Principles of Molecular Neurosurgery

Principles of Molecular Neurosurgery Progress in Neurological Surgery Vol. 18 Series Editor L. Dade Lunsford Pitts...

Author: U. J. Moore

134 downloads 2473 Views 5MB 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

Principles of Molecular Neurosurgery

Progress in Neurological Surgery Vol. 18

Series Editor

L. Dade Lunsford

Pittsburgh, Pa.

Principles of Molecular Neurosurgery

Volume Editors

Andrew Freese Minneapolis, Minn. Frederick A. Simeone Philadelphia, Pa. Paola Leone Camden, N.J. Christopher Janson Camden, N.J.

96 figures, 12 in color, and 25 tables, 2005

Basel · Freiburg · Paris · London · New York · Bangalore · Bangkok · Singapore · Tokyo · Sydney

Andrew Freese, MD, PhD

Frederick A. Simeone, MD

Department of Neurosurgery, University of Minnesota School of Medicine MMC 96, 420 Delaware St., S.E., Minneapolis, MN 55455 (USA)

University of Pennsylvania, Simeone Neuroscience Center, Pennsylvania Hospital, 800 Spruce St., Philadelphia, PA 19107 (USA)

Paola Leone, PhD

Christopher Janson, MD

Cell and Gene Therapy Center, UMDNJ-Robert Wood Johnson Medical School, Division of Neurosurgery, 401 Haddon Ave., Rm 390, Camden, NJ 08103 (USA)

Cell and Gene Therapy Center, UMDNJ-Robert Wood Johnson Medical School, Division of Neurosurgery, 401 Haddon Ave., Rm 390, Camden, NJ 08103 (USA)

Library of Congress Cataloging-in-Publication Data Principles of molecular neurosurgery / volume editors, Andrew Freese ... [et al.]. p. ; cm. – (Progress in neurological surgery, ISSN 0079-6492 ; v. 18) Includes bibliographical references and indexes. ISBN 3-8055-7784-2 (hard cover : alk. paper) 1. Nervous system–Diseases–Treatment. 2. Molecular neurobiology. 3. Nervous system–Surgery. [DNLM: 1. Nervous System Diseases–therapy. 2. Gene Therapy–methods. WL 140 P9573 2005] I. Freese, Andrew. II. Series. RC349.8.P75 2005 616.8–dc22 2004024871 Bibliographic Indices. This publication is listed in bibliographic services, including Current Contents® and Index Medicus. Drug Dosage. The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any change in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug. All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher. © Copyright 2005 by S. Karger AG, P.O. Box, CH-4009 Basel (Switzerland) www.karger.com Printed in Switzerland on acid-free paper by Reinhardt Druck, Basel ISSN 0079–6492 ISBN 3–8055–7784–2

Contents

IX Dedication X Acknowledgements XI Series Editor’s Note Lunsford, L.D. (Pittsburgh, Pa.) XIII Foreword Anderson, W.F. (Los Angeles, Calif.) 1 Introduction Freese, A. (Minneapolis, Minn.); Janson, C.; Leone, P. (Camden, N.J.); Simeone, F.A. (Philadelphia, Pa.) The Spine and Spinal Cord 5 The Molecular Basis of Intervertebral Disc Degeneration Leo, B.M.; Walker, M.H. (Charlottesville, Va.); Anderson, D.G. (Philadelphia, Pa.) 30 Genetics of Degenerative Disc Disease Kurpad, S.N.; Lifshutz, J. (Milwaukee, Wisc.) 37 Gene Therapy for Degenerative Disc Disease Kim, J.; Gilbertson, L.G.; Kang, J.D. (Pittsburgh, Pa.)

V

52 Bone Morphogenetic Proteins. Spinal Fusion Applications Bomback, D.A.; Grauer, J.N. (New Haven, Conn.) 65 Cellular and Gene Therapy Approaches to Spinal Cord Injury Steinmetz, M.P.; Liu, J.K.; Boulis, N.M. (Cleveland, Ohio) 104 Neural Stem Cell Transplantation for Spinal Cord Repair Iwanami, A.; Ogawa, Y.; Nakamura, M.; Kaneko, S. (Tokyo/Osaka); Sawamoto, K. (Osaka); Okano, H.J.; Toyama, Y.; Okano, H. (Tokyo/Osaka) Functional and Restorative Molecular Neurosurgery 124 Contemporary Applications of Functional and Stereotactic Techniques for Molecular Neurosurgery House, P.A.; Rao, G.; Couldwell, W. (Salt Lake City, Utah) 146 Xeno-Neurotransplantation Schumacher, J.M. (Miami, Fla.) 154 Adeno-Associated Viral Vectors for Clinical Gene Therapy in the Brain Samulski, R.J.; Giles, J. (Chapel Hill, N.C.) 169 Molecular Mechanisms of Epilepsy and Gene Therapy Telfeian, A. (Lubbock, Tex.); Celix, J. (Seattle, Wash.); Dichter, M. (Philadelphia, Pa.) 202 Emerging Treatment of Neurometabolic Disorders Brady, R.O.; Brady, R.O., Jr. (Bethesda, Md.) 213 Gene Therapy for Parkinson’s Disease Hadaczek, P.; Daadi, M.; Bankiewicz, K. (San Francisco, Calif.) 246 Simplifying Complex Neurodegenerative Diseases by Gene Chip Analysis Scherzer, C.R.; Gullans, S.R. (Cambridge, Mass.); Jensen, R.V. (Middletown, Conn.) 258 Molecular Pathology of Dementia. Emerging Treatment Strategies Gouras, G.K. (New York, N.Y.)

Contents

VI

270 Expanding the Role of Deep Brain Stimulation from Movement Disorders to Other Neurological Diseases Leone, M.; Franzini, A.; Broggi, G.; Bussone, G. (Milano) 284 Molecular Mediators of Pain Chaudhary, P.; Burchiel, K. (Portland, Oreg.) 322 Gene Transfer in the Treatment of Pain Fink, D.; Mata, M. (Ann Arbor, Mich.); Glorioso, J.C. (Pittsburgh, Pa.) Neurovascular Disorders 336 Gene Discovery Underlying Stroke Barone, F.C. (King of Prussia, Pa.); Read, S.J. (Macclesfield) 377 Molecular Mediators of Hemorrhagic Stroke Macdonald, R.L. (Chicago, Ill.) 413 Advances towards Cerebrovascular Gene Therapy Watanabe, Y.; Heistad, D.D. (Iowa City, Iowa) 439 Ex vivo Gene Therapy and Cell Therapy for Stroke Kondziolka, D. (Pittsburgh, Pa.); Sheehan, J. (Charlottesville, Va.); Niranjan, A. (Pittsburgh, Pa.) Neuro-Oncology 458 Neurosurgical Applications for Polymeric Drug Delivery Systems Wang, P.P.; Brem, H. (Baltimore, Md.) 499 Immunotherapy Strategies for Treatment of Malignant Gliomas Harshyne, L.; Flomenberg, P.; Andrews, D.W. (Philadelphia, Pa.) 521 Glioma-Genesis. Signaling Pathways for the Development of Molecular Oncotherapy Kapoor, G.S.; O’Rourke, D.M. (Philadelphia, Pa.) 557 Oncolytic Viral Therapy for Glioma Lamfers, M.L.M.; Visted, T. (Charlestown, Mass.); Chiocca, E.A. (Columbus, Ohio) 580 Molecular Neurosurgery in the Pituitary Gland. Gene Transfer as an Adjunctive Treatment Strategy Castro, M.G.; Jovel, N.; Goverdhana, S.; Hu, J.; Yu, J.; Ehtesham, M.; Yuan, X.; Greengold, D.; Xiong, W.; Lowenstein, P.R. (Los Angeles, Calif.).

Contents

VII

624 Stem Cells for Targeting CNS Malignancy Yip, S. (Vancouver); Sidman, R.L. (Boston, Mass.); Snyder, E.Y. (Boston, Mass./La Jolla, Calif.) 645 Author Index 646 Subject Index

Contents

VIII

Dedication

This book is dedicated to the memory of Drs. Ernst and Elisabeth Freese, brilliant scientists and wonderful parents who deciphered the chemical basis of mutagenesis, the engine of evolution and God’s way of making us better. We also dedicate the section on functional and restorative molecular neurosurgery to the memory of Anne Janson, a victim of Alzheimer’s Disease, and all the other patients who suffer from this terrible disease which robs the mind of its memories and dignity, as well as all children affected by neurodegenerative diseases – for those who have died, they are not forgotten; and for those who are living, may they soon have the promise of a cure. The section on Oncology is dedicated to the memory of Jack Geary, a victim of cancer. Finally, we dedicate the section on the spine and spinal cord to the memory of Anthony Simeone, M.D., whose devotion to his patients and family will always be remembered. The Editors Andrew Freese Frederick A. Simeone Paola Leone Christopher Janson

IX

Acknowledgements

We gratefully acknowledge the dedicated, diligent work of Jackie Alutis in helping produce this book, and the valuable contributions of Joanne Coughlin and Marcia Freese as well. Without them, this book would not have been possible. In addition, we acknowledge the support of the CNS Gene Therapy Consortium in the production of this book. Otto Freese produced the front cover illustration, for which we are grateful as well. The Editors Andrew Freese Frederick A. Simeone Paola Leone Christopher Janson

X

Series Editor’s Note

Changing the Paradigm of Neurosurgery

During the last decade of the 20th century and the first years of the 21st century, neurosurgery has been part of an enormous paradigm shift. While we previously concentrated on dealing with the removal or management of structural masses (blood clots, aneurysms, brain tumors, spinal bone spurs, ruptured discs), the future of neurosurgery lies in the application of a wide variety of new knowledge. Loosely termed ‘molecular’, this new knowledge can be applied widely to the diagnosis, management, and possible prevention of serious neurological illness. As such, we practitioners and surgeons must embrace this new knowledge and attempt to implement it in the current practice of neurosurgery. The future of neurosurgery is molecular, minimally invasive, and multidisciplinary. For the first time, we will be bringing the emerging data from the laboratory and beginning to apply it to clinical problems, rather than the reverse, the old paradigm of trying something in the operating room and then going back to the laboratory to see why it did or didn’t work. Volume 18 of Progress in Neurological Surgery is an elegant compilation of current and emerging knowledge related to the influence of molecular neurosurgery on both the present and the future. Drs. Freese, Simeone, Leone, and Janson have accumulated a wealth of information which can be applied to the spinal column and spinal cord disorders, functional and restorative brain surgery, neurovascular disorders and neuro-oncology. The author list is impressive, and the story that is presented should entice the reader to glimpse the

XI

future of neurosurgery, which is rapidly descending on us. I congratulate the current authors for their excellent collaboration, and believe the readership will enjoy this new volume which does, indeed, represent progress in neurological surgery. L. Dade Lunsford, MD The University of Pittsburgh School of Medicine, Pittsburgh, Pa., USA

Series Editor’s Note

XII

Foreword

There are not many advantages to becoming ‘senior’, an euphemism for being an old man. But there are a few, and two of them are exemplified for me by this volume. The first is in seeing how scientific/medical progress can mushroom over a span of decades. As I think back on my training, in the 1950s and early 1960s, and then look through this fascinating book, I am in awe at the progress. Yes, I have been intimately involved in gene therapy and the genetic basis of disease my whole career, but the laying out of the application of these technologies to the single field of neurosurgery leaves me in wonder. Using gene therapy to treat malignant gliomas is revolutionary enough, but scanning the chapters herein reveals the use of gene therapy and genetic approaches for degenerative disc disease, for spinal cord injuries, for epilepsy, for Parkinson’s and other neurodegenerative diseases, for the treatment of pain, for stroke as well as for malignancies. And not just genetic approaches: neurosurgery now embraces protein therapy, cell therapy, oncolytic viral therapy, immunotherapy, stem cell transplantation, and xeno-neurotransplantation. Truly a glorious story of scientific/medical progress! The second advantage of growing old is seeing the successes of the many young physician investigators that one has trained. Everywhere that I travel, there seems to be someone who had passed through my laboratory over the past 40 years. The joy of hearing the stories of successful, productive careers is wonderful. But this book brings an added touch. Andrew Freese is the son of one of the men that played a significant role in my career. When I interviewed at NIH in 1963 for a position of Research Associate (following my medical

XIII

training), I narrowed my choices to two: Marshall Nirenberg (who was in the middle of deciphering the genetic code) and Ernst Freese, who was doing fascinating work in genetic model systems. This was one of the toughest decisions of my career. I found Dr. Freese (I still cannot call him by his first name although he has asked me to for years!) to be a brilliant scientist and marvelous human being. So much so that, although I joined Marshall Nirenberg and helped in the final genetic code decipherment, I maintained a long and fruitful friendship and scientific mentorship with Dr. Freese. Thus, it was with extraordinary pleasure that I received a letter from Dr. Freese’s son, Andrew, to write the Foreword for this book. As I think back to all the advice and knowledge that I received from Dr. Freese over so many years, I am honored and humbled to add my name to a volume that is dedicated to his memory by his son. W. French Anderson, MD USC Keck School of Medicine, Los Angeles, Calif., USA

Foreword

XIV

Freese A, Simeone FA, Leone P, Janson C (eds): Principles of Molecular Neurosurgery. Prog Neurol Surg. Basel, Karger, 2005, vol 18, pp 1–4

Introduction Andrew Freesea, Christopher Jansonb, Paola Leoneb, Frederick A. Simeonec a Department of Neurosurgery, University of Minnesota School of Medicine MMC 96, Minneapolis, Minn., bCell and Gene Therapy Center, UMDNJ-Robert Wood Johnson Medical School, Division of Neurosurgery, Camden, N.J., cDepartment of Neurosurgery, University of Pennsylvania, Philadelphia, Pa., USA

Until recently, Neurosurgery has been a surgical discipline largely focused on ablative approaches to diseases affecting the nervous system – clipping aneurysms, evacuating hematomas, removing disc herniations, decompressing the stenotic spine, extirpating tumors, lesioning the neostriatum, and others. However, as physician-scientists have begun to unravel the molecular basis of many neurological disorders, a paradigm shift has begun to occur – one that promises to dramatically alter the face and texture of Neurosurgery, and convert it into a field that not only ablates diseased tissues and structures, but also restores and improves function within the nervous system. Thus, we believe the advent of a variety of molecular and cellular technologies will have a marked impact on what neurosurgeons, neurologists, psychiatrists, neuroradiologists, and other medical practitioners focused on the nervous system can offer patients. Indeed, distinctions among these medical disciplines will begin to fade, as psychiatric diseases increasingly find neurosurgical solutions, and neurosurgical diseases ‘molecularize’ into nonsurgical, neurological approaches. As the field of medicine in general continues to hone in on molecular interventions, in parallel, the field of Neurosurgery will convert itself from a macromolecular discipline to one that relies increasingly on molecular approaches to improve the outcomes of patients. To hopefully assist our colleagues in medical and surgical disciplines dealing with the nervous system in anticipating this future, we believe that a book focused on the current principles of molecular Neurosurgery is needed, and we have, therefore, attempted to bring together a

variety of superb contributors who can further shed light on this topic in this compilation of chapters. In this edition, several categories of neurological diseases are examined. It is important to note that the chapters are not meant to be an exhaustive compilation of facts detailing the entire field of molecular neurosurgery, but instead are meant to highlight some of the most exciting advances. Thus, this book attempts to spark the imagination of its readers as much as it tries to recite and explain exhaustive details of the most recent research work and discoveries. If there is one fact of which we are certain, it is that this book will be obsolete within several years of its publication, and a number of promising technologies outlined therein will have proven to be failures and subsequently will have been abandoned. Similarly, undoubtedly a number of approaches that are only briefly skimmed or even inadvertently omitted in this book will prove to be extraordinarily successful and important. We cannot predict which will succeed and which will fail. However, we are certain that Neurosurgery faces an exciting future as these types of approaches continue to evolve and ultimately allow us to better care for our patients. The first section of the book, encompassing the first six chapters, addresses disorders of the spine and spinal cord. The chapter by Leo et al. focuses on understanding the molecular and biochemical alterations associated with degenerative disc disease, a problem that has a huge financial and emotional impact on a large proportion of our society. Only by understanding the molecular basis of disc disease can we ever hope to meaningfully intervene prior to the development of symptoms and the progressive cycle of incapacitation that results from degenerative disease. Kurpad and Lifshutz discuss the genetics of disc and spinal degeneration, as there is clear evidence of a genetic predisposition to this disease process, and a variety of gene defects and altered protein profiles have been associated with it. Once a better understanding of the genetic alterations, and subsequent molecular and biochemical disruptions that occur in degenerative spine disease can be garnered, then one can begin to envisage meaningful interventions. The chapter by Kim et al. outlines one exciting area of investigation, focused on genetic intervention using different gene transfer technologies to introduce therapeutic gene products directly into the disc to retard degeneration or promote restoration of normal function. Although ideally one would want to intervene in the spine before the emergence of axial back pain and/or radicular symptoms, surgical intervention will still be required in a number of patients who need fusion procedures. To optimize the success of this type of surgery, however, molecular approaches to enhance fusion need to be developed and promise to significantly improve outcomes from fusion procedures. The chapter by Bomback and Grauer enumerates these approaches. Although degenerative spine disorders have an enormous impact

Freese/Janson/Leone/Simeone

2

on our society, it is those patients who have traumatic spinal cord injuries that frequently suffer the most. Relatively little can be done currently for these patients, but enormous progress is being made in understanding the molecular basis of spinal cord injury and hopes for restoration of function. The last two chapters (Steinmetz et al. and Iwanami et al.) in this section on the spine thus discuss a variety of molecular and cellular interventions that are being developed to improve outcomes in patients with spinal cord injury. The second section of the book, encompassing 11 chapters, addresses functional and metabolic disorders of the nervous system, and restorative approaches to these diseases. In the chapter by House et al., an overview of the field is provided with an emphasis on targeting structures within the nervous system using contemporary neurosurgical techniques. Schumacher as well as Samulski and Giles identify some exciting technological advances that allow molecular and functional intervention in the nervous system using cellular transplantation and gene therapy approaches, selecting two unique and promising systems, one based on xenotransplantation, and the other, a promising gene transfer system based on adeno-associated virus, both of which have already been used in clinical trials for human neurological diseases. Although these two chapters focus on a specific tissue source and viral vector, respectively, they also discuss alternatives and principles underlying cell transplantation and viral vectormediated gene transfer. Telfeian et al. discuss current concepts regarding the molecular basis of epilepsy and seizure disorders, identifying promising research approaches to these diseases. Since pharmacological approaches to epilepsy are often suboptimal, it is likely that cellular and genetic intervention will allow more directed and efficacious therapy to the tissue source of the aberrant electrical activity. Brady and Brady Jr. discuss the advent of molecular intervention for neurometabolic diseases, starting with the successful development of enzyme replacement therapy, and then branching out to exciting new advances in gene and cell therapy. The following four chapters address a variety of neurodegenerative diseases, including Parkinson’s disease, Alzheimer’s disease, and others, with a focus on further elucidating the molecular bases of these diseases, and then improving neurosurgical approaches to them. Chaudhary and Burchiel as well as Fink et al. then focus on pain, and include a thorough evaluation of the molecular mediators of pain, and subsequently novel genetic and surgical interventions for patients with intractable pain disorders. The third section of the book, encompassing four chapters, focuses on neurovascular disorders. The first two chapters (Barone/Read and Macdonald) identify different genes and molecular mediators associated with stroke, an understanding of which may identify potential genetic and molecular targets for improved clinical intervention. The chapter by Watanabe and Heistad focuses on genetic intervention in stroke, including an evaluation of preclinical studies,

Introduction

3

with the hope of developing neurosurgical or endovascular delivery systems to minimize damage to neural tissue associated with stroke. Kondziolka et al. then discuss the advent of cellular transplant therapy for stroke, including some results of a recent clinical neurosurgical trial. The fourth and final section, encompassing the last six chapters, addresses the field of neuro-oncology, with a neurosurgical perspective. The chapter by Wang and Brem provides a thorough overview of novel polymeric and related intracranial drug delivery systems for chemotherapeutic agents as adjuvant therapy following surgical extirpation of brain tumors. Harshyne et al. discuss immunotherapy strategies for malignant brain tumors, and their likely impact on developing a more global approach to a disease process that usually extends beyond the typical resection margins of surgery, virtually assuring recurrence. The chapter by Kapoor and O’Rourke identifies common molecular signaling events involved in gliomagenesis, and their relevance for developing targeted approaches to intervene in these biochemical sequences and their role in progression to a more malignant tumor phenotype. Lamfers et al. then focus on viral gene therapy approaches to malignant brain tumors, including an evaluation of failures of clinical trials of gene therapy in the past, and opportunities for improvement in the future. The chapter by Castro et al. further provides an overview of the potential for genetic intervention in brain tumors, but in a particular subset, pituitary adenomas, in which focal gene expression may well have a significant impact clinically, and may someday be the primary modality of treatment for these typically benign tumors which nonetheless cause significant morbidity. The final chapter by Yip et al. gives an exciting glimpse into the potential for stem cell technology to seek out malignancies in the nervous system, and selectively destroy them. It is our hope that by providing an overview of the developing interface between molecular biology and clinical Neurosurgery, we will further stimulate the imaginations of clinicians and scientists, and provide additional impetus for aggressive investigation of these and related technologies. Andrew Freese, MD, PhD Professor and Vice Chairman of Neurosurgery Director of Spinal Neurosurgery University of Minnesota 420 Delaware Street, S.E. Suite D429, Mayo Memorial Building Minneapolis, MN 55455 (USA) Tel. ⫹1 612 624 2471, Fax ⫹1 612 624 0644, E-Mail [email protected]

Freese/Janson/Leone/Simeone

4

The Spine and Spinal Cord Freese A, Simeone FA, Leone P, Janson C (eds): Principles of Molecular Neurosurgery. Prog Neurol Surg. Basel, Karger, 2005, vol 18, pp 5–29

The Molecular Basis of Intervertebral Disc Degeneration Brian M. Leoa, Matthew H. Walker a, D. Greg Andersonb a Department of Orthopaedic Surgery, University of Virginia, Charlottesville, Va., bDepartment of Orthopaedic Surgery, Thomas Jefferson University, Philadelphia, Pa., USA

Introduction

Low back pain is an endemic problem in Western societies leading to significant morbidity [1–3]. Not only does back pain account for much individual suffering, but the societal costs of time lost from work, for treatment, and for compensation of lost wages numbers in the billions of dollars annually. It has been estimated that up to 80% of the population experiences some form of back pain over the course of their lives, making this a leading health concern [1–5]. Although the etiologies are many, intervertebral disc degeneration appears to be the leading cause for chronic axial low back pain [6]. Significant strides have been made in understanding the molecular basis for disc degeneration. Despite this, the currently available treatment modalities for disc-related spinal pain continue to focus on alleviating symptoms rather than addressing the underlying cause of degeneration. It is likely that clinical outcomes for patients with painful intervertebral disc degeneration would improve if therapies were developed that could slow, halt, or reverse this process. Histopathological changes classified as ‘degeneration’ have been recognized as early as the second decade of life and are known to progress through a series of stages that have been quantified histologically [7–10] and characterized noninvasively using magnetic resonance imaging [7, 9, 11, 12]. Initially, the degenerative process begins asymptomatically in the nucleus pulposus (NP) with cell loss and matrix alteration. This leads to an inability to retain water and results in slight disc space narrowing. As this process progresses, the outer annulus fibrosis (AF) becomes increasingly disorganized, losing its normal lamellar arrangement and leading to diminished mechanical strength. Tissue fissures

and clefts begin in the inner AF and progress outwards towards the periphery, ultimately culminating in a loss of mechanical integrity [13]. Mechanical stresses are progressively transferred to the surrounding vertebral endplates causing microfractures and marginal osteophyte formation. Cytokines, produced within the disc, leach out, leading to the ingrowth of nerve and vascular elements that may play a role in the etiology of spinal pain [14–17]. The secondary bony changes serve to stiffen the spinal segment and restabilize the spine. Interestingly, although this sequence happens throughout life in essentially all humans, there are significant variations in the degree of symptoms noted by different people. Equally perplexing is the poor correlation between the degree of degenerative changes noted on imaging studies and the presence of symptoms of spinal pain [18]. Clearly, we have much to learn regarding the relationship between spinal degenerative disease and various pain syndromes. In recent years, researchers have attempted to understand the genetic and molecular aspects of disc degeneration in order to determine the etiology of degeneration and to identify stages in this process where therapeutic intervention would be beneficial. There has been a growing enthusiasm in the concept of developing tissue engineering strategies that can alter the course of the degenerative process and address the underlying disease process. This chapter will review the normal disc development and structure as well as the changes that occur during degeneration on the biomechanical, biochemical and ultrastructural levels. By gaining a better understanding of the cellular and molecular biology of the disc, the various strategies that are being discussed or studied to combat disc degeneration can be evaluated in a rational manner.

Embryological Development of the Disc

Although the embryology of the human disc has not been well studied directly, animal models have contributed significantly to our understanding of the development of the intervertebral disc and the axial skeleton. The development of the spine begins during the third week of gestation in a phase termed ‘gastrulation.’ It is during this phase that the ectoderm, mesoderm, and endoderm are formed; embryological layers, which eventually give rise to all tissues of the body. Differentiation of these tissues begins with the formation of the primitive streak. Associated with the primitive streak is a primitive node surrounding a small invagination known as the primitive pit, located slightly caudally from the midline of the embryo. Cells in the primitive pit, known as prenotochordal cells, migrate cephalad to the prechordal plate and contribute to the cell layers forming the notochordal plate. As these cells proliferate and detach from the endoderm, they form a solid cord of cells that becomes

Leo/Walker/Anderson

6

The developing embryo Primitive node Primitive streak

Prenotochordal cells migrating

Fig. 1. The developing embryo has a primitive streak extending caudad from the primitive node. Arrows approximate the migration of the prenotochordal cells cephalad to the prechordal plate.

the definitive notochord and forms the basis of the axial skeleton (fig. 1). Concurrently, the vertebral column and outer portion of the intervertebral discs form from an aggregation of mesenchyme surrounding the notochord. This process involves the medial migration of somatic mesoderm from the ventromedial portions of the somites, the sclerotomes. In a process called ‘segmentation,’ a pattern of alternating light and dark bands becomes evident in the mesenchymal column by the time the embryo reaches 5 mm in length [19] (fig. 2). The dark bands contain cells with a greater nuclear density than those of the light bands and are the precursors of the intervertebral discs, termed the perichordal discs. The light bands are the precursors of the vertebral bodies [19]. By the time the embryo reaches 12.5 mm, the perichordal disc becomes trilaminar with a denser middle region surrounded by lighter regions both cranially and caudally [20]. At this point, the outer mesenchymal cells begin to arrange themselves in a lamellar manner with their long axis parallel to the long axis of the embryo [19]. This outer lamellar zone darkens as it becomes populated with fibroblasts forming the primitive AF [19]. The lighter inner cell mass is formed primarily from notochordal cells and embryonic cartilage and can be well seen by the 40–50 mm stage [19]. Cells in the AF begin to deposit collagen fibers in the outer region of the perichordal disc by the 20–40 mm stage [19]. With continued growth, the outer AF becomes progressively more fibrous and less cellular, while the inner AF becomes fibrocartilaginous and retains a high cell density [21]. Growth of the disc is hypothesized to occur due to increasing lamellar thickness, as opposed to an increase in the number of lamellar layers; the number of lamellae remain

Molecular Basis of Intervertebral Disc Degeneration

7

Precursor of vertebral body

Notochord Precursor of intervertebral disc – perichordal disc

Fig. 2. The mesenchymal column segments into light and dark bands. The dark bands contain cells with a greater nuclear density and represent the precursors to the intervertebral discs, the perichordal discs. The light bands represent the precursors to the vertebral bodies.

fairly constant throughout development, and in lumbar discs constitutes twelve to sixteen layers [19, 21]. The lamellar fibers in these alternating layers of the AF are oriented at oblique angles to each other, a fashion designed to optimally dissipate multidirectional stresses on the disc. The NP continues to develop from both the intervertebral expansion of the notochord and the growth of primitive cartilage. As the NP region develops, the ground substance softens leading to a loosely arranged matrix [19, 22]. The notochordal cells of the NP play a key role in cell division and matrix formation. As the matrix of the NP becomes looser, the once compact mass of notochordal cells is broken up into cellular clusters in a loose network known as the ‘chorda reticulum’ [23]. The NP continues to expand in size during fetal development and early postnatal life with an 18-fold expansion in the number of notochordal cells before becoming quiescent [19]. Although the appearance of the NP gradually changes to become more fibrous, resembling the transition zone, the embryological notochordal cells remain active producers of matrix material until the end of the first decade of life, at which time a ‘notochordal’ NP can no longer be defined due to the increased collagen content and loss or metaplasia of notochordal cells [19, 24]. Insults to the developing fetus can have severe consequences to organ systems undergoing rapid development at the time of the insult. The time period for the most rapid development of the lumbar vertebral canal is between 12 and 32 weeks in utero [116]. A recent retrospective study by Jeffrey et al. [25]

Leo/Walker/Anderson

8

Atlas 7 Cervical vertebrae

Axis Cervical curvature

Thoracic curvature

12 Thoracic vertebrae

5 Lumbar vertebrae Lumbar curvature Sacrum (5 fused vertebrae) Coccyx (4 fused vertebrae) Sacral curvature

Fig. 3. Schematic of the human spine. (Reproduced with permission [117].)

showed that low birth weight, low placental weight, low socioeconomic class, and smoking during pregnancy can have detrimental effects on the size of the lumbar vertebral canal and may predispose these people to future spinal problems. Structure and Anatomy

The human spine contains twenty-three intervertebral discs (fig. 3). The size of the discs generally increases in both height and diameter as one moves


9

Fig. 4. Micrograph of rabbit intervertebral disc histology. # ⫽ Outer AF; * ⫽ inner AF; arrowhead ⫽ transition zone; NP ⫽ nucleus pulposus.

more caudally within the spine. The disc itself is a component of a complex biomechanical system composed of the bony vertebral endplate, the cartilaginous endplate, the AF, and the NP. This complex organ functions to allow motion and to dissipate stress within the spinal column. Similar to other connective tissues, the disc is composed mostly of extracellular matrix with a complex array of structural and water-binding proteins and a relative paucity of cells. Each disc can be further subdivided into four regions: the outer annulus, inner annulus, transition zone, and NP (fig. 4). The outer annulus is composed of highly organized, directionally oriented collagenous lamellae running at alternating 30-degree angles to the long axis of the spine. The collagen in this region of the disc is mostly type I and includes fibrils that insert into the vertebral bodies. The cell population in the outer annulus is primarily fibroblastic. The inner AF is larger and more fibrocartilaginous, containing less collagen and lacking the lamellar architecture of the outer AF. Collagen in this region is mostly type II. In addition, the inner AF contains a higher proportion of large proteoglycan aggregates. The cell population in the inner AF has characteristics of both fibroblasts and chondrocytes. The transition zone is a distinct, thin acellular fibrous layer that separates the inner AF from the NP. The NP contains an amorphous matrix of highly hydrated proteoglycans embedded in a loose network of collagen. Like the inner AF, the collagen in the NP is mostly type II. The cell population of the NP is sparse and unevenly distributed with more cells present in the central regions of the NP than at the periphery. At least two distinct cell populations are recognized in the NP in early life. The first is a small round cell resembling a chondrocyte. The second is much larger and has a vacuolated appearance described by Virchow [22] as

Leo/Walker/Anderson

10

Table 1. Collagen composition of the intervertebral disc. (Reproduced with permission from [116]) Type

Predominant location

Percent of total collagen (%)

Fibril-forming collagens I II III V XI

Annulus Annulus and nucleus Annulus Annulus and nucleus Annulus and nucleus

0–50 0–50 ⬍5 1–2 1–2

Short helix collagens VI IX XII

Annulus and nucleus Annulus and nucleus Annulus

5–20 1–2 ⬍1

‘physoliferous’ (or ‘bubble-bearing’), containing prominent cellular processes and intracellular glycogen deposits. This cell type is thought to be of notochordal origin. As mentioned, in humans, these large, notochordally derived cells tend to disappear (or become rare) by adolescence, leaving scattered chondrocyte-like cells in their place [7, 8, 26, 27]. A recent study by Kim et al. [28] demonstrated the migration of endplate chondrocytes into the NP region of the disc, which might account in part for the shift in cell populations within the NP region. The disc matrix consists primarily of collagens and proteoglycans, but varies significantly between regions of the disc. Collagen cross-linking gives the disc substantial tensile strength, while highly hydrated proteoglycans give the disc stiffness and resistance to compression [27]. Collagens account for 60% of the dry weight of the AF, with type I collagen being the most abundant type (80%); the NP, on the other hand, contains up to 20% collagen with a predominance of type II collagen [7, 29, 30]. Additional collagen types are also present throughout the disc in much smaller quantities. For example, other fibrillar collagens such as type V and type XI are found in greatest concentrations in areas with high type I collagen and type II collagen, respectively [27]. Nonfibrillar collagens found in the disc include short, helical collagens, specifically types VI (up to 10% in the annulus and 15% in the nucleus), IX, and XII [27] (table 1). Proteoglycans comprise a small proportion of the outer AF, but become increasingly abundant as one moves toward the more hydrated central regions of the disc, accounting for 50% of the dry weight of the NP. Large proteoglycan aggregates consist of central hyaluronan molecules with multiple attached


11

Fig. 5. Schematic of a proteoglycan molecule with a central core of hyaluronic acid and attached chains of aggrecan.

proteoglycan side chains predominantly of aggrecan. The connection between the proteoglycan side chains and the hyaluronan backbone is stabilized by small link proteins. The proteoglycan molecules contain numerous sulfated polysaccharide molecules leading to a strong net negative charge. It is this net negative charge that serves to attract water within the disc (fig. 5). The relative proportion of collagens and proteoglycans changes throughout life and with disc degeneration. Age-related changes lead to a decline in the amount of large proteoglycan aggregates in the NP, thus resulting in diminished water-binding capacity. In addition, there is an increase in the proportion of nonaggregated proteoglycans and a shift in the composition of sulfated polysaccaride side chains leading to the diminished structural properties of the disc [31]. The vertebral endplate is a specialized structure that contains both a bony and a cartilaginous portion. The bony endplate is made up of cortical bone, which is thicker around the periphery of the disc and thinner in the central region. Adjacent to the bony endplate are specialized capillaries, which are the primary source of nutrient exchange to the disc. The bony endplate is covered by a layer of hyaline cartilage that forms a barrier between the vertebral body and the disc and limits solute transport into and out of the disc. With age, the cartilaginous endplate undergoes progressive calcification, a process which diminishes its diffusional capacity and may lead to a nutritional crisis within the disc [32]. In addition to aging, a number of environmental and genetic factors have been linked to disc degeneration possibly by altering the disc by limiting endplate diffusion. Smoking, in particular, causes shrinkage of the vascular

Leo/Walker/Anderson

12

buds adjacent to the endplate that undoubtedly has a negative impact on disc nutrition [33–35]. The intervertebral disc is innervated only in its outermost portion. Small unmyelinated and encapsulated nerve endings have been found on the surface of the outer annulus, and small free nerve endings may penetrate the outermost layers of the annulus. The recurrent nerve of Luschka (sinuvertebral nerve), formed from small branches of the lumbar ventral ramus, provides this sensory innervation to the disc and is likely to be responsible for the discogenic pain. In addition to supplying the outer annulus, these branches also supply the posterior longitudinal ligament and ventral dural tube. Nerve endings have not been found in the inner annulus or NP region of normal human discs [27].

Disc Nutrition

Most of the normal intervertebral disc is avascular, depending on the diffusion for nutrient and waste exchange. Although small blood vessels can be found on the surface of the annulus and may penetrate a short distance into the outer layers of the disc, the central region has cells which can lie 6–8 mm from the closest blood supply, making the disc the largest avascular organ in the human body [27]. Unfortunately, the diffusional capacity of the disc is relatively poor even in the nonpathological state, and is further limited by aging and degenerative changes. Studies comparing the transport of small molecules into the disc in exercised and anesthetized dogs have shown no differences between the two groups, suggesting that simple diffusion rather than a physiological ‘pump’ mechanism is responsible for small molecule exchange within the disc [35]. These results have been confirmed in motionless [36] and mobilized spinal segments [37]. Recently it has been recognized that the supply of simple nutrients is a crucial factor regulating the density of disc cells. Diffusion chamber studies have suggested that glucose supply rather than oxygen is the major factor regulating cell viability within the disc [38]. Smoking probably promotes disc degeneration via a nutritional mechanism although a direct toxic insult to disc cells is also possible [39]. In the largely avascular environment of the disc, cells can survive with little oxygen [38]. However, under anaerobic conditions disc cells produce lactate as a by-product of metabolism (fig. 6), thus leading to an acidic pH. Normal disc tissue has a mildly acidic pH of 6.9–7.2, but under conditions of stress (degenerative disc disease) the disc pH can be as low as 6.1 [35]. It is known that an acidic pH leads to the inhibition of proteoglycan and collagen synthesis and may, therefore, contribute to matrix deterioration [40]. However, matrix degradation is likely to be dependant on many other factors, as suggested by


13

Glucose ATP 1 ADP Glucose-6-phosphate 2 Fructose-6-phosphate ATP 3

ADP

Fructose-1,6-bisphosphate 4

5

Glyceraldehyde-3-phosphate 6

Dihydroxyacetone-phosphate

2 NAD ⫹Pi

2 NADH ⫹2 H 2 1,3-bis-phosphoglycerate ADP

7

ATP

Enzyme key 2 3-Phosphoglycerate 1. Hexokinase 2. Phosphoglucoisomerase 3. Phosphofructokinase 4. Aldolase 5. Triose-P-isomerase 6. Glyceraldehyde-3-Pdehydrogenase 7. Phosphoglycerate kinase 8. Phosphoglycerate mutase 9. Enolase 10. Pyruvate kinase 11. Lactate dehydrogenase

8 2 2-Phosphoglycerate 9 2 Phosphoenolpyruvate ADP 10

ATP

11

2 Pyruvate

2 Lactate

2 NADH ⫹ 2 H

2 NAD

Fig. 6. The glycolytic pathway. Anaerobic metabolism leads to lactate as a by-product, an acid that can lower the pH in the environment of disc cells.

Leo/Walker/Anderson

14

Compressive force causes disc bulging Vertebral body

Disc

Vertebral body

Fig. 7. Disc in compression (intervertebral disc between two vertebral bodies). Compressive forces on one side of the disc lead to disc bulging, with these forces being converted to tensile hoop stresses by the AF. On the opposite side, the disc fibers stretch.

studies demonstrating a poor correlation between measured oxygen or lactate levels and the grade of degeneration [41].

Disc Biomechanics

Intervertebral discs in humans have evolved to withstand the significant forces of an upright posture. When healthy, the normal disc can withstand forces greater than the surrounding bone, which fractures prior to the disruption of the disc. By dissipating the large compressive forces in the spine, generated as the result of musculature activity and vigorous physical situations, the disc serves to protect the surrounding spine from trauma. Forces up to 17,000 N have been estimated in lumbar discs during heavy lifting activities [42] (fig. 7). To dissipate these loads, the disc converts compressive forces to tensile stresses in the outer annulus by exerting a hydrostatic pressure via the interstitial fluid within the disc. However, the tensile properties vary within different regions of the annulus leading to a ‘biphasic phenomenon’ during loading. Because fibers in the anterior/outer regions are stiffer than those in posterolateral/inner regions of the annulus, the stiffer outer layers convert compressive loads into hoop stresses while the inner layers act as a ‘shock absorber’ [27]. The high tensile modulus of the normal annulus helps to prevent disc bulging. During aging and in the degenerating disc, the swelling pressure of the NP decreases and the stiffness of


15

the AF increases [43, 44]. This results in poor load dissipation and increased stress transfer to the bony elements of the spine [43–46]. Several studies have shown that pathological loading of the spine may play a role in disc degeneration [47–49]. Hadjipavlou et al. [50] demonstrated disc degeneration following a 30-degree torsional injury to the spine in an animal model. This stress led to early degenerative changes in the disc including an increase in phospholipase A2 and a decrease in NP volume by 60–90 days. Following the onset of degeneration, increased levels of calcitonin gene-related peptide and vasoactive intestinal peptide were found within the ganglion, supporting the association of disc degeneration with spinal pain [51]. Researchers have shown, in animal models, that discs loaded statically are more prone to degeneration when compared to those loaded cyclically. However, a recent investigation by MacLean et al. [52] demonstrated that dynamic loading (12.6 N with a frequency of 0.2 Hz for 2 h) in rats led to decreased collagen type I and II gene expression in the AF, and an increase in the catabolic genes collagenase and stromelysin. In the NP, these mechanically induced changes in gene expression were less significant [52]. In addition, the magnitude and frequency of loading serve to affect the rate of degeneration [53]. For example, Kasra et al. [54] demonstrated that rabbit intervertebral disc cells in the AF and the NP respond to high-freqenecy (20 Hz) and highamplitude (1.7 MPa for annular cells and up to 3.0 MPa for nuclear cells) loading by increasing collagen synthesis and decreasing collagen degradation when subjected to 3 days of loading. Nuclear cells demonstrated significantly less collagen degradation as the load increased in amplitude from 1.0 to 3.0 MPa [54]. These contrasting studies suggest a role in mechanical loading for both degenerative and regenerative phenomenon in the disc suggesting a complex interaction between mechanical stress and metabolism within the disc. Segmental spinal instability associated with disc degeneration has been quoted as a cause of low back pain. Although there is some controversy in the literature regarding the relationship between disc degeneration, annular fissures and nonlinear segmental instability, several researchers have documented an increased range of axial rotation in degenerative discs. Mimura et al. [55] studied flexionextension forces in human lumbar cadaveric spines and found that the range of motion in flexion-extension decreased while axial rotation increased in degenerative spines. Krismer et al. [56] documented similar findings and correlated increased axial instability with fissure formation in the outer annulus.

Etiology of Disc Degeneration

Disc degeneration is undoubtedly a multifactorial process involving both environmental and genetic contributions. Some environmental factors thought

Leo/Walker/Anderson

16

Table 2. Factors that have been implicated to cause intervertebral disc degeneration.

Possible etiologies of disc degeneration 1. Heavy lifting 2. Vibration 3. Immobilization 4. Trauma (torsion) 5. Smoking 6. Diabetes 7. Vascular disease 8. Genetics 9. Infection

to contribute to disc degeneration include: demanding physical activities, such as heavy lifting; vibration (experienced while driving or operating machinery); immobilization, and repetitive torsional loads. Other environmental factors that affect disc metabolism include smoking, poor glycemic control and vascular disease [57] (table 2). In spite of the many environmental factors linked to disc degeneration, it is felt that genetic influences play the predominant role in early and perhaps more symptomatic disc disease. The genetic contribution is supported by family studies, case-control studies, and twin studies as well as the identification of certain genetic linkages and single gene defects leading to degenerative disc disease. Scapinelli [58] identified disc degeneration as a familial trait. In a case-control study, Matsui et al. [59] documented an increased incidence of disc degeneration among the family members of patients requiring lumbar surgery. Disc degeneration has been found to be significantly more prevalent in siblings of patients with disc degeneration than in random population samples [60]. Twin studies by Sambrook et al. [61] have shown a strong heritable component to disc degeneration in both the cervical and lumbar regions. Genetic studies have identified molecular defects contributing to disc degeneration in certain subgroups of patients. Videman et al. [62] compared MRI-documented disc degeneration in Finnish twins with alleles of the vitamin D receptor and found an increased risk of disc degeneration with two specific vitamin D receptor alleles. In a study of Japanese women, Kawaguchi et al. [63] found that women with smaller numbers of tandem repeats in the aggrecan gene had more severe disc degeneration. Annunen et al. [64] identified mutations in the type IX collagen ␣-2 gene which causes a single codon substitution (tryptophan for glutamine) leading to disc degeneration in 4% of the Finnish back pain population. Another defect in the ␣-3 chain of type IX collagen was shown to be associated with an elevated risk of disc degeneration [65]. Certain alleles of the matrix metalloproteinase-3 (MMP-3) gene


17

have been linked with disc degeneration in elderly Japanese patients [69]. The environmental risk factor obesity has been shown to act synergistically with an allele of the COL9A3 gene, leading to a high rate of early degenerative disc disease [70].

