see also edema) cytotoxic edema 185, 188–190, 192 vasogenic edema 185, 188, 190–192 cerebral energy metabolism 114, 115 cerebral ischemia 65–67, 87, 120, 127, 130, 136, 396 (see also ischemia) cerebral metabolic rate of oxygen (CMRO2) 112, 119, 120, 208, 210, 281 cerebral oxygen tension 111 cerebral oxygen transport 111, 118 cerebral oxygenation 121, 207, 209 cerebral perfusion pressure 112, 120, 207 cerebral tissue PO2 111, 113, 121 cerebrospinal ﬂuid (CSF) 98, 119, 185, 266, 299, 307, 320, 386, 406, 414 cerebrovascular permeability 237, 238 cerebrovascular reactivity 112 carbon dioxide (CO2) reactivity 112 Cethrins 217, 223, 224, 229 (see also Rho antagonist) c-fos 343 CGRP 99, 101, 102, 196 child abuse 253, 254, 293

437

chloride 98 chondroitin sulfate proteoglycan (CSPG) 132 citric acid cycle 111, 115, 117, 118 c-jun 343 cognitive dysfunction 253–255, 257 cognitive performance 320, 321, 345 collagen gel 385–388 community costs 6 computer tomography (CT) 119, 402 concussions 254, 264 controlled cortical impact (CCI) 67, 86, 88, 90, 127, 271, 319, 395, 405, 412 contusion volume 136, 283, 284, 395, 396 contusions 17, 27, 44, 45, 98, 120, 207, 235, 237–239, 241, 266, 270, 283, 298, 395, 396 corpus callosum 55, 126, 266, 267, 334, 342, 371, 378 cortical circuits 200 cortical contusion 144, 270 craniectomy 212, 299, 393–398 critical time frames 244 crossed cerebellar diaschisis (CCD) 329 cross-tolerance 256, 257, 353, 356, 358, 359 CSF production 185–187, 192 CT scan 4, 6, 119, 120, 239, 243, 244, 246–248, 270, 283, 294 0 0 cyclic guanosine 3 ,5 -monophosphate (cGMP) 171, 173 cGMP synthesis 174 cyclosporine A (CsA) 81, 88 cysteine proteases 43, 45, 46, 52, 53, 403 (see also calpains and caspases) cytochrome C 46, 52, 83, 122, 223 cytokines 87, 89, 127, 176, 225, 279, 358, 359, 368, 410 IL-1b 127 IL-6 127 TNFa 127 cytoskeleton 31, 46, 52, 53, 87, 272, 274, 275, 277, 279, 285, 318 decompression craniectomy 299, 393–398 dementia pugilistica 254, 255, 258, 306 demyelination 125, 128, 131, 132, 220, 221, 227, 228, 419, 420, 422 depolarization 33, 55, 62, 67, 68, 155–157, 164, 274, 276, 278–280, 331, 403 dexanabinol 244, 246, 247

diffuse axonal injury (DAI) 17, 27, 31, 49, 53, 55, 126, 264, 266, 298, 306, 402 diffuse head injuries 270 diffuse injuries 17, 27, 29, 43, 44, 55, 56, 90, 335 (see also DTBI) diffuse traumatic brain injury (DTBI) 43–49, 53, 55, 56 (see also diffuse injuries) drug delivery 369, 385–388 delivery strategies 385 injectable delivery 386 intrathecal drug delivery 385 dynamic brain injury 404 dynamic loading 15, 16, 265 edema 28, 65, 66, 86, 88, 97–99, 101–103, 119, 120, 126, 131, 136, 185–192, 235, 237–240, 246, 296, 356, 359, 393, 395, 396, 402, 405, 419 (see also cerebral edema) cytotoxic edema 65, 98, 119, 120, 126, 185, 188, 189, 190, 192 vasogenic edema 97–99, 102, 104, 185, 190–192, 237, 238, 396 electrophysiology 62, 143, 144, 156, 158, 159, 165, 327 energy 45, 65, 82, 84, 85, 87, 90, 111, 114, 115, 117, 118, 121, 268–270, 280, 281, 304, 319, 358, 404, 406 energy metabolism 111, 114, 118, 121, 319 enzyme-linked immunosorbant assays (ELISA) 413 epidemiology 3, 264, 303, 328 EURODERM study 309 MIRAGE study 309, 310 epidermal growth factor 385 erythropoietin (Epo) 353, 357, 358 European Union Directive 244, 247 excitatory amino acid (EAA) 46, 90, 97, 175, 279, 299, 342, 402 excitatory neurotransmitters 279 excitotoxicity 33, 65, 67, 218, 223, 330, 346, 403 experimental models 13, 27–29, 90, 115, 187, 188, 222, 248, 253–256, 285, 296–298, 300, 356, 357, 381, 394, 419 cellular injury models 14 compression injury 18, 385–388 contusion 18 ﬂuid percussion injury 18, 72, 144, 271, 277, 319, 320, 327, 405 of stroke 369