Genesis of Back Pain

Exactly how disc degeneration relates to back pain is poorly understood. Some patients with minimal morphological changes in the disc complain of chronic back pain while others with significant changes note minimal symptoms. Many factors including structural changes in the spine, soluble mediators and nerve/vessel ingrowth into the outer annulus have all been hypothesized to be a cause of chronic spinal pain [57]. Studies have shown that mononuclear cells infiltrating along the margins of herniated discs express inflammatory mediators such as interleukin-1 (IL-1), intracellular adhesion molecule-1, lymphocyte function-associated antigen and basic fibroblast growth factor [68]. These mediators may contribute to persistent inflammation and pain and the induction of neovascularization [68]. Sang-Ho et al. [69] studied the expression of mRNA of various cytokines and chemokines in herniated lumbar discs and noted an association between the expression of IL-8 and radicular pain produced by back extension, suggesting IL-8 as a possible target for symptomatic treatment. Alterations in the disc can lead to changes in the alignment and the mechanical milieu of the vertebral bodies, facet joints, spinal ligaments and muscles, producing a complex and poorly understood biomechanical environment that may contribute to spinal pain. It is known that the degenerating disc is capable of producing a spectrum of cytokines and chemical mediators that are capable of stimulating pain in the surrounding nerve endings and inducing blood vessel and nerve ingrowth into annular defects in the disc [57]. For example, tumor necrosis factor-␣ (TNF-␣), a proinflammatory cytokine, has been shown to be a key pain mediator in neurogenic pain following disc herniation. Onda et al. [70] demonstrated, using electrophysiological testing, that an antibody to tumor necrosis factor-␣ partially blocked the neurogenic response suggesting that tumor necrosis factor-␣ blockage may have a therapeutic role in treating sciatica. In addition, Nygaard et al. [71] studied leukotriene and thromboxane levels in herniated intervertebral discs and found significantly higher levels of leukotrienes B4 and thromboxanes B2 in noncontained as compared to contained disc herniation; they suggested that these inflammatory mediators play a role in discogenic pain and sciatica and thus were possible targets for therapeutic intervention.

Leo/Walker/Anderson

18

Degenerative Events: Apoptosis, Degradative Enzymes and Inflammatory Cytokines

In order to develop rational therapies that can slow or reverse the degenerative process, it is first necessary to understand the molecular events leading to disc degeneration. In recent years, investigators have identified candidate molecules and cellular processes such as apoptosis, or programmed cell death, which appear to play a prominent role in disc degeneration. Although the current molecular understanding of disc degeneration is relatively crude, knowledge in this area is expanding rapidly. Cell viability has been shown to be affected by age, with a larger proportion of necrotic cells present in older individuals. In fetal and infantile intervertebral discs approximately 2% of NP cells are necrotic; this number increases steadily to 50% by adolescence and 80% in elderly humans [57]. In addition, apoptosis, or programmed cell death, appears to play a prominent role in the disc, with higher rates of apoptosis present in older individuals [72]. High rates of apoptosis have also been recognized in herniated disc fragments [73, 74]. This process appears to be related in part to the expression of Fas and the Fas ligand, which trigger an intercellular cascade leading to programmed cell death when a Fas-bearing cell comes into contact with a Fas ligand-carrying cell. Normal disc cells do not appear to express the Fas receptor, but do up-regulate this membrane-bound protein shortly after the onset of experimental disc degeneration [75]. As mentioned previously, static compression of intervertebral discs can lead to degeneration. A recent study by Ariga et al. [76] showed that static loading of mouse intervertebral discs resulted in higher numbers of apoptotic cells in the cartilaginous endplate; the number of apoptotic cells increased with the load. Inhibitors of mitogen-activated protein kinase and p38 significantly increased the number of apoptotic cells in the loaded discs. Certain growth factors, such as insulin-like growth factor-1 (IGF-1) and platelet-derived growth factor can exert an anti-apoptotic effect on cultured disc cells suggesting a possible mechanism for the apoptosis of cells within the disc [76, 77]. Some researchers have suggested that degeneration represents an alteration in the disc cellular homeostasis, with the balance tipping away from anabolic events and towards disc catabolism. Antoniou et al. [78] found decreased levels of aggrecan and type II collagen production by old and degenerated disc cells when compared to younger nondegenerated disc cells. Aguiar et al. [79] found that NP cells were able to up-regulate their production of proteoglycans when cocultured with notochordal cells, which are found in high concentrations only in younger humans. The effect appeared to be due to the presence of a soluble mediator produced by the notochordal NP cells and may explain the onset of degeneration shortly after these notochordal cells disappear within the disc.


19

In addition to soluble mediators, direct cell-to-cell communication probably contributes to the behavior of the disc cells, as gap junctions and connexin proteins that have been identified within the disc [80]. A host of inflammatory, degradative, and catabolic factors have been identified that may play a role in disc degeneration. These include proteolytic and degradative enzymes, oxygen free radicals, nitric oxide, ILs, and prostaglandins. Proteolytic enzymes involved in disc degeneration include cathepsin, lysozyme, aggrecanase, and several MMPs [81–86]. A positive correlation between the level of MMPs-1, -2, -3 and -9 and the grade of disc degeneration has been documented [81, 83, 86]. Similarly, Melrose et al. [85] found higher levels of lysozyme in older and degenerative discs. Membrane-damaging oxygen-derived free radicals and nitric oxide have been observed in cultured disc cells [87]. Herniated discs have also been shown to express inducible nitric oxide synthetase and produce nitric oxide [82]. These molecules have the potential to cause direct chemical injury to cell membranes and matrix proteins. Collagens and fibronectin, for example, are known to undergo cleavage or form high-molecular-weight complexes following exposure to superoxides and other oxygen-derived free radicals [87] and may accumulate as lipoprotein complexes within the matrix [88, 89]. Other disc macromolecules undergo complex glycation reactions to form sugar-amino acid by-products that may interfere with normal cell-matrix interactions [90]. Inflammatory cytokines such as IL-1 have been shown to play a major role in articular cartilage degeneration and may play a role in disc degeneration as well. Cell culture experiments have demonstrated that rabbit disc cells increase their rate of caseinolytic activity in response to IL-1 [91]. IL-1 has been shown to decrease the rate of proteoglycan synthesis by the disc, an effect that could be blocked by an IL-1 receptor antagonist [92]. Other effects of IL-1 include inducing increased expression of stromelysin-1, a matrix degradation protease, and an increased production of prostaglandin E2, an inflammatory mediator [93]. Other mediators produced by disc cells include IL-6, nitric oxide, and prostaglandin E2 [94, 95]. Herniated disc fragments are capable of producing very high levels of phospholipase A2, an enzyme critical in the production of prostaglandins and leukotrienes, which are important mediators of inflammation and pain [96–98]. As the matrix of the disc undergoes degeneration, many of the macromolecules are only partially broken down. These nonfunctional molecules accumulate within the disc matrix and may be seen as lipofusion or amyloid by light microscopy [88] or dense granular material when viewed with the electron microscope [99–101]. Some by-products, such as fibronectin, appear to build up within the disc during degeneration. Of interest is the role of bioactive fragments of fibronectin which may promote tissue degeneration [102–104] (fig. 8).

Leo/Walker/Anderson

20

a

b

c Fig. 8. a Degenerative disc histology. b Degenerative changes in the AF with chondroid nests 16 weeks after injection of FN-f (the amino terminal fragment of fibronectin). c Higher powered micrograph of NP degeneration with a loss of cells and matrix disarray 4 weeks after FN-f injection.

Growth factor decline has also been implicated in disc degeneration. For instance, the IGF-1 receptor appears to decrease in older animals leading to a decrease in the IGF-1-dependent proteoglycan synthesis and perhaps the expression of an IGF-1-binding protein [105]. Therapeutic interventions with growth factors remain an active area of interest. Biological Approaches to Disc Degeneration

A thorough understanding of the molecular aspects of intervertebral disc degeneration is fundamental to the development of rational biological therapies


21

Replication-deficient virus carrying a therapeutic gene

Growth factor RNA

Cell

Fig. 9. Schematic of virally mediated gene transfer. A desired gene is inserted into a viral carrier that is taken up by the cell. Viruses depend on host cells for replication; this process increases the expression of the desired gene.

aimed at slowing or stopping this process. A large number of factors have been elucidated that contribute to the process of degeneration and it is likely that the interplay of many different pathways contribute to the end result, making therapeutic intervention a challenge. However, by understanding the molecular basis of the disc degeneration, researchers hope to target specific steps in the degenerative process, thus producing a therapeutic benefit for patients with degenerative disc disease. Because growth factors and stimulatory cytokines can stimulate both cell division and metabolism, these molecules have been seen as good targets for therapeutic intervention. Epidermal growth factor and transforming growth factor-␤ (TGF-␤) were shown to produce a 5-fold increase in the metabolic activity of cultured NP cells [106], and were more effective than fibroblast growth factor or IGF-1. Osteogenic protein-1 has been shown to overcome the degradative effects of IL-1 on cultured disc cells [107]. TGF-␤ also has the added benefit of promoting the ‘chondrogenic phenotype’ by increasing the production of type II collagen and proteoglycan, an effect that is desirable in the inner regions of the disc [108]. Despite the promising results in vitro, direct injections of recombinant proteins into the disc does not appear to be a viable solution to disc degeneration due to the limited half-life of these molecules in vivo [109]. Gene therapy has generated a high level of interest for achieving a longterm solution to disc degeneration. Nishida et al. [108] transduced rabbit NP cells in vivo with an adenovirus carrying the TGF-␤ gene driven by a cytomegalovirus promotor and observed a 2-fold increase in proteoglycan

Leo/Walker/Anderson

22

production at the one week time point (fig. 9). Although encouraging, the authors noted that the long-term expression of a therapeutic gene might be difficult using the viral vector systems due to the possibility of immunological activity against viral antigens. Additionally, it is unclear whether a degenerating disc in metabolic disarray would be able to respond to growth factors in the same manner as the normal disc cells. Investigations are currently ongoing to determine the optimal genes and transduction mechanisms for gene therapy within the disc. Another attractive therapeutic option being investigated currently is the use of a cell-based therapy using either mature cells (chondrocytes or disc cells) or pluripotent cells (stem cells). Gruber et al. [110] successfully studied autologous disc cells transplanted into the intervertebral disc of the sand rat (Psammomys obesus). These authors demonstrated that autografted cells were able to exhibit morphologies similar to native disc cells and were able to survive at least 33 weeks in vivo. Others have suggested that stem cells, which are able to differentiate into multiple cell types, including chondrocytes, may be an ideal vehicle for cell-based therapy. These cells could be treated with therapeutic genes ex vivo prior to implantation and thus used to deliver therapeutic genes to the disc and participate in the repair process. Although encouraging, such strategies have yet to be successfully achieved in a reasonable model of degenerative disc disease. Early tissue engineering approaches have evaluated cellular scaffolds for the delivery of cells to the disc. The advantage of a cell scaffold is that the therapeutic cells are maintained at the implantation site and are provided with a three-dimensional environment necessary for division and migration within the disc [111]. Perka et al. [112] found that NP cells suspended in fibrin-alginate beads and fibrin beads were capable of proliferating and producing extracellular matrix. Lee et al. utilized cells transduced with the TGF-␤ gene in a ‘pellet culture’ system as an alternative to alginate bead microspheres and achieved a native cell phenotype capable of producing type II collagen and proteoglycan [111]. Some scaffolds may not be as effective as previously hoped. Alini et al. [113] demonstrated problems with proteoglycan retention when using scaffolds of type I collagen and hyaluronan seeded with bovine NP and AF cells. In contrast, Sato et al. [114] found that allografted AF cells placed in an atelocollagen honeycomb-shaped scaffold with a membrane seal (ACHMS-scaffold) were able to proliferate, retain type II collagen mRNA, and actually decrease the narrowing of intervertebral disc space in Japanese white rabbits [115]. Much more research is needed to determine the role of cell scaffolds and cellular therapies in treating intervertebral disc degeneration.


23

Conclusion

The intervertebral disc is a part of a complex mechanical system, pivotal in the dissipation of forces in the spinal column. Aging, environmental and molecular-genetic factors all contribute to the degeneration of the disc and its ultimate mechanical failure. Recently, researchers around the world have begun to understand the molecular basis of intervertebral disc degeneration. This understanding forms the basis for the design of biological therapies aimed at slowing or reversing the degenerative process. With time, these early efforts may lead to a shift in treatment options from those focusing on symptomatic relief to those aimed at correcting the underlying disease process. Although promising, these early attempts are far from achieving the goal of tissue regeneration. Much work remains to define the ability of gene- and cell-based strategies to produce a biologically desirable result and to apply these clinically for the benefit of patients. Fortunately, the underlying molecular basis of this ubiquitous disease process is rapidly becoming clear.

References 1 2 3 4 5 6 7 8

9 10 11 12

13

McMeeken J, Tully E, Stillman B, Nattrass C, Bygott IL, Story I: The experience of back pain in young Australians. Man Ther 2001;6:213–220. Waddell G: Low back disability. A syndrome of Western civilization. Neurosurg Clin N Am 1991;2:719–738. Waddell G: Low back pain: A twentieth century health care enigma. Spine 1996;21:2820–2825. Fischgrund JS, Montgomery DM: Diagnosis and treatment of discogenic low back pain. Orthop Rev 1993;22:311–318. Frymoyer JW, Cats-Baril WL: An overview of the incidences and costs of low back pain. Orthop Clin North Am 1991;22:263–271. Schwarzer AC, Aprill CN, Derby R, Fortin J, Kine G, Bogduk N: The relative contributions of the disc and zygapophyseal joint in chronic low back pain. Spine 1994;19:801–806. Buckwalter JA: Aging and degeneration of the human intervertebral disc. Spine 1995;20: 1307–1314. Boos N, Weissbach S, Rohrbach H, Weiler C, Spratt KF, Nerlich AG: Classification of age-related changes in lumbar intervertebral discs: 2002 Volvo Award in Basic Science. Spine 2002;27: 2631–2644. Miller JA, Schmatz C, Schultz AB: Lumbar disc degeneration: Correlation with age, sex, and spine level in 600 autopsy specimens. Spine 1988;13:173–178. Boos N, Dreier D, Hilfiker E et al: Tissue characterization of symptomatic and asymptomatic disc herniations by quantitative magnetic resonance imaging. J Orthop Res 1997;15:141–149. Antoniou J, Pike GB, Steffen T, et al: Quantitative magnetic resonance imaging in the assessment of degenerative disc disease. Magn Reson Med 1998;40:900–907. Boos N, Wallin A, Schmucker T, Aebi M, Boesch C: Quantitative MR imaging of lumbar intervertebral disc and vertebral bodies: Methodology, reproducibility, and preliminary results. Magn Reson Imaging 1994;12:577–587. Berlemann U, Gries NC, Moore RJ: The relationship between height, shape and histological changes in early degeneration of the lower lumbar discs. Eur Spine J 1998;7:212–217.

Leo/Walker/Anderson

24

14 15 16 17 18 19 20 21 22 23 24 25 26 27

28

29 30 31

32 33

34 35 36 37 38

39

Andersson GB: What are the age-related changes in the spine? Baillieres Clin Rheumatol 1998; 12:161–173. Freemont AJ, Watkins A, Le Maitre C, et al: Nerve growth factor expression and innervation of the painful intervertebral disc. J Pathol 2002;197:286–292. Jones MD, Pais MJ, Omiya B: Bony overgrowths and abnormal calcifications about the spine. Radiol Clin North Am 1988;26:1213–1234. O’Neill TW, McCloskey EV, Kanis JA, et al: The distribution, determinants, and clinical correlates of vertebral osteophytosis: A population based survey. J Rheumatol 1999;26:842–848. Videman T, Battie MC, Gibbons LE, Maravilla K, Manninen H, Kaprio J: Associations between back pain history and lumbar MRI findings. Spine 2003;28:582–588. Taylor JR, Twomey LT: The development of the human intervertebral disc; in Ghosh P (ed): The Biology of the Intevertebral Disc. Boca Raton, CRC Press, 1988, pp 39–82. Wyburn GM: Observations on the development of the human intervertebral column. J Anat 1944;78. Walmsley R: The development and growth of the intervertebral disc. Edinburgh Med J 1953;60. Virchow R: Untersuchungen über die Entwicklung des Schädelgrundes. Berlin, Reimer, 1857. Peacock A: Observations on the prenatal development of the intervertebral disc in man. J Anat 1951;85. Peacock A: Observations on the postnatal structure of the intervertebral disc in man. J Anat 1952;86. Jeffrey J, Campbell D, Golden M, et al: Antenatal factors in the development of the lumbar vertebral canal. Spine 2003;28:1418–1423. Freemont AJ, Peacock TE, Goupille P, Hoyland JA, O’Brien J, Jayson MI: Nerve ingrowth into diseased intervertebral disc in chronic back pain. Lancet 1997;350:178–181. Buckwalter JA, Mow VC, Boden SD, Eyre DR, Weidenbaum M: Intervertebral disk structure, composition, and mechanical function; in Buckwalter JA, Einhorn TA, Simon SR (eds): Orthopaedic Basic Science – Biology and Biomechanics of the Musculoskeletal System, ed 2. Rosemont, American Academy of Orthopaedic Surgeons, 2000, pp 548–555. Kim KW, Lim TH, Kim JG, Jeong ST, Masuda K, An HS: The origin of chondrocytes in the nucleus pulposus and histologic findings associated with the transition of a notochordal nucleus pulposus to a fibrocartilaginous nucleus pulposus in intact rabbit intervertebral discs. Spine 2003; 28:982–990. Eyre DR, Muir H: Types I and II collagens in intervertebral disc. Interchanging radial distributions in annulus fibrosus. Biochem J 1976;157:267–270. Eyre DR: Biochemistry of the intervertebral disc. Int Rev Connect Tissue Res 1979;8: 227–291. Buckwalter JA, Martin J: Intervertebral disk degeneration and back pain; in Weinstein JN, Gordon SL (eds): Low Back Pain: A Scientific and Clinical Overview. Rosemont, American Academy of Orthopaedic Surgeons, 1996. Roberts S, Urban JP, Evans H, Eisenstein SM: Transport properties of the human cartilage endplate in relation to its composition and calcification. Spine 1996;21:415–420. Iwahashi M, Matsuzaki H, Tokuhashi Y, Wakabayashi K, Uematsu Y: Mechanism of intervertebral disc degeneration caused by nicotine in rabbits to explicate intervertebral disc disorders caused by smoking. Spine 2002;27:1396–1401. Uematsu Y, Matuzaki H, Iwahashi M: Effects of nicotine on the intervertebral disc: An experimental study in rabbits. J Orthop Sci 2001;6:177–182. Holm S, Maroudas A, Urban JP, Selstam G, Nachemson A: Nutrition of the intervertebral disc: Solute transport and metabolism. Connect Tissue Res 1981;8:101–119. Urban JP, Holm S, Maroudas A, Nachemson A: Nutrition of the intervertebral disc: Effect of fluid flow on solute transport. Clin Orthop 1982;296–302. Katz MM, Hargens AR, Garfin SR: Intervertebral disc nutrition. Diffusion versus convection. Clin Orthop 1986;243–245. Horner HA, Urban JP: 2001 Volvo Award Winner in Basic Science Studies: Effect of nutrient supply on the viability of cells from the nucleus pulposus of the intervertebral disc. Spine 2001;26: 2543–2549. Holm S, Nachemson A: Nutrition of the intervertebral disc: Acute effects of cigarette smoking. An experimental animal study. Ups J Med Sci 1988;93:91–99.


25

40 41 42 43 44 45 46 47

48 49 50 51

52 53

54 55 56

57

58 59 60 61 62 63 64

Ohshima H, Urban JP: The effect of lactate and pH on proteoglycan and protein synthesis rates in the intervertebral disc. Spine 1992;17:1079–1082. Bartels EM, Fairbank JC, Winlove CP, Urban JP: Oxygen and lactate concentrations measured in vivo in the intervertebral discs of patients with scoliosis and back pain. Spine 1998;23:1–7. Cholewicki J, McGill SM, Norman RW: Lumbar spine loads during the lifting of extremely heavy weights. Med Sci Sports Exerc 1991;23:1179–1186. Acaroglu ER, Iatridis JC, Setton LA, Foster RJ, Mow VC, Weidenbaum M: Degeneration and aging affect the tensile behavior of human lumbar anulus fibrosus. Spine 1995;20:2690–2701. Panagiotacopulos ND, Knauss WG, Bloch R: On the mechanical properties of human intervertebral disc material. Biorheology 1979;16:317–330. Fujita Y, Duncan NA, Lotz JC: Radial tensile properties of the lumbar annulus fibrosus are site and degeneration dependent. J Orthop Res 1997;15:814–819. Ishihara H, Urban JP: Effects of low oxygen concentrations and metabolic inhibitors on proteoglycan and protein synthesis rates in the intervertebral disc. J Orthop Res 1999;17:829–835. Kroeber MW, Unglaub F, Wang H, et al: New in vivo animal model to create intervertebral disc degeneration and to investigate the effects of therapeutic strategies to stimulate disc regeneration. Spine 2002;27:2684–2690. Natarajan RN, Ke JH, Andersson GB: A model to study the disc degeneration process. Spine 1994;19:259–265. Sandover J: Dynamic loading as a possible source of low-back disorders. Spine 1983;8:652–658. Hadjipavlou AG, Simmons JW, Yang JP, et al: Torsional injury resulting in disc degeneration. I. An in vivo rabbit model. J Spinal Disord 1998;11:312–317. Hadjipavlou AG, Simmons JW, Yang JP, Bi LX, Simmons DJ, Necessary JT: Torsional injury resulting in disc degeneration in the rabbit. II. Associative changes in dorsal root ganglion and spinal cord neurotransmitter production. J Spinal Disord 1998;11:318–321. MacLean JJ, Lee CR, Grad S, Ito K, Alini M, Iatridis JC: Effects of immobilization and dynamic compression on intervertebral disc cell gene expression in vivo. Spine 2003;28:973–981. Ching CT, Chow DH, Yao FY, Holmes AD: The effect of cyclic compression on the mechanical properties of the intervertebral disc: An in vivo study in a rat tail model. Clin Biomech (Bristol, Avon) 2003;18:182–189. Kasra M, Goel V, Martin J, et al: Effect of dynamic hydrostatic pressure on rabbit intervertebral disc cells. J Orthop Res 2003;21:597–603. Mimura M, Panjabi MM, Oxland TR, Crisco JJ, Yamamoto I, Vasavada A: Disc degeneration affects the multidirectional flexibility of the lumbar spine. Spine 1994;19:1371–1380. Krismer M, Haid C, Behensky H, Kapfinger P, Landauer F, Rachbauer F: Motion in lumbar functional spine units during side bending and axial rotation moments depending on the degree of degeneration. Spine 2000;25:2020–2027. Buckwalter JA, Boden SD, Erye DR, Mow VC, Weidenbaum M: Intervertebral disk aging, degeneration, and herniation; in Buckwalter JA, Einhorn TA, Simon SR (eds): Orthopaedic Basic Science – Biology and Biomechanics of the Musculoskeletal System, ed 2. Rosemont, American Academy of Orthopaedic Surgeons, 2000, pp 557–566. Scapinelli R: Lumbar disc herniation in eight siblings with a positive family history for disc disease. Acta Orthop Belg 1993;59:371–376. Matsui H, Kanamori M, Ishihara H, Yudoh K, Naruse Y, Tsuji H: Familial predisposition for lumbar degenerative disc disease. A case-control study. 1998. Bijkerk C, Houwing-Duistermaat JJ, Valkenburg HA, et al: Heritabilities of radiologic osteoarthritis in peripheral joints and of disc degeneration of the spine. Arthritis Rheum 1999;42:1729–1735. Sambrook PN, MacGregor AJ, Spector TD: Genetic influences on cervical and lumbar disc degeneration: A magnetic resonance imaging study in twins. Arthritis Rheum 1999;42:366–372. Videman T, Leppavuori J, Kaprio J, et al: Intragenic polymorphisms of the vitamin D receptor gene associated with intervertebral disc degeneration. Spine 1998;23:2477–2485. Kawaguchi Y, Osada R, Kanamori M, et al: Association between an aggrecan gene polymorphism and lumbar disc degeneration. Spine 1999;24:2456–2460. Annunen S, Paassilta P, Lohiniva J, et al: An allele of COL9A2 associated with intervertebral disc disease. Science 1999;285:409–412.

Leo/Walker/Anderson

26

65 66

67

68 69 70

71 72 73 74 75 76 77 78

79 80

81 82

83

84 85 86

87

Paassilta P, Lohiniva J, Goring HH, et al: Identification of a novel common genetic risk factor for lumbar disk disease. JAMA 2001;285:1843–1849. Takahashi M, Haro H, Wakabayashi Y, Kawa-uchi T, Komori H, Shinomiya K: The association of degeneration of the intervertebral disc with 5a/6a polymorphism in the promoter of the human matrix metalloproteinase-3 gene. J Bone Joint Surg Br 2001;83:491–495. Solovieva S, Lohiniva J, Leino-Arjas P, et al: COL9A3 gene polymorphism and obesity in intervertebral disc degeneration of the lumbar spine: Evidence of gene-environment interaction. Spine 2002;27:2691–2696. Doita M, Kanatani T, Harada T, Mizuno K: Immunohistologic study of the ruptured intervertebral disc of the lumbar spine. Spine 1996;21:235–241. Sang-Ho A, Yoon-Woo C, et al: mRNA expression of cytokines and chemokines in herniated lumbar intervertebral discs. Spine 2002;27:911–917. Onda A, Yabuki S, Kikuchi S: Effects of neutralizing antibodies to tumor necrosis factor-alpha on nucleus pulposus-induced abnormal nociresponses in rat dorsal horn neurons. Spine 2003;28: 967–972. Nygaard O, Mellgren S, Osterud B: The inflammatory properties of contained and noncontained lumbar disc herniation. Spine 1997;22:2484–2488. Gruber HE, Hanley EN Jr: Analysis of aging and degeneration of the human intervertebral disc. Comparison of surgical specimens with normal controls. Spine 1998;23:751–757. Park JB, Chang H, Kim KW: Expression of Fas ligand and apoptosis of disc cells in herniated lumbar disc tissue. Spine 2001;26:618–621. Park JB, Kim KW, Han CW, Chang H: Expression of Fas receptor on disc cells in herniated lumbar disc tissue. Spine 2001;26:142–146. Anderson DG, Izzo MW, Hall DJ, et al: Comparative gene expression profiling of normal and degenerative discs: Analysis of a rabbit annular laceration model. Spine 2002;27:1291–1296. Ariga K, Yonenobu K, Nakase T, et al: Mechanical stress-induced apoptosis of endplate chondrocytes in organ-cultured mouse intervertebral discs. Spine 2003;28:1528–1533. Gruber HE, Norton HJ, Hanley EN Jr: Anti-apoptotic effects of IGF-1 and PDGF on human intervertebral disc cells in vitro. Spine 2000;25:2153–2157. Antoniou J, Steffen T, Nelson F, et al: The human lumbar intervertebral disc: Evidence for changes in the biosynthesis and denaturation of the extracellular matrix with growth, maturation, aging, and degeneration. J Clin Invest 1996;98:996–1003. Aguiar DJ, Johnson SL, Oegema TR: Notochordal cells interact with nucleus pulposus cells: Regulation of proteoglycan synthesis. Exp Cell Res 1999;246:129–137. Gruber HE, Ma D, Hanley EN Jr, Ingram J, Yamaguchi DT: Morphologic and molecular evidence for gap junctions and connexin 43 and 45 expression in annulus fibrosus cells from the human intervertebral disc. J Orthop Res 2001;19:985–989. Crean JK, Roberts S, Jaffray DC, Eisenstein SM, Duance VC: Matrix metalloproteinases in the human intervertebral disc: Role in disc degeneration and scoliosis. Spine 1997;22:2877–2884. Furusawa N, Baba H, Miyoshi N, et al: Herniation of cervical intervertebral disc: Immunohistochemical examination and measurement of nitric oxide production. Spine 2001;26: 1110–1116. Kanemoto M, Hukuda S, Komiya Y, Katsuura A, Nishioka J: Immunohistochemical study of matrix metalloproteinase-3 and tissue inhibitor of metalloproteinase-1 human intervertebral discs. Spine 1996;21:1–8. Konttinen YT, Kaapa E, Hukkanen M, et al: Cathepsin G in degenerating and healthy discal tissue. Clin Exp Rheumatol 1999;17:197–204. Melrose J, Ghosh P, Taylor TK: Lysozyme, a major low-molecular-weight cationic protein of the intervertebral disc, which increases with ageing and degeneration. Gerontology 1989;35:173–180. Roberts S, Caterson B, Menage J, Evans EH, Jaffray DC, Eisenstein SM: Matrix metalloproteinases and aggrecanase: Their role in disorders of the human intervertebral disc. Spine 2000;25: 3005–3013. Oegema T: The role of proteinases and other degradative mechanics in idiopathic low back pain; in Weinstein JN, Gordon SL (eds): Low Back Pain: A Scientific and Clinical Overview. Rosemont, American Academy of Orthopaedic Surgeons, 1996.


27

88 Ishii T, Tsuji H, Sano A, Katoh Y, Matsui H, Terahata N: Histochemical and ultrastructural observations on brown degeneration of human intervertebral disc. J Orthop Res 1991;9:78–90. 89 Yasuma T, Koh S, Okamura T, Yamauchi Y: Histological changes in aging lumbar intervertebral discs. Their role in protrusions and prolapses. J Bone Joint Surg Am 1990;72:220–229. 90 Hormel SE, Eyre DR: Collagen in the ageing human intervertebral disc: An increase in covalently bound fluorophores and chromophores. Biochim Biophys Acta 1991;1078:243–250. 91 Shinmei M, Kikuchi T, Yamagishi M, Shimomura Y: The role of interleukin-1 on proteoglycan metabolism of rabbit annulus fibrosus cells cultured in vitro. Spine 1988;13:1284–1290. 92 Maeda S, Kokubun S: Changes with age in proteoglycan synthesis in cells cultured in vitro from the inner and outer rabbit annulus fibrosus. Responses to interleukin-1 and interleukin-1 receptor antagonist protein. Spine 2000;25:166–169. 93 Rannou F, Corvol MT, Hudry C, et al: Sensitivity of anulus fibrosus cells to interleukin 1 beta. Comparison with articular chondrocytes. Spine 2000;25:17–23. 94 Kang JD, Georgescu HI, McIntyre-Larkin L, Stefanovic-Racic M, Evans CH: Herniated cervical intervertebral discs spontaneously produce matrix metalloproteinases, nitric oxide, interleukin-6, and prostaglandin E2. Spine 1995;20:2373–2378. 95 Kang JD, Georgescu HI, McIntyre-Larkin L, Stefanovic-Racic M, Donaldson WF III, Evans CH: Herniated lumbar intervertebral discs spontaneously produce matrix metalloproteinases, nitric oxide, interleukin-6, and prostaglandin E2. Spine 1996;21:271–277. 96 Franson RC, Saal JS, Saal JA: Human disc phospholipase A2 is inflammatory. Spine 1992;17: S129–S132. 97 Gronblad M, Virri J, Tolonen J, et al: A controlled immunohistochemical study of inflammatory cells in disc herniation tissue. Spine 1994;19:2744–2751. 98 Saal JS, Franson RC, Dobrow R, Saal JA, White AH, Goldthwaite N: High levels of inflammatory phospholipase A2 activity in lumbar disc herniations. Spine 1990;15:674–678. 99 Trout JJ, Buckwalter JA, Moore KC: Ultrastructure of the human intervertebral disc. II. Cells of the nucleus pulposus. Anat Rec 1982;204:307–314. 100 Yasuma T, Makino E, Saito S, Inui M: Histological development of intervertebral disc herniation. J Bone Joint Surg Am 1986;68:1066–1072. 101 Yasuma T, Arai K, Suzuki F: Age-related phenomena in the lumbar intervertebral discs. Lipofuscin and amyloid deposition. Spine 1992;17:1194–1198. 102 Oegema TR Jr, Johnson SL, Aguiar DJ, Ogilvie JW: Fibronectin and its fragments increase with degeneration in the human intervertebral disc. Spine 2000;25:2742–2747. 103 Trout JJ, Buckwalter JA, Moore KC, Landas SK: Ultrastructure of the human intervertebral disc. I. Changes in notochordal cells with age. Tissue Cell 1982;14:359–369. 104 Hollander AP, Heathfield TF, Webber C, et al: Increased damage to type II collagen in osteoarthritic articular cartilage detected by a new immunoassay. J Clin Invest 1994;93:1722–1732. 105 Okuda S, Myoui A, Ariga K, Nakase T, Yonenobu K, Yoshikawa H: Mechanisms of age-related decline in insulin-like growth factor-I dependent proteoglycan synthesis in rat intervertebral disc cells. Spine 2001;26:2421–2426. 106 Thompson JP, Oegema TR Jr, Bradford DS: Stimulation of mature canine intervertebral disc by growth factors. Spine 1991;16:253–260. 107 Takegami K, Thonar EJ, An HS, Kamada H, Masuda K: Osteogenic protein-1 enhances matrix replenishment by intervertebral disc cells previously exposed to interleukin-1. Spine 2002;27:1318–1325. 108 Nishida K, Kang JD, Gilbertson LG, et al: Modulation of the biologic activity of the rabbit intervertebral disc by gene therapy: An in vivo study of adenovirus-mediated transfer of the human transforming growth factor beta 1 encoding gene. Spine 1999;24:2419–2425. 109 Nishida K, Gilbertson LG, Robbins PD, Evans CH, Kang JD: Potential applications of gene therapy to the treatment of intervertebral disc disorders. 110 Gruber HE, Johnson TL, Leslie K, et al: Autologous intervertebral disc cell implantation: A model using Psammomys obesus, the sand rat. Spine 2002;27:1626–1633. 111 Sato M, Asazuma T, Ishihara M, et al: An experimental study of the regeneration of the intervertebral disc with an allograft of cultured annulus fibrosus cells using a tissue-engineering method. Spine 2003;28:548–553.

Leo/Walker/Anderson

28

112 Perka C, Arnold U, Spitzer RS, Lindenhayn K: The use of fibrin beads for tissue engineering and subsequential transplantation. Tissue Eng 2001;7:359–361. 113 Alini M, Li W, Markovic P, Aebi M, Spiro RC, Roughley PJ: The potential and limitations of a cell-seeded collagen/hyaluronan scaffold to engineer an intervertebral disc-like matrix. Spine 2003;28:446–454. 114 Sato M, Asazuma T, Ishihara M, et al: An atelocollagen honeycomb-shaped scaffold with a membrane seal (ACHMS-scaffold) for the culture of annulus fibrosus cells from an intervertebral disc. J Biomed Mater Res 2003;64A:248–256. 115 Yung LJ, Hall R, Pelinkovic D, et al: New use of a three-dimensional pellet culture system for human intervertebral disc cells: Initial characterization and potential use for tissue engineering. Spine 2001;26:2316–2322. 116 Buckwalter JA, Einhorn TA, Simon SR (eds): Orthopaedic Basic Science – Biology and Biomechanics of the Musculoskeletal System, ed 2. Rosemont, AAOS, 2000, p 550. 117 Ashton-Miller JA, Schultz AB: Biomechanics of the human spine; in Mow BC, Hayes WC (eds): Basic Orthopaedic Biomechanics. Philadelphia, Lippincott-Raven, 1997, pp 353–393.

D. Greg Anderson, MD Department of Orthopaedic Surgery Thomas Jefferson University 925 Chestnut St., 5th Floor Philadelphia, PA 19107 (USA) Tel. ⫹1 267 339 3623, E-Mail [email protected]


29


Genetics of Degenerative Disc Disease Shekar N. Kurpad, Jason Lifshutz Department of Neurosurgery, Medical College of Wisconsin, Milwaukee, Wisc., USA

Introduction

Back and neck pain are some of the most common conditions for which patients seek medical attention. 80% of the population has reported this complaint at some point during their lifetime, with 5% having chronic pain [8, 14]. Back pain is the most frequent cause of activity limitation in patients under the age of 45 and is a common cause of disability and medical cost [8, 14]. While there are several causes of back pain, degenerative and mechanical disorders of the spine and intervertebral discs encompass the most common reasons. Traditionally, degeneration has been considered to be a mechanical phenomenon; current studies suggest that genetic and biochemical mechanisms may play a much larger role than is generally appreciated. Topics examined in this review include a brief description of the normal anatomy and biochemistry of the intervertebral disc. In addition, the disc complex is examined from a cellular, biochemical, and genetic basis. We conclude with a look toward the future, and how these new developments may be used in translational research in going from bench to bedside.

Biochemistry of the Intervertebral Disc

The adult intervertebral disc is an avascular fibrocartilaginous complex that links the adjacent vertebrae of the spine. Each disc is made of a gelatinous nucleus pulposus surrounded by a laminated annulus fibrosus [3, 5–8, 14]. There is no definitive interface between these two regions, and it is referred to in the literature as the ‘transition zone’ [8, 14]. The adult disc is comprised of

poorly characterized cells surrounded by an exhaustive extracellular matrix. There are generally two types of cells, fibrocytes in the outer annulus and chondrocytes in the remaining layers [3, 5–8, 14], which are thought to be better equipped to withstand the avascular environment of the disc. The role of these cell populations is to synthesize, maintain and repair the matrix of the disc. The matrix of the disc is a framework of polar macromolecules bound with water, mainly collagen fibrils and proteoglycans, which are glycosaminoglycans such as chondroitin sulfate and keratin sulfate attached to a protein core. Collagens are important in conferring tensile strength to the disc, and appear to play a role in the genetic predisposition toward spinal degeneration. They make up approximately 70% of the annulus but only 20% of the nucleus pulposus. In the annulus, collagen is found in tightly packed fibrils arranged in specific lamellae. The majority of collagen found in this region is type I, with smaller amounts of types II, III, V, and IX also being present. Within the nucleus pulposus, 85% of the collagen is of type II, with smaller amounts of VI and IX also being found [3, 7, 8, 11, 14]. Proteoglycans absorb water, conferring both stiffness and resilience to the disc; they are present within the lamellae of the collagen fibrils and are found in the nucleus pulposus in their greatest concentrations. The proteoglycan aggregate is made up of a central glycosaminoglycan hyaluronate filament that is attached to various proteoglycans via linker proteins [7, 8, 14]. Intervertebral discs are situated between the cartilaginous endplates of adjacent vertebrae. The endplates are initially composed of hyaline cartilage, produced by chondrocytes; later in life, this is replaced in part by calcified cartilage. Collagen fibers of the annulus attach directly to the endplates. This is a site susceptible to mechanical failure, especially when exposed to shear forces [8, 11, 14]. At birth the intervertebral discs have an abundant vascular supply. As one ages, the discs generally lose this vascular supply (usually by 2 years of age), a process accelerated by the degeneration and calcification of the vertebral endplate, and subsequently derive their nutrients and eliminate their waste through the process of diffusion, which is driven by the high osmotic pressure within the discs plus the hydrostatic pressure acting on the discs. This process requires a high intrinsic water content, provided by the proteoglycan attraction.

Molecular Biology of Disc Degeneration

Diffusion The main characteristic of disc degeneration is the loss of the hydrostatic properties of the disc. At birth, the nucleus pulposus contains 85–90% water. With aging, this percentage drops to approximately 70%. This drop in water content occurs with changes in the proteoglycan extracellular matrix of the disc.

Genetics of Degenerative Disc Disease

31

A decrease in the turgor of the disc causes the disc to lose its ability to resist compression and provides resilience to the spine. The diffusion of disc homeostatic processes may become impaired in this dehydrated state due to quantitative and qualitative changes in the extracellular matrix [8, 14]. Diffusion may become impaired by other means that decrease blood flow to the spine/disc. Blood flow to the disc space regresses as aging proceeds and other concomitant factors such as diabetes and vascular disease may impair the vascular supply to the vertebral endplate [8, 14]. Calcification of the vertebral endplate may have a detrimental effect on the diffusive processes necessary for the nutrition of resident cells. Lack of adequate oxygen supply to cellular structures of the disc may lead to anaerobic glycolysis, lactate production, a decrease in the pH, and ultimately breakdown and damage to the extracellular matrix. This pH change causes poor hydrostatic pressure, and hence further deterioration in the diffusion process, with its related consequences [3, 5, 7, 8, 14]. Cellular Activity During the early phases of degeneration, there is thought to be a proliferation of cells within the annulus, with a metaplasia of these cells into chondrocytes [7, 8, 14]. As degeneration continues, diffusion through the disc declines, leading to anaerobic metabolism and cell death. With the loss of support cells, synthesis and maintenance of the disc matrix disappears with decreased water content and hence a decrease in the diffusion of essential nutrients. Biochemical Changes Proteoglycans Both quantitative and qualitative changes are seen in the extracellular matrix of the aging disc as it degenerates. First, the amount of proteoglycans decreases in the degenerating disc. This decrease in proteoglycans is generally an early sign of degeneration. The mucopolysaccharide complexes in young discs are made up primarily of chondroitin sulfate A and C side chains, which are strongly hydrophilic [8, 14]. As the disc ages, these large molecules break down into smaller ones, such as chondroitin sulfate B and keratin sulfate, which do not have the water-storing capacity of types A and C chondroitin sulfate, and subsequently lead to disc dehydration [8, 14]. Keratin sulfate, in particular, has been found to be a marker of disc degeneration in surgical and pathological specimens [5]. Disc dehydration eventually results in impaired diffusion and further disc degeneration [14]. Collagen As previously mentioned, the intervertebral disc is composed mostly of type I and type II collagen [3, 5–8, 14]. Type I collagen is found primarily in

Kurpad/Lifshutz

32

the annulus fibrosus and type II collagen in the nucleus pulposus. In early degeneration, the location and types of collagen do not change. However, there is an increase in the amount of collagen found within both the annulus and nucleus. There also may be an increase in collagen type III, V, and VI [3, 7, 8, 14]. As degeneration continues, there are qualitative changes seen in the type of collagen in various portions of the disc. For example, there is an increase in the amount of collagen type I within the nucleus. In addition, there is a loss of collagen type II at the endplates. Further changes that are seen include the formation of collagen types IV and X, and changes in the post-translational modification of these new fibrils. These larger collagen fibrils are thought to be weaker than their narrow counterparts, which may lead to tearing and disruption [8, 14]. Young and healthy discs have been found to have active matrix formation and denaturation of collagen type II. Studies on the role of collagen in disc degeneration show a decrease in the levels of aggrecan and type II procollagen formation and a general increase in type II collagen degeneration and type I synthesis. Type IX collagen has recently been reported to be present in both the aging and degenerated disc, whose formation may represent a compensatory repair process. A molecular defect in type IX collagen has been discovered as a contributing factor toward disc degeneration, specifically a conversion of the codon for glutamate to tryptophan in the COL9A2 gene. This genetic polymorphism was found in approximately 10% of patients with disc disease, and MRI correlated an association of the genetic defect with radial tears within the disc. In addition, another genetic mutation in collagen formation, the TRP3 allele, has been found in 12% of patients studied with disc degeneration. Inflammation There are several inflammatory mediators which are present within the degenerative disc. Some of those that are present include nitric oxide, interleukin-6 (IL-6), IL-8 [2] (which has been associated with radiculopathy), prostaglandin E2, and a family of enzymes known as matrix metalloproteinases (MMPs) [3–13]. IL-6 and Prostaglandin E2 seem to exert their effects by inhibiting proteoglycan synthesis. They appear to be under the influential control of IL-1, which is both a central mediator of the inflammatory process, and a direct toxin to the proteoglycan matrix. The principal cause of the release of IL-1 is yet to be determined [4, 7, 9]. The role of extracellular matrix degeneration in the disc disease process is becoming better understood. MMPs are proteinases that degrade at least one component of the extracellular matrix; they are secreted in a latent form and


33

require activation for proteolytic activity. Their activity is inhibited by specific tissue inhibitors. This family of enzymes can be divided into four groups; collagenases, stromelysins, gelatinases and membrane metalloproteinases [3, 4, 7, 8, 14]. Recent studies suggest that this family of proteases, in particular MMP-1, MMP-2, MMP-3 and MMP-9, play a significant role in the degradation and degeneration of the intervertebral disc [3–5, 7–12]. Recent studies showed that increases in the mRNA in MMP-1 and MMP-3 have been reported in cell populations of discs cultured with IL-1, IL-12 and tumor necrosis factor-␣. Further evidence of the role of MMPs is demonstrated in their activity in the degeneration of aggrecan in degenerated discs. Finally, changes in the role of cathepsins, fibromodulin, and fibronectin are being described and elucidated in the degenerative disease process [8, 14]. A therapeutic goal is to find potential inhibitors of these enzymes. Some experimental work has targeted the enzymes with specifically designed inhibitors. Other inhibitors being investigated include hydroxamic acid derivatives, tetracyclines and quinolones [8, 14]. Vertebral Endplates There are a variety of biological changes that occur in the degeneration of the vertebral endplates, which contribute, at least in part, to this disease phenomenon. As the endplate ages, it becomes calcified and replaced by bone. This calcification and bone formation impedes the vascular supply to the disc and accelerates the negative diffusion process described above. In addition, these bony changes within the endplate lead to an unequal distribution of loadsharing forces across the disc complex, which further accelerates the degenerative process. There is currently a paucity of biological study on endplate changes in degenerative disease. However, recent work has described differences in the proteoglycan composition on degenerated endplates, the contribution of MMP-3 in endplate degeneration, and a possible role for apoptosis in this disease process.