438

piglet subdural injection models 298 rodent subdural hematoma model 296 weight drop 18, 30, 155, 156, 254, 255, 257, 405, 429 extracellular signal regulated kinase (ERK) 34, 69 ERK1 343 ERK2 343 Fampridines 220, 221, 229 (see also 4-aminopyridine or 4-AP) fas ligand 84 ﬁeld extracellular postsynaptic potentials (fEPSP) 147 ﬁnite element analysis (FEA) 13, 21, 23, 24 ﬂuid percussion injury 18, 144, 271, 277, 279, 319, 320, 327, 405 (see also lateral ﬂuid percussion injury) focal head injuries (FHI) 270 focal injuries 16, 27, 44, 270, 331 free radicals 35, 46, 47, 85, 136, 159, 177, 279, 342, 357, 402 funding 7, 8, 24, 259, 391 fura-2 158 g-aminobutyric acid (GABA) 33, 62, 146, 148, 150, 156, 162, 163, 171 GABAA receptor 147, 149–151, 158, 162, 342 gap junctions 63, 66, 72, 73 gene expression 30, 34, 47, 98, 100, 329, 331, 398 GFP 371, 420, 422, 427, 429, 430 Glasgow Coma Score (GCS) 281, 413 Glasgow outcome scale (GOS) 240, 328 Glasgow coma scale (GCS) 240 glia 30, 33, 34, 62–64, 67, 81–83, 86, 98, 103, 117, 118, 144, 158, 160, 186, 225, 256, 284, 331–333, 342, 428 glial cells 30, 61, 64, 98, 114, 144, 160, 228, 236, 279, 283, 304, 317, 333, 340, 342, 358 glial ﬁbrillary acidic protein (GFAP) 61, 64, 67, 134, 196, 282, 422 glial scar 64, 226, 331, 379 glial swelling 44, 66, 98, 126 glucose 65–67, 111, 115, 117, 120, 122, 278, 280, 281, 319 glucose metabolism 120, 278, 281

GLUT transporters 115 GLUT-1 115 GLUT-3 115 glutamate 17, 32, 46, 64–65, 67, 117, 148, 159–162, 218, 223, 254, 279, 296, 318, 319, 332, 342, 402, 403 glutamate receptors 31, 32, 63, 65, 160, 162, 279, 318 N-methyl D-aspartate receptors (NMDAR) 31–35, 159, 160, 161 a-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors (AMPAR) 32, 33, 160–162, 164 metabotropic glutamate receptors (mGluRs) 32, 63, 160, 318 glutamate transporters 65, 332 GLT-1 65 GLAST 65 glycolysis 65, 84, 111, 115–117, 121, 279, 280 GM-1 218, 222, 228 granulocyte colony stimulating factor (G-CSF) 367, 371, 375 growth factor 135, 226, 320, 358, 368, 379, 385, 391 Hagen–Poiseuille equation 112, 119 head injury 4, 6, 17, 18, 30, 47, 87, 88, 111, 118–122, 257, 264–268, 271, 274, 280–284, 293, 294, 298, 299, 306, 307, 309, 310, 329, 356–357, 402, 404 minor 6 moderate 6 severe 6 heat acclimation 257, 353–360 heat shock proteins 331, 355 helmets 7, 14, 97, 263, 264 bicycle helmets 7 hockey helmets 268 motorcycle helmets 7 hematoma 16, 27, 44, 66, 67, 118–120, 209, 210, 228, 239, 266, 268, 270, 293–296, 298, 300 hemicraniectomy 299, 300 high-throughput immunoblotting 411 hippocampal 20, 63, 66, 85, 86, 144–148, 156, 160, 162, 164, 253, 256, 304, 306, 308, 319–321, 341, 343, 355, 411 (see also hippocampus)

439

hippocampus 44, 45, 48, 62, 64, 67, 72, 83, 86, 89, 99, 101, 126, 136, 145–147, 151, 153, 161, 162, 225, 254–257, 307, 308, 310, 320, 321, 327, 329, 341–343, 346 CA1 145 CA3 145 dentate gyrus (DG) 145 perforant pathway 153 histopathology of cerebral contusion 235 hormonal cycling 339, 346 hormonal status 339–341, 344–348 hyaluronan and methylcellulose (HAMC) 385, 388–391 hyaluronan 385, 388 hydrogel 18, 367, 379, 385, 386, 391 hydrolytic enzymes 43, 47, 56 hyperglycolysis 120, 280 hyperventilation 98, 112–114, 119, 120, 210–212 hypotension and hypoxia 118 hypothermia 81, 90, 91, 98, 135, 136, 211, 212, 247 hypoxia 65–67, 90, 102, 115, 118, 119, 121, 122, 210, 211, 296, 298, 300, 301, 343, 357–359 hypoxia inducible factor-1 353, 357, 358 ice hockey 263, 265, 267–271, 275, 286 ICP and cerebral circulation 119, 120 immunoreactivity 64, 67, 87, 89, 99, 100, 103, 177, 255 impact loading 16, 266 impulsive loading 16, 266, 267 in vitro models 18, 21, 24, 27–30, 34, 35, 68, 143, 144, 158, 253, 257 cell/substrate stretch model 30 scratch model/laceration 30 weight drop/compression 30 in vitro TBI models 14, 18, 159 (see also experimental models) in vivo animal models 17, 278 (see also experimental models) in vivo imaging 35, 377, 378 in vivo models 17, 18, 21, 27–29, 31, 33, 143, 144, 253, 354, 358 cortical impact model 29, 86 ﬂuid percussion model 29 incapacity 244, 246 acute incapacity 247

incidence of spinal cord injury 4 incidence of traumatic brain injury 4 infancy 293, 294, 298, 344 inﬂammation 98, 127, 128, 135, 160, 176, 178, 188, 218, 224, 300, 340, 346, 353, 357, 359, 388, 410 neurogenic inﬂammation 97, 100–104 inﬂammatory processes 127, 130, 134, 348 inﬂammatory response 45, 46, 101, 128, 130, 136, 176, 177, 224, 284, 357, 389, 390, 402 informed consent 243–248 inhibitory postsynaptic current (sIPSC) 150, 156 inositol-1,4,5 trisphosphate (IP3) 32, 34, 68, 69, 72 IP3 receptor 63 intensive care 6, 118, 243, 249, 281, 299, 415 interleukin-1 127 interleukin-1b (IL-1b) 223, 359 international neurotrauma society (INTS) 3, 8 intracranial pressure (ICP) 6, 28, 49, 66, 98, 112, 118, 120, 186, 207, 235, 281, 296, 299, 300, 393, 403 monitoring 111 ionic ﬂuxes 278, 281 ionic homeostasis 31, 33, 45, 66, 82, 111, 144, 164, 218, 263 ischemia 64–68, 72, 87, 88, 102, 120, 121, 127, 130, 136, 158, 207, 209, 210, 222, 236, 254, 256, 282, 284, 296, 298, 300, 322, 329, 343, 346, 354, 356, 357, 359, 396, 419 ischemic preconditioning (IPC) 256, 343 jugular bulb oximetry 208 jugular bulb oxygen saturation (SjvO2) 111, 120, 121, 207–212 kainate receptors 148 laceration 30, 83, 97, 131, 266 lactate 65, 84, 111, 113, 115, 117, 120, 121, 210, 211, 280, 319 lactate-to-pyruvate ratio 121, 211 laser doppler ﬂowmetry 111 lateral ﬂuid percussion injury (LFPI) 72, 144, 145, 147, 148, 150–153, 155, 156, 164, 279 (see also ﬂuid precursion injury) lesion progression 131, 235