Genetics in Disc Disease

The role of genetics in degenerative disc disease is still largely unknown. Any genetic defect affecting collagen synthesis, proteoglycan synthesis, or growth and development of resident support cells (e.g., chondrocytes) would be expected to impact the rate of disc degeneration. As described above, Annunen et al. reported on the role of the COL9A2 gene in disc degeneration in a cohort of patients in Finland. In their report, this gene, which codes for a portion of the type IX collagen of the disc, was screened and found to have common polymorphisms in patients with intervertebral disc disease, associated with

Kurpad/Lifshutz

34

abnormal formation of the collagen 3 chain. Gruber and Hanley have described the role of apoptosis in both age and degenerative changes which related to a decrease in the amount of cells in the disc, and it appears likely that genetic factors could contribute toward a susceptibility to this process. Fibroblast-derived growth factors may regulate proteolytic activity in herniated discs. Nagano et al. studied degenerated disc in the rat model and showed that that normal structures were replaced with scattered chondrocytes with fibroblast growth factor-like activity and receptors [7]. New studies of the canine disc have demonstrated that both epidermal growth factor and transforming growth factor are associated with matrix synthesis and cell proliferation within the disc, and could represent another genetic risk factor if such factors are deficient.

Conclusion

Degenerative spine disease is a common problem for which neurosurgical patients seek treatment. New technology and advances in basic science has led to a better and more thorough understanding of the molecular biology of this disease process. With this new information, we may be able to develop novel and minimally invasive therapeutics for this very common disease entity. The role of genetics has until recently been underemphasized in the etiology of disc disease, but as more genetic influences are recognized, the number of targets for in vivo and ex vivo gene transfer will increase and new therapies will become available.

References 1 2 3 4 5 6 7 8

Ahn S-H, et al: mRNA expression of cytokines and chemokines in herniated lumbar intervertebral discs. Spine 2002;27:911–917. Burke JG, et al: Spontaneous production of monocyte chemoattractant protein-1 and interleukin-8 by the human lumbar intervertebral disc. Spine 2002;27:1402–1407. Crean JKG, et al: Matrix metalloproteinases in the human intervertebral disc: Role in disc degeneration and scoliosis. Spine 1997;22:2877–2884. Doita M, et al: Influence of macrophage infiltration of herniated disc tissue on the production of matrix metalloproteinases leading to disc resorption. Spine 2001;26:1522–1527. Fujita K, et al: Neutral proteinases in human intervertebral disc – Role in degeneration and probable origin. Spine 1993;18:1766–1773. Furusawa N, Baba H, et al: Herniation of cervical intervertebral disc – Immunohistochemical examination and measurement of nitric oxide production. Spine 2001;26:1110–1116. Goupille P, Jayson M, et al: Matrix metalloproteinases: The clue to intervertebral disc degeneration? Spine 1998;23:1612–1626. Guiot B, Fessler R: Molecular biology of degenerative disc disease. Neurosurgery 2000;47: 1034–1040.


35

9

10 11 12 13 14

Kang J, et al: Toward a biochemical understanding of human intervertebral disc degeneration and herniation – Contributions of nitric oxide, interleukins, prostaglandin E2 and matrix metalloproteinases. Spine 1997;22:1065–1073. Matsui Y, et al: The involvement of matrix metalloproteinases and inflammation in lumbar disc herniation. Spine 1998;23:863–869. Nemoto O, et al: Matrix metalloproteinase-3 production by human degenerated intervertebral disc. J Spinal Disord 1997;10:493–498. Roberts S, Caterson B, et al: Matrix metalloproteinases and aggrecanase – Their role in disorders of the human intervertebral disc. Spine 2001;25:3005–3013. Takao T, Iwaki T: A comparative study of localization of heat shock protein 27 and heat shock protein 72 in the developmental and degenerative intervertebral discs. Spine 2002;27:361–368. Vaccaro A, Betz R, Zeidman S (eds): Principles and Practice of Spine Surgery. Mosby Publisher, 2003, chaps 6, 28.

Shekar N. Kurpad, MD, PhD Department of Neurosurgery, Medical College of Wisconsin 9200 West Wisconsin Avenue, Milwaukee, WI 53226 (USA) Tel. ⫹1 414 805 3666, E-Mail [email protected]

Kurpad/Lifshutz

36


Gene Therapy for Degenerative Disc Disease Joseph Kim, Lars G. Gilbertson, James D. Kang Department of Orthopaedic Surgery, University of Pittsburgh Medical Center, Pittsburgh, Pa., USA

Introduction

Degenerative disc disease (DDD) is a chronic process that can clinically manifest in multiple disorders, such as idiopathic back pain, disc herniation, radiculopathy, myelopathy, and spinal stenosis. It is a significant source of patient pain and morbidity, utilizing a large portion of health care resources [2, 3, 5, 15, 34]. The available treatment options for the clinical manifestations of DDD include conservative measures such as bed rest, anti-inflammatory drugs, analgesia, and physical therapy. This approach is usually effective in alleviating symptoms within 2 months in the majority of cases. However, when conservative methods fail, invasive surgical procedures such as discectomy, instrumentation, or fusion, with their inherent complication risks and expense, may be required. These treatment modalities focus on the clinical symptoms of intervertebral disc (IVD) degeneration without addressing the underlying pathological processes occurring early in the course of degeneration. However, recent advances in molecular biology may result in the development of novel therapies that target the ongoing physiological changes that occur in this disease. Although the pathophysiology of DDD is not completely understood, an insult to the disc or its supporting structures initially leads to a cascade of cellular changes that may promote either healing or further disc degeneration. One major contributing factor includes the progressive decline in aggrecan, the primary proteoglycan of the nucleus pulposus [1, 7, 24]. At the biochemical level, aggrecan homeostasis is altered by various combinations of decreased synthesis and increased breakdown. Kang et al. [13] demonstrated increased levels of matrix metalloproteinases in degenerated human discs compared to

normal, nondegenerated controls. These enzymes are known to contribute to net proteoglycan loss by increasing its degradation. With reductions in proteoglycan content of the intervertebral matrix, the nucleus pulposus dehydrates, decreasing both disc height and its load-bearing capacity [8, 32, 33]. This may directly affect biomechanical function by altering the loads experienced by the facet joints, leading to degenerative changes. Although disc degeneration most probably evolves in response to a complex interplay of multiple biochemical and biomechanical factors [10], the ability to restore proteoglycan content may have therapeutic benefit by increasing disc hydration and potentially improving biomechanics. The ability to increase proteoglycan synthesis in the IVD was demonstrated by Thompson et al. [30] who showed that the exogenous application of human transforming growth factor (TGF)-1 to canine disc tissue in culture stimulated in vitro proteoglycan synthesis. The authors suggested that growth factors might be useful for the treatment of disc degeneration. Subsequent studies with other growth factors such as insulin-like growth factor-1 (IGF-1), bone morphogenic protein-2 (BMP-2), and osteogenic protein-1 also exhibited the ability to up-regulate proteoglycan content in IVD cells [23, 29]. However, due to the relatively brief half-life of these factors, practical application of growth factor therapy to chronic conditions such as DDD would necessitate repeated administrations. Consequently, efforts were directed at developing approaches to induce endogenous synthesis of growth factors via gene therapy such that genetically modified disc cells manufacture the desired growth factors on a continuous basis.

Overview of Gene Therapy

The definition of gene therapy has become quite broad. The term was previously used to describe replacement of a defective gene with a functional copy by means of gene transfer. The diseases originally targeted for gene therapy were classic, heritable genetic disorders. The term now defines therapy involving the transfer of genes encoding therapeutic proteins into cells to treat any disease [26]. Genetically altered cells are made into factories producing disease-altering proteins. These proteins affect not only the metabolism of the cells from which they were made, but they can also affect the metabolism of adjacent nongenetically altered cells via paracrine mechanisms (fig. 1). Successful gene therapy for DDD will depend on efficient transfer of genes to target cells with sustained expression. With few exceptions, naked DNA is not taken up and expressed by cells and consequently, vectors are necessary to package and insert genes into cells in such a way that the genetic

Kim/Gilbertson/Kang

38

DNA coding for growth factor

Growth factor released from cell

Viral vector carrying growth factor gene Ribosomes making growth factors

Cell

Nucleus

Fig. 1. The DNA encoding the growth factor of interest is constructed into a viral vector that is rendered incapable of replication. The vector is then exposed to host cells, attaches to their surface, and is then internalized. The released genetic information can then either travel to the nucleus, where it may become integrated into the host genome or remain episomal. It then commandeers the normal protein-making machinery of the cell and produces large quantities of the transgene.

information can be expressed. The two broad categories of vectors are viral and nonviral. The most commonly used nonviral vectors are liposomes. These phospholipid vesicles deliver genetic material into a cell by fusing with the cell’s phospholipid membrane. Liposome vectors are simple, inexpensive, and safe. Their drawbacks are transient expression of the transgene, cytotoxicity at higher concentrations, and low efficiency of transfection. Other nonviral methods of gene delivery include DNA-ligand complexes and the biolistics or penetration of cells with a ‘gene gun.’ These vectors are nonpathogenic and relatively inexpensive to construct. However, the overall transfection efficiency of nonviral vectors is generally inferior to that of viral-mediated gene transfer. Thus, most current studies involving gene therapy employ viral vectors. The most commonly used viral vectors are retroviruses, herpes simplex viruses, adeno-associated viruses, and adenoviruses; although, there are likely to be many other naturally occurring viruses which could be adapted for gene transfer. Viruses are frequently rendered incapable of replication prior to gene therapy application in an effort to make them less pathogenic. The various viral vectors and their advantages and disadvantages are discussed elsewhere in this volume, and will not be discussed further.

Gene Therapy for Degenerative Disc Disease

39

There are two fundamental approaches to delivering exogenous genes with vectors to target cells within the body. The first is the direct, in vivo method in which the gene-carrying vector is directly injected into the patient. The second approach, known as ex vivo gene therapy, involves removing target cells from the body, genetically altering them in vitro, and then reimplanting them in the body. There are advantages and disadvantages to both of these approaches that depend on the anatomy and physiology of the target organs, the pathophysiology of the disease being treated, the vector of choice, and safety considerations [9].

Biology of the IVD

The IVD is an avascular organ consisting of nucleus pulposus cells scattered within an extracellular matrix. This gel-like inner core is encircled by an outer annulus fibrosis. The highly differentiated, nondividing cells of the nucleus pulposus are responsible for matrix synthesis. The matrix of a healthy nucleus pulposus is normally rich in proteoglycans and type II collagen, with a water content of over 85% by volume in juveniles, decreasing to approximately 70–75% in adults, and decreasing even further with aging and degeneration [11, 12]. This progressive loss of matrix and water content reflects the inability of the IVD for self-repair, as demonstrated by Bradford et al. [6]. The avascular disc receives the majority of its nutrition via passive diffusion through the cartilaginous endplates. The lack of a direct vascular supply results in low oxygen tension within the disc and causes the cells of the nucleus pulposus to undergo anaerobic metabolism. The ensuing high lactate concentration and subsequent low environmental pH most likely inhibit matrix repair. The avascularity of the IVD, though limiting its potential for repair and regeneration, does confer a distinct advantage in the context of gene therapy application. Early research attempting to characterize the IVD demonstrated an autoimmune response when subjects were exposed to their own nucleus pulposus tissue, suggesting that the IVD is an immune-privileged site exempt from prior exposure to the host immune system [4].

Adenoviral (Ad) Vectors for Gene Therapy to the IVD

Vectors based on adenoviruses have been frequently used in gene therapy studies for the IVD due to their ability to efficiently transduce highly differentiated, nondividing cells such as the cells of the nucleus pulposus. However, successful gene therapy depends not only on efficient gene transfer, but also on the expression of transgene for sufficiently long periods of time. The duration

Kim/Gilbertson/Kang

40

of gene expression following adenoviral (Ad) transfer to an immunocompetent animal is limited in most organs and tissues by immune reactions to viral proteins and to foreign proteins encoded by the transgene [17, 31, 36]. Specifically, expression for longer than 12 weeks has been difficult to achieve in musculoskeletal tissues following Ad delivery due to brisk immune responses. On the other hand, sustained gene expression can occur for longer periods of time in an immune-privileged site such as the IVD. Kang et al. [14] demonstrated positive expression of the marker genes lacZ and luciferase in the rabbit lumbar disc one year after adenovirus-mediated transduction. A corollary experiment attempting to characterize the immune response in the same study showed no production of neutralizing antibody to viral proteins in 3 of 6 rabbits after intradiscal injections of Ad vectors carrying the luciferase marker gene. The positive immune response in the other 3 rabbits was presumably due to the leakage of virus from an injected disc. Importantly, all 6 rabbits, including the 3 with circulating antibodies, had positive transgene expression several weeks after injections. On histological review, there was no evidence of cellular infiltration, increased vascularity, fibrosis, or other hallmarks of an immune response. Furthermore, intradiscal expression of luciferase was apparent for up to at least 42 days in rabbits whose immune system had been deliberately primed by subcutaneous inoculations with Ad proteins 2 weeks prior to intradiscal vector-gene injections (fig. 2). This implied that circulating antibodies do not reach the IVD in significant quantities to exert an immune response against the transduced disc cells. These findings further confirmed the IVD as being an immune-privileged organ and demonstrated the feasibility of using Ad vectors to achieve efficient, sustained expression of foreign genes.

Comparative Studies of Intradiscal Gene Therapy

With the success of in vitro studies demonstrating an increase in proteoglycan synthesis in IVD cells treated with the exogenous administration of growth factors, efforts were directed towards stimulating endogenous synthesis of these proteins by IVD cells with the use of gene therapy. Nishida et al. [22] reported the first successful in vivo gene transfer to the IVD in 1998, using an Ad vector to deliver the lacZ marker gene to the rabbit lumbar disc. The authors were able to demonstrate sustained transgene production with no significant reduction in the expression for up to 3 months after transduction (fig. 3). Followup studies revealed evidence of continued foreign gene expression even at one year [14]. Notably, the rabbits used in these studies showed no signs of systemic illness in response to the Ad vector and its transgene synthesis. In addition, no histological changes suggesting a cellular immune response were observed.


41

0.3

O.D. 405

O.D. 405

0.4

Group A

0.4

0.2

0.3 0.2 0.1

0.1

0

0 0

a

Group B

7

14

Intradiscal injection

0.4

21

28

35

42

2

0

4

6

8

0

(log10 RLU) @42 days

Time (days)

b

7

14

21

28

Intradiscal and subcutaneous injection

35

42

0

2

4

6


Group C

O.D. 405

0.3 0.2 0.1 0 14 7

c

Subcutaneous injection

0

7

14

21

28

35

42

Intradiscal injection

0

2

4

6

8


Fig. 2. a–c Plots on the left show sequential production of specific antibodies for Ad proteins in peripheral blood. Bar charts on the right show luciferase activity at 42 days postintradiscal injection of Ad-luciferase. a In Group A, 3 rabbits produced little or no antibody to Ad proteins in the peripheral blood after injection of Ad-luciferase into the lumbar IVD. In the remaining 3 rabbits, antibody was produced within 3 weeks, presumably due to the leakage of virus from injected discs. b In Group B, all rabbits exhibited significantly increased production of antibody within 2 weeks after simultaneous subcutaneous and intradiscal injection of Ad-luciferase. c In Group C, the rabbits were immunized by subcutaneous injections of Ad-luciferase 2 weeks prior to the intradiscal gene therapy. All rabbits exhibited increased production of antibody in the peripheral blood by the time of the intradiscal injection. a–c As shown in the bar charts (right), all rabbits from the three groups exhibited significant amounts of intradiscal transgene expression. There was no correlation between the neutralizing antibody titer and intradiscal transgene expression at 6 weeks postintradiscal injection by Pearson correlation analysis (p 0.395).

Encouraged by these results with marker proteins, successful in vivo transduction of the IVD with a putative therapeutic gene was soon accomplished [21]. Using Ad vector, the gene for human TGF-1 was delivered. This study demonstrated a 30-fold increase in active TGF-1 synthesis and a 5-fold increase in total TGF-1 production in discs injected with the Ad-growth factor construct (fig. 4). Biological modulation was also documented by a 100%

Kim/Gilbertson/Kang

8

42

increase in proteoglycan synthesis (fig. 5). Assays for TGF-1 production and proteoglycan synthesis were performed with experimental discs that had been injected with Ad vector carrying only the luciferase marker gene. These viral control discs demonstrated no increase in production of TGF-1 or proteoglycan, indicating that the increases in the TGF-1 experimental group were a result of transgene expression and not a nonspecific response to the Ad vector. As in the previous studies, no signs of local or systemic immune response were noted. Additional in vitro studies with cultured human nucleus pulposus cells yielded similar results. Successful transduction of the lacZ marker gene delivered via Ad vectors was achieved with human cells from degenerated discs [19]. Similar experiments with retroviral delivery of marker genes resulted in a smaller percentage of transduction [25], perhaps due to the minimal mitotic activity of the IVD cells. The response of human cells from degenerated discs to Ad-mediated delivery of TGF-1 was assessed; increased expression of TGF-1, as well as increased proteoglycan and collagen synthesis, was demonstrated in cells receiving gene therapy as compared to controls [20]. Of note, cells receiving the Ad-TGF-1 construct showed increased proteoglycan and collagen synthesis when compared to cells receiving exogenous TGF-1 protein, presumably in response to the sustained expression of this growth factor. Interestingly, the viral dose required to increase proteoglycan synthesis was significantly less than that required for 100% transduction of the cells, perhaps highlighting the ability of a transduced cell to influence the biological activity of nongenetically altered neighboring cells. The concept that successfully transduced cells appear to exert a paracrine-like effect on their nontransduced neighboring cells implies that significant alteration in protein synthesis can be achieved with a small number of transduced cells [9]. A better understanding of this paracrine effect with TGF-1 gene transfer may enable the use of decreased viral loads to achieve a therapeutic effect, thereby minimizing potential viral toxicity. These experiments were also performed with a viral control, which further established that the increase in biological activity was the result of the delivered genetic material and not of the Ad vector. Subsequent in vitro studies with other growth factors such as BMP-2 and IGF-1 documented the potential of Ad delivery of these factors to increase the proteoglycan synthesis in a viral dose-dependent manner [18, 35]. Tissue inhibitor of metalloproteinase-1 also demonstrated the same ability [34] following Ad vector delivery. Tissue inhibitor of metalloproteinase-1 is an endogenous inhibitor of matrix metalloproteinases, which are enzymes capable of degrading the extracellular matrix of the IVD [27]. This finding established a second gene therapy strategy to modify the disrupted balance of synthesis and catabolism occurring in the degenerated IVD, namely, inhibition of


43

a

b

c

d

e

f

g

Kim/Gilbertson/Kang

44

140 120

N.S.

350

1 week 6 weeks

300

100 80

*

60 40

pg/ml/mg (wet weight)

pg/ml/mg (wet weight)

*

*

250

*

200 150 100 50

20

a

N.S. 1 week 6 weeks

0

0 Intact

Saline

Ad/luciferase

Ad/TGF-1

b

Intact

Saline

Ad/luciferase

Fig. 4. a Active TGF-1 production in rabbit nucleus pulposus tissue one and 6 weeks after in vivo injection of Ad-TGF-1, as compared to intact control discs and discs injected with saline and Ad-luciferase. b Total (active and latent) TGF-1 production in rabbit disc tissue one and 6 weeks after in vivo injection. There were no significant differences in either active or total TGF-1 production at one week and 6 weeks, indicating that the therapeutic gene expression was sustained. * A significant increase over corresponding intact, saline, and viral (Ad-luciferase) control groups (p 0.05).

matrix degradation with ensuing net increases or stabilization of proteoglycan content. Considering the potential adverse effects of viral vectors, studies have been undertaken to develop strategies to minimize viral loads while maintaining the same biological effects. Experiments with combination gene therapy involving TGF-1, IGF-1, and BMP-2 suggested that these growth factors are synergistic in amplifying matrix synthesis [18]. Ad delivery of a single growth factor increased proteoglycan synthesis by a range of 180–295%, whereas combination gene therapy with two agents resulted in increases of 322–398%. When all three growth factors were combined, proteoglycan synthesis was increased by 471% (fig. 6). It remains to be determined if combination gene Fig. 3. a–g Qualitative analysis of intradiscal lacZ transgene expression up to and including one year after injection of Ad-lacZ into lumbar intervertebral discs of adult New Zealand white rabbits. Serial histological sections were stained with X-Gal and counter-stained with eosin. Representative sections of lumbar discs at 3 weeks (a, b), 6 weeks (c, d), and 24 weeks (e, f ) postinjection are shown. All of the discs injected with Ad-LacZ exhibited positive X-Gal staining. [Original magnifications: a, c, e. 40; b, d, f. 200.] At 52 weeks postinjection, positive X-Gal staining was observed in the discs from two of three rabbits. However, the intensity of positive staining was less than in discs from the other time periods (g). [Original magnification: g. 600].


45

Ad/TGF-1

N.S. 3.5

1 week 6 weeks

*

Ratio (experimental/intact)

3.0

*

2.5

2.0

1.5

1 0.5

0 Intact

Saline

Ad/luciferase

Ad/TGF-1

Fig. 5. Proteoglycan synthesis in rabbit nucleus pulposus tissue one and 6 weeks after in vivo injection of Ad-TGF-1, as compared to intact control discs and discs injected with saline and Ad-luciferase. There were no significant differences in proteoglycan synthesis at one week and 6 weeks, indicating that the biologic effects of transgene synthesis was sustained. * A significant increase over corresponding intact, saline, and viral (Ad-luciferase) control groups (p 0.05).

therapy with both an anabolic growth factor and a catabolic inhibitor such as tissue inhibitor of metalloproteinase-1 will have a similar synergistic effect.

Areas of Ongoing Research

For the continued progress of gene therapy for DDD toward successful human clinical trials, it is critical to rigorously test the proposed gene therapy strategies in animal models of disc degeneration that closely simulate the human condition. A number of models have been proposed. Disc degeneration occurs spontaneously in some species, such as the nonchondrodystrophic beagle and the sand rat [28]. Other species require artificial interventions to bring about degenerative changes within a reasonable time frame. The annular stab model of degeneration in the New Zealand white rabbit has been well described in the literature by Lipson and Muir [16]. In previous gene therapy studies with this model, our group found that a 3 mm incision of the anterior annulus allowed escape of nuclear material from the disc, with subsequent loss of viral

Kim/Gilbertson/Kang

46

700

Triple

600 Single gene

Percent control

500 400

*

Double

*

* *

*

300

*

*

200 100 0 Saline Ad/luciferase Ad/TGF-1 Ad/IGF-1 Ad/BMP-2 Ad/TGF1-1 Ad/TGF-1 Ad/IGF-1 Ad/TGF-1 Ad/BMP-2 Ad/BMP-2 Ad/IGF-1 IGF-1 Ad/BMP-2

Fig. 6. Proteoglycan synthesis in human intervertebral disc cells treated with different combinations of therapeutic Ad vectors (Ad-TGF-1, Ad-IGF-1, Ad-BMP-2). All groups showed significant increase in synthesis compared to saline and viral (Ad-luciferase) control groups (* p 0.05).

injections directed at the nucleus pulposus. There was also concern that degenerative changes induced by the full thickness 3 mm annular incision were too abrupt, in contrast to the gradual changes that occur in the human condition. For these reasons, we modified this technique to produce a puncture injury using a 16-gauge hypodermic needle. Extensive MRI and histological data have shown that the needle puncture model produces gradual and consistent degenerative changes that closely parallel human disc degeneration. The MRI analysis revealed progressive loss of mean nucleus pulposus signal intensity of stabbed lumbar discs as a function of time from puncture surgery (fig. 7). Importantly, there was no MRI evidence of spontaneous recovery of any of the degenerated discs. Histological examinations of punctured discs revealed cracks and clefts within the nucleus as well as delamination and infolding of the annulus. In addition, clusters of notochordal cells were readily apparent in healthy discs but were sparse in discs that had been punctured (fig. 8). Further validation of the puncture model was achieved by the demonstrated loss of mean water content from 85 to 70% twenty-four weeks after needle injury.


47

L1-L2 (control)

20

NP signal intensity

*

*

15

L2-L3

10 5 Gray value on NP * Area of NP

L2-L3 (stab)

L3-L4 (stab)

L4-L5 (stab)

0

Pre-op

12 weeks

*

15

24 weeks

* L3-L4

10 5 0

Pre-op

12 weeks

24 weeks

20 15

L4-L5

10 5 Pre-op

1 wk

2 wks

3 wks

6 wks

12 wks

24 wks

0

Pre-op

12 weeks

24 weeks

Fig. 7. Representative T2-weighted sagittal MRI of a New Zealand white rabbit lumbar spine: immediately before and 1, 2, 3, 6, 12 and 24 weeks after puncture surgery (left). Signal intensity of nucleus pulposus decreased with time after surgery (right). n 8 rabbits.

a

b Pre-operative

24 weeks after stab

Fig. 8. a Healthy L5–6 rabbit disc. Clusters of notochordal cells are apparent. b Degenerated L4–5 rabbit disc 24 weeks after puncture surgery. Nuclear displacement occurred, accompanied by infolding of the contralateral inner annulus towards the direction of nuclear displacement. Clusters of notochordal cells are sparse.

Kim/Gilbertson/Kang

48

Future Directions

The potential of gene therapy to alter the biological processes occurring in the degenerated IVD disc has been clearly established. The next step in this development will be to assess the feasibility of transducing degenerated rabbit discs, using needle-stab modeling as described above, with marker and therapeutic genes. Additional in vivo studies in this model will help to clarify the potential benefits and toxicity of gene therapy. Because other vectors are now available which can transduce cartilagenous cells, most notably adeno-associated vector, expanding this experimental methodology to other vector systems will be important. The basic science of the effects of growth factors and catabolic inhibitors in the biological processes and mechanical functioning of the spine also needs to be further elucidated, to determine which factors are best to promote growth of collagen and proteoglycans. Biochemical studies are necessary to delineate the relationship between viral concentration, transgene synthesis, and protein expression in the disc space. Despite the hurdles that remain, gene therapy to alter the course of IVD degeneration holds much clinical promise, and will continue to stimulate future investigations. References 1 2 3 4 5 6 7 8 9 10 11 12 13

Adler JH, Schoenbaum M, Silberberg R: Early onset of disk degeneration and spondylosis in sand rats (Psammomys obesus). Vet Pathol 1983;20:13–22. Anderson J: Back pain and occupation; in Jayson MIV (ed): The Lumbar Spine and Back Pain. London, Churchill Livingstone, 1987, pp 2–36. Anderson JA: Epidemiological aspects of back pain. J Soc Occup Med 1986;36:90–94. Bobechko: Auto-immune response to nucleus pulposus in the rabbit. J Bone Joint Surg Br 1965; 47:574–580. Borenstein D: Epidemiology, etiology, diagnostic evaluation, and treatment of low back pain. Curr Opin Rheumatol 1992;4:226–232. Bradford DS, Cooper KM, Oegema TR Jr: Chymopapain, chemonucleolysis, and nucleus pulposus regeneration. J Bone Joint Surg Am 1983;65:1220–1231. Buckwalter JA: Aging and degeneration of the human intervertebral disc. Spine 1995;20: 1307–1314. Butler D, Trafinow JH, Andersson GB, McNeil TW: Discs degenerate before facets. Spine 1990; 15:111–113. Evans CH, Robbins P: Possible orthopaedic applications of gene therapy. J Bone Joint Surg Am 1995;77:1103–1113. Garfin SR: The intervertebral disc: Disc disease – Does it exist? in Weinstein JN (ed): The Lumbar Spine. Philadelphia, W.B. Saunders, 1990, pp 369–380. Hallen A: Hexosamine and ester suphate content of the human nucleus pulposus at different ages. Acta Chem Scand 1958;12:1869–1872. Hallen A: The collagen and ground substance of the human nucleus pulposus at different ages. Acta Chem Scand 1962;16:705–709. Kang JD, Georgescu HI, McIntyre-Larkin L, Stefanovic-Racic M, Donaldson WF 3rd, Evans CH: Herniated lumbar intervertebral discs spontaneously produce matrix metalloproteinases, nitric oxide, interleukin-6, and prostaglandin E2. Spine 1996;21:271–277.


49

14 15 16 17

18

19

20

21

22

23

24 25

26 27

28 29

30 31

32 33 34 35

Kang JD, Boden SD: Orthopaedic gene therapy. Spine. Clin Orthop 2000;379S:256–259. Kraemer J: Natural course and prognosis of intervertebral disc diseases. International Society for the Study of the Lumbar Spine, Seattle, Washington, June 1994. Spine 1995;20:635–639. Lipson SJ, Muir H: 1980 Volvo Award in Basic Science: Proteoglycans in experimental intervertebral disc degeneration. Spine 1981;6:194–210. McCoy RD, Davidson BL, Roessler BJ: Expression of human interleukin-1 receptor antagonist in mouse lungs using recombinant adenovirus: Effects on vector induced inflammation. Gene Ther 1995;2:437–442. Moon S, Nishida K, Gilbertson LG, Hall RA, Robbins PD, Kang JD: Biologic response of human intervertebral disc cell to gene therapy cocktail. San Francisco, Orthopaedic Research Society, 2001. Moon SH, Gilbertson LG, Nishida K, Knaub M, Muzzingro T, Robbins PD, Evans CH, Kang JD: Human intervertebral disc cells are genetically modifiable by adenovirus-mediated gene transfer. Spine 2000;25:2573–2579. Moon SH, Nishida K, et al: Proteoglycan synthesis in human intervertebral disc cells cultured in alginate beads; exogenous TGF-1 vs adenovirus-mediated gene transfer of TGF 1 cDNA. Orlando, Florida Orthopaedic Research Society 2000; (Abstr 1061). Nishida K, Kang JD, Gilbertson LG, Moon SH, Suh JK, Vogt MT, Robbins PD, Evans CH: Modulation of the biologic activity of the rabbit intervertebral disc by gene therapy: An in vivo study of adenovirus-mediated transfer of the human transforming growth factor beta 1 encoding gene. Spine 1999;24:2419–2425. Nishida K, Kang JD, Suh JK, Robbins PD, Evans CH, Gilbertson LG: Adenovirus-mediated gene transfer to nucleus pulposus cells. Implications for the treatment of intervertebral disc degeneration. Spine 1998;23:2437–2442; discussion 2443. Osada R, Oshima H, Ishihara H: Autocrine/paracrine mechanism of insulin-like growth factor-1 secretion, and the effect of insulin-like growth factors-1 on proteoglycan synthesis in bovine intervertebral discs. J Orthop Res 1996;14:690–699. Pearce RH, Grimmer BJ, Adams ME: Degeneration and the chemical composition of the human lumbar intervertebral disc. J Orthop Res 1987;5:198–205. Reinke J, et al: Transfer of therapeutic genes to human chondrocytes-like cells of lumbar disc prolapse (abstract 56). Annual Meeting of International Society for the Study of the Lumbar Spinel, Singapore, 1997. Robbins PD, Ghivizzani SC: Viral vectors for gene therapy. Pharmacol Ther 1998;80:35–47. Roberts S, Caterson B, Menage J, Evans EH, Jaffray DC, Eisenstein SM: Matrix metalloproteinases and aggrecanase: Their role in disorders of the human intervertebral disc. Spine 2000; 25:3005–3013. Silberberg R, Aufdermaur M, Adler JH: Degeneration of the intervertebral disks and spondylosis in aging sand rats. Arch Pathol Lab Med 1979;103:231–235. Takegami K, Thonar EJ, An HS, Kamada H, Masuda K: Osteogenic protein-1 enhances matrix replenishment by intervertebral disc cells previously exposed to interleukin-1. Spine 2002;27: 1318–1325. Thompson JP, Oegema TR Jr, Bradford DS: Stimulation of mature canine intervertebral disc by growth factors. Spine 1991;16:253–260. Tripathy SK, et al: Immune responses to transgene-encoded proteins limit the stability of gene expression after injection of replication-defective adenovirus vectors. Nat Med 1996;2: 545–550. Urban JP, McMullin JF: Swelling pressure of the intervertebral disc: Influence of proteoglycan and collagen contents. Biorheology 1985;22:145–157. Urban JP, McMullin JF: Swelling pressure of the lumbar intervertebral discs: influence of age, spinal level, composition, and degeneration. Spine 1988;13:179–187. Waddell G: Low back pain: A twentieth century health care enigma. Spine 1996;21: 2820–2825. Wallach CJ, Sobajima S, Watanabe Y, Gilbertson LG, Kang JD: Gene transfer of the catabolic inhibitor TIMP-1 increases measured proteoglycans in human intervertebral disc cells. International Society for the Study of the Lumbar Spine, Cleveland, Ohio, 2002.

Kim/Gilbertson/Kang

50

36

Yang Y, et al: Cellular immunity to viral antigens limits E1-deleted adenoviruses for gene therapy. Proceedings of the National Academy of Sciences of the United States of America 1994; 91:4407–4411.

James D. Kang, MD Assistant Professor of Orthopaedic Surgery and Neurological Surgery Division of Spinal Surgery, University of Pittsburgh Medical Center Department of Orthopaedic Surgery, Liliane Kaufmann Building 3471 Fifth Avenue, Suite 1010, Pittsburgh, PA 15213 (USA) Tel. 1 412 605 3241, Fax 1 412 687 3724, E-Mail [email protected]


51


Bone Morphogenetic Proteins Spinal Fusion Applications

David A. Bomback, Jonathan N. Grauer Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, New Haven, Conn., USA

Overview of Bone Morphogenetic Proteins (BMPs)

The clinical track record of autogenous iliac crest bone graft makes it the current ‘gold standard’ for spinal arthrodesis. However, autograft utilization is accompanied by a number of limitations. For example, pseudoarthrosis rates vary from 5 to 35% [1], which prolongs recovery from surgery and leads to potential complications such as graft migration, instability, or even spinal cord impingement. Moreover, chronic donor site pain has been reported in up to 25% of patients who undergo removal of iliac crest material for spinal autografting. The availability of donor bone from a given patient may also be limited secondary to prior graft harvest or poor bone quality. Finally, an additional operative site increases blood loss, operative time, and cost [1]. Such limitations have prompted investigations into a variety of bone graft alternatives. The goals of such efforts are aimed at eliminating donor-site pain and increasing union rate with a product that is virtually limitless in supply. Such alternatives can be classified as either bone graft extenders or substitutes. Graft extenders, when added to autogenous bone, allow for arthrodesis of a greater number of levels or the use of less autograft and yield a fusion rate equal to or superior to that of autograft alone. Graft substitutes completely replace autogenous bone yet allow for comparable or increased fusion rates compared with the autograft [2]. In order to understand the biological application of bone graft alternatives, one must be familiar with the basic terminology. Osteoconduction is the ability of a material to behave as a scaffold for the ingrowth of new host bone. Osteoinduction is defined as the capability of initiating de novo bone formation

by inducing osteoblastic precursor stem cells to differentiate into mature boneforming cells. An ideal bone graft substitute must possess both of these characteristics. Osteogenesis simply refers to the ability of graft cells to directly form bone. Only autogenous bone graft and bone marrow aspirates possess osteogenic properties [3]. In 1965, Urist [4] made the observation that implanted devitalized bone was capable of inducing a cellular response resulting in new bone formation. His laboratory subsequently demonstrated that proteins extracted from the organic component of bone were responsible for such a behavior [5, 6]. Implantation of this bone matrix protein mixture into animals resulted in a multitude of cellular events including mesenchymal cell infiltration, cartilage formation, vascular ingrowth, bone formation, and bony remodeling [7]. Urist thus coined the term ‘bone morphogenetic protein’ (BMP). Over time, such extracts and proteins have become exploited and modified to induce fusions.

Biology of Spinal Fusion

The goal of spinal arthrodesis using decortication and autogenous bone graft is the development of a well-formed fusion mass bridging one bony surface to another. In order to achieve such an endpoint, a specific set of events needs to occur. First, osteoprogenitor cells must enter the fusion bed. Decortication of host bone enables cells to exit the bone marrow and enter the fusion environment. Next, osteoprogenitor cells differentiate into osteoblast precursors and ultimately mature osteoblasts, depositing new bone matrix. Finally, bony remodeling of the fusion mass occurs according to Wolff’s law (i.e., remodeling occurs in response to physical stresses; bone is deposited in sites subjected to stress and resorbed from sites of little stress), resulting in a stable fusion mass able to withstand physiological stress [3]. To study the many variables which affect bone formation and fusion, animal models have been developed. One such approach is the New Zealand white rabbit posterolateral lumbar fusion model, which has been validated and extensively studied [8]. Histological analysis reveals that maturation of the spine fusion mass occurs first in the ‘outer zone’ (adjacent to the transverse processes) followed by the ‘central zone’ (between transverse processes). This temporal and spatial sequence would be expected postdecortication because osteoprogenitor cells from the marrow must travel a longer distance to reach the central zone. In each region, inflammatory, reparative, and remodeling histological phases of bone healing can be observed [3]. In further evaluating the process, mRNA expression of various BMPs has been

Bone Morphogenetic Proteins

53

shown to occur at different times and at different locations during fusion. Such findings suggest unique roles for specific BMPs during spine fusion and indicate the potential for clinical applications of these proteins. A full understanding of the process of BMP expression during fusion has not yet been achieved, although limited gene expression studies following induction with BMP have been performed.

Molecular and Cellular Mechanisms of Action of BMPs

The BMPs are dimeric molecules belonging to the transforming growth factor-␤ superfamily based on amino acid homology [9, 10]. BMP molecules are multifunctional proteins that exhibit both autocrine and paracrine effects. They act by binding to specific serine-threonine kinase receptors present on the surface of undifferentiated mesenchymal stem cells. The receptors then transduce a signal via a group of G-proteins known as Smads, which in turn activate genes in the nucleus of the cell related to the osteoblast phenotype [11]. When applied in vivo, BMPs induce undifferentiated mesenchymal stem cells to switch from a fibrogenic to an osteogenic pathway of development, culminating in mature bone with normal marrow cavities [12]. The activity of BMPs is tightly controlled and self-limiting. Outside the cell, inhibitory proteins (e.g., noggin, chordin, follistatin) can bind specific BMPs, thus preventing their binding to cell surface receptors [11, 13, 14]. Furthermore, intracellular BMP transcription and translation is regulated by a combination of signal-transducing and inhibitory Smad proteins. BMPs can themselves up-regulate the expression of these extracellular antagonists and intracellular inhibitors, suggesting a negative feedback autoregulation cycle. As a result of all of these regulatory mechanisms, bone induction is tightly limited and bone overgrowth is avoided [11].

Research and Clinical Use of BMPs

Several BMP preparations have been, and are currently being investigated preclinically and clinically for use in spinal arthrodesis. These BMP products include recombinant human BMPs (rhBMPs) and demineralized bone matrices (DBMs). The two rhBMPs which have been most investigated are rhBMP-2 (Medtronic Sofamore Danek, Memphis, Tenn., USA) and rhBMP-7 also known as osteogenic protein-1 or OP-1 (Stryker Biotech, Hopkinton, Mass., USA). They are highly purified single proteins produced by recombinant DNA biotechnology. Others include BMP-9 [9, 15] and growth and differentiation

Bomback/Grauer

54

factor-5 [16]. Small amounts of BMP may also be present in some DBM preparations. Examples of these formulations, which are derived from human allograft bone, include, but are not limited to, Grafton (Osteotech, Eatontown, N.J., USA), Dynagraft (Gen-Sci Regeneration Laboratories, Calif., USA), and Osteofil (Regeneration Technologies, Fla., USA). However, concentration of BMPs in DBM is not thought to be sufficient for it to be a complete bone graft substitute in spinal applications [17, 18]. In addition, a more concentrated DBM preparation, bovine bone-derived BMP extract or bBMPx (Sulzer Orthopedics Biologics, Denver, Colo., USA), has been investigated [10, 19]. Theoretically, this may offer more osteogenic potential than the standard human DBM preparations.

Preclinical Studies with Specific Classes of BMP

rhBMP-2 The first preclinical interbody cage study using rhBMP-2 was by Sandhu et al. [20]. L4-L5 retroperitoneal anterior lumbar interbody fusions were performed in a sheep model. Cylindrical, threaded titanium fusion cages were filled with either iliac crest autograft or rhBMP-2. The 6-month follow-up results demonstrated a 100% fusion rate for the rhBMP-2 group compared with a 37% fusion rate for the autograft controls. A similar study in a goat model using titanium BAK fusion cages (Spinetech, Minneapolis, Minn., USA) packed with either autograft or rhBMP-2 yielded a 95% fusion rate in the rhBMP-2 and a 48% fusion rate in the autograft group [21]. Finally, another study utilizing allograft bone dowels for anterior interbody fusion in nonhuman primates showed that dowels filled with rhBMP-2 resulted in a 100% fusion rate at 6 months, as opposed to the 33% fusion rate seen in the autograft-filled dowels [22]. Numerous posterolateral fusion studies have been reported in the past decade, with similar conclusions. Schimandle et al. [23] noted a 100% fusion rate for rhBMP-2 in a rabbit posterolateral fusion model, while autograft controls fused only 42% of the time. In addition, fusions in the rhBMP-2 animals were biomechanically stronger and stiffer than autograft fusions. A canine model demonstrated 100% rhBMP-2 fusions and 0% autograft fusions at 12 weeks [24]. Use of rhBMP-2 as an autograft enhancer also has been studied in a canine model; in that model, gross specimens and CT scans demonstrated significantly increased fusion mass volume 6 months after surgery in rhBMP-2 autograft dogs when compared with autograft alone dogs [25, 26]. Martin et al. [27] established the systemic effects of ketorolac on posterolateral spine fusion and then tested rhBMP-2’s ability to overcome such inhibition. First, the investigators demonstrated an autograft fusion rate of 35% with


55

IV ketorolac pump infusion as compared to an autograft fusion rate of 75% with IV saline infusion. Fusion rates subsequently increased to 100% in an autograft/rhBMP-2 group with IV ketorolac infusion. In a related study, Silcox et al. [28] demonstrated that the inhibitory effect on fusion of systemic nicotine could be overcome with rhBMP-2. These investigators administered nicotine to rabbits via mini-osmotic pumps. They subsequently performed single-level posterolateral fusions on these rabbits comparing autograft alone to autograft mixed with rhBMP-2 to allogeneic DBM mixed with BMP-2. They achieved a 100% fusion rate in the autograft/rhBMP-2 group, a 64% fusion rate in the DBM/rhBMP-2 group and a 0% fusion rate in the autograft alone group. rhBMP-7 (OP-1) Cook et al. [29] reported the first spinal application of OP-1 using a canine posterolateral fusion model. Radiographical and histological examination revealed solid fusion for the OP-1 group 6 weeks after surgery. The autograft group attained comparable fusion rates but not until 26 weeks postsurgery. No fusions were observed for negative controls (i.e., no implant material or carrier alone). Grauer et al. [30] have studied comparative intertransverse lumbar fusion in the New Zealand white rabbit model. Study groups were autograft alone, carrier alone, or OP-1 with carrier. Manual palpation and biomechanical testing at 5 weeks confirmed a 0% fusion rate in the carrier group, a 63% fusion rate in the autograft group, and a 100% fusion rate in the OP-1 group. At 5 weeks, histology revealed more mature bone in the OP-1 group. Cunningham et al. [31] studied a skip-level posterolateral canine model using autograft alone, autograft and OP-1, or OP-1 alone. Statistically significant differences in the rate of fusion between the autograft alone and the OP-1-containing specimens were noted at all timepoints studied (4, 8 and 12 weeks postoperatively). Only a few other studies reporting the use of OP-1 in interbody fusion have been reported. Magin and Delling [32] compared OP-1, autograft alone, and an osteoconductive hydroxyapatite bone graft alternative using a posterior lumbar interbody fusion sheep model. At 4 months time, OP-1 animals had an 80% fusion rate with a 60% increase in bone formation compared to the other groups. Cunningham et al. [33] studied a sheep thoracic spine model using threaded fusion cages (BAK devices) placed thoracoscopically. BAK cages packed with OP-1 had fusion rates equivalent to those packed with autograft and to autograft bone dowel alone, suggesting that OP-1 is as effective as autograft in obtaining interbody fusion. Similar to the rhBMP-2 nicotine study above, Patel et al. [34] tested the ability of OP-1 to overcome the inhibitory effects of nicotine in a rabbit posterolateral fusion model. L5-L6 fusions were performed using either iliac crest autograft or OP-1. Nicotine was administered to all animals via a subcutaneous

Bomback/Grauer

56

mini-osmotic pump. The fusion rate was 25% for the nicotine-exposed autograft group and 100% for the nicotine-exposed OP-1 group at 5 weeks, demonstrating the latter’s ability to overcome the negative effects of nicotine in this model. DBM

Grafton DBM was shown to be effective as a graft extender with autograft in a rabbit posterolateral spine fusion model. It demonstrated equal effectiveness with autograft in a 1:1 or 3:1 dosing ratio. However, it never demonstrated fusion rates greater than the autograft alone [1]. In a canine posterior fusion model, Cook et al. [35] evaluated fusion in 9 adult mongrel dogs at 6, 12, and 26 weeks. Four sites on each animal received implants consisting of DBM gel, DBM gel with allograft, allograft alone, or autograft alone. Radiographical studies demonstrated that the autograft sites had achieved fusion by 26 weeks postoperatively. Conversely, the DBM gel alone and with the allograft demonstrated some new bone formation but did not achieve fusion by 26 weeks. Mechanically, the autograft sites demonstrated torsional stability significantly greater than all other fusion sites. Histological analysis confirmed the radiographical and mechanical findings. The results indicate that the DBM gel alone or with the allograft is inferior to the autograft. Only sparse reports are available regarding utility of bBMPx in spinal arthrodesis, but the limited information appears encouraging. Boden et al. [36] demonstrated a dose response with bBMPx in the rabbit intertransverse process fusion model. bBMPx mixed with collagen and DBM achieved fusion rates of 50–100% depending on dose. Autograft fusion rates were 62%, and DBM with collagen alone were only 17%. More recently, a posterolateral fusion model was studied in nonhuman primates comparing bBMPx delivered in DBM to the autograft alone. Efficacy data demonstrated an autograft fusion rate of 21%; the bBMPx displayed a dose response in which 3 mg per side gave twice the fusion rate as that of the autograft [19]. Clinical Studies Using BMPs

The preclinical data cited above paved the way for completed, ongoing, and future human trials with these three differentiation factors. rhBMP-2 In 1996, 14 patients with single-level lumbar degenerative disc disease were enrolled in a prospective randomized nonblinded controlled trial to test interbody cage with BMP treatment versus bone autograft. All patients received


57

Bridging bone

Fig. 1. Patient one-year status postinterbody fusion with Infuse (BMP-2 on absorbable collagen sponge) in an LT Cage (image courtesy of Medtronic Sofamor Danek, Memphis, Tenn., USA).

a tapered titanium interbody fusion device (LT Cage, Medtronic Sofamor Danek, Memphis, Tenn., USA) filled with either rhBMP-2 or iliac crest autograft. All 11 patients randomized to the rhBMP-2 group were fused radiographically at 6 months, while one of the 3 patients in the autograft group had a nonunion at one-year follow-up. Given the small numbers, the differences were not statistically significant [37], although the clinical results were considered excellent with rhBMP-2 compared to other interventions. A variety of clinical trials followed this pilot study, which have demonstrated efficacy in varied clinical applications of BMPs. Anterior lumbar interbody fusion rates performed either open or laparoscopically using rhBMP-2-filled interbody fusion cages have been shown to be equivalent to those with autograft-filled cages [37] (fig. 1). When rhBMP-2 was used with machined allografts (bone dowels) for anterior lumbar interbody fusion, it yielded higher fusion rates, superior improvement in pain and function, and a greater likelihood of returning to work compared with autograft-filled dowel controls [38]. It should be noted, however, that heterotopic bone within the spinal canal was noted in patients enrolled in an rhBMP-2 posterior lumbar interbody fusion or PLIF trial. There were no neurological sequelae reported, but the study was halted prior to completion [39]. Further investigation is warranted to define appropriate safety parameters, given the concern about bony overgrowth. Finally, the first human trial investigating rhBMP-2 as an adjunct to posterolateral intertransverse arthrodesis has recently been reported. Twentyfive patients undergoing lumbar arthrodesis were randomized based on the arthrodesis technique: autograft/Texas Scottish Rite Hospital (TSRH) pedicle

Bomback/Grauer

58

screw instrumentation (n ⫽ 5), rhBMP-2/TSRH (n ⫽ 11), and rhBMP-2 only without internal fixation (n ⫽ 9). On each side, 20 mg of rhBMP-2 were implanted. The patients had single-level disc degeneration, grade 1 or less spondylolisthesis, mechanical low back pain with or without leg pain, and failure of nonoperative treatment for at least 6 months. The radiographical fusion rate was 40% (2/5) in the autograft/TSRH group and 100% (20/20) with rhBMP-2 group with or without TSRH internal fixation (p ⫽ 0.004). In addition, statistically greater and quicker improvement in patient-derived clinical outcome was measured in the rhBMP-2 groups [40]. These results strongly suggest that BMP treatment augments the efficacy of spinal fusion in the setting of autographs, with or without internal fixation. All interbody clinical trials with rhBMP-2 demonstrated less blood loss compared to autograft controls. In addition, the incidence of donor site pain in those patients who underwent bone graft harvest was 30–40% at 2 years [37]. rhBMP-7 (OP-1) All OP-1 human trials have involved uninstrumented posterolateral intertransverse process fusion in the setting of degenerative spondylolisthesis. An early Australian study [41] placed autograft on one side and OP-1 on the contralateral side. The 6-month follow-up noted bone formation to be equal or greater on the OP-1 side (assessed by CT scan) as compared with the autograft side. Although this was encouraging, it is difficult to interpret the results of such studies with different bone graft materials on different sides, as one side may affect the other. An initial safety and efficacy study in the United States compared autograft alone to autograft augmented with OP-1 for posterolateral arthrodesis. Sixteen patients with degenerative lumbar spondylolisthesis and spinal stenosis were randomized to each treatment arm. At 6-month follow-up, 75% of patients in the OP-1/autograft group were radiographically fused, whereas only 50% in the autograft only group were fused [42] (fig. 2). A subsequent study of similar design is underway, in which patients receive either iliac crest autograft alone or OP-1 alone [43]. The 6-month results for 36 enrolled patients have shown a clinical success rate 32% higher in the OP-1 group than in the autograft group [44]. No OP-1-related adverse events have been observed to date. DBM At the present time, there have been no prospective clinical trials with human DBM products in the spine literature. A clinical trial is underway in Switzerland testing bBMPx in a posterolateral lumbar fusion but results are not yet available [19].


59

Bridging bone

Fig. 2. Patient one-year status postdecompression and posterolateral fusion with OP-1 alone for degenerative spondylolisthesis.

Future Directions in Spinal Fusion Research

Gene therapy may play an active role in future preclinical and clinical trials with various BMPs. Gene-based therapies attempt to deliver specific genes, known as transgenes, to target cells to change the existing physiological state or disease process [45]. Genes encoding for factors in the osteogenic cascade are either inserted into the patient’s own cells that exist at the fusion site (in vivo) or into cells that have been removed and will be reimplanted at the site of fusion (ex vivo) [3]. Once these cells are in place, the transgene produces a protein that initiates the bone-formation cascade. Hence, it is the activity and half-life of the transgene itself that is the limiting temporal factor for the presence of osteoinductive stimulatory signals at the fusion site [3]. This therapy might allow for potentially longer expression of BMP activity and thus perhaps an increased window of time for bone formation. Boden et al. [46] have reported successful use of gene therapy techniques in an athymic rat posterolateral spine fusion model. The gene encoding for anosteoinductive intracellular signaling protein named LIM mineralization protein-1 was identified and cloned [47]. It appears to be regulated by BMP-6 and to function very early in the cascade of events leading to de novo bone formation [46].

Bomback/Grauer

60

Using an ex vivo gene therapy strategy, the LIM mineralization protein-1 gene was transfected into the harvested bone marrow cells of athymic rats and then reimplanted at appropriate posterolateral fusion sites. Successful arthrodesis was achieved in 100% of the sites receiving cells containing the LIM mineralization protein-1 gene and 0% of the sites receiving control cells [46]. In vivo gene therapy methodology has also been reported. In two separate studies, an adenovirus vector containing either rhBMP-2 or rhBMP-9 was injected percutaneously into the lumbar paraspinal musculature of athymic rats. Both studies reported successful arthrodesis at the experimental sites without any evidence of canal or neuroforaminal compression [9, 48]. Future techniques may include delivery of genetic material via nonviral means (e.g., liposomes, gene gun therapy) or with newer viral vectors (e.g., adeno-associated virus) demonstrating less immunogenicity. All of these promising techniques, however, are certainly associated with greater expense and there are concerns of oncogenesis and bony overgrowth. Although gene therapy may allow for higher levels of osteoinductive proteins to be expressed for longer time periods, it is unknown if high fusion rates can be achieved with one-time applications. In addition, potentially increased bone production might compromise the safety of these implants. Vector design must incorporate gene regulation techniques and spine-specific targeting strategies before human clinical trials can be safely conducted [45]. In terms of local application of recombinant BMP products at the time of surgery, the use of carrier systems with BMP recombinant proteins is still in evolution. Carriers for BMP in spine fusion are used to increase the retention of these differentiation factors at the fusion site while at the same time providing an osteoconductive matrix on which bone formation can occur. Four major categories of carriers are used for BMP delivery: inorganic materials (e.g., hydroxyapatite, tricalcium phosphate), synthetic polymers (e.g., polylactide, polyglycolide), natural polymers (e.g., collagen formulations), and composites of the above three materials [49]. Carrier efficacy is both site specific and species specific. The dosing of BMP products with their associated carriers is currently under investigation. At the present time, it is unclear if the optimal dose or the optimal delivery system has been established, and if delayed-release products can successfully compete with the theoretical advantages of gene transfer.

Conclusions

The ability of BMPs to promote, extend, or enhance spinal fusion is attracting interest in both the basic science and clinical settings. Although autograft currently remains the ‘gold standard’ for initiating spine fusion in the clinical


61

arena, osteoinductive differentiation factors combined with osteoconductive matrices are being investigated in preclinical and clinical trials. Results of these early clinical investigations indicate that rhBMP may be an acceptable, safe bone graft alternative. However, current numbers are small and follow-up is still of relatively short duration. Longer follow-up and additional studies are, therefore, needed to test acute application and long-term application of osteoinductive factors. Newer gene therapy techniques have not yet been introduced into clinical trials, but preliminary animal study results are promising. References 1 2

3

4 5 6 7 8 9

10 11 12 13 14 15

16 17

18

Morone MA, Boden SD: Experimental posterolateral lumbar spinal fusion with a demineralized bone matrix gel. Spine 1998;23:159–167. Martin GJ Jr, Boden SD, Titus L, Scarborough NL: New formulations of demineralized bone matrix as a more effective graft alternative in experimental posterolateral lumbar spine arthrodesis. Spine 1999;24:637–645. Boatright KC, Boden SD: Biologic enhancement of spinal arthrodesis: Past, present, and future; in Fardon DF, Garfin SR, North American Spine Society, Abitbol J-J (eds): Orthopaedic Knowledge Update: Spine 2, ed 2. Am Acad Orthopaedic Surgeons, 2002, pp 459–468. Urist MR: Bone formation by autoinduction. Science 1965;150:893. Urist MR, Iwata H, Ceccotti PL, Dorfman RL, Boyd SD, McDowell RM, Chien C: Bone morphogenesis in implants of insoluble bone gelatin. Proc Natl Acad Sci USA 1973;70:3511–3515. Urist MR, Mikulski A, Leitze A: A solubilized and insolubilized bone morphogenetic protein. Proc Natl Acad Sci USA 1979;76:1828. Reddi AH: Cell biology and biochemistry of endochondral bone development. Coll Relat Res 1981;1:209–226. Boden SD, Schimandle JH, Hutton WC: An experimental lumbar intertransverse process spinal fusion model: Radiographic, histologic, and biomechanical healing characteristics. Spine 1995;20:412. Helm GA, Alden TD, Beres EJ, Hudson SB, Das S, Engh JA, Pittman DD, Kerns KM, Kallmes DF: Use of bone morphogenetic protein-9 gene therapy to induce spinal arthrodesis in the rodent. J Neurosurg (Spine 2) 2000;92:191–196. Wozney JM: Overview of bone morphogenetic proteins. Spine 2002;27:S2–S8. Ebara S, Nakayama K: Mechanism for the action of bone morphogenetic proteins and regulation of their activity. Spine 2002;27:S10–S15. Urist MR, DeLange RJ, Finerman GAM: Bone cell differentiation and growth factors: Induced activity of chondro-osteogenetic DNA. Science 1983;220:680. Piccolo S, Sasai Y, Lu B, DeRobertis EM: Dorsoventral patterning in Xenopus: Inhibition of ventral signals by direct binding of chordin to BMP-4. Cell 1996;86:589–598. Zimmerman LB, DeJesus-Escobar JM, Harland RM: The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell 1996;86:599–606. Varady P, Li JZ, Cunningham M, Beres EJ, Das D, Engh J, Alden TD, Pittman DD, Kerns KM, Kallmes DF, Helm GA: Morphologic analysis of BMP-9 gene therapy-induced osteogenesis. Hum Gene Ther 2001;12:697–710. Spiro RC, Liu L, Heidaran MA, Thompson AY, Ng CK, Pohl J, Poser JW: Inductive activity of recombinant human growth and differentiation factor-5. Biochem Soc Trans 2000;28:362–368. Boden SD, Andersson GBJ, Anderson DG, Damien C, Ebara S, Helm G, Lane JM, McKay B, Sandhu HS, Seeherman H, Wozney J: Summary statement: Overview of bone morphogenetic proteins for spine fusion. Spine 2002;27:S1. Bomback DA, Grauer JN, Lugo R, Troiano N, Patel TC, Friedlaender GE: Comparison of posterolateral lumbar fusion rates of Grafton Putty and OP-1 Putty in an athymic rat model. Spine 2004;29:1612–1617.

Bomback/Grauer

62

19 20

21

22

23 24

25

26

27

28

29

30 31

32

33

34 35 36

37 38

Damien CJ, Grob D, Boden SD, Benedict JJ: Purified bovine BMP extract and collagen for spine arthrodesis. Spine 2002;27:S50–S58. Sandhu HS, Turner S, Kanim LEA, Kabo JM, Toth JM: Augmentation of threaded titanium fusion cages with rhBMP-2 for experimental anterior lumbar fusion. Presented at the 11th Annual Meeting of the North American Spine Society, October 23–26, Vancouver, Canada, 1996. Zdeblick TA, Ghanayem AJ, Rapoff AJ, Swain C, Bassett T, Cooke ME, Markel M: Cervical interbody fusion cages: An animal model with and without bone morphogenetic protein. Spine 1998; 23:758–765. Hecht BP, Fischgrund JS, Herkowitz HN, Penman L, Toth JM, Shirkhoda A: The use of recombinant human bone morphogenetic protein 2 (rhBMP-2) to promote spinal fusion in a nonhuman primate anterior interbody fusion model. Spine 1999;24:629. Schimandle JH, Boden SD, Hutton WC: Experimental spinal fusion with recombinant human bone morphogenetic protein-2. Spine 1995;20:1326. Sandhu HS, Kanim LE, Kabo JM, Toth JM, Zeegan EN, Liu D, Seeger LL, Dawson EG: Evaluation of rhBMP-2 with an OPLA carrier in a canine posterolateral spinal fusion model. Spine 1995;20:2669. Fischgrund JS, James SB, Chabot MC, Hankin R, Herkowitz HN, Wozney JM, Shirkhoda A: Augmentation of autograft using rhBMP-2 and different carrier media in the canine spinal fusion model. J Spinal Disord 1997;10:467. Helm GA, Sheehan JM, Sheehan JP, Jane JA Jr, diPierro CG, Simmons NE, Gillies GT, Kallmes DF, Sweeney TM: Utilization of type I collagen gel, demineralized bone matrix and bone morphogenetic protein-2 to enhance autologous bone lumbar spinal fusion. J Neurosurg 1997;86:93–100. Martin GJ, Boden SD, Titus L: Recombinant human bone morphogenetic protein-2 overcomes the inhibitory effect of ketorolac, a nonsteroidal anti-inflammatory drug (NSAID), on posterolateral lumber intertransverse process spine fusion. Spine 1999;24:2188–2193. Silcox DH, Boden SD, Schimandle JH, Johnson P, Whitesides TE, Hutton WC: Reversing the inhibitory effect of nicotine on spinal fusion using an osteoinductive protein extract. Spine 1998; 23:291–296. Cook SD, Dalton JE, Tan EH, Whitecloud TS III, Rueger DC: In vivo evaluation of recombinant human osteogenic protein (rhOP-1) implants as a bone graft substitute for spinal fusions. Spine 1994;19:1655. Grauer JN, Patel TC, Erulkar JS, Troiano NW, Panjabi MM, Friedlaender GE: Evaluation of OP-1 as a graft substitute for intertransverse process lumbar fusion. Spine 2001;26:127–133. Cunningham BW, Shimanoto N, Sefter JC, McAfee A: Posterolateral spinal arthrodesis using osteogenic protein-1: An in vivo time course study using a canine model. Presented at the 15th Annual Meeting of the North American Spine Society, New Orleans, LA, 2000. Magin MN, Delling G: Improved lumbar vertebral interbody fusion using rhOP-1: A comparison of autogenous bone graft, bovine hydroxylapatite (Bio-Oss), and BMP-7 (rhOP-1) in sheep. Spine 2001;26:469–478. Cunningham BW, Kanayama M, Parker LM, Weis JC, Sefter JC, Fedder IL, McAfee PC: Osteogenic protein versus autologous interbody arthrodesis in the sheep thoracic spine: A comparative endoscopic study using the Bagby and Kuslich interbody fusion device. Spine 1999;24:509. Patel TC, Erulkar JS, Grauer JN, Troiano NW, Panjabi MM, Friedlaender GE: Osteogenic protein-1 overcomes the inhibitory effect of nicotine on posterolateral lumbar fusion. Spine 2001;26:1656. Cook SD, Dalton JE, Prewett AB, Whitecloud TS III: In vivo evaluation of demineralized bone matrix as a bone graft substitute for posterior spinal fusion. Spine 1995;20:877–886. Boden SD, Schimandle JH, Hutton WC: Lumbar intertransverse-process spinal arthrodesis with use of a bovine bone-derived osteoinductive protein. A preliminary report. J Bone Joint Surg Am 1995;77:1404–1417. McKay B, Sandhu HS: Use of recombinant human bone morphogenetic protein-2 in spinal fusion applications. Spine 2002;27:S66–S85. Burkus JK, Transfeldt EE, Kitchel SH, Watkins RG, Balderston RA: Clinical and radiographic outcomes of anterior lumbar interbody fusion using recombinant human bone morphogenetic protein-2. Spine 2002;27:2396–2408.


63

39 40

41 42

43 44 45 46

47 48

49

Poynton AR, Lane JM: Safety profile for the clinical use of bone morphogenetic proteins in the spine. Spine 2002;27:S40–S48. Boden SD, Kang J, Sandhu H, Heller JG: Use of recombinant human bone morphogenetic protein-2 to achieve posterolateral lumbar spine fusion in humans: A prospective, randomized clinical pilot trial: 2002 Volvo Award in Clinical Studies. Spine 2002;27:2662–2673. Speck G: Posterolateral fusion using OP-1: A model using degenerative spondylolisthesis. Australian Spine Meeting, Adelaide, Australia, 2000. Patel TC, McCullough JA, Vaccaro AR: A pilot safety and efficacy study of OP-1 (rhBMP-7) in posterolateral lumbar fusion as a replacement for iliac crest autograft. Seattle, North Am Spine Society, 2001. Vaccaro AR, Anderson G, Toth CA: Recombinant human osteogenic protein-1 (bone morphogenetic protein-7) as an osteoinductive agent in spinal fusion. Spine 2002;27:S59–S65. Patel TC, Vacarro AR, Truumees E: A safety and efficacy study of OP-1 (rhBMP-7) as an adjunct to posterolateral lumbar fusion. Seattle, North Am Spine Society 2001. Alden TD, Varady P, Kallmes DF, Jane JA, Helm GA: Bone morphogenetic protein gene therapy. Spine 2002;27:S87–S93. Boden SD, Titus L, Hair G, Liu Y, Viggeswarapu M, Nanes MS, Baranowski C: Lumbar spine fusion by local gene therapy with cDNA encoding a novel osteoinductive protein (LMP-1). Spine 1998;23:2486–2492. Liu Y, Hair G, Titus L: BMP-6 induces a novel LIM protein involved in bone mineralization and osteocalcin secretion. J Bone Min Res 1997;12:S115. Alden TD, Pittman DD, Berer EJ, Hankins GR, Kallmes DF, Wisotsky BM, Kerns KM, Helm GA: Percutaneous spinal fusion using bone morphogenetic protein-2 gene therapy. J Neurosurg (Spine 1) 1999;90:109–114. Seeherman H, Wozney J, Li R: Bone morphogenetic delivery systems. Spine 2002;27:S16–S23.

David Bomback, MD 400E 71st Street Apt. #51 New York, NY 10021 (USA) Tel. ⫹1 212 472 2143, Fax ⫹1 212 774 2779, E-Mail [email protected]

Bomback/Grauer

64


Cellular and Gene Therapy Approaches to Spinal Cord Injury Michael P. Steinmetz, James K. Liu, Nicholas M. Boulis Department of Neurosurgery S31, Cleveland Clinic Foundation, Cleveland, Ohio, USA

Introduction

Acute spinal cord injury (SCI) is relatively uncommon, affecting about one in 40 patients who present to a major trauma center [1]. It is estimated that there are 11,000 new cases per year or 40 cases per million population (National Spinal Cord Injury Databank, 2001). Despite this relatively low incidence, these injuries pose serious problems for the patients, their families, and society in general [2]. Mortality from SCI is estimated to range from 4.4 to 16.7% for those that survive the initial injury and receive treatment [3]. Aside from the obvious physical damage, there may be serious psychological effects on both the patients and their families. The financial burden to the patient, the health care system, and society is great in terms of both direct and indirect costs (i.e., lost income and productivity) [4]. In 1990, it was estimated that the cost to the United States of caring for all SCI patients was USD 4 billion annually [5]. Clinical therapy for acute SCI is sparse and often disappointing. The clinician is limited to surgical decompression (if appropriate), IV methylprednisolone, and acute and long-term SCI rehabilitation. Various research protocols are also in progress. Despite these therapies, the prognosis for SCI remains dismal. The failure of functional recovery following SCI is multifactorial. Axonal regeneration requires neuroprotection, neuronal cell-body stimulation, the need to overcome local inhibitors at the injury site, and finally reconnection of neuronal pathways (both ascending and descending) essential for functional recovery. Current research is focused on each of these areas. Both genetic and

cellular therapies are emerging as strategies to overcome barriers to neural regeneration. This chapter will review past and current research with genetic and cellular therapeutic options.

Cellular Therapies

There are many evolving cellular therapeutic strategies for SCI. The focus of this therapy is on neuroprotection, remyelination, and regeneration. Cellular therapies include endogenous and transplanted stem cells, fetal tissue transplants, allo- and xenografted Schwann cells and olfactory ensheathing cells (OECs) as well as autologous macrophages. Each area of research has uncovered difficulties unique to individual cellular approaches. These include the ethical and moral concerns raised by embryonic stem (ES) cells and fetal grafts, as well as the potential need for immunosuppression after cellular transplants.

Models of SCI

There are many animal models of SCI available to the researcher. The most popular animals for these injury paradigms include the rat and mouse. Both are fairly inexpensive and are readily available. The mouse model has the further advantage of being used for transgenic experiments. Methods of experimental SCI entail complete spinal cord transection, partial transection, contusion, and compression [6]. Complete transection involves the complete disruption of the spinal cord. The main advantage of this model is that all tracts are transected; therefore, any axons demonstrated on retrograde labeling (see below) are due to regeneration and not from sparing (not destroyed during the experimental injury). Partial transection models utilize animals in which only certain tracts are cut (e.g., the rubrospinal tract). This leaves the contralateral tract available for comparison [6]. The absence of complete paraparesis and urinary retention renders these animals easier to care for (i.e., one does not need to manually express the bladder in those rats who have only a unilateral rubrospinal transection). A serious disadvantage of these models is the potential for confusing spared and regenerating axons. Contusion and compression models reflect more accurately the SCIs that generally occur in humans [6]. Methods available for contusion and compression models include weight-drop and clip-compression strategies. Because the lesions in these models are even less discrete, they have an even greater potential for confusion between axon sparing and true regeneration.

Steinmetz/Liu/Boulis

66

Recovery Assays

Immunohistochemical, electrophysiological, and behavioral assays can all be used to measure spinal cord recovery [6]. Axons tracers are used to follow tracts in the spinal cord (both anterograde and retrograde). Anterograde tracers include biotinylated dextran amine and the cholera toxin B subunit. These tracers are applied to the cell body and then transported anterograde and may be used to identify axons regenerating at the injury site. Retrograde tracers are taken up by axons and transported back to the cell body. These may then be placed distal to the injury site to assess regeneration at a site of injury. Examples include Flouro-Gold and the cholera toxin B subunit. Electrophysiological tests may assess axon integrity, both in vivo and in vitro. Examples of in vivo tests include somatosensory evoked and motor potentials. Isolated spinal cords (in vitro) may also be assessed neurophysiologically [6]. Finally, behavioral tests are available to evaluate the neurological recovery of the injured animal. These include open field test of locomotion, such as the BBB; sensory tests, such as sensing and removing a piece of tape from a paw; and other tests or motor and skilled behavior, such as walking across a wire grid or narrow beam.

Cellular Therapies for SCI

Stem Cells Neural stem cells (NSCs) are defined by their ability to generate neural tissue (both neuronal and glial), their ability to self-renew, and their pluripotency. Pluripotency refers to a cell’s ability to generate a variety of lineages through cell division [7]. Progenitor cells have a more restricted fate (fig. 1). For example, neural progenitors differentiate into all the neuronal cells of the central nervous system (CNS), but not the glia. NSCs may be isolated from adult and fetal brain and also embryonic tissue. Cells may be harvested from the adult subventricular zone, hippocampus [7], the fetal telencephalon, or the inner cell mass of blastocyst-stage embryos [8]. These cells are then grown in cell culture in the presence of a high concentration of mitogens such as fibroblast growth factor (FGF) or epidermal growth factor [7]. After several rounds of division, the cells are exposed to either media with the mitogens withdrawn or to a new substrate. Different substrates can drive stem cell differentiation into specific lineages (e.g., oligodendrocyte). Immunostaining for specific marker antigens can be used to identify these lineages. Cells may be infected with a replication-incompetent retroviral vector

Cellular and Gene Therapy Approaches

67

Cell type Pluripotent embryonic stem cell self-renewing

Multipotent stem cells self-renewing

Neural progenitor cells limited self-renewal

Committed progenitor cells no self-renewal Neuronal

Glial

Differentiated cells no self-renewal

Neuron

Glial

Fig. 1. The progression of cell development from a pluripotent, self-renewing, embryonic stem cell to differentiated neurons and glia.

encoding lacZ in order to assess clonal relationships of progeny and identify them in situ following transplantation [9]. NSCs The adult CNS has a limited capacity to repair itself after injury. This is due in part to the inability to generate new neurons and the inability to initiate


68

Exogenous stem cells

1.

2.

Syrinx

Syrinx

a Exogenous factor

2.

1.

Central canal

Syrinx

Potential stem cells

Central canal

Syrinx

b Fig. 2.a Exogenous stem cell transplantation. 1. Exogenous stem cells have been transplanted into a syrinx cavity following a chronic SCI. 2. Following transplantation, the stem cells have populated the cavity and migrated into the spinal cord parenchyma. b Endogenous stem cells. 1. The endogenous stem cells of the spinal cord probably reside in the region of the central canal. 2. Following injury or factor injection, these cells are stimulated to divide and migrate into the area of injury (in this case a syrinx cavity). Some cells also migrate into the cord parenchyma.

functional axonal regrowth. The ability to transplant multipotent NSCs may overcome the former. As such, significant enthusiasm has focused on the application of NSCs for the repair of focal neural tissue destruction, including that which is seen with SCI. These stem cells may be isolated from embryonic or adult brain tissue of a variety of species, including mouse, rat, and human [7, 10]. They may also be derived from the mouse and human ES cells [8, 11–13] derived from nonneural embryonic tissue. These cells are stable through multiple passages in vitro without loss of their multipotentiality [14]. Multipotentiality or ‘pluripotency’ refers to the NSC’s ability to differentiate into a variety of lineages, including neuronal, oligodendrocytic, or astrocytic phenotypes. These stem cells have been shown to survive transplantation into the CNS and also have the ability to migrate. Thus, these cells may be transplanted into the injured CNS with the potential to repair specific regions. In addition to the ability to transplant exogenous stem cells, it may be possible to induce endogenous multipotential cells to ‘self-repair’ after injury or disease [14] (fig. 2).


69

Recent evidence also suggest that NSCs may be derived from bone marrow and umbilical cord blood, thus serving as another source of both exogenous and endogenous stem cells, but residing outside the CNS [15]. Both in vivo and in vitro studies have demonstrated neuronal and glial differentiation; furthermore, these cells (i.e., bone marrow or umbilical cord blood) have the potential to be delivered systemically for the treatment of CNS pathology as opposed to direct transplantation into the CNS [15]. The use of marrow-derived stem cells may permit the clinician to harvest autologous stem cells, amplify them in vitro and then transplant them back into the patient. This may obviate the need for chronic immunosuppression and its inherent morbidity. Experimentally, autologous bone marrow-derived stem cells have been used to regenerate infarcted myocardium [16]. Further progress may provide the same opportunity for the nervous system. Exogenous NSCs Experiments have demonstrated that NSCs demonstrate significant survival, migration, and differentiation. NSCs undergo area-specific differentiation following transplantation into the CNS [9, 17, 18]. It appears that these cells have the capacity to respond appropriately to local signals in the developing CNS [14]. It may be that the local environment is the predominant determinant of the differentiated fate of the engrafted cells. When pluripotent NSCs are transplanted into the injured spinal cord, the engrafted cells differentiate only into astrocytes, and the temporal progression of that differentiation is markedly retarded [19, 20]. The mechanism regulating this transformation is unknown. Successful neuronal replacement may, therefore, require transplanting NSCs already committed to a neuronal lineage to avoid local environmental cues defining a glial lineage. Neural restricted precursors (NRPs) are an exciting alternative to multipotent NSCs. These cells are committed to a neuronal lineage at the time of isolation, and have been isolated from embryonic CNS tissue, ES cells, and multipotential NSCs [11] (fig. 1). These cells have been transplanted into the adult rat spinal cord. Neuronal maturation was observed, but was significantly retarded. It appears that additional modification of the grafts and/or the host environment will be needed for mature neuronal differentiation [14]. The intrinsic state of NSCs at the time of transplant, like the host environment, may also be important [21]. There appear to be differences between neural progenitor cells isolated from different brain regions [22, 23]. When NRPs derived from embryonic spinal cord are grafted into the subventricular zone (SVZ), the cells were observed to migrate extensively and generate mature neurons of various neurochemical and morphological phenotypes [24]. When NRPs isolated from the SVZ are engrafted back to the SVZ, they demonstrate less migratory and differentiation potential [25]. Spinal cord-derived NRPs were


70

also observed to differentiate into a neuronal phenotype expressing choline acetyltransferase in brain areas where endogenous choline acetyltransferasepositive neurons have not been found. In this instance, therefore, intrinsic characteristics of the transplanted NRPs committed them to a restricted phenotype even after ectopic engraftment [14]. This finding suggests that one may need to isolate precursors from specific CNS regions for proper functional neural replacement [14]. Many studies have demonstrated successful transplantation of NSCs and NRPs into the CNS, but the physiological function of these grafts has not been completely elucidated. McDonald et al. [26] demonstrated functional improvement (locomotor) after transplantation of NSCs derived from ES cells into a rat spinal cord in a contusion injury model. ES cell embryoid bodies were derived from the D3 cell line (mouse) [27] at the 4⫺/4⫹ stage (i.e., 4 days without, then 4 days with retinoic acid) for transplantation. These cells were transplanted as cell aggregates directly into the syrinx (fig. 2a) 9 days after the experimental SCI. Two weeks after transplantation, labeled ES cells [Bromodeoxyuridine (BrdU), or mouse-specific antibodies M2, EMA, or Thy 1.1/1.2] were identified in situ. Cells were found to be filling the syrinx cavity, but also as far as 8 mm away from the syrinx edge in either the rostral or caudal direction. 43% of these cells were found by immunohistochemistry to be oligodendrocytic and 19% were found to be astrocytic. Many of the ES-derived oligodendrocytes were immunoreactive for myelin basic protein. 8% demonstrated neuronal staining (neuron-specific nuclear protein, NeuN). One month following transplantation, a difference of 2 points on the BBB scale was observed (7.9 vs. 10) between treated animals and sham controls. The difference in the BBB score reflected the ability to mobilize with partial hind limb weight-bearing and coordination as opposed to no hind limb weight-bearing or coordination. It is unclear what factors were responsible for the improved functional score following transplantation. The ES cells may have remyelinated the injured axons or provided neurotrophic or tissue-sparing effects. Furthermore, ES-derived neurons may have matured and made functional connections with injured spinal tracts. The rapidity of locomotor improvement and the observation that most ES cell-derived cells were oligodendrocytes positive for myelin basic protein make remyelination the most probable cause of recovery. The fact that oligodendrocyte precursors transplanted into chemical lesions have previously been associated with remyelination and improved axonal conduction creates further precedence for this explanation [28]. Various goals underlie the rationale for NSCs or NRPs spinal cord transplantation. As previously discussed, it is unlikely that transplanted stem cells will be able to completely recapitulate the injured ascending and descending tracts of the spinal cord (fig. 3e). Although stem cell-derived neurons have been


71

NSC

Trophic factors

a

NSC

Trophic factors

Native glial cell

b

Ex vivo Fibroblast gene transfer

c

Trophic factors

NSC precursor

Glia

d

NSC

e Fig. 3. Various mechanisms of NSC graft-induced recovery following spinal cord injury. a NSCs may secrete various trophic factors that induce regrowth of spinal cord axons. b NSCs may induce native CNS support cells (e.g., glial cells) to secrete trophic factors that lead to axonal regrowth. c Fibroblasts implanted following ex vivo gene transfer may secrete trophic factors, leading to axonal regrowth. d NSCs that have been stimulated to become oligodendrocyte or glial precursors are transplanted into the spinal cord. Following further differentiation, remyelination of damaged axons is initiated. e A common misconception of NSC transplantation is that NSCs completely recapitulate a new axon, replacing the damaged axon. It is highly unlikely that this occurs following NSC transplantation.

shown to possess ion channels similar to native neurons and are capable of generating action potentials, no study has clearly shown that these cells integrate into host circuitry and create functional synapses. It is more likely that these grafts provide neuroprotection (tissue sparing), trophic support, or remyelination, although direct evidence is speculative (fig. 3a). Furthermore, stem cell grafts may provide effective tissue bridges that permit or promote the passage of endogenous regenerating axons. Endogenous Stem Cells The use of exogenous stem cells necessitates a grafting procedure and possibly immunosuppression. The use and manipulation of endogenous stem cells may obviate these cumbersome and potentially hazardous interventions. It is now widely accepted that neurogenesis occurs in the adult CNS. This process has been demonstrated in the hippocampus and the SVZ [29–32]. Lois and Alvarez-Buylla [31] labeled the brains of adult male mice with [3H]thymidine. Proliferating, hence, dividing cells were localized almost exclusively to the SVZ. To test the fate of these cells, the SVZ of labeled brains were isolated and grown


72

in culture. By six cell divisions, the explants had generated an outgrowth of flat glial cells (GFAP positive) and cells containing processes growing on the glial monolayer. These cells were determined to be neurons due to their immunoreactivity to neuronal markers (MAP-2, NF, and NSE) and absence of staining for glial markers (GFAP). Explants stained with neuron-specific antibodies were processed for autoradiography to detect the presence of [3H]thymidine in the cell nuclei. It was found that approximately 84% of neurons were labeled with [3H] thymidine. These results indicate that proliferating cells do exist in the adult SVZ in vivo and these cells possess the ability to generate neurons and glia. Eriksson et al. [29] demonstrated similar results in the adult human dentate gyrus. They examined the hippocampus and SVZ of patients who had succumbed to cancer and had received an intravenous infusion of BrdU before they died. The results of these experiments demonstrated that in all BrdU-treated patients, the granule cell layer contained BrdU-positive cells, which also doublestained with neuron-specific markers (e.g., NeuN). Furthermore, the SVZ also contained BrdU-positive cells, but these did not colabel with neuron-specific markers. It is believed that these too are progenitor cells, but they must first migrate from the SVZ before they differentiate. It is believed that the cellular substrate for this neurogenesis is the endogenous stem cell [33]. The mechanisms that induce and control this process are unknown. It may be that one can manipulate these endogenous cells to replace damaged neural tissue following injury. Indeed some important observations have been made. Johansson et al. [34] demonstrated that multipotential NSCs migrate to the area of injury after dorsal funiculus sectioning. However, similar to cellular transplants, most differentiated into astrocytes. After prolonged administration of BdrU to the spinal cord of rats, a substantial number of ependymal cells lining the central canal were labeled and few were seen outside the spinal cord central canal ependyma (due to the lack of an SVZ in the spinal cord). After an incision (one day following injury) was made in the dorsal funiculus at T2, there was a 50-fold increase in the proliferating ependymal cells. Electron microscopy demonstrated the cell division to be asymmetric. To demonstrate the fate of these ependymal cells, lesions were made in animals that had received an injection with a fluorescent marker called Dil that labels the ependymal cells prior to the lesion. Dil-labeled cells were abundantly seen in the injury site within one week after the lesion. These cells demonstrated immunoreactivity to GFAP, but neither ␤-tubulin nor O4, confirming them to be astrocytes, and not neurons or oligodendrocytes. Therefore, local cues will have to be overcome so that neurons are formed rather than glia, which may exacerbate scarring. It is also postulated that since there is little neurogenesis occurring in the mature spinal cord, the number of endogenous stem cells may be inappropriate for the replacement of tissue following injury.


73

Remyelination of the Spinal Cord Glial precursors have been found throughout the developing and adult CNS. A large proportion of these are oligodendrocyte precursors. When adult rats are pulse-labeled with BrdU, 70–75% of BrdU-labeled cells found in the spinal cord and cortex express NG2, a marker for oligodendrocyte precursors [35, 36]. The exact function of these cells is unknown, but one may speculate that these cells remyelinate the CNS following injury. It has been demonstrated that the numbers of these oligodendrocyte precursors are markedly increased following SCI and demyelination [37–39]. It is hypothesized that these glial precursor cells may be transplanted into the injured spinal cord to remyelinate axons and promote functional recovery. As previously mentioned, when multipotential NSCs are grafted into the injured CNS, only a fraction differentiates into oligodendrocytes. Therefore, stimulating oligodendrocyte lineage commitment prior to transplantation may be necessary to achieve effective remyelination. Grafts of this type are called oligospheres. When oligospheres are transplanted into the myelin-deficient spinal cord, significantly larger areas of myelination were demonstrated compared to neurosphere transplantation [13, 40]. It appears that astrocytes also play an important role during myelination. Therefore, glial restricted precursors may prove more effective for remyelination than oligospheres due to their ability to differentiate into both astrocytes and oligodendrocytes after transplantation [41]. The efficacy of these glial restricted precursors for functional myelination is still in question. Although the use of endogenous and transplanted stem cells has demonstrated some remyelination of axons, no study has clearly demonstrated ‘functional’ myelination after this cellular therapy. As with other NSC grafting strategies, the therapeutic mechanism of oligospheres remains unknown. As with the other grafts, neuroprotective or trophic mechanisms may contribute (fig. 3b). In vitro studies have demonstrated partial electrophysiological recovery of remyelinated axons [42]. However, there has not been electrophysiological evidence of recovery in the live animal [14]. Although few studies have demonstrated functional recovery following stem cell therapy, the field is advancing rapidly. The ability to replace lost neural elements (i.e., neurons and glia) is paramount to neural regeneration following injury. Non-Stem Cell Strategies Fetal Tissue After SCI, a gap often exists at the site of injury. There is often a cyst or syrinx cavity in the spinal cord. Therefore, a bridge may be necessary to permit adequate neural regeneration for both spinal and supraspinal projections. Fetal


74

tissue transplants have been demonstrated to provide a permissive condition for axonal regrowth and provide such a bridge. After complete spinal cord transection in the newborn rat or kitten, fetal spinal cord tissue transplanted into the site of injury has allowed some restoration of supraspinal projections and improvement in the locomotor function [43, 44]. In a transection model of adult SCI, transplantation of fetal tissue at the lesion permitted axonal regeneration into the graft, but not beyond the graft-host interface [45–48]. The failure to achieve significant numbers of graft-spanning axons has remained an obstacle to most studies involving tissue grafts for the treatment of SCI. Some have demonstrated that the exogenous administration of brain-derived neurotrophic factor (BDNF) or neurotrophin-3 (NT-3) increases supraspinal axonal growth into the transplant fetal tissue grafts and prevents the atrophy of axotomized supraspinal neurons [49]. It is also problematic that most studies utilizing tissue bridges employ acute models of SCI. These models may not be clinically relevant. Often a syrinx cavity does not develop until the subacute or chronic phase of SCI. Coumans et al. [50] demonstrated that if a fetal tissue transplant and neurotrophin administration is delayed 2–4 weeks after a complete SCI in the rat, axonal regrowth from both propriospinal and supraspinal neurons is increased within the transplant and the host cord caudal to the lesion. These animals also demonstrated significant improvement in locomotion, including recovery of weight-supported plantar stepping on both treadmill and over-ground tasks such as stair climbing. In summary, although fetal tissue transplants have shown some success as tissue bridges, experiments remain hampered with the distal host-graft interface. Some studies have demonstrated axonal presence across the host-graft interface. The numbers of these axons is sparse and may actually represent ‘axonal sparing’ and not regeneration. OECs Axonal regrowth into a site of injury following cellular grafting is plagued by the inability of those axons, which have entered the graft, to cross the hostgraft interface. Therefore, a cell, which may enable axons to re-enter the CNS, may be useful to overcome this barrier to regeneration. Olfactory axons continue to re-enter the olfactory bulb throughout adult life. The entry point is associated with special glial cells known as OECs [51–54]. Investigators have demonstrated that OECs transplanted into the spinal cord mediate the re-entry of regenerating dorsal root axons into the spinal dorsal horn and the injections also increased axon growth into Schwann cell-filled guidance channels [55, 56]. As opposed to other cellular grafts, transplantation of OECs facilitated axonal growth past the host-graft interface. This may be due to the migratory capacity of these cells. In a study by Li and Raisman [57], regenerating axons were


75

demonstrated to re-enter the distal host corticospinal tract up to 10 mm caudal to the injection site. These regenerating axons are covered by peripheral myelin formed by the OEC cell. Schwann Cells Because the environment of the peripheral nervous system (PNS) is permissive for regeneration, Schwann cell transplants have also been used as a strategy for the treatment of experimental SCI. These cells may be neuroprotective and have been demonstrated to secrete various growth factors. Theoretically, these cells are also able to form myelin around spared and regenerating axons. As with other cellular grafts, investigators have found that following the transplantation of cultured Schwann cells, the cells integrate into the host tract glial structure [57, 58]. These cellular grafts greatly increase axon sprouting in lesions of the corticospinal tract, but few axons were found to re-enter the distal tract [59]. In contrast, other investigators have demonstrated axonal growth beyond the graft. Schwann cells transplanted after a moderate contusion of the rat thoracic spinal cord permitted propriospinal and supraspinal axons reaching 5–6 mm beyond the graft. A modest improvement in hind limb locomotor performance was detected at 8–11 weeks after injury [60]. Nonetheless, the limited growth beyond the graft, even in this experiment, suggests that recovery was likely to be due to neuroprotection or remyelination of spared axons rather than axonal regeneration. Schwann cells have also been seeded into mini-channels that have been used as bridges. When this transplantation technique is combined with exogenous neurotrophin administration, axonal growth was demonstrated into the graft and into the distal spinal cord, albeit for a limited distance [61]. In summary, these studies demonstrate that both OECs and Schwann cell transplants may be useful to induce axonal regeneration and remyelination after SCI. Although some studies have demonstrated some functional improvement following Schwann cell or OEC transplants, it is unclear if the improvement is due to neuroprotection, trophic factor secretion, or remyelination (see above). The myelination that has been observed has not been demonstrated to be functional nor has it been quantified. While some studies have demonstrated axonal growth into the distal spinal cord, the amount of regrowth is not highly significant, and is unlikely to account for a significant amount of functional improvement following SCI. Macrophages Macrophage recruitment and stimulation are among the earliest events in the multifactorial process of tissue healing. This observation has led to the


76

hypothesis that stimulating an appropriate inflammatory response could encourage a cascade of events necessary for regeneration and repair [62]. Because of the immune-privileged status of the CNS, a restricted inflammatory response is seen following injury. This restriction may contribute to the poor regenerative capacity of the CNS compared to the PNS [63–65]. Macrophages, previously exposed in vitro to regenerating peripheral nerve segments, have been shown to induce axonal regrowth in completely transected rat optic nerves [64]. This observation drew attention to the potential for stimulated macrophages to play a role in spinal cord repair. Macrophages stimulated via exposure to peripheral nerve segments in vitro and then re-implanted at the site of a complete SCI in the rat induce partial recovery of motor function [62]. The fact that this function was lost after retransection proved that the recovery was not due to intrinsic spinal cord reflex pathways. Anterograde labeling demonstrated continuity of nerve fibers across the transection site. The authors of this study hypothesize that activated macrophages may provide cytokines, growth factors, and other wound-healing factors [41, 66, 67]. These factors may control the astrocytic response seen after injury, thus reducing the glial scar known to inhibit axonal regeneration [68]. Stimulated macrophages may accelerate processes that normally occur relatively slowly in the injured CNS [62]. Because the activated macrophage strategy is aimed at upstream processes in the injury cascade, this one intervention may then affect numerous downstream events. Given the complexity of SCI pathophysiology, multifactorial therapies of this kind may ultimately prove the most effective. Furthermore, because autologous cells are utilized, many of the ethical and immunological difficulties inherent in other cellular therapies are absent.