440

loads 14–19, 266 dynamic loading 15, 16, 265 impact loading 16, 266 impulsive loading 16, 266, 267 static loading 15 long-term depression (LTD) 151, 153 long-term potentiation (LTP) 149, 151–153, 161, 257, 320, 341 low molecular weight antioxidants (LMWA) 357 lucifer yellow 19, 153 lysosomal enzymes 47, 48 (see also hydrolytic enzymes) cathepsin 47 macrophage 98, 128, 130, 134, 176–178, 224, 225, 227, 370, 373, 381, 389, 420 magnesium 31, 98, 102, 103 magnetic resonance imaging (MRI) 127, 134, 227, 228, 235, 270, 283, 296, 341, 348, 367, 368, 377, 381, 402, 414 diffusion-weighted MR 236 marker (see biomarker) mass spectrometry 401, 403, 406, 407, 409, 411, 415 tandem mass spectrometry 409, 410 matrix metalloproteases 85 mechanical permeability 32 mechanical response 13–16, 21, 23, 24 mechanoporation 32, 48, 49, 84, 88, 98 dextrans 48, 50, 51, 53 horseradish peroxidase 48 membrane resealing 43, 49 mediators 61, 87, 98, 102, 130, 224, 226, 279, 359, 360 metabolic cascade 278, 357 metabotropic glutamate receptors (mGluRs) 32, 34, 63, 69, 70, 160, 318 mGluR1 160 mGluR5 160 methylcellulose 385, 388 methylprednisolone 135, 178, 217, 218, 221, 222, 228, 229, 385 MicroBeads 367, 377, 378, 381 microdialysis 46, 111, 120, 121, 175, 211, 278, 296 microglia 31, 35, 89, 98, 130, 176, 196, 304, 329, 359, 370, 373, 389, 402 microtubules 271–274, 277, 279, 280, 285

mild head injury 6, 274, 309 mild injury 73, 144, 158, 160, 253, 254, 263 miniature inhibitory postsynaptic currents (mIPSCs) 150, 151, 156 minocycline 217, 223 minor traumatic brain injury (mTBI) 263 (see also mild injury) minor head injury 6 (see also mild head injury) mitochondria 33, 34, 44, 52, 68, 72, 84–86, 90, 113–115, 117, 118, 121, 271, 275, 279, 280 electron transport chain 85, 117 mitochondrial function 111, 121, 122, 159 mitochondrial permeability transition pore (MPTP) 84, 122 mitogen activated protein kinase (MAPK) 33–35, 69 MK-801 159 Monro–Kellie doctrine 119 morphological injury 97 morris water maze (MWM) 145, 254, 255, 320 multi modal monitoring 207 myelin 129, 136, 137, 220, 223–226, 228, 274, 282, 342, 371, 420–427, 429 (see also remyelination, demyelination) Na+-K+ ATPase 65, 164 Na+/K+ ATPase activity 164 nanoparticles 369, 373, 375, 377, 381 iron oxide 367, 368, 371, 379 near infra red spectroscopy (NIRS) 111, 207, 209–211 necrosis (see necrotic) aponecrosis 45 necrotic 34, 43–49, 67, 82, 85–88, 90, 126, 218, 235, 238–241, 395, 403, 419 nerve growth factor (NGF) 226, 343 NeuN 322, 370 neural progenitor cells (NPCs) 225, 226, 369 neural stem cells (NSCs) 136, 225, 226, 320 neurodegenerative disease 45, 125, 253, 255, 256, 403 (see also Alzheimer’s disease, Parkinson’s disease) neuroexcitation 46 neuroﬁbrillary tangles 304, 307, 308, 310 neuroﬁlaments 46, 88, 271–274, 277, 278, 280, 285, 379 neurogenesis 225, 226, 320–322, 340 neurokinin A (NKA) 99–102