Molecular Therapies for SCI

Concepts of Gene Transfer A variety of investigators have pursued gene transfer as a means of inducing neuroprotection and axonal regeneration in the injured spinal cord. As with cellular therapies, a variety of potential strategies exist for spinal cord gene transfer. Transfer can be affected with viral and nonviral vectors. In vivo strategies, which entail gene transfer directly into injured cord parenchyma, and ex vivo strategies, which entail gene transfer into cells that are subsequently grafted, have both been proposed. The best method for gene delivery remains debated. An effective method must accomplish four basic steps. Gene delivery is initiated when a vector binds to the host cell. The cell membrane constitutes the first barrier to gene delivery.


77

Efforts are ongoing in our laboratory to create neurotrophic vectors that will target specific cell types. Once past the cell membrane, a vector must provide protection from degradation in the cytoplasm. If the vector enters through endocytosis, lysosomal fusion may result in enzymatic degradation of the transgene. An effective delivery method must either avoid entry into the cell via endocytosis thus preventing lysosomal degradation, or allow entry into the cell via an alternative pathway. The next barrier to gene delivery is the nuclear membrane. In order for most transgenes to be expressed, they must enter the nucleus. The nuclear membrane serves as a relatively effective barrier against foreign entry into the nucleus [69]. Early in the development of gene therapy strategies, entry into the nucleus was limited to mitotically active host cells, and thus naturally occurring breakdown of the nuclear membrane [70–72]. Recent advances have involved the use of nuclear localization signals and viral vectors to overcome this barrier. The final step of gene delivery is expression of the transgene. Because recovery from SCI will require significant amounts of time, gene-based approaches to SCI require long-term gene expression. The duration of expression can vary depending on the vector being used, from transient to extended expression. Long-term expression is usually accomplished through integration of the transgene into the host DNA. Integration may be accomplished through a variety of methods depending on the vector being used. While an effective gene therapy delivery system is able to accomplish these four basic steps, an ideal vector should not be a source of pathogenicity to the host cell. Thus an ideal vector must be nontoxic and elicit little, if any, immune response in the host. Vectors for gene therapy can be divided into two main groups: viral and nonviral gene therapy. Here we will discuss the advances that have been made in each category of gene therapy delivery. Non-Viral Gene Therapy Nonviral gene therapy poses several advantages over viral gene therapy. The main advantage is the lack of pathogenicity of nonviral vectors. The simplest method to provide a transgene for a host is the delivery of naked DNA. In 1980, Capecchi [69] was able to successfully microinject DNA via glass micropipettes directly into the nuclei of host cells in vitro, although DNA expression could not be detected when DNA was injected into the cytoplasm. These results emphasize the importance of overcoming the nuclear membrane as a barrier to transgene delivery. In 1990, Wolff et al. [73] showed that both DNA and RNA transgenes could be effectively expressed when injected into mouse skeletal muscle. However, injection into other major organs, such as the liver, spleen and brain, resulted in relatively inefficient transgene


78

expression. In order to increase the efficiency of gene delivery while retaining the simplicity of naked DNA delivery, researchers have developed methods to augment transgene uptake into the host cell. One such method is utilizing electropermeabilization to enhance the uptake of plasmid DNA. Electrically mediated gene transfer has proven effective for gene delivery to murine melanoma cells [74]. Lin et al. [75] designed an intrathecal electroporation probe to be used following intrathecal injection of plasmid DNA. This device greatly enhanced transgene uptake into the spinal cord. Unfortunately, expression of the transgene was transient, greatly diminishing after 2 weeks. To further increase the uptake of DNA in the host cell, researchers have combined plasmid DNA with nonviral carriers. One method utilized by Yang et al. [76] involved coating the transgene onto fine gold particles. In vivo particle bombardment was shown to be effective in a variety of major organs in both rats and mice. Another method of delivery involved combining plasmid DNA with cationic lipid-forming lipoplexes [77]. Variant forms of lipoplexes can improve the efficiency of transfer. The addition of cationic polymers, such as poly-L-lysine or protamine, to the DNA/liposome complex can greatly enhance transgene delivery through a number of methods. The polymers enhance lipoplex endocytosis, provide protection from nuclease activity and enhance transgene entry into the host cell nucleus [78]. Another common addition to lipoplexes is dioleoylphosphatidylethanolamine. Dioleoylphosphatidylethanolamine is a neutral lipid that is capable of destabilizing the lysosomal membrane, permitting the release of the plasmid into the host cell cytoplasm, thus reducing lysosomal degradation of the transgene [79, 80]. An alternative to combining plasmid DNA with cationic lipids is the use of cationic polymers. Cationic polymers have been found to be far more effective than their lipid counterparts at condensing DNA. One such polymer being used is polyethylenimine. Polyethylenimine also acts as a proton sponge, which causes osmotic disruption of the lysosome, rescuing the transgene from enzymatic degradation. Protection of the delivered gene allows for the greater transfection efficiency, which has allowed for polyethylenimine to become one of the most efficient synthetic delivery systems available [81]. Transposons have also emerged as an effective method of delivering plasmid DNA into the host cell. Transposons are naturally occurring elements capable of integrating foreign plasmids into the host cell DNA with the help of two enzymes, integrase and transposase. Transposons have proven to be an extremely effective transgene delivery system for their ability to integrate into the host genome and allow for long-term expression. Transposons can also specifically direct the site of transgene integration [82].


79

Viral Vectors Given the limitations present in using nonviral vectors, viral vectors have emerged as the most efficient form of gene therapy delivery. Viral vectors present a variety of advantages over nonviral vectors, including increased efficiency of transfection as well as extended transgene expression. The drawbacks of using viral vectors include the immune response induced as well as constraints on transgene size. Nonetheless, the overall transfection efficiency of viral over nonviral vectors has led researchers to develop novel viral vectors that minimize their limitations while maintaining their effectiveness. Here, we will discuss the four main types of viral vectors being used for gene therapy delivery. Retroviral Vectors Researchers have used retroviral vectors for the purpose of gene delivery for a relatively long time. Retroviruses are advantageous for gene delivery because they allow integration of the transgene into the host genome. This allows for the expression of the transgene for the life of the host cell. Early retroviral studies were successful in using oncoretroviruses such as the Moloney murine leukemia virus for in vivo transfection [83, 84]. The problem posed by these early retroviral vectors was the inability to infect nondividing cells; a problem also faced when using nonviral vector delivery [85, 86]. Thus, focus has turned to lentiviruses, a form of retroviruses that are able to infect nondividing cells. Retroviruses replicate with the help of a preintegration complex that replicates viral RNA through a DNA intermediate, which allows for integration into the host genome. The preintegration complex in oncoretroviruses is believed to be excluded by the nuclear membrane while the matrix protein in lentiviruses contains a nuclear-targeting component which allows for transport of the transgene into the host nucleus, explaining the ability of lentiviruses to infect nondividing cells [87–89]. Human immunodeficiency virus type 1 (HIV-1) is a well-known member of the lentivirus family, which was found to be able to infect nondividing macrophages [90, 91]. HIV-based lentivirus vectors are capable of transfecting liver and muscle tissue, sustaining expression of the transgene for over 6 months [92]. In order to increase the range of transfection, the membranes of lentivirus vectors were modified to contain envelop proteins from different viruses. One common virus whose envelope was used for this purpose was the vesicular stomatitis virus G (VSV-G) [93]. An obvious concern in the use of HIV-1 as vehicle for gene therapy delivery is the inadvertent infection of the host. This has led to the development of attenuated forms of the virus. Attenuation of lentiviruses can be accomplished by eliminating accessory viral genes without hindering the transfection efficiency of the vector. HIV requires several basic genes for function. In addition to structural genes gag, pol, and env, HIV-1 genome contains two regulatory genes, tat and


80

rev, and four accessory genes. nef, vif, vpr, vpu [94]. Researchers have attempted to create attenuated HIV vectors that can eliminate as many regulatory and accessory genes as possible to allow for minimal pathogenicity while maintaining transfection efficiency. Zufferey et al. [95] first introduced the attenuated HIV vector by deleting accessory genes vif, vpr, vpu, and nef as well as the structural gene env. They showed that this first generation-attenuated vector was able to retain its transfection ability in nondividing cells. Further studies by Kim et al. [96] demonstrated that the tat gene is not necessary for HIV-1 transfection in nondividing cells in vitro. Tat is a strong transcriptional activator of HIV-1, but Kim showed that this function might be replaced by inserting a human cytomegalovirus promoter into the HIV genome. A third generation lentivirus vector was created which, containing only the gag, pol, and env genes, was shown to be successful in transfecting neurons in vivo [97]. In addition to these attenuated viruses, Zufferey et al. [98] also developed self-inactivating lentiviruses through deletions in the 3⬘ long terminal repeats of the HIV genome. Using self-inactivating viruses decreases the possibility of recombination with wild-type virus, further rendering the vector safe for gene delivery. An alternative to attenuation is the use of nonprimate lentiviruses such as feline immunodeficiency virus. Feline immunodeficiency viruses are unable to infect human cells, but when pseudotyped with VSV-G, transfection of nondividing human cells is possible [99]. Herpes simplex Virus Vectors Herpes simplex virus 1 (HSV-1) is a member of the human herpes viruses. HSV-1 became an attractive vector for the delivery of therapeutic transgenes for several reasons. Like lentivirus, HSV-1 is able to infect nondividing cells. The HSV-1 genome is composed of 152-kb double-stranded DNA, which allows the insertion of large transgenes and general ease of genetic manipulation. Lastly, HSV-1’s most distinguishing characteristic as a viral vector is its ability to establish latent infection in neurons [100]. Like lentivirus, HSV-1 must be attenuated to prevent viral replication in the host. One method to accomplish this was the creation of defective viral vectors. One example is amplicons, which contain the transgene to be delivered flanked by viral recognition sequences. The absence of any genes encoding viral proteins reduces the potential for an inflammatory response to these proteins and prevents replication. However, in order for the transgene to be packaged in an HSV-1 coat, the viral genes for replication and packaging must be provided in trans through a helper virus. Thus HSV-1 vectors may be produced by transfecting cells with amplicon and either cotransfecting with helper virus DNA or superinfecting with HSV [101]. Geller and Breakefield [102] demonstrated the use of such a defective HSV-1 vector to deliver the Escherichia coli


81

lacZ gene to neurons in vitro. Originally, Geller used an HSV-1 temperaturesensitive mutant, ts K, as the helper virus in HSV-1 vector production to avoid cell damage. However, because ts K was later found to revert to wild type, an HSV-1 deletion mutant which also effectively packages the amplicon was substituted [103]. An alternative to amplicons is the creation of recombinant viruses. The HSV-1 genome contains three main classes of genes; immediate early genes (IE), early (E) and late (L) genes. Mutational analysis revealed that most of these genes were nonessential for viral replication in cell cultures. The development of HSV-1 deletion mutants has allowed the effective delivery of reporter genes into postmitotic cells [104]. Research is continuing to define deletions capable of further reducing the potential for wild-type reversion of HSV recombinants. Adenovirus Adenovirus holds several advantages over its viral counterparts as a transgene delivery vehicle. Adenovirus is comprised of a 36-kb double-stranded DNA genome, allowing for a large area of transgene manipulation [105]. Several generations of attenuated adenovirus have been created in an attempt to decrease viral toxicity while maintaining efficiency of infectivity. First generation adenovirus was created by deleting the E1 gene, which is necessary for viral gene expression and replication. These viruses were used to successfully deliver the cystic fibrosis transmembrane conductance regulator gene into the lungs of nonhuman primates [106]. The attenuated virus was able to induce transgene expression, though only for a limited time. Transient gene expression is one of the major obstacles to the application of adenovirus. The lack of extended expression may be secondary to the immune response initiated by the low level expression of the remaining viral genes. Such an immune response may precipitate the destruction of transfected cells eliminating transgene expression [107–109]. Immunosuppression in parallel with adenoviral administration prolongs transgene expression [110]. An alternative approach was to completely eliminate viral gene expression. Thus, subsequent generations of adenoviruses were developed each with a larger deletion of viral genes. Second and third generation viruses include deletions in the E1, E2a, as well as the E4 genes. These further-attenuated forms of adenovirus caused less inflammation and allow for longer transgene expression [111–116]. The most advanced generation of adenoviral vectors involve removal of all viral genes. These ‘gutless’ vectors lack all viral genes with the exception of the inverted terminal repeats and packaging sequences required for inclusion into the vector. The lack of viral genes allows insertion of transgenes up to 28 kb in size. In order for this vector to be packaged, a helper virus is needed to provide viral genes in trans. This helper virus lacks the inverted terminal repeats and


82

packaging sequences, thus inhibiting it from being packaged into the virus particle [117, 118]. These gutless vectors have allowed for high levels of gene expression with little immune response. Adeno-Associated Virus Adeno-associated virus (AAV) is a 4.7-kb single-stranded DNA virus. Unlike the other systems discussed, AAV is not infectious to humans in its wildtype form. In fact, wild-type AAV is naturally attenuated requiring a helper virus to provide the necessary genes before initiating replication [119]. The AAV vector itself contains no viral genes with the exception of the 125-bp AAV terminal repeats flanking the transgene in question. In order for the vector to be packaged into the AAV vector, viral genes rep and cap are provided in trans through the use of a helper plasmid. This helper plasmid, much like the helper plasmid used in gutted adenoviral production, contains the necessary viral genes for replication and encapsidation, but lacks the terminal repeats necessary to package the helper plasmid itself into the viral vector. AAV is capable of gene delivery to terminally differentiated cells, with minimal inflammatory response and resulting long-term gene expression. More recent studies have been able to further reduce the potential for toxicity of AAV by applying the use of a truncated adenoviral genome rather than a virus to supply the necessary helper genes. This eliminates the risk of contamination of viral preparations by infectious helper virus [120]. Another advantage of AAV is the ability to integrate into the host genome. AAV has been shown to be capable of specific integration into chromosome 19. This capacity may be responsible for the vector’s prolonged transgene expression [121]. Integration along with the lack of a significant immune response has allowed for AAV-delivered transgene expression for up to 18 months [122]. Kaplitt et al. [123] were the first to demonstrate the use of AAV for delivery of transgenes into postmitotic cells in vivo. Much recent work in viral gene therapy has focused on the use of AAV due to its nonpathogenic properties. One of the major drawbacks of AAV is the limited genome size. The AAV genome is not able to accommodate transgenes greater than 4.7 kb in size [124]. Recent work by Sun et al. [125] has provided a strategy to overcome this handicap by utilizing the ability of AAV to heterodimerize. A large single transgene is split and packaged into two separate AAV vectors and coinfected into the host cell. Once inside the cell, heterodimerization occurs which allows for expression of the original transgene. Therapeutic Transgenes The application of gene therapy to SCI depends on the existence of genes capable of stimulating neuroprotection, remyelination, or regeneration. Because


83

neurotrophic factors possess these properties, they have attracted the most enthusiasm for application to SCI. A great deal of focus has surrounded members of the ‘classic’ neurotrophin family, including nerve growth factor (NGF), BDNF, and NT-3. Different neurotrophic factors show varying degrees of effectiveness in promoting regeneration of the spinal cord. The difference in effectiveness of these neurotrophic factors can be accounted for by the presence or absence of receptors for these factors on varying types of neurons. For example, in the adult rat lumbar dorsal root ganglia (DRG), receptors for NGF, known as TrkA, are found predominantly in small, unmyelinated neurons, which enter into the dorsal horn from the DRG. Oudega and Hagg [126] have shown that continuous infusion of NGF into the spinal cord following peripheral nerve transection and insertion of a peripheral nerve graft promoted re-entry of sensory axons into the denervated dorsal columns. Further studies by Ramer et al. [127] confirmed NGF’s ability to promote sensory axon growth into the spinal cord. The axons responding to NGF are positive for calcitonin gene-related peptide, a marker for small, umyelinated peptidergic axons. In contrast, TrkB and TrkC receptors, which bind BDNF and NT-3 respectively, are found mainly in DRG neurons possessing thick, myelinated axons [128], which form most of the ascending fibers of the dorsal columns. The scarcity of TrkB receptors within the dorsal horn may explain the failure of spinal BDNF administration to promoted sensory axon growth into the spinal cord [127]. In contrast, BDNF has been shown to promote motor axon growth, illustrating that different neurotrophic factors exert their effects on different neuronal subtypes [129]. Unlike BDNF, NT-3 has been shown to be able to promote sensory axon growth into the spinal cord. Oudega and Hagg [130] showed that continuous infusion of NT-3 into the spinal cord promotes regeneration of dorsal column sensory axons into the spinal cord. In a separate experiment, NT-3 was infused into the spinal cord at the site of a crush lesion. NT-3 was shown to stimulate axonal regrowth in the region of the lesion and distally, without the use of a peripheral nerve graft. NT-3 is also the only neurotrophic factor capable of promoting the growth of corticospinal axons. Axonal sprouting has been observed following a single injection [131]. Glia cell line-derived neurotrophic factor (GDNF) is a neurotrophic factor, which belongs to the cytokine rather than the neurotrophin family. Once thought to be specific for the protection of dopaminergic neurons, GDNF has been proven effective in protecting from motor neuron death following axotomy [132]. Although GDNF was not shown to have an effect on the regrowth of lesioned dorsal column axons, it is more effective than NGF on stimulating axonal growth into the spinal cord [127, 133]. In contrast to the growth-inducing abilities of neurotrophic factors, anti-apoptotic proteins have been utilized in an effort to protect neuronal


84

degeneration following injury. The Bcl-2 family of proto-oncogenes is a group of apoptosis-regulating proteins. Two proteins from this family, Bcl-2 and Bcl-XL, play an anti-apoptotic role and are believed to exert their actions by preventing the release of cytochrome C from the mitochondria, an important step in the apoptosis pathway [134]. Overexpression of Bcl-2 into embryonic sensory neurons was able to prevent death following deprivation of certain growth factors, while Bcl-XL has been shown to prevent death in primary neuronal cultures when delivered via adenoviral gene transfer [135, 136]. In addition, Bcl-xL has also been shown in vivo to prevent apoptotic death of cholinergic neurons following axotomy [137]. Thus anti-apoptotic proteins have become a reasonable addition to the library of therapeutic transgenes capable of protecting neuron survival. Viral Gene Delivery to the Spinal Cord While direct spinal cord injection of viral vectors carrying therapeutic transgenes is the most efficient method of introducing transgenes into the spinal cord, several complications have arisen from this method of injection. Liu et al. [138] attempted to inject recombinant first generation adenovirus expressing the lacZ gene under control of the cytomegalovirus promoter directly into the T7–8 levels of spinal cord in adult Sprague-Dawley rats. Transgene expression was effective after one week, but quickly diminished thereafter, almost completely disappearing by 2 months postinjection. This down-regulation is most likely due to the intense immune response elicited with the injection of adenoviral vectors. Our laboratory observed a cellular infiltrate in spinal cords 7 days after adenoviral injection (fig. 4). Immunohistochemistry suggested that this response was predominantly gliotic although a variety of mononuclear cells were also observed. In animals immunsuppressed with cyclosporin A, ␤-gal transgene staining remained robust up to 2 months postinjection. The immune response can be partly attributed to a specific reaction in response to the early generation adenovirus and partly attributed to the nonspecific immune reaction in response to the trauma induced by spinal cord injection. The former can be partly solved using the now available ‘gutless’ adenoviral vectors. Because these vectors lack viral genes, the host cannot present the viral gene products on major histocompatibility complexes. However, an immune response can still be mounted against the viral capsid, which is itself immunogenic. In addition, the application of gutless vectors does not eliminate the problem of the trauma of direct injection. The presence of a significant immune response to direct spinal cord injection and the potential trauma of this approach has spurred the search for alternatives. Direct intraparenchymal spinal cord and brain injection of early generation adenoviral vectors results in a mononuclear inflammatory infiltrate,


85

a

b

c

Fig. 4. Photomicrographs demonstrating local inflammatory response following adenoviral injection into the T-8 spinal cord. a At 7 days post-PBS (vehicle) injection, no infiltrate surrounding the cannula tract (arrow) is seen. b At 7 days post-lacZ adenovirus injection, inflammatory infiltrate is revealed (arrow). c At 7 days post-NGF adenovirus injection, a mild inflammatory infiltrate (arrow) is found surrounding the cannula tract.

eliminating transgene expression [139, 140]. Researchers, including us, have thus turned to peripheral injections as an alternative. In theory, this approach should remove the viral capsid from the spinal cord, hence eliminating the potential for damage through an immune response to the viral coat proteins. This approach also avoids direct trauma to the spinal cord. Earlier studies have shown that replication-defective adenoviruses have the ability to undergo retrograde transport following injection into the CNS [141, 142]. Kuo et al. [142] used adenoviral vectors containing the lacZ gene under the control of the Rous sarcoma virus promoter for retrograde axonal tracing studies. Kuo demonstrated staining at the site of injection as well as several sites distal to the injection. These studies led our laboratory to evaluate the retrograde transport of vectors injected into the PNS and its projection areas. Attenuated adenoviral vectors with deletions in the E1a, E1b, and E3 viral genes expressing lacZ under the Rous sarcoma virus promoter were injected into the sciatic nerve, foot pad, and anterior tibialis muscle of adult rats. Histological examination of the spinal cord revealed ␤-gal staining (transgene expression) occurring predominantly in neuronal cells with large cell bodies (fig. 5). Staining was also present in the DRG. This phenomenon, which we call ‘remote delivery’, was significantly greater in spinal cords following injection into the sciatic nerve in comparison with foot pad and intramuscular injection [140]. These studies demonstrated that vectors based on viruses that had not previously been considered neurotrophic were capable of penetrating the CNS from peripheral inoculation sites.


86

Fig. 5. ␤-Galactosidase staining of motor neuron cell bodies from rat spinal cord. Animals were injected into the sciatic nerve with adenovirus carrying a lacZ reporter gene 7 days prior to histology.

In order to confirm that retrograde axonal transport is responsible for the remote delivery phenomenon, we studied the effects of intraneural colchicine on adenoviral vector transport. Colchicine inhibits tubulin polymerization and hence disrupts axonal transport. Intrasciatic colchicine injection inhibits remote adenoviral and AAV gene delivery following sciatic injection in a dose-dependant fashion implicating retrograde axonal transport in this process [143]. Spinal cord gene expression following peripheral injections did not trigger the inflammatory response observed following direct injections. Because the termination of gene expression is linked to the inflammatory response, this discovery led to the hope that remote delivery might prolong transgene expression. Nonetheless, spinal cord transgene expression following remote delivery followed the same time course as direct injection deteriorating within 3 weeks of injection [140]. Both chronic dexamethasone and cyclosporine treatment stimulated higher levels of gene expression in the lumbar spinal cord and DRG and prolonged gene expression following remote adenoviral injection [144]. Nonetheless, no sign of cell death could be detected in parallel with the disappearance of transgene expression. Together, these findings suggested that an inflammatory response to the vector at the site of injection was shutting off


87

gene expression through a promoter level noncytolytic mechanism. Because the neurons of interest appeared to remain healthy after the disappearance of transgene expression, we conducted repeated sciatic nerve injection of adenoviral vectors in the hope of boosting gene expression. While repeated injection resulted in prolonged gene expression at the site of injection, spinal cord gene delivery was not boosted. Following initial injection, gene expression in the nerve was found predominantly in the perineurium without a significant inflammatory response at early timepoints. In contrast, gene expression at the repeat-injection sites occurs within phagocytic infiltrative cells noted shortly after injection. This immune response is likely to reduce the available titer of vector for remote delivery and may inhibit axonal uptake. Glatzel et al. [145] were also able to use adenovirus for the delivery of transgenes into the spinal cord. Though they were unable to show spinal cord expression of the transgene following intramuscular injection, they were able to demonstrate staining following adenoviral sciatic nerve injection. In their experiment, injection directly into the DRG resulted in a much higher efficiency of gene transfer compared to sciatic nerve injections, without an increase in neurological side effects. In addition, they evaluated the role that the immune response had on eliminating gene expression. Rag-1⫺/⫺ mice, which lack differentiated B and T cells, were transfected with adenoviral vectors containing the lacZ transgene. ␤-Gal expression could be seen for up to 102 days without any signs of deterioration, which further confirms the role of the immune response as the rate-limiting step in the duration of transgene expression. Because AAV vectors induce prolonged gene expression with a minimal inflammatory response, attention has turned to their application to spinal cord gene transfer. AAV vectors were demonstrated to successfully allow for transgene expression following direct injection into the cervical enlargement of adult rat spinal cord [146]. Thus, attention turned to using AAV as a vector for delivery given the lack of an immune response generated by administration of AAV [124]. Glatzel injected recombinant AAV delivering enhanced green fluorescent protein (EGFP) into the DRG of L4/L5. Expression of EGFP was detectable up to 52 days postinjection without any signs of deterioration [145]. Our laboratory has also observed excellent transduction of cervical spinal cord neurons following direct spinal cord injection (fig. 6). The next step with AAV, as with adenovirus, was to utilize AAV’s property of retrograde transport to allow for the remote delivery of transgene into the spinal cord following indirect injection in the PNS. Kaspar et al. [147] have demonstrated retrograde transport of AAV-delivered GFP from the hippocampus and striatum to the entorhinal cortex and substantia nigra. Our laboratory has observed retrograde delivery of AAV following peripheral nerve injection in mice at a variety of locations [143].


88

Fig. 6. Motor neuron GFP expression 3 weeks after AAV.GFP injection of the cervical spinal cord in SOD1 mutant mice.

Lentiviral vectors have also proven useful for transgene delivery to the spinal cord. Mazarakis et al. [148] was able to demonstrate that equine infectious anemia virus pseudotyped either with rabies-G envelope or VSV-G can effectively deliver transgenes into the rat spinal cord following intraspinal injection. Following injection, the spinal cord failed to show any significant cell damage and only a mild degree of inflammation, confirming the low toxicity of lentiviral vector delivery. Mazarakis also showed that rabies-G pseudotyped lentivirus, but not VSV-G, is able to undergo retrograde transport following injection into rat gastrocnemius muscle. Transgene expression was detected at 3 weeks following lentiviral delivery, but is believed to persist for much longer time periods based on observations in other parts of the CNS. Prior to this study, lentiviral vectors have not been shown to undergo retrograde transport. This new capacity for retrograde transport is likely to be secondary to innate properties of the rabies-G protein to convey axonal uptake and transport [149]. Our laboratory has achieved retrograde delivery to cervical spinal cord motor neurons through injection into the brachial plexus using the rabies-G pseudotyped equine infectious anemia virus carrying a lacZ reporter gene. Spinal cord stained 3 weeks following brachial plexus injection of the virus showed transgene expression, confirming retrograde transport (fig. 7). The properties of the rabies-G envelope combined with a low immune toxicity should provide several new possibilities for using lentiviruses as potential vectors for transgene delivery into the spinal cord. In addition to making ‘remote delivery’ possible, direct injection of this vector into the spinal cord should make gene delivery to


89

Fig. 7. ␤-Galactosidase staining of motor neuron cell bodies in the ventral horn of spinal cord from mice injected with lacZ expressing rabies-G pseudotyped equine infectious anemia virus. The mice were injected with 4 ⫻ 108 PFU titer of EIAV.RabG.lacZ into the brachial plexus. The spinal cord was stained 3 weeks postinjection.

both the site of injection as well as upstream projection areas possible. This may prove useful for trophic factor delivery to upper motor neurons projecting into the site of injury. Therapeutic Animal Models Several animal models have been used to test the therapeutic potential of the viral vectors discussed above. One animal model is the delivery of GDNF into transgenic superoxide dismutase 1 (SOD1) mice. Transgenic SOD1 mice contain a mutation in the Cu/Zn SOD1 gene on chromosome 21, mimicking a form of amyotrophic lateral sclerosis (ALS) found in 20% of all ALS cases. GDNF has been shown to demonstrate an overwhelmingly potent ability to protect motor neuron survival compared to other neurotrophic factors, and is thus an ideal candidate for use in ALS animal models [132]. Acsadi et al. [150] delivered GDNF via intramuscular adenoviral vector injection into SOD1 mice. This adenoviral vector contained the rat GDNF


90

cDNA under a cytomegalovirus promoter followed by an internal ribosomal entry signal and an EGFP cDNA (AVR-GDNF). The virus was injected at a titer of 5 ⫻ 109 plaque forming units (PFUs) into the anterior tibialis, gastrocnemius, quadriceps, and paraspinal muscles of 5- to 7-day-old SOD1 mice. Control animals were injected with virus containing lacZ as the reported gene in order to confirm uptake of the virus into the muscle and retrograde transport into the spinal cord via ␤-gal staining. ELISA analysis was used to confirm the level of GDNF in the muscles, measuring the GDNF expression at 3 months postinjection to be 454 ⫾ 268 pg/mg (mean ⫾ SE), and at 4 months postinjection to be 180 ⫾ 106 pg/mg. GDNF levels in untreated mice were at undetectable levels. Importantly, injections in neonatal animals appeared to induce a longer lasting period of gene expression. This effect may be secondary to limited immune recognition in the neonate. SOD1 mice treated with AVR-GDNF showed a clear delay in the development of ALS symptoms. Untreated mice developed symptoms (e.g., hind limb tremor, slowing of movements) at 106.2 ⫾ 2.71 (mean ⫾ SD) days of age compared to treated mice, which developed symptoms at 116.1 ⫾ 8.6 days of age. Injection with AVR-GDNF also increased the lifespan of SOD1 mice following onset of symptoms by 8 days compared to untreated SOD1 mice, increasing lifespan overall by an average of 14 days. To quantitatively measure the effect of the disease on mice, the ability of the mice to maintain their balance on a rotating rod (RotaRod) was measured. RotaRod performance started to decline in SOD1 mice compared to wild-type mice following 8 weeks of age. The study showed that SOD1 mice treated with AVR-GDNF showed a significantly slower decline in performance compared to untreated SOD1 mice. Finally, the effect of GDNF was demonstrated in motor neuron counts of the spinal cord anterior horn 2, 3, and 4 months postinjection compared to untreated SOD1 mice. AVR-GDNF demonstrated an ability to prolong large motor neuron (⬎20 ␮m) survival for up to 2 months, after which motor neuron survival declined in a similar fashion to that found in untreated SOD1 mice. A similar study conducted by Manabe et al. [151] also delivered an adenoviral vector-containing GDNF into SOD1 mice. In this study, adenovirus at 108 PFU was injected into the gastrocnemius muscle of SOD1 mice, once a week starting at 35 weeks of age. Quantitative measurements of disease included evaluation of clinical grade, unilateral movement in a circular cage, and RotaRod performance at 35, 40, 42, and 46 weeks of age. Although there was not a significant difference between adenoviral vector-containing GDNFtreated mice compared to untreated mice, there was a tendency of improvement in the treated animals. In addition, large motor neuron survival was evaluated via hematoxylin and eosin staining as well as immunohistochemistry for p-Akt positive large motor neurons indicative of apoptotic death. Both means of


91

evaluation showed significant preservation of motor neurons in adenoviral vector-containing GDNF-treated SOD1 mice. The difference in the survival outcome between these two studies may be derived from the age of the animal at the time of treatment. Wang et al. [152] were able to demonstrate the neuroprotective effects of AAV vector GDNF gene delivery. Once again, SOD1 mice were used as animal models. In order to differentiate GDNF that is transgenically expressed from those being endogenously expressed, an AAV vector containing a transgene expressing a GDNF-FLAG fusion protein was developed. GDNF-FLAG can be easily determined via immunofluorescence staining with polyclonal rabbit anti-FLAG antibodies. The AAV-GDNF was injected into the gastrocnemius and triceps brachii muscles of all the four limbs of SOD1 mice at 9 weeks of age. At 110 days of age (7 weeks postinjection), GDNF levels in AAV-GDNFtreated mice was 7,985.0 ⫾ 874.0 pg/mg, an increase of ⬎120-fold compared to untreated mice. AAV-GDNF-treated mice also showed considerable preservation of the gastrocnemius muscle, showing little evidence of neutrogenic atrophy and weighing nearly one and a half times more than untreated mice. Retrograde transport of GDNF was demonstrated via FLAG staining of the spinal cord. Nissl staining of the motor neurons in the lumbar spinal cord showed significant protective effects of AAV-GDNF in the side of the cord ipsilateral to AAV-GDNF injection. Finally, AAV-GDNF demonstrated similar effects as adenovirusdelivered GDNF on RotaRod testing. Performance on the RotaRod deteriorated after 12 weeks of age in SOD1 mice compared to wild-type mice. AAV-GDNF was able to delay the onset of motor deficits as well as slow the deterioration of performance on the RotaRod. The onset of motor deficit in AAV-GDNF treated mice was 114.0 ⫾ 4.0 days compared to 101.3 ⫾ 5.4 days in untreated mice. Lifespan in treated mice was increased by a mean of 16.6 ⫾ 4.1 days; an improvement remarkably similar to the one found in AVR-GDNF-treated SOD1 mice in the study conducted by Acsadi et al. [150]. Despite the delay in the onset of symptoms and increase in lifespan, AAV-GDNF-treated mice showed no difference in the duration of disease when compared to untreated mice. The significantly higher level of expression of GDNF in these experiments compared to those found in the AVR-GDNF experiments lends a great deal of promise to the use of AAV as an ideal vector for delivery of therapeutic transgenes. Further experiments are necessary to test whether administering the transgene at an earlier age can affect the duration of disease [152]. GDNF’s ability to protect motor neuron following axotomy was also illustrated by Baumgartner and Shine [153]. Adenoviral vectors were created containing expression cassettes for BDNF, GDNF, NGF, or ciliary neurotrophic factor. Adenoviral vectors were injected into the gastrocnemius, flexor longus digitorum, and tibialis PFU titer. Adenoviruses carrying a lacZ transgene and


92

adenovirus void of the transgene, as well as a virus-free vehicle, were injected into the same hind limb muscles as controls. Retrograde transport of virally delivered transgenes was confirmed with ␤-gal staining in the lumbar spinal cord motor neurons. In order to study the neuroprotective effects of the growth factors, 2 days following injection the sciatic nerve was severed. Seven days postaxotomy, the lumbar spinal cord segments were removed and it was found that only animals treated with GNDF-containing adenovirus showed a significant difference in preserving neuron survival when compared to empty adenovirus or vehicle-treated animals. At 2 days postaxotomy, animals treated with GDNF showed preservation of 70% of its motor neurons compared to 44% seen in vehicle controls. The neuroprotective effect of GDNF proved to be transient, showing no difference from control animals at 42 days postaxotomy. These results further demonstrate the potent ability of GDNF in neuroprotection as well as setting the stage for the use of neurotrophic factors for protection following SCI. Smith and his colleagues [154] have been able to utilize adenoviral vectors for the delivery of neurotrophic factors to induce functional recovery of axons into the dorsal root entry zone. Recombinant adenoviral vectors were created containing transgenes encoding for FGF2, NGF, L1 cell adhesion molecule, or ␤-galactosidase (LacZ). Sprague-Dawley rats were treated with triple crush lesions at the L4 and L5 dorsal roots. Under natural conditions, peripheral nerve regeneration is halted at the dorsal root entry zone, the CNS border. In rats injected with a 7.5 ⫻ 106 PFU/␮l titre of adenovirus carrying either NGF or FGF2 16 days following rhizotomy, large numbers of calcitonin gene-related peptide-positive axon fibers can be seen growing into the dorsal horn compared to uninfected rats. In addition to histological analysis, Smith was able to demonstrate functional recovery following NGF or FGF2 administration. Rats treated with NGF or FGF2 showed recovery of nociception as evaluated using a plantar heat test. Furthermore, recovery of proprioception was evaluated using a grid runway test. None of the neurotrophic factors administered was able to induce any recovery of proprioception, indicating specific targeting by NGF and FGF2. Ex vivo Gene Transfer The use of cell grafts that express a therapeutic transgene is an alternative to in vivo gene transfer. This method, known as ex vivo gene transfer, involves genetically modifying cells in vitro to express a gene of interest, and then transplanting the cell graft into the host. One advantage of this method is the potential to verify transgene expression in the desired cells before transplantation into the host. Another major advantage is the ability of the transplanted cell to produce a long-term steady-state therapeutic level of transgene, a problem commonly encountered with the use of viral or nonviral vectors (fig. 8).


93

Retroviral vector

Retroviral vector

Fibroblast

Spinal cord

NGF Spinal cord

NGF

b

NGF Spinal cord

a Fig. 8.a Ex vivo gene transfer. Following in vitro infection with a retroviral (or an alternative virus) vector carrying a therapeutic transgene, the fibroblasts are grown in culture. Verification of the gene expression can be confirmed before transplantation of the fibroblast into animal spinal cords. b In vivo gene transfer. Retroviral vectors carrying the therapeutic transgene are directly injected into the spinal cord of animal models.

The first consideration in ex vivo gene transfer is the type of cell to be used for grafting. Ease of infection in vitro and the viability of the cells themselves, as well as their effects on the viability of the cells around them once grafted into the host, are all factors that can mitigate the choice of cell type. For these reasons,


94

primary fibroblasts have been a popular choice for ex vivo experiments. Not only are fibroblasts easily sustained in vitro and well accepted into the CNS, fibroblasts naturally secrete collagen and fibronectin, which provide a conducive environment for neurite growth. Also, since fibroblasts are nonneuronal and nonglial in origin, they are void of any growth-inhibiting molecules that may prohibit neuronal growth. The earliest studies using fibroblasts to provide therapeutic transgenes to the CNS came from Rosenberg et al. [155], who were able to modify fibroblasts by infecting them with mouse NGF cDNA via a retroviral vector. The cells were then grafted into the brain following surgical lesion between the fimbria and the fornix. Two weeks after graft implantation, the animals were sacrificed and stained for choline acetyltransferase, an indication of survival of cholinergic cell bodies. It was found that animals receiving the NGF-secreting cell graft preserved 92% of their cholinergic cells compared to only 49% in control animals. These experiments set the stage for using fibroblasts for ex vivo gene transfer. Fibroblasts producing NGF have been transplanted into the spinal cord in an attempt to induce neuronal growth. Such cell grafts were found to be viable and producing NGF for up to one year following transplantation [156]. The grafts heavily induced growth of sensory axons, verified by calcitonin generelated peptide staining, proving the grafts’ ability to induce growth of a specific axonal type. The ability of fibroblast grafts to induce axonal growth following acute SCI was also observed. NGF-producing fibroblasts were implanted into the spinal cord following spinal dorsal hemisection lesions. The cell grafts were able to induce growth of not only sensory axons, but also of motor axons, albeit to a lesser extent [157]. The injury-evoked responsiveness to NGF provides important insight into the selective use of NGF as a neurotrophic factor in SCI. Similar results were obtained in studies in which NGF grafts were implanted 1–3 months following spinal cord lesion to study the effects on the chronically injured spinal cord. While sensory fibers were noted to regenerate, no motor neuron response was observed in the chronically injured spinal cord. Following SCI neurotrophin receptors are hypothesized to increase [158]. This upregulation is most likely transient, explaining the lack of motor neuron response following chronic injury [159]. The above experiments were carried out using early generation retrovirus to genetically modify the cell grafts. Retroviral vectors can be effective because they integrate into the host genome and allow for long-term gene expression. Alternative vectors have also been utilized and proved to be successful. Lentiviral vectors, retroviral vectors capable of gene transfer to terminally differentiated cells, have been successfully used to modify fibroblasts [160, 161]. Liu et al. [162] were also able to use adenoviral vectors to modify fibroblasts. Though ex vivo neurotrophic gene transfer to fibroblasts has shown promise, the use of alternative cell lines as well as other neurotrophins are also


95

capable of inducing axonal growth in the CNS. Because of their possible advantage for autologous grafting, skin fibroblasts have been attempted for use as cell grafts. Early attempts at using skin fibroblasts were unsuccessful in producing prolonged transgene expression [163]. OECs have also been used as implants for ex vivo gene transfer. Primary OECs have been modified with both adenoviral and lentiviral vectors, with grafts allowing for transgene expression for up to one and 4 months respectively [164]. Cell grafts that have been modified to produce NT-3 have also been shown to promote corticospinal axon growth when grafted into the spinal cord following hemisection lesion [165]. Conclusion

Despite many years of research in the field of SCI and regeneration, the prognosis for those who have sustained an SCI remains dismal. Few therapeutic strategies are available to the clinician. The barriers to effective neural regeneration, and hence functional recovery, are multifactorial. These barriers include the lack of an intrinsic cellular response to divide and regenerate, the need to overcome inhibitory barriers at the injury site, and the need to recapitulate the native circuitry following injury. Advances in both genetic and cellular therapy for SCI have begun to unravel some of the difficulties with functional neuronal recovery. Stem cells, both exogenous and endogenous, have the capability to differentiate into various CNS lineages. These cells may, therefore, provide neurogenesis, neuroprotection, trophic support, and/or remyelination. Furthermore, they may provide effective tissue bridges for regenerating axons. Stem cell therapy may supplant existing tissue or cell transplant paradigms and their inherent shortcomings. The fields of molecular neurobiology and genetic therapy are rapidly advancing. This therapeutic strategy will grant clinicians the ability to alter the function of intrinsic or extrinsically placed cells. These cells may then produce growth factors or other growth-permissive and neural-protective proteins capable of supporting the injured spinal cord. These therapies may be delivered systemically without the need to access the CNS surgically. Despite the grim prognosis of SCI, cellular and genetic therapies continue to provide the hope of recovery following an SCI. In experimental animal models of SCI, both are providing evidence of functional recovery. This work lays the groundwork for human clinical trials. References 1

Burney R, Maio R, Maynard F: Incidence, characteristics, and outcome of spinal cord injury at trauma centers in North America. Arch Surg 1992;128:569–599.


96

2

3 4 5 6 7 8

9

10 11 12 13 14 15 16 17

18

19

20

21

22 23 24

25

Tator C: Epidemiology and general characteristics of the spinal cord injury patient; in Contempory Management of Spinal Cord Injury. Illinois, American Association of Neurological Surgeons, 1995, pp 9–13. Kraus J: Injury to the head and spinal cord: The epidemiological relevance of the medical literature published from 1960 to 1978. J Neurosurg 1980;53(suppl):S3–S10. Tator C, Duncan E, Edmonds V: Complications and costs of management of acute spinal cord injury. Paraplegia 1993;31:700–714. Stripling T: The cost of economic consequences of traumatic spinal cord injury. Paraplegia News 1990;August:50–54. Kwon D, Oxland T, Tetzlaff W: Animal models used in spinal cord regeneration research. Spine 2002;27:1504–1510. Gage F: Mammalian neural stem cells. Science 2000;287:1433–1438. Brustle I, Jones K, Learisk R, Karram K, Choudhary K, Wiestler O, Duncan I, McKay R: Embryonic stem cell-derived glial precursors: A source of myelinating transplants. Science 1999; 285:754–756. Flax J, Aurora S, Yang C, Simonin C, Wills A, Billinghurst L, Jendoubi M, Sidman R, Wolfe J, Kim S, Snyder E: Engraftable human neural stem cells respond to developmental cues, replace neurons, and express foreign genes. Nat Biotechnol 1998;16:1033–1039. McKay R: Stem cells in the central nervous system. Science 1997;276:66–71. Mujtaba T, Piper D, Kalyani A, Groves A, Lucero M, Rao M: Lineage-restricted neural precursors can be isolated from both the mouse neural tube and cultured ES cells. Dev Biol 1999;214:113–127. Reubinoff B, Itsykson P, Turetsky T, Pera M, Reinhartz E, Itzik A, Ben-Hur T: Progenitors from human embryonic stem cells. Nat Biotechnol 2001;19: 1134–1140. Zhang S, Wernig M, Duncan I, Brustle O, Thomson J: In vitro differentiation of transplantable neural precursors from human embryonic stem cells. Nat Biotechnol 2002;19:1129–1133. Cao Q, Benton R, Whittemore S: Stem cell repair of central nervous system injury. J Neurosci Res 2002;68:501–510. Sanchez-Ramos J: Neural cells derived from adult bone marrow and umbilical cord blood. J Neurosci Res 2002;69:880–893. Penn M, Francis G, Ellis S, Young J, McCarthy P, Topol E: Autologous cell transplantation for the treatment of damaged myocardium. Prog Cardiovasc Dis 2002;45:21–32. Brustle I, Choudhary K, Karrma K, Huttner A, Murray K, Dubois-Dalcq M: Chimeric brains generated by intraventricular transplantation of fetal human brain cells into embryonic rats. Nat Biotechnol 1998;16:1040–1044. Rosser A, Tyerrs P, Dunnett S: The morphological development of neurons derived from EGF and FGF-2 driven human CNS precursors depends on their site of integration in the neonatal rat brain. Eur J Neurosci 2000;12:2405–2413. Cao Q, Zhang Y, Howard R, Walters W, Tsoulfas P, Whittemore S: Pluripotent stem cells engrafted into the normal or lesioned adult rat spinal cord are restricted to a glial lineage. Exp Neurol 2001;167:48–58. Chow S, Moul J, Tobias C, Himes B, Liu Y, Obrocka M, Hodge L, Tessler A, Gischer I: Characterization and intraspinal grafting of EGF/bFGF-dependent neurospheres derived from embryonic rat spinal cord. Brain Res 2000;874:87–106. White P, Morrison S, Orimoto K, Kubu C, Verdi J, Anderson D: Neural crest stem cells undergo cell-intrinsic developmental changes in sensitivity to instructive differentiation signals. Neuron 2001;29:57–71. Rao M, Mayer-Proschel M: Precursor cells for transplantation. Prog Brain Res 2000;128:273–292. Anderson D: Stem cells and pattern formation in the nervous system: The possible versus the actual. Neuron 2001;30:19–35. Yang H, Mtjtaba T, Benkatraman G, Wu Y, Rao M, Luskin M: Region-specific differentiation of neural tube derived neuronal restricted progenitor cells after heterotopic transplantation. Proc Natl Acad Sci USA 2000;97:13366–13371. Zigova T, Pencea V, Betarbert R, Wiegand S, Alexander C, Bakay RA, Luskin MB: Neuronal progenitor cells of the neonatal subventricular zone differentiate and disperse following transplanation into the adult rat striatum. Cell Transplant 1998;7:137–156.


97

26

27 28

29 30 31 32 33 34 35

36 37 38 39 40 41

42

43

44 45 46 47 48 49 50

McDonald J, Liu X, Qu Y, Liu S, Mickley S, Turetsky D, Gottlieb D: Transplanted embryonic stem cells survive, differentiate and promote recovery in injured rat spinal cord. Nat Med 1999; 5:1410–1412. Bain G, Kitchen D, Yao M, Huettner J, Gottlieb D: Embryonic stem cells express neuronal properties in vitro. Dev Biol 1995;168: 342–357. Gimenez Y, Ribotta M, Feraboli-Lohnherr O, Privat A: Recovery of locomotion following transplantation of monoaminergic neurons in the spinal cord of paraplegic rats. Ann NY Acad Sci 1998;860:393–411. Eriksson P, Perfilieva E, Bjork-Eriksson T, Alborn A, Nordborg C, Peterson D, Gage F: Neurogenesis in the adult human hippocampus. Nat Med 1998;4:1313–1317. Gould E, Tanapat P: Stress and hippocampal neurogenesis. Biol Psychiatry 1999;46:1472–1479. Lois C, Alvarez-Buylla A: Proliferating subventricular zone cells in the adult mammalian forebrain can differentiate into neurons and glia. Proc Natl Acad Sci USA 1993;90:2074–2077. Rietze R, Poulin P, Weiss S: Mitotically active cells that generate neurons and astrocytes are present in multiple regions of adult mouse hippocampus. J Comp Neurol 2000;424:397–408. Temple S, Alvarez-Buylla A: Stem cells in the adult mammalian central nervous system. Curr Opin Neurobiol 1999;9:135–141. Johansson C, Momma S, Clarke K, Risling M, Lendahl U, Friesen J: Identification of a neural stem cell in the adult mammalian central nervous system. Cell 1999;96:25–34. Horner P, Power AE, Kempermann G, Kuhn H, Palmer T, Winkler J, Thal L, Gage F: Proliferation and differentiation of progenitor cells throughout the intact adult rat spinal cord. J Neurosci 2000;20: 2218–2228. Levine J, Reynolds R, Fawcett J: The oligodendrocyte precursor cell in health and disease. Trends Neurosci 2001;24:39–47. Keirstead H, Levine J, Blakemore W: Response of the oligodendrocyte progenitor cell population (defined by NG2 labelling) to demyelination of the adult spinal cord. Glia 1998;22:161–170. Levine J, Reynolds R: Activation and proliferation of endogenous oligodendrocyte precursor cells during ethidium bromide-induced demyelination. Exp Neurol 1999;160:333–347. McTigue D, Wei P, Stokes B: Proliferation of NG2-positive cells and altered oligodendrocyte numbers in the contused rat spinal cord. J Neurosci 2001;21:3392–3400. Zhang S, Lipsitz D, Duncan I: Self-renewing canine oligodendroglial progenitor expanded as oligospheres. J Neurosci Res 1998;54:181–190. Herrera J, Yang H, Zang S, Proschel C, Tresco P, Duncan I, Luskin M, Mayer-Proschel M: Embryonic-derived glial-restricted precursor cells (GRP cells) can differentiate into astrocytes and oligodendrocytes in vivo. Exp Neurol 2001;171:11–21. Akiyama Y, Honmou O, Kato T, Uede T, Hashi K, Kocsis J: Transplantation of clonal neural precursor cells derived from adult human brain establishes functional peripheral myelin in the rat spinal cord. Exp Neurol 2001;167:27–39. Howland D, Bregman B, Tessler A, Goldberger M: Transplants enhance locomotion in neonatal kittens whose spinal cords are transected: A behavioral and anatomical study. Exp Neurol 1995; 135:123–145. Miya D, Giszter S, Mori F, Adipudi V, Tessler A, Murrary M: Fetal transplants alter the development of function after spinal cord transection in newborn rats. J Neurosci 1997;17:4865–4872. Bregman B, Diener P, McAtee M, Dai H, James C: Intervention strategies to enhance anatomical plasticity and recovery of function after spinal cord injury. Adv Neurol 1997;72:257–275. Bregman B, Kunkel-Bagden E, McAtee M, O’Neill A: Extension of the critical period for developmental plasticity of the corticospinal pathway. J Comp Neurol 1989;282:355–370. Bregman B, McAtee M, Dai H, Kuhn P: Neurotrophic factors increase axonal growth after spinal cord injury and transplantation in the adult rat. Exp Neurol 1997;148:475–494. Jakeman L, Reier P: Axonal projections between fetal spinal cord transplants and the adult rat spinal cord: A neuroanatomical tracing study of local interactions. J Comp Neurol 1991;307:311–334. Bregman B, Broude E, McAtee M, Kelly M: Transplants and neurotrophic factors prevent atrophy of mature CNS neurons after spinal cord injury. Exp Neurol 1998;149:13–27. Coumans J, Lin T, Dai H, MacArthur L, McAtee M, Nash C, Bregman B: Axonal regeneration and functional recovery after complete spinal cord transection in rats by delayed treatment with transplants and neurotrophins. J Neurosci 2001;21:9334–9344.


98

51 52 53 54 55 56 57 58 59 60

61

62

63 64 65 66 67

68 69 70 71 72 73 74 75 76

Barnett S, Hutchins A, Noble M: Purification of olfactory nerve ensheathing cells from olfactory bulb. Dev Biol 1993;155:337–350. Raisman G: Specialized neuroglial arrangement may explain the capacity of vomeronasal axons to reinnervate central neurons. Neuroscience 1995;14:237–254. Ramon-Cueto A, Nieto-Sampedro M: Glial cells from adult rat olfactory bulb: Immunocytochemical properties of pure cultures of ensheathing cells. Neuroscience 1992;47:213–220. Valverde F, Lopez-Mascarauque L: Neuroglial arrangements in the olfactory glomeruli of the hedgehog. J Comp Neurol 1991;307:658–674. Ramon-Cueto A, Nieto-Sampedro M: Regeneration into the spinal cord of transected dorsal root axons is promoted by enhancing glia transplants. Exp Neurol 1994;127:232–244. Ramon-Cueto A, Plant G, Avila J, Bunge M: Long-distance axonal regeneration in the transected adult rat spinal cord is promoted by olfactory ensheathing glia transplants. J Neurosci 1998;18:3803–3815. Li Y, Raisman G: Integration of transplanted cultured Schwann cells into the long myelinated fiber tracts of the adult spinal cord. Exp Neurol 1997;145:397–411. Brook G, Lawrence J, Raisman G: Morphology and migration of cultured Schwann cells transplanted into the fimbria and hippocampus in adult rats. Glia 1993;9:292–304. Li Y, Raisman G: Schwann cells induce sprouting in motor and sensory axons in the adult rat spinal cord. J Neurosci 1994;14:4050–4063. Takami T, Oudega M, Bates M, Wood P, Kleitman N, Bunge M: Schwann cell but not olfactory ensheathing glia transplants improve hindlimb locomotor performance in the moderately contused adult rat thoracic spinal cord. J Neurosci 2002;22:6670–6681. Bamber N, Li H, Lu X, Oudega M, Aebisher P, Xu X: Neurotrophins BDNF and NT-3 promote axonal re-entry into the distal host spinal cord through Schwann cell-seeded mini-channels. Eur J Neurosci 2001;13:257–268. Rapalino O, Lazarov-Spiegler O, Agranov E, Velan G, Yoles E, Fraidakis M, Solomon A, Gepstein R, Katz A, Belkin M, Hadani M, Schwarta M: Implantation of stimulated homologous macrophages results in partial recovery of paraplegic rats. Nat Med 1998;4:814–821. Lazarov-Spiegler O, Rapalino O, Agranov G, Schwartz M: Restricted inflammatory response in the CNS: A key impediment to axonal regeneration? Mol Med Today 1998;4:337–342. Lazarov-Spiegler O, Solomon A, Hirschberg D, Levine V, Schwartz M: Transplantation of activated macrophages overcomes central nervous system regrowth failure. FASEB J 1996;10:1296–1302. Schwartz M, Hirschberg D, Beserman P: Central nervous system regeneration and the immune system. Mol Med Today 1995;1:60–61. Faber-Elman A, Levine V, Schvartz I, Shaltiel S, Schwartz M: Vitronectin overrides a negative effect of TNF-alpha on astrocytes migration. FASEB J 1995;9:1605–1613. Faber-Elman A, Solomon A, Abraham J, Marikovsky M, Schwartz M: Involvement of woundassociated factors in rat brain astrocytes migratory response taxonal injury: In vitro stimulation. J Clin Invest 1996;97:162–171. Mckeon R, Hoke A, Silver J: Injury-induced proteoglycans inhibit the potential for laminin-mediated axon growth on astrocytic scars. Exp Neurol 1995;136:32–43. Capecchi M: High efficiency transformation by direct microinjection of DNA into cultured mammalian cells. Cell 1980;22:479–488. Cartier R, Reszka R: Utilization of synthetic peptides containing nuclear localization signals for nonviral gene transfer systems. Gene Ther 2002;9:157–167. Tseng W, Haselton F, Giorgio T: Mitosis enhances transgene expression of plasmid delivered by cationic liposomes. Biochim Biophys Acta 1999;1445:53–64. Wilke M, Fortunati E, van den Broek M, Hoogeveen A, Scholte B: Efficacy of a peptide-based gene delivery system depends on mitotic activity. Gene Ther 1996;3:1133–1142. Wolff J, Malone R, Williams P, Chong W, Acsadi G, Jani A, Felgner P: Direct gene transfer into mouse muscle in vivo. Science 1990;247:1465–1468. Rols M, Teissie J: Electropermeabilization of mammalian cells to macromolecules: Control by pulse duration. Biophys J 1998;75:1415–1423. Lin C, Tai M, Cheng J, Chou A, Wang J, Tan P, Marsala M, Yang L: Electroporation for direct spinal gene transfer in rats. Neurosci Lett 2002;317:1–4. Yang N, Burkholder J, Roberts B, Martinell B, McCabe D: In vivo and in vitro gene transfer to mammalian somatic cells by particle bombardment. Proc Natl Acad Sci USA 1990;87:9568–9572.


99

77 Felgner P, Gadek T, Holm M, Roman R, Chan H, Wenz M, Northrop J, Ringold G, Danielson M: Lipofection: A highly efficient, lipid-mediated DNA-transfection procedure. Proc Natl Acad Sci USA 1987;84:7413–7417. 78 Gao X, Huang L: Potentiation of cationic liposome-mediated gene delivery by polycations. Biochemistry 1996;35:1027–1036. 79 Farhood H, Gao X, Son K, Yang Y, Lazo J, Huang L, Barsoum J, Bottega R, Epand R: Cationic liposomes for direct gene transfer in therapy of cancer and other diseases. Ann NY Acad Sci 1994; 716:23–34. 80 Hui S, Langner M, Zhao Y, Ross P, Hurley E, Chan K: The role of helper lipids in cationic liposome-mediated gene transfer. Biophys J 1996; 71:590–599. 81 Boussif O, Lezoualc’h F, Zanta M, Mergny M, Scherman D, Demeneix B, Behr J: A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: Polyethylenimine. Proc Natl Acad Sci USA 1995;92:7297–7301. 82 Kaminski J, Huber M, Summers J, Ward M: Design of a nonviral vector for site-selective, efficient integration into the human genome. FASEB J 2002;16:1242–1247. 83 Guild B, Finer M, Housman D, Mulligan R: Development of retrovirus vectors useful for expressing genes in cultured murine embryonal cells and hematopoietic cells in vivo. J Virol 1988;62:3795–3801. 84 Miller A, Skotzko M, Rhoades K, Belldegrun A, Tso C, Kaboo R, McBride W, Jacobs E, Kohn D, Moen R: Simultaneous use of two retroviral vectors in human gene marking trials: Feasibility and potential applications. Hum Gene Ther 1992;3:619–624. 85 Miller D, Adam M, Miller A: Gene transfer by retrovirus vectors occurs only in cells that are actively replicating at the time of infection. Mol Cell Biol 1990;10:4239–4242. 86 Roe T, Reynolds T, Yu G: Integration of murine leukemia virus DNA depends on mitosis. EMBO J 1993;12:2099–2108. 87 Bukrinsky M, Sharova N, Dempsey M, Stanwick T, Bukrinskaya A, Haggerty S, Stevenson M: Active nuclear import of human immunodeficiency virus type 1 preintegration complexes. Proc Natl Acad Sci USA 1992;89:6580–6584. 88 Bukrinsky M, Haggerty S, Dempsey M, Sharova N, Adzhubel A, Spitz L, Lewis P, Goldfarb D, Emerma M, Stevenson M: A nuclear localization signal within HIV-1 matrix protein that governs infection of non-dividing cells. Nature 1994;369:107–108. 89 Gallay P, Stitt V, Mundy C, Oettinger M, Trono D: Role of the karyopherin pathway in human immunodeficiency virus type 1 nuclear import. J Virol 1996;70:1027–1032. 90 Lewis P, Hensel M, Emerma M: Human immunodeficiency virus infection of cells arrested in the cell cycle. EMBO J 1992;11:3053–3058. 91 Lewis P, Emerman M: Passage through mitosis is required for oncoretroviruses but not for the human immunodeficieny virus. J Virol 1994;68:510–516. 92 Kafri T, Blomer U, Peterson D, Gage F, Verma I: Sustained expression of genes delivered directly into liver and muscle by lentiviral vectors. Nat Genet 1997;17:314–317. 93 Naldini L, Blomer U, Gallay P, Ory D, Mulligan R, Gage F, Verma I, Trono D: In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science 1996;272:263–267. 94 Trono D: HIV accessory proteins: Leading roles for the supporting cast. J Virol 1995;82:189–192. 95 Zufferey R, Nagy D, Mandel R, Naldini L, Trono D: Multiply attenuated lentiviral vector achieves efficient gene delivery in vivo. Nat Biotechnol 1997;15:871–875. 96 Kim V, Mitrophanous K, Kingsman S, Kingsman A: Minimal requirement for a lentivirus vector based on human immunodeficiency virus type 1. J Virol 1998;72:811–816. 97 Dull T, Zufferey R, Kelly M, Mandel R, Nguyen M, Trono D, Naldini, L: A third-generation lentivirus vector with a conditional packaging system. J Virol 1998;72:8463–8471. 98 Zufferey R, Dull T, Mandel R, Bukovsky A, Quiroz D, Naldini L, Trono D: Self-inactivating lentivirus vector for safe and efficient in vivo gene delivery. J Virol 1998;72:9873–9880. 99 Poeschla E, Won-Staal F, Looney D: Efficient transduction of nondividing human cells by feline immunodeficieny virus lentiviral vectors. Nat Med 1998;4:354–357. 100 Stevens J: Latent herpes simplex virus and the nervous system. Curr Top Microbiol Immunol 1975;70:31–50. 101 Kaplitt M, Pfaff D: Viral vectors for gene delivery and expression in the CNS. Methods 1996; 10:343–350.


100

102 Geller A, Breakefield X: A defective HSV-1 vector expresses Escherichia coli beta-galactosidase in cultured peripheral neurons. Science 1988;241:1667–1669. 103 Geller A, Keyomarsi K, Bryan J, Pardee A: An efficient deletion mutant packaging system for defective herpes simplex virus vectors: Potential applications to human gene therapy and neuronal physiology. Procl Natl Acad Sci USA 1990;87:8950–8954. 104 Glorioso J, Goins W, Meaney C, Fink D, DeLuca N: Gene transfer to brain using herpes simplex virus vectors. Ann Neurol 1994; 35(suppl):S28–S34. 105 Shenk T: Group C adenoviruses as vectors for gene therapy; in Kaplitt M, Loewy A (eds): Viral Vectors: Gene Therapy and Neuroscience Applications. San Diego, Academic Press, 1995, pp 43–54. 106 Engelhardt J, Simon R, Yang Y, Zepeda M, Weber-Pendleton S, Doranz B, Grossman M, Wilson J: Adenovirus-mediated transfer of the CFTR gene to lung of nonhuman primates: Biological efficacy study. Hum Gene Ther 1993;4:759–769. 107 Byrnes A, Wood M, Charlton H: Role of T cells in inflammation caused by adenovirus vectors in the brain. Gene Ther 1996;3:644–651. 108 Yang Y, Li Q, Ertl HC, Wilson JM: Cellular and humoral immune responses to viral antigens create barriers to lung-directed gene therapy with recombinant adenoviruses. J Virol 1995;69:2004–2015. 109 Yang Y, Nunes F, Berencsi K, Furth E, Gonczol E, Wilson J: Cellular immunity to viral antigens limits E1-deleted adenoviruses for gene therapy. Proc Natl Acad Sci USA 1994;91:4407–4411. 110 Romero M, Smith G: Adenoviral gene transfer into the normal and injured spinal cord: Enhanced transgene stability by combined administration of temperature-sensitive virus and transient immune blockage. Gene Ther 1998;5:1612–1621. 111 Engelhardt J, Litzky L, Wilson J: Prolonged transgene expression in cotton rat lung with recombinant adenoviruses defective in E2a. Hum Gene Ther 1994;5:1217–1229. 112 Armentano D, Sookdeo C, Hehir K, Gregory R, St George J, Prince G, Wadsworth S, Smith A: Characterization of an adenovirus gene transfer vector containing an E4 deletion. Hum Gene Ther 1995;6:1343–1353. 113 Gorziglia M, Kadan M, Yei S, Lim J, Lee G, Luthra R, Trapnell B: Elimination of both E1 and E2a from adenovirus vectors further improve prospects for in vivo human gene therapy. J Virol 1996; 70:4173–4178. 114 Pao C, Lin K, Zhu J, Wu G, Farmer P, Phillips L: In vitro transcription of the rat insulin-like growth factor-I gene. J Biol Chem 1996; 271:8667–8674. 115 Goldman M, Litzky L, Engelhardt J, Wilson J: Transfer of the CFTR gene to the lung of nonhuman primates with E1-deleted, E2a-defective recombinant adenoviruses: A preclinical toxicology study. Hum Gene Ther 1995;6: 839–851. 116 Yang Y, Nunes F, Berencsi K, Gonczol E, Engelhardt J, Wilson J: Inactivation of E2a in recombinant adenoviruses limits cellular immunity and improves the prospect for gene therapy of cystic fibrosis. Nat Genet 1994;7:362–369. 117 Parks R, Chen L, Anton M, Sankar U, Rudnicki M, Graham F: A helper-dependent adenovirus vector system: Removal of helper virus by Cre-mediated excision of the viral packaging signal. Proc Natl Acad Sci USA 1996;93:13565–13570. 118 Parks R, Graham F: A helper-dependent system for adenovirus vector production helps define a lower limit for efficient DNA packaging. J Virol 1997;71:3293–3298. 119 Berns K, Bohenzky R: Adeno-associated viruses: An update. Adv Virus Res 1987;32:243–306. 120 Xiao X, Li J, Samulski R: Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus. J Virol 1998;72:2224–2232. 121 Kotin R, Siniscalco M, Samulski R, Zhu X, Hunter L, Laughlin C, McLaughlin S, Muzycka N, Rocchi M, Berns K: Site-specific integration by adeno-associated virus. Proc Natl Acad Sci USA 1990;87:2211–2215. 122 Xiao X, Li J, Samulski RJ: Efficient long-term gene transfer into muscle tissue of immunocompetent mice by adeno-associated virus vector. J Virol 1996;70:8098–8108. 123 Kaplitt M, Leone P, Samulski R, Xiao X, Pfaff D, O’Malley K: Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat Genet 1994;8:148–154. 124 Xiao X, Li J, McCown TJ, Samulski RJ: Gene transfer by adeno-asociated virus vectors into the central nervous system. Exp Neurol 1997;144:113–124.


101

125 Sun L, Li J, Xiao X: Overcoming adeno-associated virus vector size limitations through viral DNA heterodimerization. Nat Med 2000;6:599–602. 126 Oudega M, Hagg T: Nerve growth factor promotes regeneration of sensory axons into adult rat spinal cord. Exp Neurol 1996;140:218–229. 127 Ramer M, Priestley J, McMahon S: Functional regeneration of sensory axons into the adult spinal cord. Nature 2000;403:312–316. 128 McMahon S, Armanini M, Ling L, Phillips H: Expression and coexpression of Trk receptors in subpopulations of adult primary sensory neurons projecting to identified peripheral targets. Neuron 1994;12:1161–1171. 129 Jakeman L, Wei P, Guan Z, Stokes B: Brain-derived neurotrophic factor stimulates hindlimb stepping and sprouting of cholinergic fibers after spinal cord injury. Exp Neurol 1998;154:170–184. 130 Oudega M, Hagg T: Neurotrophins promote regeneration of sensory axons in the adult rat spinal cord. Brain Res 1999;818:431–438. 131 Schnell L, Schneider R, Kolbeck R, Barde Y, Schwab M: Neurotrophin-3 enhances sprouting of corticospinal tract during development and after adult spinal cord lesions. Nature 1994;367:170–173. 132 Henderson C, Phillips H, Pollock R, Davies A, Lemeulle C, Armanini M, Simpson L, Moffet B, Vandlen R, Koliatsos V, Rosenthal A: GDNF: A potent survival factor for motor neurons present in peripheral nerve and muscle. Science 1994;266:1062–1064. 133 Bradbury E, Khemani S, King V, Priestley J, McMahon S: NT-3 promotes growth of lesioned adult rat sensory axons ascending in the dorsal columns of the spinal cord. Eur J Neurosci 1999;11:3873–3883. 134 Kluck R, Bossy-Wetzel E, Green D, Newmeyer D: The release of cytochrome c from mitochondria: A primary site for Bcl-2 regulation of apoptosis. Science 1997;275:1132–1136. 135 Allsopp T, Wyatt S, Paterson H, Davies A: The proto-oncogene bcl-2 can selectively rescue neurotrophic factor-dependent neurons from apoptosis. Cell 1993;73:295–307. 136 Matsuoka N, Yukawa H, Ishii K, Hamada H, Akimoto M, Hashimoto N, Miyatake S: Adenovirusmediated gene transfer of Bcl-xL prevents cell death in primary neuronal culture of the rat. Neurosci Lett 1999;270:177–180. 137 Blomer U, Kafri T, Randolph-Moore L, Verma I, Gage F: Bcl-xL protects adult septal cholinergic neurons from axotomized cell death. Proc Natl Acad Sci USA 1998;95:2603–2608. 138 Liu Y, Himes B, Moul J, Huang W, Chow S, Tessler A, Fischer I: Application of recombinant adenovirus for in vivo gene delivery to spinal cord. Brain Res 1997;768:19–29. 139 Byrnes A, Rusby J, Wood M, Charlton H: Adenovirus gene transfer causes inflammation in the brain. Neuroscience 1995; 66:1015–1024. 140 Boulis N, Turner DE, Dice J, Bhatia V, Feldman E: Characterization of adenoviral gene expession in spinal cord after remote vector delivery. Neurosurgery 1999;45:131–137. 141 Ghadge G, Roos R, Kang U, Wollmann R, Fishman P, Kalynych A, Barr E, Leiden J: CNS gene delivery by retrograde transport of recombinant replication-defective adenoviruses. Gene Ther 1995;2:132–137. 142 Kuo H, Ingram D, Crystal R, Mastrangeli A: Retrograde transfer of replication deficient recombinant adenovirus vector in the central nervous system for tracing studies. Brain Res 1995;705:31–38. 143 Boulis NM, Willmarth N, Song D, Feldman E, Imperiale M: Intraneural colchicine inhibition of adenoviral and adeno-associated viral vector remote spinal cord gene delivery. Neurosurgery 2003;52:381–387. 144 Turner D, Noordmans A, Feldman E, Boulis N: Remote adenoviral gene delivery to the spinal cord: Contralateral delivery and reinjection. Neurosurgery 2001;48:1309–1316. 145 Glatzel M, Flechsig E, Navarro B, Klein M, Paterna J, Bueler H, Aguzzi A: Adenoviral and adenoassociated viral transfer of genes to the peripheral nervous system. Proc Natl Acad Sci USA 2000; 97:442–447. 146 Peel A, Zolotukhin S, Schrimsher G, Muzyczka N, Reier P: Efficient transduction of green fluorescent protein in spinal cord neurons using adeno-associated virus vectors containing cell type-specific promoters. Gene Ther 1997;4: 16–24. 147 Kaspar B, Erickson D, Schaffer D, Hinh L, Gage F, Peterson D: Targeted retrograde gene delivery for neuronal protection. Mol Ther 2002;5:50–56. 148 Mazarakis N, Azzouz M, Rohll F, Ellard F, Wilkes F, Olsen A, Carter E, Barber R, Baban D, Kingsman S, Kingsman A, O’Malley K, Mitrophanous K: Rabies virus glycoprotein pseudotyping


102

149 150

151

152

153

154 155

156

157

158

159 160

161 162 163

164

165

of lentiviral vectors enables retrograde axonal transport and access to the nervous system after peripheral delivery. Hum Mol Genet 2001;10:2109–2121. Tsiang H: Evidence for an intraaxonal transport of fixed and street rabies virus. J Neuropathol Exp Neurol 1979;38:286–299. Acsadi G, Anguelov R, Yang H, Toth G, Thomas R, Jani A, Wang Y, Ianakova E, Mohammad S, Lewis R, Shy M: Increased survival and function of SOD1 mice after glial cell-derived neurotrophic factor gene therapy. Hum Gene Ther 2002;13:1047–1059. Manabe Y, Nagano I, Gazi M, Murakami T, Shiote M, Shoji M, Kitagawa H, Setoguchi Y, Abe K: Adenovirus-mediated gene transfer of glial cell line-derived neurotrophic factor prevents motor neuron loss of transgenic model mice for amyotrophic lateral sclerosis. Apoptosis 2002;7:329–334. Wang L, Lu Y, Muramatsu S, Ikeguchi K, Fujimoto K, Okada T, Mizukami H, Matsushita T, Hanazono Y, Kume A, Nagatsu T, Ozawa K, Nakano I: Neuroprotective effects of glial cell linederived neurotrophic factor mediated by an adeno-associated virus vector in a transgenic animal model of amyotrophic lateral sclerosis. J Neurosci 2002;22:6920–6928. Baumgartner B, Shine H: Neuroprotection of spinal motor neurons following targeted transduction with an adenoviral vector carrying the gene for glial cell line-derived neurotrophic factor. Exp Neurol 1998;153:102–112. Romero M, Rangappa N, Garry M, Smith G: Functional regeneration of chronically injured sensory afferents into adult spinal cord after neurotrophin gene therapy. J Neurosci 2001;21:8408–8416. Rosenberg M, Friedmann T, Robertson R, Tuszynski M, Wolff J, Breakefield X, Gage F: Grafting genetically modified cells to the damaged brain: Restorative effects of NGF expression. Science 1988;242:1575–1578. Tuszynski M, Peterson D, Ray J, Baird A, Nakahara Y, Gage F: Fibroblasts genetically modified to produce nerve growth factor induce robust neuritic ingrowth after grafting to the spinal cord. Exp Neurol 1994;128:1–14. Tuszynski M, Gabriel K, Gage F, Suhr S, Meyer S, Rosetti A: Nerve growth factor delivery by gene transfer induces differential outgrowth of sensory, motor, and noradrenergic neurites after adult spinal cord injury. Exp Neurol 1996; 137:157–173. Ernfors P, Henschen A, Olson L, Persson H: Expression of nerve growth factor receptor mRNA is developmentally regulated and increased after axotomy in rat spinal cord motor neurons. Neuron 1989;2:1605–1613. Grill R, Blesch A, Tuszynski M: Robust growth of chronically injured spinal cord axons induced by grafts of genetically modified NGF-secreting cells. Exp Neurol 1997;148:444–452. Englund U, Ericson C, Rosenblad C, Mandel R, Trono D, Wictorin K, Lundberg C: The use of a recombinant lentiviral vector for ex vivo gene transfer into the rat CNS. Neuroreport 2000;11: 3973–3977. Ericson C, Wictorin K, Lundberg C: Ex vivo and in vitro studies of transgene expression in rat astrocytes transduced with lentiviral vectors. Exp Neurol 2002;173:22–30. Liu Y, Himes B, Tyron B, Moul J, Chow S, Jin H, Murray M, Tessler A, Fischer I: Intraspinal grafting of fibroblasts genetically modified by recombinant adenoviruses. Neuroreport 1998;9:1075–1079. Palmer TD, Rosman G, Osborne W, Miller A: Genetically modified skin fibroblasts persist long after transplantation but gradually inactivate introduced genes. Proc Natl Acad Sci USA 1991; 88:1330–1334. Ruitenberg M, Plant G, Christensen C, Blits B, Niclou S, Harvey A, Boer G, Verhaagen J: Viral vector-mediated gene expression in olfactory ensheathing glia implants in the lesioned rat spinal cord. Gene Ther 2002;9:135–146. Grill R, Murai K, Blesch A, Gage F, Tuszynski M: Cellular delivery of neurotrophin-3 promotes corticospinal axonal growth and partial functional recovery after spinal cord injury. J Neurosci 1997;17:5560–5572. Nicholas M. Boulis, MD Department of Neurosurgery S31, Cleveland Clinic Foundation 9500 Euclid Ave, Cleveland, OH 44195 (USA) Tel. ⫹1 216 444 5188, Fax ⫹1 216 445 1466, E-Mail [email protected]


103


Neural Stem Cell Transplantation for Spinal Cord Repair Akio Iwanami a–c, Yuto Ogawa a–c, Masaya Nakamura b,c, Shinjiro Kaneko a–c, Kazunobu Sawamoto c,d, Hirotaka James Okano a,c, Yoshiaki Toyama b,c, Hideyuki Okano a,c,d Departments of aPhysiology and bOrthopaedic Surgery, Keio University School of Medicine, Shinjuku, Tokyo, cCore Research for Evolutional Science and Technology, Japan Science and Technology Corporation, Osaka and dDepartment of Neuroanatomy, Osaka University Graduate School of Medicine, Suita, Osaka, Japan

Introduction

Ever since the neuroanatomist Cajal [1] reported in the 1920s that the mature central nervous system (CNS) cannot generate new neurons, it was believed that the mammalian CNS does not have the capacity for repair after injury. Nevertheless, in the 1970s, preliminary work in neural transplantation was done which indicated that the neural tissue obtained from fetal rat brain survived, reconstructed neuronal networks, and reversed motor abnormalities when grafted into the animal model of Parkinson’s disease. With the discovery of the potential for functional brain transplantation, interest in neural transplantation escalated sharply and the field of ‘Functional Neurosurgery’ was born. During the 1980s, clinical transplantation trials for Parkinson’s disease attempted to replace dopaminergic neurons with transplants of dopamine-producing cells of various derivations. Clinical application of neural tissue transplantation is still practically limited by the lack of availability of donor fetal brain or spinal cord tissue. In recent years, however, it has become evident that the developing and even the adult mammalian CNS contains self-renewing, multipotent neural stem cells (NSCs), which can be harvested as a source of material for grafting. Recent technological advances developed for the identification, isolation, and expansion of

NSCs raises the enormous potential of therapeutic applications for nervous system disorders [2]. Studies of neural progenitor cells or NSCs are broadly divided into studies on the activation of endogenous NSCs in situ, or studies involving transplantation of NSCs isolated from the brain or spinal cord. At least within the spinal cord, therapeutic strategies using activation of endogenous NSCs are not expected to be practical, because endogenous NSCs appear to proliferate but differentiate only into astrocytes after spinal cord injury (SCI) [3–5]. On the other hand, there are many reports demonstrating the transplantation potential of neural progenitor cells for various CNS disorders or trauma. These cells have not only shown the reconstruction of neuronal networks, but have also rendered functional recovery in animal models. In this chapter, we will introduce recent progress in transplantation, especially as it pertains to practical issues of timing. In addition, we will discuss the future prospects for their clinical application.

Definition and Selective Culture of NSCs

NSCs have been defined as neural cells with the potential to self-renew and generate all three major cell types of the CNS: neurons, astrocytes, and oligodendrocytes. The existence of mammalian NSCs was first suggested by researchers such as Altman, Bayer, Kaplan and others starting in the 1960s, but it was not until the 1990s that these cells were demonstrated in humans and were isolated in culture. Enormous progress has been made in studies on the biological properties of NSCs and their localization in the body over the last decade since Reynolds and Weiss and coworkers [6, 7] developed the neurosphere technique, a selective culture technique for NSCs in 1992. They cultured CNS cell populations including NSCs derived from the mouse embryonic striate body and spinal cord in serum-free culture medium containing insulin, transferrin, selenium, progesterone, and cell division-promoting epidermal growth factor or fibroblast growth factor 2. Although many of these cells could not survive in serum-free medium, the cells surviving in this unusual environment could be grown as floating cell aggregates or ‘neurospheres’. When the neurosphere was dissociated into single cells and cultured in the same medium, they formed the neurospheres again, indicating a self-renewing ability. Moreover, the differentiation into the three neural cell types was driven by growth factor withdrawal, demonstrating multipotency. Thus, the neurosphere technique was shown to expand multipotent, self-renewing NSCs for many passages without apparent phenotypic change (fig. 1) [8]. Subsequently, other culture techniques, such as a low-density monolayer technique, were developed by

Neural Stem Cell Transplantation

105

Neural stem cell

Other cells

Growth factors (EGF, FGF2) Removal of growth factors 1. Self-renewal

Neuron Astrocyte Oligodendrocyte

2. Multipotency

Repeat

Fig. 1. Neurosphere method. The major breakthrough for research on stem cell biology of the CNS was the development of the clonogenic expansion of NSCs in floating culture, called neurosphere culture, within a serum-free medium containing epidermal growth factor EGF and/or fibroblast growth factor 2 or FGF2. A neurosphere derived from a single cell is capable of generating the major three lineages of the CNS, i.e., neurons, astrocytes, and oligodendrocytes, indicating the multipotency of the neurosphere-initiating cell, upon the differentiation assay. If the neurosphere is dissociated into single cells, each cell starts to form a secondary neurosphere again with high frequency. From [6].

Davis and Temple [9] and a high-density monolayer technique was developed by Gage and his coworkers [10]. Selective Markers of NSCs

Even though selective culture methods for NSCs have been established, specific markers for NSCs have not been identified. Instead, highly selective markers of NSCs are known. These include the intermediate filament Nestin [11], the RNA-binding protein Musashi1 [12] identified by our group, and RC2 (i.e., marker of radial glia). These markers are strongly expressed in NSCs; however, they are also expressed in intermediate progenitor cells such as neuronal and glial progenitor cells. Therefore, they are not 100% specific for NSCs. Since cell populations expanded by the neurosphere technique include neural progenitor cells that have differentiated to some degree, it is currently impossible to completely discriminate NSCs from partially differentiated progenitor cells using a positive marker.

Iwanami/Ogawa/Nakamura/Kaneko/Sawamoto/Okano/Toyama/Okano

106

Existence and Localization of NSCs in Adult Mice and Humans

Reynolds and Weiss and coworkers [6, 7, 13] demonstrated the existence of NSCs in the embryonic mouse striate body and spinal cord and successfully cultured them. In 1996, Gritti et al. [14] showed that multipotent, self-renewing NSCs also exist in the adult mouse striatum, indicating that NSCs exist not only in the embryonic but also in the adult mouse brain. Furthermore, Eriksson et al. [15] demonstrated that neurogenesis also occurred in the adult human brain. That study stained postmortem brain samples from cancer patients with 5-bromodeoxyuridine (BrdU) treatment and demonstrated that neurons incorporating BrdU were present in the hippocampal dentate gyrus. Using Nestin and Musashi1 as markers, the collaborative team of Goldman’s group [16] at Cornell University and our own research group showed that NSC-like cells were present around the lateral ventricles of intractable epileptics who had undergone temporal lobectomy. In support of the studies by Eriksson et al. [15], these observations also indicate the existence of NSCs in the adult brain. Their locations correspond to sites of neurogenesis in the granule cell layer of the hippocampal dentate gyrus, the subventricular zone facing the lateral ventricles, and/or the ependymal layer. Recent studies have also suggested the existence of NSCs in the parenchyma of the adult cerebral cortex and spinal cord [10, 17]. Although CNS injury leads to the proliferation of endogenous NSCs, these cells are not usually capable of self-repair. While recent results suggest that forebrain damage due to ischemia could be recovered by activating endogenous NSCs to induce de novo neurogenesis [18, 19], such a strategy has not yet been successful in the injured spinal cord. This failure is presumably because there are few endogenous NSCs in the adult spinal cord or because their differentiation into neuronal cells is inhibited by some mechanism, which remains to be elucidated. Studies are in progress throughout the world in two major areas of research to develop therapeutic strategies for CNS injuries and disorders: first, the activation of endogenous NSCs and second, the transplantation of harvested NSCs. This chapter primarily addresses transplantation therapy using neuronal precursor cells or NSCs.

In situ Identification and Effective Isolation of NSCs

The effective isolation, culture, and expansion of NSCs are essential in considering the clinical application of transplantation therapy, and it is necessary to develop appropriate methods to achieve these purposes. As described above, it is feasible to expand NSCs by the methods represented by the


107

neurosphere technique. However, we can define the neurosphere-initiating cells as NSCs retrospectively only after the formation of the neurosphere. It has been difficult to prospectively identify them in the early stages of culture and impossible to do so in situ. Thus, for experimental purposes we have made transgenic mice that express enhanced green fluorescent protein (EGFP) under the control of the Nestin gene promoter or enhancer to isolate Nestin-positive cells by a fluorescence-activated cell sorter according to the intensity of EGFP expression and fluorescence [20]. We found that the activity of the isolated cells as NSCs correlates well with the intensity of fluorescence; a more intensely fluorescent group of cells had a higher formation rate of neurospheres, showing self-renewing ability and multipotency even in low-density culture. This finding is significant not only because it became possible to prospectively identify NSCs in a ‘living state’ by GFP fluorescence using fluorescence-activated cell sorter, but also because NSCs can be concentrated by this method without using growth factors as in conventional methods [8]. Recently, it is feasible to perform transplantation therapy with a new source of NSCs in place of embryonic tissue transplantation.

Shifting from Neural Tissue Transplantation to Neural Precursor Cell Transplantation

Studies on Transplantation for Parkinson’s Disease Studies on transplantation therapy for CNS disorders have been more advanced for Parkinson’s disease than for other diseases because of the earlier establishment of animal models. In 1979, Björkund and Stenevi [21] reported that rats with experimental Parkinson’s disease recovered from symptoms after transplantation of embryonic rat midbrain tissue into their striata. Later, numerous studies reported results including symptomatic recovery following transplantation of fetal cells of different derivations, and clinical trials also started. In fact, some patients transplanted with fetal tissue have achieved symptomatic relief for more than 10 years and have been demonstrated by PET to have cell transplants functioning effectively [22]. However, fetal tissue transplantation posed many problems such as low engrafting rates and the need for as many as five to ten fetal midbrains for a unilateral striatum transplant. Because of these practical and ethical problems, it was hoped that new donor cells would be developed. Against this background, in 1996, Svendsen et al. [23, 24] transplanted rat neural progenitor cells and human-derived cells shortly afterwards into the striata of model rats with Parkinson’s disease (6-OH-dopamine-administered rats), and reported successful engrafting of the transplants. They reported that although many of the transplanted cells


108

differentiated into glia, a few tyrosine hydroxylase-positive cells were observed, with modest functional improvement. By taking advantage of the NestinEGFP system, we tried to isolate neural progenitor cells from the fetal ventral mesencephalic region. In fact, we obtained a strongly GFP-positive undifferentiated cell population from the fetal ventral mesencephalon of the NestinEGFP mouse by fluorescence-activated cell sorter, which was transplanted into the striatum body of rat models of Parkinson’s disease. We demonstrated the differentiation of the cell transplants into dopaminergic neurons, with recovery from symptoms of Parkinson’s disease [25]. Studies on Transplantation for SCI Experiments on transplantation of nervous system cells for SCI started in 1980 with peripheral nerve transplantation by Aguayo and his coworkers [26]. Then in 1993, Bregman et al. [27] reported that both immature and adult rats in which the thoracic spinal cords had been partially transected and which were transplanted with a fetal spinal cord showed elongation of injured axons with functional recovery, which was predominant in the immature rats. These studies indicated that the introduction of an appropriate environment into the injured site could cause injured axons to regenerate. In addition, other reports described limited spinal cord regeneration including the promotion of the regeneration of injured axons by neurotrophic factors [28] and the identification of axonal growth inhibitors [29]. These studies indicated that regeneration of the injured spinal cord might really be possible. Although researchers first focused on the effectiveness of fetal spinal cord transplantation for SCI [30–32], as with Parkinson’s disease, donor tissue shortage and ethical problems precluded the practical clinical application of this approach. As a result of remarkable advances in neuroscience in recent years, NSCs also have stepped into the limelight as a new transplant material in the field of the spinal cord repair. In 1999, McDonald et al. [33] developed elaborate sequential culture conditions that differentiated mouse ES cells into NSCs in vitro, and transplanted them into the traumatic cavity of rat models of thoracic spinal cord contusion injury. They reported that the engrafted cells could differentiate into neurons, astrocytes, and oligodendrocytes, and that the model rats improved in lower limb motor function to a greater degree than the control group. More recently, Vacanti et al. [34] transplanted gels packed with adult rat-derived NSCs into rat models of thoracic spinal cord transection, with similar results. Our group has looked specifically at time-dependent changes in the cavity microenvironment after SCI and achieved excellent results. On the ninth post-traumatic day during the subacute stage between the immediate post-traumatic stage and the chronic stage during which glial scarring of the injured site progressed, we transplanted fetal rat spinal cord-derived


109

NSCs into the cavity region of rat models of cervical spinal cord contusion injury and observed differentiation of the transplanted cells into neurons, forming synapses with host axons, with recovery of motor function [35, 36] (see Strategies of Transplantation of Fetal Spinal Cord-Derived NSCs for SCI). Other Reports on NSC Transplantation Other studies reported that the transplantation of NSCs into the cerebral ventricle of myelin basic protein-deficient dysmyelinated shiverer mice caused engrafting of myelin basic protein-positive cells [37]. Moreover, transplantation of rat hippocampus-derived NSCs into the growing retina of rats resulted in the appearance of cells expressing molecular markers and having the morphology of Mueller, amacrine, bipolar, horizontal, and photoreceptor cells and astrocytes [38]. These experiments, unlike the above-described ones, have a drawback in that there was no testing of functional improvement. However, based on these results, hopes are mounting for future experiments on their clinical application. Aims of Neural Progenitor Cell Transplantation We consider that the above-described studies on neural precursor cell transplantation had two broad aims: first, to allow NSC transplants to appropriately proliferate and differentiate, replace lost neurons, reform synapses, and induce remyelination; second, to activate endogenous NSCs by the trophic effects of the grafted cell, to induce the differentiation in the desired direction and repair injured neural tissue. As described above, NSCs have now been shown to exist in a number of separate locations, such the fetal and adult brain, spinal cord, and retina [39]. As reported by Kempermann et al. [40], some stimuli appear to activate endogenous NSCs to increase the generation of neurons and glia. Unfortunately, the manner and mechanism of this activation remain to be elucidated. In contrast, the transplantation of exogenous NSCs is aimed at activating endogenous NSCs through neurotrophic factors or some signal to participate in the mechanism of repair and regeneration of lost tissue. In line with these aims, we describe below the present status and future prospects of NSC transplant-based regenerative medicine for the injured spinal cord.