441

neurokinin receptors 100 (see also tachykinin receptors) neuron speciﬁc enolase (NSE) 68, 256, 282, 283 neuronal cascades 33 neuronal transmitter 62, 171, 173 neuropeptides 97–102, 104 neuroprotection 3, 61, 64, 217, 223, 342–344, 353, 354, 356–360, 428, 430 neuroproteomics 401, 403, 409, 411, 413, 415 neurotransmitter receptors 62, 158 NF200 333 NF-kappaB 130 nicotinamide adenine dinucleotide phosphate diaphorase (NADPHd) 171–173, 177 nitric oxide 46, 88, 89, 98, 101, 113, 136, 171–179, 223 nitric oxide synthase (NOS) 46, 88, 171, 318 brain nitric oxide synthase (bNOS) 172 calcium-dependent nitric oxide synthase (cNOS) 176 inducible NOS (iNOS) 136, 176 neuronal nitric oxide synthase (nNOS) 175 NK1 receptor antagonists 103, 104 NMDA antagonists 34 NMDA receptor 46, 64, 68, 84, 148, 159, 160, 162, 175, 279, 341 N-methyl-D-aspartate (NMDA) 148, 175, 279 nodal reconstruction 422 nogo 223, 224 anti-NOGO monoclonal antibodies 217 olfactory ensheathing cells (OECs) 217, 225, 228, 369, 419, 420, 422, 424, 425, 428–430 oligodendrocytes 35, 125, 128, 132, 176, 224, 226, 227, 342, 371, 420, 425 oligodendroglial precursor cells (OPCs) 225, 227 osmotic potential 239 oxidative phosphorylation 65, 111, 115, 117, 118, 280, 281 oxidative stress 97, 122, 308, 356, 357 oxygen and hemoglobin 113, 114 pain 99, 101, 104, 134, 135, 195–201, 229 Parkinson’s disease 81, 130, 371, 403 (see also neurodegenerative disease) PARP-1 86

patch clamping 158, 333 pathogenesis 45, 51, 53, 56, 67, 83, 87, 89, 130, 187, 190, 303, 304, 357 pathomechanisms 125, 126, 128, 131, 136 pathophysiology 13, 23, 61, 65, 70, 73, 97, 99, 111, 118, 125–127, 131, 185, 195, 235, 275, 283, 293, 296, 299, 300, 327, 335, 359, 393, 395, 397, 419 patient protection 243, 248 peptide chromatography 407, 409 peroxynitrite (ONOO) 175 pH 66, 72, 117, 120, 356, 368 brain pH 65, 121, 209 pH paradox 121 pharmacological therapies 157, 218, 229, 248 phospholipases 68, 279 phospholipase A2 280 phospholipase C 32, 63, 280, 305 polypharmacia 91 poly-hydroxypropylmethacrylamide 367, 379 population spike amplitudes (PSA) 343, 344 positron emission tomography (PET) 120, 207, 208, 210–212, 278 post-tetanic potentiation (PTP) 151 potassium channel blockers 217, 221 potassium channels 220, 221, 227, 422, 424 antagonism of 220 potassium 33, 98, 117, 217, 218, 220, 279 preconditioning 256, 257, 344, 353, 355 ischemic preconditioning 256, 343 presenilin 304, 305, 306 pressure changes 28, 398 pressure volume index 119 presynaptic hyperexcitability 329–331 primary damage 13, 19, 23, 395, 396 primary phase 14, 19 progenitor cells 136, 225, 226, 320, 367–369, 372, 377, 420, 425 programmed cell death 82, 126 (see also apoptosis) progressive injury 125, 127, 128, 130, 132, 135, 136 propidium iodide 70, 72, 158, 256 prostaglandins 113, 136, 279 proteases 43, 45, 46, 52, 53, 55, 82, 84, 86, 100, 122, 280, 305, 306, 403 protein kinase C (PKC) 46, 160, 162, 305 proteomics 401, 406, 411 (see also neuroproteomics)

442

proxy consent 243, 244, 247 Prussian blue staining 367, 372, 373, 376, 379, 380 purinergic receptors 69, 72 P2X receptors 69 P2Y receptors 69 Purkinje cell 327–333 randomized controlled trial (RCT) 219, 243,244 enrollment criteria 244 reactive oxygen species (ROS) 34, 84, 86, 87, 90, 218, 357, 358 rehabilitation 97, 202, 217, 219, 220, 310, 328, 340, 346 remyelination 131, 132, 134, 226, 342, 419, 420, 422, 430 repeated injuries in vitro 256 repeated TBI 253, 254, 257, 258 (see also repetitive traumatic brain injury) repetitive traumatic brain injury, 253 Rho 223, 224, 386 Rho antagonist 217, 224 risk–beneﬁt ratio 244, 246, 248 road trafﬁc accident mortality 5 rotational head injury 307 S-100B protein 256, 317, 320 S100B 317–322 neurotrophic properties of S100B 318 neuroprotective effect of S100B 319 S-100a protein 282 Schwann cells 132, 225, 226, 228, 282, 284 second impact syndrome 285, 286 second messengers 72 secondary insults 118, 119, 126, 207, 254, 281, 296, 321, 403, 414, 415 secondary phase 14, 21, 23 secretases 304, 305 a-secretase 304, 305 b-secretase 304, 305 g-secretase 304, 305 seizure threshold 143, 156 seizures 64, 145, 299, 300, 341, 347 sex differences 339 sex steroids 86, 345 estradiol 342, 346 estrogen 86, 309, 340–344, 346, 348 progesterone 309, 342, 344–346, 348 testosterone 86, 342, 344, 348