Strategies of Transplantation of Fetal Spinal Cord-Derived NSCs for SCI

Properties of Endogenous NSCs of the Spinal Cord When the spinal cord is injured, Nestin-positive cells, derived from vigorously proliferating cells near the central canal adjacent to the injured site,


110

Spinal cord injury

Reactive astrocytes

Glial scar formation

Cystic formation

Endogenous neural stem cells

Fig. 2. Properties of endogenous NSCs of the spinal cord. Recent studies have shown that there are endogenous NSCs in the adult spinal cord near the central canal. However, these cells differentiate into astrocytes, but not into neurons or oligodendrocytes after SCI.

migrate to the area of the injured site and differentiate into astrocytes [3]. The study of Johansson et al. [4] in 1999 showed the existence of NSCs in the adult spinal cord near the central canal. After SCI, these endogenous NSCs proliferate [5] and differentiate mostly into astrocytes, but not neurons or oligodendrocytes. Since endogenous astrocytes eventually form a glial scar around the wound cavity in a time-dependent manner after injury, the regeneration, elongation and remyelination of damaged axons is entirely disturbed (fig. 2). For repairing injured spinal cord, therefore, neuronal replacement therapy remains the most promising therapeutic strategy. Optimal Timing of NSPCs Transplantation For the purpose of transplantation into the injured spinal cord, we cultured NSCs obtained from 14-day gestational age rat spinal cord, using the neurosphere technique. When these cells were induced to differentiate in vitro, they differentiated into neurons, astrocytes, and oligodendrocytes. About 50% of the cells formed astrocytes and 5% of the total into neurons [Nakamura, pers. commun.]. Thus, it seems highly unlikely that the transplantation of NSCs without any contrivance results in the repair and regeneration of neurons and oligodendrocytes that have been lost through axonotemesis or apoptosis.


111

NSCs obtained from adult CNS showed extensive differentiation potential when transplanted into an adult neurogenic site (i.e., the hippocampal dentate gyrus) [41]. It is well known that the surrounding microenvironment greatly influences the survival and differentiation of engrafted neural progenitor cells. To address the optimal time of transplantation, we investigated the post-traumatic changes of the microenvironment within the injured spinal cord. In the injured spinal cord, the expression of mRNA for various proinflammatory cytokines [e.g., tumor necrosis factor-, interleukin (IL)-1, IL-1, IL-6] peaked 6–12 h after injury and remained elevated until the fourth day [42]. Hart and coworkers [43] have reported the same results. Since these proinflammatory cytokines are known to have biphasic actions, both neurotoxic and neurotrophic, their action within the injured spinal cord requires careful interpretation. Extremely high expression at least within 7 days after injury is thought to be neurotoxic, representing a microenvironment unfit for the survival of NSC transplants. Johe and coworkers [44] have reported that platelet-derived growth factor, ciliary neurotrophic factor, and thyroid hormone (T3) instructively induced the fetal rat hippocampus-derived NSCs to differentiate into neurons, astrocytes, and oligodendrocytes, respectively. Taga and coworkers [45] reported that leukemia inhibitory factor and bone morphogenic protein-2 promote the differentiation of fetal mouse neuroepithelium-derived NSCs into astrocytes. These reports both described members of the IL-6 superfamily (e.g., ciliary neurotrophic factor, leukemia inhibitory factor) and suggest that a signal mediated by the gp130 subunit of the cytokine receptors induces the differentiation of NSCs into astrocytes. During the acute inflammatory phase immediately after SCI, under the condition in which there are high levels of IL-6, NSC transplants are difficult to engraft, but also easy to differentiate into astrocytes if they engraft. We have found that the expression of the anti-inflammatory cytokine transforming growth factor -1 (TGF-1) did not increase immediately after injury, but gradually increased with a peak on the fourth day after injury [42], suggesting that TGF-1 acts to alleviate the inflammatory situation. These observations on the survival and differentiation of NSC transplants indicate that the optimal time of the transplantation is probably not immediately after injury. However, if too much time passes after injury, a glial scar forms around the injured site and inhibits the regeneration of axons; therefore, we considered the optimal time of transplantation to be 7–14 days after trauma (fig. 3). In addition, the benefits of NSC transplantation at this timepoint could also result from microvascular regeneration in the host, considering previous findings from fetal neural tissues transplanted into the cerebral cortex [46, 47]. Correspondingly, a recent report indicates that the formation of new


112

Acute phase Subacute – chronic phase Inflammation SCI


Hemorrhage BBB destruction Infiltration of inflammatory cells


0 1 4

Cystic cavity

7 Pro-inflammatory cytokine Free radical NO Excitatory amino acid

Time

14 windo

w for

trans

plant

ation

?

(Posti njury days) Anti-inflammatory cytokine

Fig. 3. Optimal timing of neural stem/progenitor cells (NSPCs) transplantation based on changes in the microenvironment within the injured spinal cord. We consider the optimal timing of NSPCs transplantation to be 7–14 days after trauma, when the lesion site is neither inflammatory and neurotoxic, nor surrounded by glial scar.

vessels occurs most actively 7–14 days after a contusion injury to the rat spinal cord [48]. Fetal Rat Spinal Cord-Derived Cells for Rat SCI: Delayed Transplantation Based on the above considerations, we made models of quantitative cervical cord contusion injury by performing C4–5 laminectomy on adult rats and allowing a 35-gram weight to stand still on the exposed dura for 15 min. On the ninth day after injury, we transplanted fetal rat spinal cord-derived neural progenitor cells that had been cultured and expanded by the neurosphere technique and labeled with BrdU in and around the injured site (fig. 4). The transplanted cells survived in the host spinal cord and differentiated into neurons, astrocytes, and oligodendrocytes at the 5-week timepoint after transplantation (fig. 5a–c). To investigate the properties of new neurons derived from donors in more detail, we took advantage of the fact that the 1.1-kb promoter element of the T-1 tubulin gene is only active in cells of the neuronal lineage (including neuronal progenitors and postmitotic neurons), and not those of the glial lineage (fig. 6a, b) [49–52]. Here, we used rats that had been treated with transplanted neurospheres derived from the fetal spinal cords (E14.5) of T-1-EYFP transgenic rats. By


113

9 days after injury Fetal spinal cord-derived NSPCs transplantation

Cervical cord of an adult rat

C4 C5

Fig. 4. Delayed transplantation of NSPCs after SCI. We made models of quantitative cervical cord contusion injury by performing C4–5 laminectomy on adult rats and allowing a 35-gram weight to stand still on the exposed dura for 15 min. On the ninth day after injury, we transplanted fetal rat spinal cord-derived NSPCs that had been cultured and expanded by the neurosphere technique and labeled with BrdU, in and around the injured site.

a

b

HU

c

CNP

GFAP

Fig. 5. Survival and differentiation of the engrafted NSPCs. The NSPC transplants had survived in the host spinal cord and differentiated into neurons, astrocytes, and oligodendrocytes at the 5-week timepoint after transplantation. a Hu (neuronal marker) and 5-bromodeoxyuridine (BrdU) double-positive cells (brown: Hu; blue: BrdU). b GFAP (marker of astrocytes) and BrdU double-positive cells (brown: GFAP; blue: BrdU). c CNPase (marker of oligodendrocytes) and BrdU double-positive cells (brown: CNPase; blue: BrdU). Scale bar 5 m.

injecting BrdU intraperitoneally, we could label cells that had divided after the BrdU injection. The presence of postmitotic neurons that were double positive for BrdU-labeling and EYFP expression demonstrated that donor-derived progenitor cells underwent mitotic neurogenesis within the host spinal cord (fig. 6c).


114

T-1-EYFP transgene T-1 promoter

EYFP

Neural progenitor cells derived from T-1-EYFP Tg rats

SV40 polyA

Transplantation

a

1 kb of the 5 flanking region of the rat T-1 tubulin gene SCI rats

Transplantation

-III tubulin

Intraperitoneal injection of BrdU (50g/g B.W.) once a day from 3 to 14 days after transplantation

EYFP

b

c

Fig. 6. The advantage of using T-1-EYFP Tg rats. a k-1 transgene. b The expression of EYFP in the fetal T-1-EYFP Tg rat brain. As the promoter element of T-1 tubulin gene is only active in cells of the neuronal lineage (including neuronal progenitors and postmitotic neurons), and not those of the glial lineage, the EYFP-positive region is coincident with the region stained with -III tubulin (neuronal marker). c The proof of neurogenesis by the transplanted NSPCs. By injecting BrdU intraperitoneally, we could label cells that had divided after the BrdU injection. The presence of postmitotic neurons that were double positive for BrdU labeling and EYFP expression demonstrated that donor-derived progenitor cells underwent mitotic neurogenesis within the host spinal cord.

Five weeks after transplanting neurospheres derived from the fetal spinal cords of T-1-EYFP transgenic rats, donor-derived EYFP-positive neurons extended their axons within the host spinal cord (fig. 7a). We found that T-1-EYFP-positive neurons were surrounded by synaptophysin-immunopositive sites, a presynaptic marker (fig. 7b). Furthermore, we observed EYFPpositive presynaptic structures with presynaptic vesicles that were connected with EYFP-negative postsynaptic structures with postsynaptic densities by immnoelectron microscopic studies (fig. 7c). We also found EYFP-negative presynaptic structures that were connected with EYFP-positive postsynaptic structures. Interestingly, we found some cases in which EYFP-positive neurons had formed a synapse with host motor neurons at the injury site. In addition, compared with the control group (which received an injection of only culture fluid on the ninth day after injury), the transplantation group showed a greater degree of functional recovery as demonstrated by the pellet retrieval test (fig. 8) [32]. These results indicate that if NSCs are transplanted in the subacute phase, neither in the acute phase after SCI nor in the chronic phase characterized by marked glial scarring, they can engraft and contribute to


115

T-1-EYFP

Synaptophysin

a

T-1-EYFP

b

Presynapse

PSD

Presynapse

PSD

c Fig. 7. Neurons derived from transplanted cells extend their axon into the host spinal cord. a Five weeks after transplanting neurospheres derived from the fetal spinal cords of T-1-EYFP transgenic rats, donor-derived EYFP-positive neurons extended their axons within host spinal cord. Scale bar 50 m. b T-1-EYFP-positive neurons were surrounded by synaptophysin-immunopositive sites that are well-characterized pre-synaptic markers. Scale bar 5 m. c EYFP-positive presynaptic structures with presynaptic vesicles that were connected with EYFP-negative postsynaptic structures with postsynaptic densities. Immunoelectron microscopic studies. Scale bar 0.2 m.

some degree to the repair of the injured site. It is important to consider the following three possibilities to explain these data: (1) the transplanted cells may have differentiated into neurons, which formed synapses with neurons above and below the injured site; (2) the transplanted cells may have differentiated into oligodendrocytes, which might have remyelinated the axons that had been demyelinated by the injury; (3) the transplanted cells may have released some neurotrophic factors, which inhibited neuronal death, induced neuronal protection, or activated endogenous NSCs to repair the injured site (fig. 9). The actual functions of the transplanted NSCs are yet to be elucidated and require further study. Bregman and coworkers [53] transplanted fetal rat spinal cord tissue into two groups of rats with SCI immediately after and 2 weeks after injury, and compared the two groups in terms of their anatomical features and the degree of functional recovery. They found that the group receiving transplants 2 weeks after injury showed better regeneration of the injured axons and better lower limb functional recovery compared with the group receiving transplants


116

Number of pellets 90 80 70 60 50 40 30 20 10 0

*

ope ()

SCI

*

*

SCImed SCITP

b

a

Fig. 8. Transplantation of NSPCs improved functional recovery. a Pellet retrieval test. Rats could obtain pellets only with their forelimbs (arrows). b Results of pellet retrieval test. *p 0.01. The p value was determined using a Mann-Whitney U-test.

Synapse formation

a

b

Remyelination of growing axons

c

Trophic support for axonal regeneration

Fig. 9. Mechanism of functional recovery by the grafts. a The transplanted NSPCs differentiated into neurons, which formed synapses with neurons above and below the injured site. b The transplanted NSPCs differentiated into oligodendrocytes, which might have remyelinated the axons that had been demyelinated by the injury. c The transplanted NSPCs released some neurotrophic factors, which inhibited neuronal death, induced neuronal protection, or activated endogenous NSCs to repair the injured site.


117

immediately after injury. These results also strongly support the effectiveness of the delayed transplantation paradigm.

Clinical Applications of Human NSC Transplantation

As described above, transplant experiments with NSCs have been conducted throughout the world. Needless to say, the ultimate goal of these experiments is the human clinical application of NSC transplantation. As the culture techniques for human NSCs become established, transplant experiments in the clinic are under consideration. After the first demonstration by Svendsen et al. of transplantation into Parkinson’s models (see Studies on Transplantation for Parkinson’s Disease), Flax et al. [54] expanded NSCs from the human fetal telencephalon by the neurosphere method, transplanted them into the brains of newborn mice, and reported that they differentiated into neurons, astrocytes, and oligodendrocytes. Brüstle et al. [55] similarly transplanted NSCs cultured from the human fetal brain into the fetal rat cerebral ventricles and reported that differentiation and engrafting occurred in the rat forebrain, midbrain, and hindbrain. In 2001, Ourednik et al. [56] transplanted human fetal NSCs into the fetal monkey cerebral ventricles, and reported that some of the engrafted cells differentiated into neurons and glia, while the remaining cells engrafted and remained undifferentiated. This report holds promise in the sense that the transplantation of human NSCs into primates is essential as a preliminary step toward clinical application. It is hoped that, for SCI, experiments in transplantation of human NSCs in primate models will also yield good results.

Future Prospects

There are still many problems to solve before NSC transplantation finds clinical application. For further improvement of the transplantation therapy, more efficient techniques to isolate NSCs are needed. In contrast to the experiments with mice, for the purpose of clinical application, the collaborative team of Goldman’s group [51, 57] and our group introduced the gene for a fluorescent protein into adult human hippocampal cells by the lipofection method, and successfully isolated human NSCs. We also prepared adenoviruses expressing EGFP under the control of the Nestin enhancer or the Musashi1 promoter for gene transfer, and succeeded in effectively isolating NSCs from human fetal brain tissue [58]. Using the same strategy as for hematopoietic stem cells, an effective method was developed for effectively isolating NSPCs using antibodies to cell surface antigens [59].


118

Another problem is how to further examine the mechanisms of the autoregulation of NSC differentiation. If a favorable environment were created for endogenous NSCs, they might migrate rapidly to injured or degenerated sites to self-repair these sites, and the need for heterologous neural cell transplantation would be reduced. The key to solving this problem may be neurotrophic factors or activation of the immune system. Recently, new concepts on the plasticity of NSCs have emerged. For example, Kondo and Raff [60] have shown that oligodendrocyte precursor cells acquire multipotency similar to NSCs after the manipulation of culture conditions. Clarke et al. [61] have shown that adult ROSA26 mouse-derived NSCs, which have been transplanted into the embryonic chicken amniotic cavity and the mouse blastocyst, differentiate into ectodermal, mesodermal, and endodermal tissues and cells. These observations suggest that NSCs have the potential for differentiation similar to ES cells, and depending on their environment, sometimes differentiate into non-neural cells. On the other hand, other studies reported that non-neural bone marrow stromal cells, which are mesenchymal stem cells, differentiated into neurons in vitro [62], or migrated to the cerebrum and cerebellum to differentiate into astrocytes after transplantation into the neonatal mouse cerebral ventricle [63]. Furthermore, a study has reported that bone marrow stromal cells, which have been transplanted into the injured rat spinal cord one week after injury, bridge the epicenter of the injury in association with immature astrocytes, thus serving as guiding strands for regenerating axons, causing significant functional recovery [64]. More intriguingly, marrow stromal cells, previously thought to differentiate into mesenchymal lineages such as osteocytes, chondrocytes, and adipocytes, could be induced to generate ectoderm-derived CNS cells. However, it remains in doubt whether this so-called transdifferentiation actually occurs constantly in vivo. Many other related problems with SCI remain unsolved. In addition to considering NSC transplantation, it is necessary to create a permissive microenvironment within the site of SCI. This may entail making a biological or other physical scaffold to facilitate the elongation of regenerating axons into the traumatic cavity [65], eliminating axonal growth inhibitors (e.g., Nogo [26], myelin-associated glycoprotein [66], semaphorin [67], chondroitin sulfate) that continue to be released from post-traumatic glial scar tissue, or concomitantly using neurotrophic factors (e.g., neurotrophin-3, brain-derived neurotrophic factor) [68–70] to create a permissive environment for grafting. The most important fundamental problem continues to be how to regenerate the chronically injured spinal cord, in terms of neuronal cell body and axonal growth in patients with existing or long-standing injuries. More than 99% of patients with SCI, numbering more than 100,000 in Japan and almost 250,000


119

in the USA, are patients with long-standing injuries. Without their recovery, there can be no success in the treatment of SCI. References 1 2 3

4 5 6 7 8 9 10

11 12

13

14

15 16

17 18

19 20

Ramon y Cajal S: Degeneration and Regeneration of the Nervous System (Translated by RM Day from the 1913 Spanish edition). Oxford, Oxford University Press, 1928. Okano H: Stem cell biology of the central nervous system. J Neurosci Res 2002;69:698–707. Frisen J, Johansson CB, Torok C, Risling M, Lendahl U: Rapid, widespread, and long-lasting induction of nestin contributes to the generation of glial scar tissue after CNS injury. J Cell Biol 1995;131:453–464. Johansson CB, Momma S, Clarke DL, Risling M, Lendahl U, Frisen J: Identification of a neural stem cell in the adult mammalian central nervous system. Cell 1999;96:26–34. Namiki J, Tator CH: Cell proliferation and nestin expression in the ependyma of the adult rat spinal cord after injury. J Neuropathol Exp Neurol 1999;58:489–498. Reynolds BA, Weiss S: Generation of neurons and astrocytes from isolated cells of the adult mammalian central nervous system. Science 1992;255:1707–1710. Reynolds BA, Tetzlaff W, Weiss S: A multipotent EGF-responsive striatal embryonic progenitor cell produces neurons and astrocytes. J Neurosci 1992;12:4565–4574. Okano H: Neural stem cells: Progression of basic research and perspective for clinical application. Keio J Med 2002;51:115–128. Davis AA, Temple S: A self-renewing multipotential stem cell in embryonic rat cerebral cortex. Nature 1994;372:263–266. Palmer TD, Markakis EA, Willhoite AR, Safar F, Gage FH: Fibroblast growth factor-2 activates a latent neurogenic program in neural stem cells from diverse regions of the adult CNS. J Neurosci 1999;19:8487–8497. Lendahl U, Zimmerman LB, McKay RD: CNS stem cells express a new class of intermediate filament protein. Cell 1990;60:585–595. Sakakibara S, Imai T, Hamaguchi K, Okabe M, Aruga J, Nakajima K, Yasutomi D, Nagata T, KuriharaY, Uesugi S, Miyata T, Ogawa M, Mikoshiba K, Okano H: Mouse-Musashi-1, a neural RNAbinding protein highly enriched in the mammalian CNS stem cell. Dev Biol 1996;176:230–242. Weiss S, Dunne C, Hewson J, Wohl C, Wheatley M, Peterson AC, Reynolds BA: Multipotent CNS stem cells are present in the adult mammalian spinal cord and ventricular neuroaxis. J Neurosci 1996;16:7599–7609. Gritti A, Parati EA, Cova L, Frolichsthal P, Galli R, Wanke E, Faravelli L, Morassutti DJ, Roisen F, Nickel DD, Vescovi AL: Multipotential stem cells from the adult mouse brain proliferate and selfrenew in response to basic fibroblast growth factor. J Neurosci 1996;16:1091–1100. Eriksson PS, Perfilieva E, Bjork-Eriksson T, Alborn AM, Nordborg C, Peterson DA, Gage FH: Neurogenesis in the adult human hippocampus. Nat Med 1998;4:1313–1317. Pincus DW, Keyoung HM, Harrison-Restelli C, Goodman RR, Fraser RA, Edgar M, Sakakibara S, Okano H, Nedergaard M, Goldman SA: Fibroblast growth factor-2/brain-derived neurotrophic factor-associated maturation of new neurons generated from adult human subependymal cells. Ann Neurol 1998;43:576–585. Yamamoto S, Yamamoto N, Kitamura T, Nakamura K, Nakafuku M: Proliferation of parenchymal neural progenitors in response to injury in the adult rat spinal cord. Exp Neurol 2001;172:115–127. Nakatomi H, Kuriu T, Okabe S, Yamamoto S, Hatano O, Kawahara N, Tamura A, Kirino T, Nakafuku M: Regeneration of hippocampal pyramidal neurons after ischemic brain injury by recruitment of endogenous neural progenitors. Cell 2002;110:429–441. Arvidsson A, Collin T, Kirik D, Kokaia Z, Lindvall O: Neuronal replacement from endogenous precursors in the adult brain after stroke. Nat Med 2002;8:963–970. Kawaguchi A, Miyata T, Sawamoto K, Takashita N, Murayama A, Akamatsu W, Ogawa M, Okabe M, Tano Y, Goldman SA, Okano H: Nestin-EGFP transgenic mice: Visualization of the self-renewal and multipotency of CNS stem cells. Mol Cell Neurosci 2001;17:259–273.


120

21 22

23

24

25

26 27

28

29

30 31

32 33

34 35

36 37

38 39 40 41

Björkund A, Stenevi U: Reconstruction of the nigrostriatal dopamine pathway by intracerebral nigral transplants. Brain Res 1979;177:555–560. Piccini P, Brooks DJ, Bjorklund A, Gunn RN, Grasby PM, Rimoldi O, Brundin P, Hagell P, Rehncrona S, Widner H, Lindvall O: Dopamine release from nigral transplants visualized in vivo in a Parkinson’s patient. Nat Neurosci 1999;2:1137–1140. Svendsen CN, Clarke DJ, Rosser AE, Dunnett SB: Survival and differentiation of rat and human epidermal growth factor-responsive precursor cells following grafting into the lesioned adult central nervous system. Exp Neurol 1996;137:376–388. Svendsen CN, Caldwell MA, Shen J, ter Borg MG, Rosser AE, Tyers P, Karmiol S, Dunnett SB: Long-term survival of human central nervous system progenitor cells transplanted into a rat model of Parkinson’s disease. Exp Neurol 1997;148:135–146. Sawamoto K, Nakao N, Kakishita K, Ogawa Y, Toyama Y, Yamamoto A, Yamaguchi M, Mori K, Goldman SA, Itakura T, Okano H: Generation of dopaminergic neurons in the adult brain from mesencephalic precursor cells labeled with a nestin-GFP transgene. J Neurosci 2001;21: 3895–3903. Richardson PM, McGuinness UM, Aguayo AJ: Axons from CNS neurons regenerate into PNS grafts. Nature 1980;284:264–265. Bregman BS, Kunkel-Bagden E, Reier PJ, Dai HN, McAtee M, Gao D: Recovery of function after spinal cord injury: Mechanisms underlying transplant-mediated recovery of function differ after spinal cord injury in newborn and adult rats. Exp Neurol 1993;123:3–16. Cai D, Shen Y, De Bellard M, Tang S, Filbin MT: Prior exposure to neurotrophins blocks inhibition of axonal regeneration by MAG and myelin via a cAMP-dependent mechanism. Neuron 1999;22:89–101. Chen MS, Huber AB, van der Haar ME, Frank M, Schnell L, Spillmann AA, Christ F, Schwab ME: Nogo-A is a myelin-associated neurite outgrowth inhibitor and an antigen for monoclonal antibody IN-1. Nature 2000;403:434–439. Miya D, Giszter S, Mori F, Adipudi V, Tessler A, Murray M: Fetal transplants after the development of function after spinal cord transection in newborn rats. J Neurosci 1997;17:4856–4872. Diener PS, Bregman BS: Fetal spinal cord transplants support the development of target reaching and coordinated postural adjustments after neonatal cervical spinal cord injury. J Neurosci 1998; 18:763–778. Diener PS, Bregman BS: Fetal spinal cord transplants support growth of supraspinal and segmental projections after cervical spinal cord hemisection in the neonatal rat. J Neurosci 1998;18:779–793. McDonald JW, Liu XZ, Qu Y, Liu S, Mickey SK, Turetsky D, Gottlieb DI, Choi DW: Transplanted embryonic stem cells survive, differentiate and promote recovery in injured rat spinal cord. Nat Med 1999;5:1410–1412. Vacanti MP, Leonard JL, Dore B, Bonassar LJ, Cao Y, Stachelek SJ, Vacanti JP, O’Connell F, Yu CS, Farwell AP, Vacanti CA: Tissue-engineered spinal cord. Transplant Proc 2001;33:592–598. Ogawa Y, Sawamoto K, Miyata T, Miyao S, Watanabe M, Toyama Y, Nakamura M, Bregman BS, Koike M, Uchiyama Y, Toyama Y, Okano H: Transplantation of in vitro expanded fetal neural progenitor cells results in neurogenesis and functional recovery after spinal cord contusion injury in rats. J Neurosci Res 2002;69:925–933. Okano H, Ogawa Y, Nakamura M, Kaneko S, Iwanami A, Toyama Y: Transplantation of neural stem cells into the spinal cord after injury. Semin Cell Dev Biol 2003;14:191–198. Yandava BD, Billinghurst LL, Snyder EY: ‘Global’ cell replacement is feasible via neural stem cell transplantation: Evidence from the dysmyelinated shiverer mouse brain. Proc Natl Acad Sci USA 1999;96:7029–7034. Takahashi M, Palmer TD, Takahashi J, Gage FH: Widespread integration and survival of adultderived neural progenitor cells in the developing optic retina. Mol Cell Neurosci 1998;12:340–348. Tropepe V, Coles BL, Chiasson BJ, Horsford DJ, Elia AJ, McInnes RR, van der Kooy D: Retinal stem cells in the adult mammalian eye. Science 2000;287:2032–2036. Kempermann G, Kuhn HG, Gage FH: More hippocampal neurons in adult mice living in an enriched environment. Nature 1997;386:493–495. Shihabuddin LS, Horner PJ, Ray J, Gage FH: Adult spinal cord stem cells generate neurons after transplantation in the adult dentate gyrus. J Neurosci 2000;20:8727–8735.


121

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

Nakamura M, Houghtling RA, MacArthur L, Bayer BM, Bregman BS: Differences in cytokine gene expression profile between acute and secondary injury in adult rat spinal cord. Exp Neurol 2003;184:313–325. Pan JZ, Ni L, Sodhi A, Aguanno A, Young W, Hart RP: Cytokine activity contributes to induction of inflammatory cytokine mRNAs in spinal cord following contusion. J Neurosci Res 2002;68: 315–322. Johe KK, Hazel TG, Muller T, Dugich-Djordjevic MM, McKay RD: Single factors direct the differentiation of stem cells from the fetal and adult central nervous system. Genes Dev 1996;10: 3129–3140. Nakashima K, Yanagisawa M, Arakawa H, Kimura N, Hisatsune T, Kawabata M, Miyazono K, Taga T: Synergistic signaling in fetal brain by STAT3-Smad1 complex bridged by p300. Science 1999;284:479–482. Miyoshi Y, Date I, Ohmoto T: Three-dimensional morphological study of microvascular regeneration in cavity wall of the rat cerebral cortex using the scanning electron microscope: Implications for delayed neural grafting into brain cavities. Exp Neurol 1995;131:69–82. Miyoshi Y, Date I, Ohmoto T: Neovascularization of rat fetal neocortical grafts transplanted into a previously prepared cavity in the cerebral cortex: A three-dimensional morphological study using the scanning electron microscope. Brain Res 1995;681:131–140. Casella GT, Marcillo A, Bunge MB, Wood PM: New vascular tissue rapidly replaces neural parenchyma and vessels destroyed by a contusion injury to the rat spinal cord. Exp Neurol 2002; 173:63–76. Gloster A, Wu W, Speelman A, Weiss S, Causing C, Pozniak C, Reynolds B, Chang E, Toma JG, Miller FD: The T-1-tubulin promoter specifies gene expression as a function of neuronal growth and regeneration in transgenic mice. J Neurosci 1994;14:7319–7330. Wang S, Wu H, Jiang J, Delohery TM, Isdell F, Goldman SA: Isolation of neuronal precursors by sorting embryonic forebrain transfected with GFP regulated by the T-1 tubulin promoter. Nat Biotechnol 1998;16:196–201. Roy NS, Wang S, Jiang L, Kang J, Restelli C, Fraser RAR, Couldwell WT, Kawaguchi A, Okano H, Nedergaard M, Goldman S: In vitro neurogenesis by neural progenitor cells isolated from the adult human hippocampus. Nat Med 2000;6:271–278. Sawamoto K, Yamamoto A, Kawaguchi A, Yamaguchi M, Mori K, Goldman SA, Okano H: Visualization and direct isolation of neuronal progenitor cells by dual-color flow cytometric detection of fluorescent proteins. J Neurosci Res 2001;65:220–227. Coumans JV, Lin TT, Dai HN, MacArthur L, McAtee M, Nash C, Bregman BS: Axonal regeneration and functional recovery after complete spinal cord transection in rats by delayed treatment with transplants and neurotrophins. J Neurosci 2001;21:9334–9344. Flax JD, Aurora S, Yang C, Simonin C, Wills AM, Billinghurst LL, Jendoubi M, Sidman RL, Wolfe JH, Kim SU, Snyder EY: Engraftable human neural stem cells respond to developmental cues, replace neurons, and express foreign genes. Nat Biotechnol 1998;16:1033–1039. Brustle O, Choudhary K, Karram K, Huttner A, Murray K, Dubois-Dalcq M, McKay RD: Chimeric brains generated by intraventricular transplantation of fetal human brain cells into embryonic rats. Nat Biotechnol 1998;16:1040–1044. Ourednik V, Ourednik J, Flax JD, Zawada WM, Hutt C, Yang C, Park KI, Kim SU, Sidman RL, Freed CR, Snyder EY: Segregation of human neural stem cells in the developing primate forebrain. Science 2001;293:1820–1824. Roy NS, Benraiss A, Wang S, Fraser RA, Goodman R, Couldwell WT, Nedergaard M, Kawaguchi A, Okano H, Goldman SA: Promoter-targeted selection and isolation of neural progenitor cells from the adult human ventricular zone. J Neurosci Res 2000;59:321–331. Keyoung HM, Roy NS, Benraiss A, Louissaint A Jr, Suzuki A, Hashimoto M, Rashbaum WK, Okano H, Goldman SA: High-yield selection and extraction of two promoter-defined phenotypes of neural stem cells from the fetal human brain. Nat Biotechnol 2001;19:843–850. Uchida N, Buck DW, He D, Reitsma MJ, Masek M, Phan TV, Tsukamoto AS, Gage FH, Weissman IL: Direct isolation of human central nervous system stem cells. Proc Natl Acad Sci USA 2000;97:14720–14725.


122

60 61 62 63

64

65

66

67

68

69

70

Kondo T, Raff M: Oligodendrocyte precursor cells reprogrammed to become multipotential CNS stem cells. Science 2000;289:1754–1757. Clarke DL, Johansson CB, Wilbertz J, Veress B, Nilsson E, Karlstrom H, Lendahl U, Frisen J: Generalized potential of adult neural stem cells. Science 2000;288:1660–1663. Woodbury D, Schwarz EJ, Prockop DJ, Black IB: Adult rat and human bone marrow stromal cells differentiate into neurons. J Neurosci Res 2000;61:364–370. Kopen GC, Prockop DJ, Phinney DG: Marrow stromal cells migrate throughout forebrain and cerebellum, and they differentiate into astrocytes after injection into neonatal mouse brains. Proc Natl Acad Sci USA 1999;96:10711–10716. Hofstetter CP, Schwarz EJ, Hess D, Widenfalk J, El Manira A, Prockop DJ, Olson L: Marrow stromal cells form guiding strands in the injured spinal cord and promote recovery. Proc Natl Acad Sci USA 2002;99:2199–2204. Teng YD, Lavik EB, Qu X, Park KI, Ourednik J, Zurakowski D, Langer R, Snyder EY: Functional recovery following traumatic spinal cord injury mediated by a unique polymer scaffold seeded with neural stem cells. Proc Natl Acad Sci USA 2002;99:3024–3029. Cai D, Deng K, Mellado W, Lee J, Ratan R, Filbin M: Arginase I and polyamines act downstream from cyclic AMP in overcoming inhibition of axonal growth MAG and myelin in vitro. Neuron 2002;35:711–719. De Winter F, Oudega M, Lankhorst AJ, Hamers FP, Blits B, Ruitenberg MJ, Pasterkamp RJ, Gispen WH, Verhaagen J: Injury-induced class 3 semaphorin expression in the rat spinal cord. Exp Neurol 2002;175:61–75. Olson L, Widenfalk J, Josephson A, Greitz D, Klason T, Kiyotani T, Lipson A, Ebendal T, Cao Y, Hostetter C, Schwartz E, Prockop D, Manson S, Jurban M, Lindqvist E, Lundströmer K, Nosrat C, Brene S, Spenger C: Experimental spinal cord injury models: Prospective and repair strategies; in Ikada Y, Oshima N (eds): Tissue Engineering for Therapeutic Use, ed 5. Amsterdam, Elsevier Science BV, 2001, pp 21–36. Grill R, Murai K, Blesch A, Gage FH, Tuszynski MH: Cellular delivery of neurotrophin-3 promotes corticospinal axonal growth and partial functional recovery after spinal cord injury. J Neurosci 1997;17:5560–5572. Liu Y, Kim D, Himes BT, Chow SY, Schallert T, Murray M, Tessler A, Fischer I: Transplants of fibroblasts genetically modified to express BDNF promote regeneration of adult rat rubrospinal axons and recovery of forelimb function. J Neurosci 1999;19:4370–4387.

Dr. Hideyuki Okano Department of Physiology, Keio University School of Medicine, Shinjuku Tokyo 160-8582 (Japan) Tel. 81 3 5363 3747, Fax 81 3 3357 5445, E-Mail [email protected]


123

Functional and Restorative Molecular Neurosurgery Freese A, Simeone FA, Leone P, Janson C (eds): Principles of Molecular Neurosurgery. Prog Neurol Surg. Basel, Karger, 2005, vol 18, pp 124–145

Contemporary Applications of Functional and Stereotactic Techniques for Molecular Neurosurgery Paul A. House, Ganesh Rao, William Couldwell Department of Neurological Surgery, University of Utah Medical Center, Salt Lake City, Utah, USA

Introduction

Tremendous advances have been made in understanding the molecular basis of many neurological diseases. Although the molecular biology of brain tumors and neurodegenerative diseases has become better understood, utilizing this information to achieve improved therapeutic results remains a challenge. In some neurological diseases, the dysfunction of specific neuroanatomic sites is primarily responsible for a disease process, for example, degeneration of dopaminergic neurons of the substantia nigra pars compacta in Parkinson’s disease. Other diseases are more diffuse; for example, a glioblastoma multiforme (GBM) may have tumor cells several centimeters away from the primary tumor focus. Each of these scenarios calls for a different type of treatment approach, either delivering therapy locally or diffusely. In this chapter, we discuss strategies that are currently under investigation for delivery of drug or molecular-cellular treatments.

Basic Treatment Approaches to Neurological Disease

Neuro-Oncology Treatment of brain tumors typically involves some combination of surgical resection, radiotherapy, and chemotherapy. Surgical resection has been shown to improve survival in certain tumors such as GBM, whereas others are definitively treated by radiation, such as germinoma. Still others, such as

oligodendroglioma, respond very favorably to chemotherapy. Despite advances in these classes of treatment, the survival for patients with most primary brain tumors remains poor. The survival rate for GBM, the most common primary brain tumor, has not improved in over a decade. Improved strategies for treating GBM must address the diffuse nature of intrinsic brain tumors. Tumor cells spread widely along white matter tracts and can be found on the contralateral hemisphere from a primary tumor focus. Definitive therapy for primary brain tumors will require treatments such as well-targeted inactivation of aberrantly expressed oncogenes or re-establishment of the activity of lost or nonfunctioning tumor suppressor genes. Neurodegenerative Diseases Neurodegenerative diseases such as Alzheimer’s disease involve a continuous loss of neurons. In Alzheimer’s disease, diffuse loss of neurons in the cortex as well as basal structures such as the locus ceruleus and nucleus basalis is characteristic. The same widespread loss of neurons holds true for Parkinson’s disease, in which dopaminergic neurons are depleted, and Huntington’s disease in which striatal neurons are lost. In the case of these disorders, dysfunction of specific neuroanatomic structures must be addressed. Other neurodegenerative diseases such as amyotrophic lateral sclerosis involve degeneration of anterior horn cells throughout the spinal cord, providing a unique therapeutic challenge. Definitive therapy for all of these disorders is likely to involve grafting of cells to restore function, along with approaches to deliver local trophic and growth factors. Spinal Cord Injury Molecular underpinnings of the normal healing process following spinal cord injury suggest that a variety of steps in the healing cascade may be amenable to intervention. While some forms of intervention for spinal cord injury, such as corticosteroids, can be delivered systemically, future therapies will involve proteins and small molecules that need to be delivered locally. Although not classically considered in the realm of functional neurosurgery, the need for targeted local therapy in the spinal cord may expand the role of the functional neurosurgeon. At least three types of axon-inhibiting molecules present in the myelin of the injured spinal cord have now been characterized. Two separate types of myelin-associated glycoproteins named Nogo and MAG, along with oligodendrocyte-myelin glycoprotein, have been shown to inhibit growth cones [1–3]. It has been demonstrated that blocking Nogo-A with a monoclonal antibody (IN-1) leads to enhanced long fiber tract regeneration in spinal cord injury models [4]. Because diffusely blocking Nogo-A leads to sprouting of uninjured

Techniques for Molecular Neurosurgery

125

axons in the intact central nervous system (CNS) [5], it is likely that blocking antibodies will need to be delivered in a focused manner to patients with injured spinal cord to avoid diffuse axonal sprouting. Similarly, blocking the receptors to these glycoproteins or blocking the signaling cascade they induce will likely be treated best in a focused manner. Neurosurgeons will be required to develop and implement these technologies in the clinic.

New Approaches to Drug Delivery in the Brain

Gene Therapy Advances in understanding the genetic aberrations leading to GBM have provided new therapeutic targets. For example, loss of the tumor suppressor phosphatase and tensin homolog has been shown to be an important event in the development of GBM [6–9]. Restoration of the function of such tumor suppressor genes has shown promise in vitro. Typically, these therapeutic genes are delivered via adenoviral or retroviral vectors. Transduction of functional phosphatase and tensin homolog into glioma cell lines has been shown to reduce the proliferation of tumor cells [10]. Other successes have been obtained by introducing herpes simplex virus type 1 thymidine kinase into glioma cell lines and inducing cytotoxicity with gancyclovir or other prodrugs [10–15]. Although many promising therapies have been successful in the laboratory, improvement in patients has been significantly less dramatic. Translating in vitro successes to human patients remains a challenge and there are risks inherent to this type of treatment [16–18]. Antisense gene therapy is being developed for therapy in which overexpression of cancer-promoting genes (e.g., oncogenes) plays a role in tumor progression. The simplest antisense constructs utilize an oligonucleotide sequence in complementary orientation to a target gene, and the antisense cDNA binds to a target DNA or mRNA and prevents transcription or translation. Antisense therapy has been used on glioma targets with some success. For example, matrix metalloproteinase-9 (MMP-9) has been shown to be important for cell migration and invasion of gliomas. Antisense constructs targeted against MMP-9 in both in vitro and in vivo models have shown regression of tumor growth [19]. Similar preclinical results have been obtained with the delivery of antisense constructs directed against epidermal growth factor receptor gene, which is upregulated in gliomas. Another new antisense therapy involves the use of short interfering RNA, which can bind to mRNA and cause degradation, known as RNA interference [20]. The major obstacle to gene therapy relates to inefficient delivery to the CNS. While many gene therapy techniques utilizing adenoviral or retroviral

House/Rao/Couldwell

126

Fig. 1. Localization of GPi with CED.

vectors have been successful in cell lines, targeted delivery in the clinical setting remains quite difficult. The past failures of cancer gene therapy are mainly due to the poor delivery of the gene to tumor cells, and the method of manual injection of vector-producing cells limits the distribution of these cells [14]. The development of newer vectors such as recombinant adeno-associated virus (AAV) has provided hope for in vivo delivery. AAV is based on a nonpathogenic, replication-defective virus and has been used successfully for efficient and sustained gene transfer to proliferating and differentiated cells without a detectable immune response or toxicity [21–23]. AAV has been shown to be effective for long-term delivery of genes at biologically relevant levels in both the CNS and intramuscularly [22, 24–27]. There may be limitations to this vector, although it has safety advantages over other adenoviral or retroviral vectors [22, 23, 28]. Convection-Enhanced Delivery One of the more promising techniques for clinical drug delivery to the brain is convection-enhanced delivery (CED) or high-flow microinfusion. CED involves placement of an infusion catheter directly into the brain parenchyma and relies on bulk flow (as opposed to passive diffusion) through the CNS parenchyma, thus bypassing the blood brain barrier. Although technical issues remain (e.g., cannula size, location of the infusion pump, cellular damage caused by high flow rates), there are distinct advantages over conventional


127

Fig. 2. Targeted brainstem delivery with CED.

intravenous chemotherapy, intraventricular delivery, or drug-impregnated polymer-based therapy. CED allows for significantly higher concentrations of pharmacotherapy to be delivered to a larger volume of brain tissue, with important applications in neurodegenerative and neuro-oncological diseases. Stereotactically implanted catheters may be targeted at specific structures. For example, high-flow microinfusion of the caudate with biotinylated dextran has been performed successfully with very little spillover into the adjacent structures [29]. Similar perfusion of the globus pallidus has also been achieved (fig. 1) [30]. Indirect targeting through axonal tracts also has been shown by infusion of the striatum in rats, with subsequent identification of the infusate in the substantia nigra pars compacta [29]. These findings have therapeutic implications for the treatment of neurodegenerative diseases such as Parkinson’s or Huntington’s. CED also shows promise for diffuse neoplastic disease as there is spread of the infusate along white matter tracts. It can be used relatively safely in the cerebral hemispheres, brainstem, and spinal cord (fig. 2) [31]. Preclinical testing has demonstrated a survival advantage in C6 glioma-bearing rats treated with BCNU or toptecan delivered via CED [32, 33]. This observation has been extended to human trials, which take advantage of the overexpression of transferrin, interleukin-13, or interleukin-14 receptors on glioma cells. By linking a toxic compound (such as Pseudomonas exotoxin) to a ligand specific for these receptors, glioma cells can be targeted for specific destruction [34, 35]. CED lends itself nicely to this technique, and recent trials have shown some promise for this route of delivery.