sexual dimorphism 340, 341, 343 shaken baby syndrome 254 signal transduction 171, 173, 174, 318, 398 small interfering RNA (siRNA) 189 sodium 32–34, 55, 61, 65, 68–70, 98, 135, 195–200, 218, 221, 227, 422 sodium channel 32, 135, 196, 221, 430 Nav1.3 sodium channel 195, 197–199, 201 sodium/calcium exchangers 55, 67 spectrin 46, 47, 52, 87, 88, 273, 274, 280 sports 3, 24, 253, 254, 263–265, 269, 275, 401 American football 264 boxing 268 ice hockey 263 return-to-play guidelines 263, 285 soccer 264 spreading depression 30, 279 static injury 403 stem cells 136, 217, 225–227, 320, 367, 369, 371, 372, 379, 381, 387 bone marrow mesenchymal stem cells 367 embryonic stem cells 367, 369, 371 strain-injury 68 (see also stretch) strains 14, 16, 17, 19, 21, 23, 28, 55 stretch 30–32, 35, 68, 143, 144, 151, 158–164, 253, 256, 257, 271, 275, 277, 279, 318 stretch injury 35, 143, 158–162, 164, 253, 256, 271, 277, 279, 318 striatum 191, 343 stroke 31, 89, 97, 103, 136, 185, 188, 217, 248, 339, 346, 348, 354, 358, 367, 369, 371, 394, 412 subcellular 43–46, 51, 52, 55, 56, 153 subdural hematoma (SDH) 27, 66, 67, 118, 120, 210, 239, 266, 280, 293–296, 298, 300 (see also acute subdural hematoma) substance P 97–104, 196 superoxide 46, 175, 177 surgery 178, 218, 219, 224, 238–241, 344, 354, 394 surgical decompression 217–219, 393, 394, 396, 398 surgical treatment 217, 219, 235, 239 synaptic circuits 143, 144 synaptic plasticity 32, 69, 148, 149, 151–153, 257, 333, 335 (see also long-term potentiation, long-term depression)

443

tachykinin receptors 100, 103 (see also neurokinin receptors) tensile strain 68, 156, 158, 266–268, 273, 274, 276 (see also stretch) thalamus 44, 48, 130, 144, 195, 199–201 ventroposterior lateral (VPL) nucleus 195, 199 VPL neurons 195, 199–201 (see also ventroposterior lateral nucleus) therapeutic window 33, 88–90, 218, 227, 246, 381, 396 therapies 27, 34, 52, 97, 127, 137, 178, 217, 218, 220, 229, 300, 305, 369, 402, 405, 415 time frames 243–245 time window 243, 244, 246–248, 335, 368, 420 (see also critical time frames) tissue deformation 15, 16, 28, 31 tissue oxygen response 211 tissue strain 18, 28 tissue tolerances 14, 19, 21 tolerance criteria 13, 14, 18, 23, 24 transcranial Doppler ﬂowmetry 111

transplantation 136, 226–228, 367–369, 377, 379, 419–422, 424, 426, 427–430 cellular transplantation 136, 419 OEC transplantation 228, 420, 426, 428–430 traumatic axonal injury (TAI) 48, 85, 271, 327 treatment strategies 14, 17, 90, 111, 125, 300, 353 TrkA receptor 343 tumor necrosis factor (TNF) 84, 89, 177, 223, 359 tumor necrosis factor alpha (TNF-a) 127, 223, 225, 229, 359, 360, 410 TUNEL 82, 87, 89, 306, 428 two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) 407, 408, 409 voltage gated calcium channels 32, 55 voltage-gated sodium channels 32, 195, 196, 422 Nav1.3 195, 197–199, 201 weight drop model 155, 156, 254, 255, 257, 405, 406 whole-cell patch-clamp recordings 198 whole-cell patch voltage-clamp recordings 148

This page intentionally left blank