House/Rao/Couldwell

128

Stem Cell Therapy Stem cell therapy has been touted as a potential treatment for neurodegenerative diseases such as Parkinson’s disease, Huntington’s disease, and Alzheimer’s disease. It is now well established that neural stem cells possess the ability to differentiate into any of the various cell types in the CNS. Transplantation of fetal dopamine cells into rat models of Parkinson’s disease was successfully started decades ago [36–39]. These experiments showed that transplanted dopaminergic cells could survive and function in vivo. Fetal cell transplantation has been performed in humans with some success, and in some cases caused a huge improvement in Parkinsonian symptoms [40]. There are, however, serious difficulties with current fetal dopamine cell transplants. First, recovery of these cells from aborted fetuses is expensive and although it is legal, there are ethical considerations in ramping up production of these cells from primary sources. Further, the survival rate of these transplants can be quite poor with the majority of cells undergoing apoptosis in the absence of immunosuppression. Embryonic stem cells (ES cells) offer a more attractive source of dopaminergic cells, as these are available from any number of established ES cell lines and could be genetically engineered to match a host through new techniques of somatic cell nuclear transfer. Transplantation of ES cells is still problematic regarding control of cell growth and differentiation, as well as having a sufficient quantity of cells to transplant. In rat models of Parkinson’s disease many ES cells will not survive in situ, and up to 20% will differentiate into lethal teratomas [41]. These problems are being addressed by allowing some cellular differentiation to occur in vitro prior to transplantation. Investigators are currently attempting to develop cell lines of dopamine-producing ES cells using gene transfer with prodopaminergic genes (e.g., Nurr1) or treatment with soluble signaling factors (e.g., epidermal growth factor, insulin-like growth factor). Also, it may be possible to develop specific neuronal cells from other stem cells located elsewhere in the body (e.g., blood, bone marrow, skin) for purposes of autotransplantation. Growth Factors as a Therapeutic Strategy Growth factors used to promote functional restoration of neurons that are affected in neurodegenerative diseases include glial-derived neurotrophic factor (GDNF), ciliary neurotrophic factor, brain-derived growth factor, and insulinlike growth factor-1. GDNF in particular has shown clinical promise for neural repair. Described as the most potent neurotrophic factor for motoneurons, it was tested heavily beginning in the mid 1990s. In vitro studies showed that GDNF promoted the survival of motoneurons [42–44]. However, because of difficulties with delivery, short half-life of the recombinant protein, and various inflammatory


129

effects, clinical trials met with poor success and side effects [22, 45]. The combination of growth factor genes and viral vectors such as AAV has restored hope for its use in neurodegenerative disorders. For example, an AAV-GDNF construct injected intramuscularly into mice has been shown to result in sustained expression of transgenic GDNF and is delivered via retrograde transport to spinal motoneurons [22]. This type of therapy holds promise for the treatment of motor neuron diseases such as amyotrophic lateral sclerosis, because GDNF expression in the muscles of transgenic amyotrophic lateral sclerosis mice has improved their survival [24]. Molecular Therapies for Spinal Cord Injury Neurotrophins have been investigated for their ability to allow regenerating axons to cross the area of an injured spinal cord. For example, neurotrophin-3 promotes axon sprouting through the gray matter in lesioned spinal cords when delivered continuously via a fibroblast-producing cell line [46]. Functional neurosurgeons will need to become involved in the development of new drug delivery systems to provide such sustained levels of neurotrophins to the sites of injury. As with the blocking of inhibiting epitopes in the injured spinal cord, delivery of neurotrophins to the injured spinal cord could possibly utilize CED [47]. Osmotic pumps have been used in experimental animals to deliver some of these small molecules and could perhaps be adapted for the purpose in human subjects. A variety of biomaterials are being developed that might not only deliver the needed concentration gradients but also provide a permissive substrate for axon regrowth [48]. The transplantation of stem cells of multiple lineages also holds promise in providing permissive microenvironments for spinal cord regeneration.

Novel Surgical Techniques for Spinal Cord Injury

While a progressive understanding of the molecular biology of spinal cord injury will provide new avenues to aid recently injured patients, those with preexisting spinal cord deterioration suffer from a host of secondary complications for which neuro-augmentative surgery could provide functional improvement. Many augmentative technologies and techniques would also benefit patients with progressive degenerative disease. For example, loss of bowel and bladder control is often cited as one of the most disabling complications of diplegia or tetraplegia. Indeed, chronic hydronephrosis and secondary infections are often the ultimate cause of death for those who are disabled. Anterior sacral root stimulation combined with dorsal rhizotomy to treat the neurogenic bladder is the most widely used neurosurgical method employing a neuroprosthetic device to

House/Rao/Couldwell

130

augment bladder function. This implantation technique yields full continence in the majority of patients [49] and has been found to be cost effective [50]. This system has limitations, however, including side effects on sexual function, prompting further investigations into mechanisms that allow specific stimulation of coordinated bladder emptying as well as continence. In the decerebrate cat, stimulation of the dorsolateral funiculus within the lower thoracic spinal cord (T9–T13) has been shown to produce coordinated bladder contraction with decreased urethral sphincter tone [51]. If a coordinated control center can be found in the human spinal cord, it may be possible to produce a microelectrode system for bladder control. The functional neurosurgeon, already adept at microelectrode recording and targeting, will be needed to help overcome technical hurdles and make such a system possible. Although therapy to treat sequelae of spinal cord injury may seem like a small goal compared to the ultimate goal of restoring total spinal cord function, it may be a more achievable goal and would definitely enhance quality of life until true spinal cord repair is possible, which may take decades of further work and refinement. Advancing bionic technologies also present an area where neurosurgical expertise could lead to the development of new human-prosthetic interfaces, which would greatly augment the functional capacities of those with disabilities. Commercial examples of advancements in this area include the ‘iBot Mobility System’ (Independence Technology, LLC) which incorporates gyroscopic guidance into a wheelchair design, allowing the system to climb stairs and balance on two wheels. A similar capability-expanding peripheral device in development is called ‘Robowalker’ (Yobotics, Inc.). This exoskeleton-type device could greatly enhance the mobility of those with lower extremity weakness or limited leg control. These technologies demonstrate how sophisticated movements can be controlled using only a limited amount of input information, namely, the leaning body weight of a patient lacking full mobility. Other devices, which rely on tracking eye movement to initate and control movements, are in development for the patient with spinal cord injury.

Neuro-Prosthetic Therapies

For patients with severe neurodegenerative disorders or severe CNS injury, one factor limiting the utilization of new forms of augmentative technology is the difficulty in providing communication between the device and the injured CNS. Providing simple two-dimensional directional control has been achieved by direct implantation of a microelectrode into the human motor cortex (fig. 3) [52]. This system provides a brain-computer interface by having the computer learn to recognize firing patterns of motor cortex associated


131

I II III

IVA IVB IVC ␣ ␤ V VI

2 mm Fig. 3. One device which has been designed to serve as a brain-machine interface is the Utah Microelectrode Array. A 100 electrode array is pictured above next to a penny for size comparison. The electrode tips have been machined to penetrate neocortex to layer IV. An electron micrograph of the 4 ⫻ 4 mm array is below.

House/Rao/Couldwell

132

with direction-specific movement. A similar system utilizing a vastly increased number of microelectrodes allowed a monkey to remotely control a robot arm in a three-dimensional reaching task [53]. These experiments outline a strategy that may be employed in the future to extend cognitive control of artificial limbs or mobility-extending devices to patients with very high cervical cord injuries. A host of technical difficulties must still be overcome to allow even rudimentary cognitive control of artificial limbs, but there is certainly reason for hope [54]. Current technology allows the simultaneous recording of approximately one hundred neurons through implanted microelectrodes or microelectrode arrays [55]. Perhaps surprisingly, real-time analysis of only fifty to one hundred motoneurons was sufficient to reproduce the three-dimensional arm movements previously discussed. The first devices capable of directly recording electrical information from the brain were developed in the 1950s as external electroencephalography recorders. These devices provide limited spatial resolution but are noninvasive. Through ‘bio-feedback’, patients can operate simple one-variable devices in controlled situations. The signal resolution from surface recordings, however, is too limited to provide driving information for even the simplest artificial limb system. Subdural electrode grids, already widely used as monitoring devices in the evaluation of epilepsy, deliver finer spatial resolution than superficial devices. The possibility of implanting subdural grids was a strategy utilized to provide the first cortical stimulation systems designed to deliver visual information to the blind [56]. Such systems are able to provide enhanced communication with the CNS compared with surface EEG. Yet, each subdural electrode is affected by many thousands of neurons and spatial resolution is still too limited to drive most useful artificial limb systems. One benefit of limited resolution is the need to transfer only a limited amount of information to a recording/interpreting computer. Progress on telemetry systems now allows continuous neuronal recordings to be obtained from an entirely implanted system, although limited to only two electrodes [57]. Specially designed integrated circuits and telemetry devices continue to be developed, but so far, such devices are not able to provide continuous telemetry of implanted multielectrode subdural systems. To provide the cortical spatial resolution needed to drive an artificial limb or replace sensory information in a detailed manner, several varieties of highdensity microelectrode systems are in development. The Utah microelectrode array provides 100 electrode contacts at 400-micron spacing [58]. A microelectrode system developed at the University of Michigan also provides multiple sites of electrode contacts along each electrode [59]. Each system utilizes silicon semiconductor fabrication processes producing a robust interface with high biocompatibility.


133

As the ability to communicate with the CNS becomes progressively refined, a host of prosthetic devices will be introduced for clinical application. The cochlear implant system is a good example of a device that has been developed in the past to restore a sense in a functional way [60]. Prosthetics designed to restore vision to the blind are in development and often receive much attention from the popular press [61]. It may also become possible to restore somatosensory function to those utilizing artificial limbs. Interpreting neuronal coding signals offers hope to those with little ability to control their bodies or outside situations. Another application of these devices is implantable stimulators, which have been designed to abort epileptic activity [62]. It may even become possible to artificially replicate portions of the CNS function, such as hippocampal input [63], through implantable/programmable neurostimulator devices, or to simulate axonal connections between brain regions with coordinated stimulators in more than one brain region.

Intraoperative Navigation and Imaging

The application of intraoperative navigation has continued to evolve with new technology. Routine intraoperative navigation is now employed at most surgical centers, and many systems are commercially available. Most of these systems utilize archived data sets (CT/MRI/angiography) to provide target localization. The most important advances made over the past decade have been with the application of intraoperative imaging to provide real-time feedback to the operating surgeon. This is especially important for those instances in which shifts of important structures occur which render archived data inaccurate, such as in the resection of large intra-axial tumors. Refinements in real-time imaging of intracranial tumors are valuable to neurosurgeons in maximizing resections in a safe manner. The most contemporary method of imaging refinement is in the application of CT [64] and MRI to the operating room environment. The superior soft-tissue resolution of MRI over CT has made it the preferred intraoperative imaging machine in most institutions. The advantages of intraoperative MRI (iMRI) will be likely to make this an ubiquitous feature in neurosurgical operating rooms. Since their introduction into surgical practice in the mid 1990s, iMRI systems have allowed the delineation of the lesion, including ‘under the surface’ vision, and obtained real-time feedback of the extent of resection and the position of any residual tumor tissue. High-performance computing has extended the capabilities of iMRI with multimodal information and three-dimensional reconstructions [65]. One of the major issues surrounding the use of intraoperative magnets is the safety and ease-of-use considerations for the surgeon, nurses,

House/Rao/Couldwell

134

anesthesiologist, and patient. Safe working environment demands the use of MR-compatible instruments, head holders, and anesthesia equipment with most machines. These safety issues have been well reviewed recently by Russell [66]. Improvements in Surgical Outcomes In contemporary series, both low- and high-field iMRI have had a positive impact on patient care, maximizing tumor resection, and shortening length of stay. In a report by Schulder and Carmel [67], iMRI-guided resection of tumors in 112 patients resulted in additional tumor removal in 36%. In another 31%, imaging confirmed that the goals of surgery had been attained, so potentially harmful further dissection in and around the brain was avoided. iMRI offers the possibility of further tumor removal during the same surgical procedure in case of tumor remnants, increasing the rate of complete tumor removal. The effects of brain shift can be compensated by using intraoperative imaging data for updating. This capability is especially important in cases involving intrinsic tumor surgery (especially low-grade tumors), and in skull-base tumors in which direct surgical view is not possible (e.g., large pituitary tumors with suprasellar extension). Most current systems combine the advantages of intracranial computerassisted cranial navigation with real-time or intermittent intraoperative imaging to verify location and tumor resection status. Various other indications for the use of iMRI as a surgical adjunct include iMRI-guided instillation of phosphorous-32 for cystic craniopharyngiomas [68], monitoring resection of epilepsy foci, and resection of vascular lesions (AVM and cavernous malformations). Combining iMRI with a comprehensive neuronavigation environment with the use of ultrasound, cortical stimulation, and navigation system-guidance of biopsy probes, instruments, and endoscopy has been described [69]. iMRI has been used with planned adjuvant radiosurgical treatment [70]. The emerging use of combining functional MRI with diffusion-weighted imaging to provide the anatomical detail of cortical and subcortical white matter tracts will enhance safe and complete resections of tumors adjacent to eloquent regions of the brain. Costs of Imaging Technology A concern of many centers is the cost involved with the establishment of an iMRI program. The Department of Neurosurgery at the University of Minnesota recently published a retrospective cost comparison of the costs and benefits of brain tumor resection in a conventional operating room and those associated with the iMRI suite [71]. A comparison of the length of stay,


135

Fig. 4. The low-field Odin 0.12 T Polestar system. This unit may be swung into the operative field at any time imaging is required. Photo courtesy of Odin company.

hospital charges and payments, hospital direct and indirect costs, readmission rates, repeat resection interval, and net health outcome was performed between the patients cared for in the two operating environments. The authors noted a reduced length of stay, reduced repeat-resection interval, and reduced hospital charges and costs. Other centers, such as the University of Cincinnati College of Medicine in Ohio, have utilized a shared-resource MRI, in which the suite functions to provide both neurosurgical and diagnostic procedures in a single unit [72]. The open low-field (0.3 T) Hitachi unit is used for diagnostic studies when not being used for neurosurgical cases. The ability to perform diagnostic procedures in a shared unit has been a cost-effective solution for this particular institution. iMRI System Options There are several different options for iMRI application in the contemporary operating room environment. These include the use of low- or high-magnetic field strength units. There are also different solutions to the layout of the operating room and the concessions made to be able to image in the OR environment. The most common iMRI systems are designated low-field systems. Systems include a Siemens (Erlangen, Germany) 0.2-Tesla Magnetom Open unit [73, 74] or Odin 0.12-Tesla Polestar system [67, 69, 75, 76] (fig. 4). These systems have

House/Rao/Couldwell

136

Fig. 5. High-field intraoperative MRI machine in use at the University of Calgary is a ceiling-mounted system on rails. Photo courtesy of University of Calgary, Department of Neurosurgery.

gained widespread use and experience in multiple centers, largely for reasons of reduced cost and ease of implementation with minimal operating room renovations necessary. Room shielding requirements are minimized, and some systems require no modification with the use of a portable shielding apparatus that may be brought over the patient and machine when in use. One of the early iterations of a low-field designated intraoperative unit was the GE Signa SP system, which enables operating within the open magnet [77]. Such a system has the advantage of performing continuous real-time or periodic imaging. The openconfiguration MRI installed at the Brigham and Women’s Hospital in Boston has been in use since the mid 1990s [78]. Since that time neurosurgeons at that center have gained experience with over 500 craniotomies and 100 biopsies. The advantage of such a system is that it allows real-time imaging; disadvantages are somewhat restricted surgeon and patient positioning and the necessity to utilize MRI-compatible instruments. Higher field strength magnets are increasing in popularity, to enable acquisition of improved quality images. They also enable the use of expanded MR capabilities such as MR spectroscopy and functional MRI. There are several systems available. One well-designed system is the moveable high-field (1.5 T) magnet that is located on a roller system fixed to the ceiling as employed at the University of Calgary, Canada [79] (fig. 5). This configuration is similar to the operating microscope and other surgical adjuncts, with


137

MR technology moved to and from the patient as needed. Their system has been used to monitor a variety of neurosurgical procedures, including tumors, epilepsy, AVMs and other vascular malformations, and some cervical spine disorders. Another clever adaptation for the use of a high-field unit is the recent introduction of the Siemens 1.5 T intraoperative magnet (fig. 6). The machine is a standard 1.5 T MRI with functional MRI and MR spectroscopy capability. The room is designed to accommodate the 1.5 T machine with minimal disruption to the standard neurosurgical operating environment, including the use of standard operating instruments and microscope, which are located outside the 5 Gauss line. This enables the use of standard neurosurgical instrumentation, microscope, and image guidance systems. The patient is placed on a mobile operating table which is then rotated to fit on to the gantry of the MRI when imaging is desired. The machine has capabilities for intraoperative MR spectroscopy and functional MRI. There is no impediment to the operating surgeon, and operative positioning is independent of the scanning. Future Considerations Future developments in imaging will include more use of advanced MRI capabilities such as spectroscopy and functional MRI for intraoperative decision-making [80]. Also several centers are now planning for the adaptation of higher field strength magnets (e.g., 3 T) to the operating room environment. The introduction of MR-compatible robotic surgery with integration of robotic technology to the MR environment is an area that will help to revolutionize the future of neurosurgery, including the ability to locate and target a variety of deep structures in the brain. When combined with advances in viral gene Fig. 6.a The current high-field MRI Siemens system, demonstrating the ability of the surgeons to operate outside the 5 G line and use standard surgical instruments and microscope. Photo courtesy of Christopher Nimsky, MD, Department of Neurosurgery, Erlangen, Germany. b Schematic representation of the operating room layout for use of the Siemens 1.5 T machine with surgeons, nurses, and anesthetists positioned outside the 5 G line. c Pictures of the operating table rotating into position for intraoperative image acquisition. d(i) An example of image quality. Pre- and postresection T1-weighted images of a patient with a pituitary macroadenoma. d(ii) The use of iMRI facilitates glioma resection. Shown are comparative pre- and intraoperative images demonstrating the use of both T1-enhanced and T2 images to assess extent resection. e Standard image-guidance systems may also be employed in conjunction with intraoperative imaging. In this picture, the registration fiducials for a standard image guidance system are pictured.

House/Rao/Couldwell

138

6a

Anesthesia

Anesthesia preparation

e

-lin

5G era

Cam

4

NC RF-cabin

MRconsole

6b

6c


139

6d(i)

6d(ii)

House/Rao/Couldwell

140

6e

transfer, stem cell engineering, drug delivery devices, and neuroprosthetics, these complimentary technologies will allow precise targeting and delivery of molecular neurosurgical drugs to the brain and spinal cord. References 1 2 3 4 5

6

7 8

Wang KC, Koprivica V, Kim JA, Sivasankaran R, Guo Y, Neve RL, He Z: Oligodendrocyte-myelin glycoprotein is a Nogo receptor ligand that inhibits neurite outgrowth. Nature 2002;417:941–944. Mukhopadhyay G, Doherty P, Walsh FS, Crocker PR, Filbin MT: A novel role for myelin-associated glycoprotein as an inhibitor of axonal regeneration. Neuron 1994;13:757–767. Caroni P, Schwab ME: Antibody against myelin-associated inhibitor of neurite growth neutralizes nonpermissive substrate properties of CNS white matter. Neuron 1988;1:85–96. Schnell L, Schwab ME: Axonal regeneration in the rat spinal cord produced by an antibody against myelin-associated neurite growth inhibitors. Nature 1990;343:269–272. Buffo A, Zagrebelsky M, Huber AB, Skerra A, Schwab ME, Strata P, Rossi F: Application of neutralizing antibodies against NI-35/250 myelin-associated neurite growth inhibitory proteins to the adult rat cerebellum induces sprouting of uninjured purkinje cell axons. J Neurosci 2000;20: 2275–2286. Smith JS, Tachibana I, Passe SM, Huntley BK, Borell TJ, Iturria N, O’Fallon JR, Schaefer PL, Scheithauer BW, James CD, Buckner JC, Jenkins RB: PTEN mutation, EGFR amplification, and outcome in patients with anaplastic astrocytoma and glioblastoma multiforme. J Natl Cancer Inst 2001;93:1246–1256. Fan X, Aalto Y, Sanko SG, Knuutila S, Klatzmann D, Castresana JS: Genetic profile, PTEN mutation and therapeutic role of PTEN in glioblastomas. Int J Oncol 2002;21:1141–1150. Frisk T, Foukakis T, Dwight T, Lundberg J, Hoog A, Wallin G, Eng C, Zendenius J, Larsson C: Silencing of the PTEN tumor-suppressor gene in anaplastic thyroid cancer. Genes Chromosomes Cancer 2002;35:74–80.


141

9 10

11

12

13

14

15

16 17 18 19

20

21 22

23

24

25

26

27

Fults D, Pedone C: Immunocytochemical mapping of the phosphatase and tensin homolog (PTEN/MMAC1) tumor suppressor protein in human gliomas. Neuro-oncol 2000;2:71–79. Cheney IW, Johnson DE, Vaillancourt MT, Avanzini J, Morimoto A, Demers GW, Wills KN, Shabram PW, Bolen JB, Tavitgian SV, Bookstein R: Suppression of tumorigenicity of glioblastoma cells by adenovirus-mediated MMAC1/PTEN gene transfer. Cancer Res 1998;58: 2331–2334. Ikenaka K, Sasaki M, Tamura K, Tamura M, Miyao Y, Nanmoku K, Kawano Y, Nakahira K, Yoshimine T, Shimizu K: Treatment of glioblastoma by direct inoculation of concentrated high titer-recombinant retrovirus carrying the herpes simplex virus thymidine kinase gene. Hum Cell 2001;14:49–58. Glenn GM, Chatterjee S: Generation of adenoviruses encoding the herpes simplex virus vhs gene: A novel strategy to generate adenoviruses expressing genes toxic to producer cells. Cancer Gene Ther 2001;8:566–572. Maleniak TC, Darling JL, Lowenstein PR, Castro MG: Adenovirus-mediated expression of HSV1-TK or Fas ligand induces cell death in primary human glioma-derived cell cultures that are resistant to the chemotherapeutic agent CCNU. Cancer Gene Ther 2001;8:589–598. Rainov NG: A phase III clinical evaluation of herpes simplex virus type 1 thymidine kinase and ganciclovir gene therapy as an adjuvant to surgical resection and radiation in adults with previously untreated glioblastoma multiforme. Hum Gene Ther 2000;11:2389–2401. Voges J, Weber F, Reszka R, Sturm V, Jacobs A, Heiss WD, Wiestler O, Kapp JF: Clinical protocol. Liposomal gene therapy with the herpes simplex thymidine kinase gene/ganciclovir system for the treatment of glioblastoma multiforme. Hum Gene Ther 2002;13:675–685. Wade N: Death leads to concerns for future of gene therapy. NY Times (Print) 1999:A22. Smith L, Byers JF: Gene therapy in the post-Gelsinger era. JONAS Health Law Ethics Regul 2002;4:104–110. Dettweiler U, Simon P: Points to consider for ethics committees in human gene therapy trials. Bioethics 2001;15:491–500. Lakka SS, Rajan M, Gondi C, Yanamandra N, Chandrasekar N, Jasti SL, Adachi Y, Siddique K, Gujrati M, Olivero W, Dinh DH, Kouraklis G, Kyritsis AP, Rao JS: Adenovirus-mediated expression of antisense MMP-9 in glioma cells inhibits tumor growth and invasion. Oncogene 2002;21: 8011–8019. Hemmings-Mieszczak M, Dorn G, Natt FJ, Hall J, Wishart WL: Independent combinatorial effect of antisense oligonucleotides and RNAi-mediated specific inhibition of the recombinant rat P2X(3) receptor. Nucleic Acids Res 2003;31:2117–2126. Bueler H: Adeno-associated viral vectors for gene transfer and gene therapy. Biol Chem 1999; 380:613–622. Lu YY, Wang LJ, Muramatsu S, Ikeguchi K, Fujimoto K, Okada T, Mizukami H, Matsushita T, Hanazono Y, Kume A, Nagatsu T, Ozawa K, Nakano I: Intramuscular injection of AAV-GDNF results in sustained expression of transgenic GDNF, and its delivery to spinal motoneurons by retrograde transport. Neurosci Res 2003;45:33–40. Okada T, Nomoto T, Shimazaki K, Lijun W, Lu Y, Matsushita T, Mizukami H, Urabe M, Hanazono Y, Kume A, Muramatsu S, Nakano I, Ozawa K: Adeno-associated virus vectors for gene transfer to the brain. Methods 2002;28:237–247. Wang L, Muramatsu S, Lu Y, Ikeguchi K, Fujimoto K, Okada T, Mizukami H, Hanazono Y, Kume A, Urano F, Ichinose H, Nagatsu T, Nakano I, Ozawa K: Delayed delivery of AAV-GDNF prevents nigral neurodegeneration and promotes functional recovery in a rat model of Parkinson’s disease. Gene Ther 2002;9:381–389. Kirik D, Rosenblad C, Bjorklund A, Mandel RJ: Long-term rAAV-mediated gene transfer of GDNF in the rat Parkinson’s model: Intrastriatal but not intranigral transduction promotes functional regeneration in the lesioned nigrostriatal system. J Neurosci 2000;20:4686–4700. Kirik D, Georgievska B, Rosenblad C, Bjorklund A: Delayed infusion of GDNF promotes recovery of motor function in the partial lesion model of Parkinson’s disease. Eur J Neurosci 2001; 13:1589–1599. Mandel RJ, Snyder RO, Leff SE: Recombinant adeno-associated viral vector-mediated glial cell line-derived neurotrophic factor gene transfer protects nigral dopamine neurons after

House/Rao/Couldwell

142

28 29

30

31 32 33

34 35

36

37

38 39

40

41 42 43

44

45 46 47 48 49

onset of progressive degeneration in a rat model of Parkinson’s disease. Exp Neurol 1999;160: 205–214. Wright JF, Qu G, Tang C, Sommer JM: Recombinant adeno-associated virus: Formulation challenges and strategies for a gene therapy vector. Curr Opin Drug Discov Devel 2003;6:174–178. Lieberman DM, Laske DW, Morrison PF, Bankiewicz KS, Oldfield EH: Convection-enhanced distribution of large molecules in gray matter during interstitial drug infusion. J Neurosurg 1995; 82:1021–1029. Lonser RR, Corthesy ME, Morrison PF, Gogate N, Oldfield EH: Convection-enhanced selective excitotoxic ablation of the neurons of the globus pallidus internus for treatment of parkinsonism in nonhuman primates. J Neurosurg 1999;91:294–302. Laske DW, Youle RJ, Oldfield EH: Tumor regression with regional distribution of the targeted toxin TF-CRM107 in patients with malignant brain tumors. Nat Med 1997;3:1362–1368. Bruce JN, Falavigna A, Johnson JP, Hall JS, Birch BD, Yoon JT, Wu EX, Fine RL, Parsa AT: Intracerebral clysis in a rat glioma model. Neurosurgery 2000;46:683–691. Kaiser MG, Parsa AT, Fine RL, Hall JS, Chakrabarti I, Bruce JN: Tissue distribution and antitumor activity of topotecan delivered by intracerebral clysis in a rat glioma model. Neurosurgery 2000;47:1391–1398; discussion 1398–1399. Recht L, Torres CO, Smith TW, Raso V, Griffin TW: Transferrin receptor in normal and neoplastic brain tissue: Implications for brain-tumor immunotherapy. J Neurosurg 1990;72:941–945. Joshi BH, Leland P, Asher A, Prayson RA, Varricchio F, Puri RK: In situ expression of interleukin-4 (IL-4) receptors in human brain tumors and cytotoxicity of a recombinant IL-4 cytotoxin in primary glioblastoma cell cultures. Cancer Res 2001;61:8058–8061. Gage FH, Brundin P, Strecker R, Dunnett SB, Isacson O, Bjorklund A: Intracerebral neuronal grafting in experimental animal models of age-related motor dysfunction. Ann NY Acad Sci 1988;515:383–394. Perlow MJ, Freed WJ, Hoffer BJ, Seiger A, Olson L, Wyatt RJ: Brain grafts reduce motor abnormalities produced by destruction of nigrostriatal dopamine system. Science 1979;204: 643–647. Bjorklund A, Stenevi U: Reconstruction of the nigrostriatal dopamine pathway by intracerebral nigral transplants. Brain Res 1979;177:555–560. Freed WJ, Perlow MJ, Karoum F, Seiger A, Olson L, Hoffer BJ, Wyatt RJ: Restoration of dopaminergic function by grafting of fetal rat substantia nigra to the caudate nucleus: Long-term behavioral, biochemical, and histochemical studies. Ann Neurol 1980;8:510–519. Freed CR, Greene PE, Breeze RE, Tsai WY, DuMouchel W, Kao R, Dillon S, Winfield H, Culver S, Trojanowski JO, Eidelberg D, Fahn S: Transplantation of embryonic dopamine neurons for severe Parkinson’s disease. N Engl J Med 2001;344:710–719. Freed CR: Will embryonic stem cells be a useful source of dopamine neurons for transplant into patients with Parkinson’s disease? Proc Natl Acad Sci USA 2002;99:1755–1757. Yan Q, Matheson C, Lopez OT: In vivo neurotrophic effects of GDNF on neonatal and adult facial motor neurons. Nature 1995;373:341–344. Oppenheim RW, Houenou LJ, Johnson JE, Lin LF, Li L, Lo AC, Newsome AL, Prevette DM, Wang S: Developing motor neurons rescued from programmed and axotomy-induced cell death by GDNF. Nature 1995;373:344–346. Henderson CE, Phillips HS, Pollock RA, Davies AM, Lemeulle C, Armanini M, et al: GDNF: A potent survival factor for motoneurons present in peripheral nerve and muscle. Science 1994;266: 1062–1064. Yuen EC: The role of neurotrophic factors in disorders of peripheral nerves and motor neurons. Phys Med Rehabil Clin N Am 2001;12:293–306, viii. Grill RJ, Blesch A, Tuszynski MH: Robust growth of chronically injured spinal cord axons induced by grafts of genetically modified NGF-secreting cells. Exp Neurol 1997;148:444–452. Lonser RR, Gogate N, Morrison PF, Wood JD, Oldfield EH: Direct convective delivery of macromolecules to the spinal cord. J Neurosurg 1998;89:616–622. Hench LL, Polak JM: Third-generation biomedical materials. Science 2002;295:1014–1017. Brindley GS, Polkey CE, Rushton DN, Cardozo L: Sacral anterior root stimulators for bladder control in paraplegia: The first 50 cases. J Neurol Neurosurg Psychiatry 1986;49:1104–1114.


143

50

51

52 53

54 55 56 57 58 59 60 61 62 63 64 65

66 67 68 69

70 71 72

73 74 75

Wielink G, Essink-Bot ML, van Kerrebroeck PE, Rutten FF: Sacral rhizotomies and electrical bladder stimulation in spinal cord injury. 2. Cost-effectiveness and quality of life analysis. Dutch Study Group on Sacral Anterior Root Stimulation. Eur Urol 1997;31:441–446. Fedirchuk B, Shefchyk SJ: Effects of electrical stimulation of the thoracic spinal cord on bladder and external urethral sphincter activity in the decerebrate cat. Exp Brain Res 1991;84: 635–642. Kennedy PR, Bakay RA, Moore MM, Adams K, Goldwaithe J: Direct control of a computer from the human central nervous system. IEEE Trans Rehabil Eng 2000;8:198–202. Wessberg J, Stambaugh CR, Kralik JD, Beck PD, Laubach M, Chapin JK, Kim J, Biggs SJ, Srinivasan MA, Nicolelis MA: Real-time prediction of hand trajectory by ensembles of cortical neurons in primates. Nature 2000;408: 361–365. Nicolelis MA: Actions from thoughts. Nature 2001;409(suppl):403–407. Guillory KS, Normann RA: A 100-channel system for real time detection and storage of extracellular spike waveforms. J Neurosci Methods 1999;91:21–29. Dobelle WH, Quest DO, Antunes JL, Roberts TS, Girvin JP: Artificial vision for the blind by electrical stimulation of the visual cortex. Neurosurgery 1979;5:521–527. Nieder A: Miniature stereo radio transmitter for simultaneous recording of multiple single-neuron signals from behaving owls. J Neurosci Methods 2000;101:157–164. Nordhausen CT, Rousche PJ, Normann RA: Optimizing recording capabilities of the Utah Intracortical Electrode Array. Brain Res 1994;637:27–36. Hoogerwerf AC, Wise KD: A three-dimensional microelectrode array for chronic neural recording. IEEE Trans Biomed Eng 1994;41:1136–1146. Rauschecker JP, Shannon RV: Sending sound to the brain. Science 2002;295:1025–1029. Maynard EM: Visual prostheses. Annu Rev Biomed Eng 2001;3:145–168. Fanselow EE, Reid AP, Nicolelis MA: Reduction of pentylenetetrazole-induced seizure activity in awake rats by seizure-triggered trigeminal nerve stimulation. J Neurosci 2000;20:8160–8168. Berger T: World’s first brain prosthesis revealed. New Scientist 2003, March 12. Broggi G, Ferroli P, Franzini A, Dones L, Marras C, Marchetti M, Maccagnano E: CT-guided neurosurgery: Preliminary experience. Acta Neurochir Suppl 2003;85:101–104. Jolesz FA, Talos IF, Schwartz RB, Mamata H, Kacher DF, Hynynen K, McDannold N, Saivironporn P, Zao L: Intraoperative magnetic resonance imaging and magnetic resonance imaging-guided therapy for brain tumors. Neuroimaging Clin N Am 2002;12:665–683. Russell L: Intraoperative magnetic resonance imaging safety considerations. Norton Healthcare, Louisville, KY. Schulder M, Carmel PW: Intraoperative magnetic resonance imaging: Impact on brain tumor surgery. Cancer Control 2003;10:115–124. Hall WA, Liu H, Truwit CL: Intraoperative MR-guided instillation of phosphorus-32 for cystic craniopharyngiomas: Case report. Technol Cancer Res Treat 2003;2:19–24. Tuominen J, Yrjana SK, Katisko JP, Heikkila J, Koivukangas J: Intraoperative imaging in a comprehensive neuronavigation environment for minimally invasive brain tumour surgery. Acta Neurochir Suppl 2003;85:115–120. Schulder M, Jacobs A, Carmel PW: Intraoperative MRI and adjuvant radiosurgery. Stereotact Funct Neurosurg 2001;76:151–158. Hall WA, Kowalik K, Liu H, Truwit CL, Kucharezyk J: Costs and benefits of intraoperative MR-guided brain tumor resection. Acta Neurochir Suppl 2003;85:137–142. McPherson CM, Bohinski RJ, Dagnew E, Warnick RE, Tew JM: Tumor resection in a sharedresource magnetic resonance operating room: Experience at the University of Cincinnati. Acta Neurochir Suppl 2003;85:39–44. Nimsky C, Ganslandt O, Gralla J, Buchfelder M, Fahlbusch R: Intraoperative low-field magnetic resonance imaging in pediatric neurosurgery. Pediatr Neurosurg 2003;38:83–89. Nimsky C, Ganslandt O, Tomandl B, Buchfelder M, Fahlbusch R: Low-field magnetic resonance imaging for intraoperative use in neurosurgery: A 5-year experience. Eur Radiol 2002;12:2690–2703. Kanner AA, Vogelbaum MA, Mayberg MR, Weisenberger JP, Barnett GH: Intracranial navigation by using low-field intraoperative magnetic resonance imaging: Preliminary experience. J Neurosurg 2002;97:1115–1124.

House/Rao/Couldwell

144

76 77 78 79 80

Schulder M, Sernas TJ, Carmel PW: Cranial surgery and navigation with a compact intraoperative MRI system. Acta Neurochir Suppl 2003;85:79–86. Vitaz TW, Hushek S, Shields CB, Moriarty T: Intraoperative MRI for pediatric tumor management. Acta Neurochir Suppl 2003;85:73–78. Nabavi A, Gering DT, Kacher DF, Talos IF, Wells WM, Kikinis R, Black PM, Jolesz FA: Surgical navigation in the open MRI. Acta Neurochir Suppl 2003;85:121–125. Sutherland GR, Kaibara T, Louw DF: Intraoperative MR at 1.5 Tesla – Experience and future directions. Acta Neurochir Suppl 2003;85:21–28. Liu H, Hall WA, Truwit CL: The roles of functional MRI in MR-guided neurosurgery in a combined 1.5 Tesla MR-operating room. Acta Neurochir Suppl 2003;85:127–135.

William T. Couldwell, MD, PhD Department of Neurological Surgery, University of Utah Medical Center Suite 3B409, 30 North 1900 East Salt Lake City, UT 84132-2303 (USA) Tel. ⫹1 801 581 6908, Fax ⫹1 801 581 4385, E-Mail [email protected]


145


Xeno-Neurotransplantation James M. Schumacher Center for Movement Disorders, University of Miami School of Medicine, Miami, Fla., USA

Introduction

Though limited, the human nervous system has some capability to regenerate or repair damaged or degenerated cells. This inherent biological mechanism is not sufficient to reverse the devastating effects of neurodegenerative conditions such as Parkinson’s, Alzheimer’s, and Huntington’s disease. Strategies to replace degenerative neuronal systems have included transplantation of human embryonic fetal cells, embryonic stem cells, adult ‘stem’ cells, genetically modified somatic cells, viral-assisted gene transfer, and crossspecies cell transplants (i.e., xenotransplantation). Cell replacement therapies for neurodegenerative diseases were considered for human application after Parkinson’s disease (PD)-like motor deficits in animal models of the disease were ameliorated using transplanted embryonic dopamine cells. Fetal allogenic neuronal transplants have been shown to effect functional and behavioral recovery in a variety of animal models of neurodegenerative disease [1, 12, 15]. Using PD as a target syndrome, several investigations have been performed in humans. The first human neurotransplantation for PD was performed in 1988 [17]. Since then, over 300 patients worldwide have been transplanted with human tissue. Open-label studies suggested efficacy of transplantation and resulted in many cases in graft survival and increased dopamine utilization in the striatum [16]. Until recently, however, none of these surgical studies were done with adequate controls. Recently, two controlled studies have been performed using solid grafts in patients with advanced PD. Neither study demonstrated statistically powered efficacy [6, 7]. In the first study the endpoint was the Global Rating Scale. This scale is a subjective account of how the patient feels after surgery. The second study looked at the Unified Parkinson’s Disease Rating Scale (UPDRS), Part III motor ‘off scores.’

Table 1. Milestones of xeno-neurotransplantation 1890

Thompson. Cat cerebral cortex into brains of adult dogs. (No survival.)

1917

Dunn. Rat neonatal cerebral cortex into adult rat brain.

1921

Shirai. First description of brain immunoprivilege and xenografts.

1979

Bjorklund. Fetal rat brain to adult rat brain.

1985

Isacson. Fetal rat brain into adult rat model of Huntington’s and Parkinson’s.

1995

Schumacher. Fetal pig dopaminergic cells into a Parkinson’s patient.

Although results were uneven, there were patients within these studies and in previous open-label studies who have shown remarkable improvement in their condition and decreased need for pharmacological dopamine replacement. PET fluorodopa studies have also confirmed restoration of dopamine in the striatum of previously depleted areas. Patients who improved the most from transplantation were those who had a large difference between their ‘on’ and ‘off ’ UPDRS scores. Meta-analyses of the published open-label studies demonstrate that neurotransplantation is a very promising work in progress [12–14]. In order to demonstrate benefit and adequate dopamine cell survival, nearly 10 fetuses (3–5 human fetuses per putamen of embryonic age 8–10 weeks) are required for transplantation in any given patient. The logistics of obtaining this quantity of human tissue are prohibitive, notwithstanding the ethical considerations involved. Hence, the search for alternate cell sources has led investigators to cross-species transplants (xenotransplantation) and embryonic stem cells. In the case of xenotransplantation, neuroblasts from other species such as pigs could provide unlimited, screenable, precisely incubated cells for transplantation. Early attempts at cross-species neurotransplantation were unsuccessful due to immune rejection. With the introduction of cyclosporine and other immunosuppressive agents, however, successful xenotransplantation has become possible (table 1). Xenotransplantation is a beneficial laboratory tool in animal models of neurodegenerative disease. When host animals undergo immunosuppression, cross-species transplantation of embryonic dopaminergic cells has shown similar results to allografts [11, 19]. In addition, the unique antigenicity of the graft and host allows specific antibody cell labeling. In this

Xeno-Neurotransplantation

147

manner, graft survival, outgrowth and target specificity can be clearly defined. Several studies using embryonic porcine dopaminergic transplants into animal models of PD have been successful. Porcine tissue has been chosen for most studies because it is plentiful (up to 20 fetuses/liter), similar phenotypically to human, and has been used extensively in human medicine (i.e., cardiac valves, insulin). Human trials using porcine tissue have now been performed to treat PD, Huntington’s disease and epilepsy. Recent trials of neurotransplantation and the safety and immunological concerns of xeno-neurotransplantation are discussed below.

Immunology of Neural Xenografts

Although the central nervous system has a higher degree of immunoprivilege than other systems such as the heart, lung, liver and kidney, immunological reactions are a concern in allografts and especially xenografts. Factors that determine successful graft-host integration include the donor tissue embryonic stage, phylogenetic distance between donor and host, method of transplantation, preparation of graft (i.e., solid vs. suspension), host site, and method of immunosuppression. Several factors contribute to the immunological privilege in the host. In xenotransplantation, the lack of major histocompatibility complex (MHC) class I and II antigens is probably most important in graft survival. Cytokines are an important factor in graft cell death. These antigens can be induced in either system by influx of cytokines in the face of the inflammatory response of transplantation trauma [22]. Subsequent T-cell and macrophage activation and cell death is deactivated to a major extent by treatment with cyclosporine immunosuppression. Continued immunosuppression is important in that several other cytokines such as interleukin-2, -4, and -10 are induced for up to 30 days after transplantation [5, 18]. Anti-C5 (complement) antibody treatment has been found to inhibit cell death in xenografts. Further graft survival is seen with the combination of C5 inhibitor, cyclosporine methylprednisolone and azathioprine [2]. In the adult brain, MHC antigens are restricted to endothelial cells. Solid grafts contain intact endothelial cells and supporting cells and for this reason cell suspension grafts are favored over solid tissue pieces. The role of MHC 1 in tolerance induction has been shown to be an effective strategy in animals of xenotransplantation [18]. The antibody against the graft cell surface antigen is thought to promote tolerance by inhibiting T-cell induction. This technique is not as effective as cyclosporine but may provide an adjunct to immunosuppression and graft protection.

Schumacher

148

Safety Issues in Xenotransplantation

Acute, type I graft rejection is generally not seen in neurotransplantation or xeno-neurotransplantation. Patterns of rejection are mild when compared with that seen in major organ transplantation. In neurotransplantation, T-cell and macrophage-mediated rejection is seen over days and weeks. This response is greatly reduced with cyclosporine and other modifiers of immune response. In experimental animal models and human studies, no adverse side effects in the host have been seen. Either in animals or in human safety studies, no systemic immune effect has been documented. Local pathological effects of the inflammatory cell response have been observed. Immunosuppressants carry their own hazards of use including increased risk of infection and renal damage. Of special concern in cross-species transplantation is the risk of transmission of viral nucleic acid sequences (i.e., porcine endogenous retrovirus). To date, no transmission of porcine endogenous retrovirus from animal to man has been observed. Risk of transmission of virus or bacteria between species (xenozooinosis) is possible. Animals and their tissues must be carefully screened and monitored to avoid this problem. Prophylactic antibiotics are given before and after transplantation.

Xeno-Neurograft Procedures in Humans

Clinical trials of xenotransplantation in humans have been quite limited. Our study in transplanting porcine dopaminergic mesencephalic cells into patients with PD was the first human study of xeno-neurotransplantation [20], and will be discussed below. Subsequently, other studies for Huntington’s disease and epilepsy have been performed and these also will be briefly discussed. Patient Selection Patients selected for xenotransplantation were affected with advanced PD, were failing medical treatment with L-dopa, and were having ‘on-off’ motor fluctuations but were still responsive to L-dopa. Patients were screened using the Core Assessment Protocol in Intracerebral Transplantation (CAPIT protocol; UPDRS ⫹ time testing). Patients with dementia, poor medical condition, or serious comorbidity were excluded. Fluorodopa PET scans and cranial MRIs were performed before surgery and at 6 months and one year after transplantation. Preparation of Embryonic Porcine Ventral Mesencephalon Tissue Donor animals from a Yorkshire porcine herd were screened by serology for pathogen exposure, tested for parasites, and isolated. Embryonic tissue was


149

prepared by dissection of the ventral mesencephalon region from embryonic day 25–28 fetuses. Cells were then trypsonized to prepare a cell suspension at 50,000 cells/␮l. In our initial study, some cells were treated with an F(ab⬘)2 fragment of a monoclonal antibody directed against MHC I. This technique has been shown to promote xenograft survival without pharmacological immunosuppression [8, 18]. Viability prior to transplantation was assessed by acridine orange staining and screen by gram stain for bacteria. Aliquots of cell suspensions were cultured for dopaminergic cells using antibody to tyrosine hydoxylase. Patient’s blood mononuclear cells were obtained, archived and tested for porcine endogenous retrovirus. Preoperative Preparation In our study, 6 patients were loaded preoperatively with cyclosporine (15 mg/kg) 12 h prior to surgery. Six other patients were transplanted with cells that had been treated with the monoclonal antibody. All patients received perioperative antibiotics. Surgical Procedure Patients underwent the procedure with local anesthesia and MRI/CT-guided stereotaxy. Eighty microliter volumes of suspension were transplanted unilaterally in the striatum along three separate 5 mm tracts. One tract was placed in the caudate head and two in the mid and posterior putamen. Postoperative Evaluation MRI after one week showed evidence of the tracts in the striatum. Standard adverse event reporting, chemistry and blood testing was done per protocol. Cyclosporin levels were followed in those immunosuppressed patients. CAPIT testing was performed at 6 months and one year after surgery as was PET scanning and MRI. Clinical Results in Xenotransplantation In the original human study of CNS xenotransplantation of porcine cells, no adverse effects were seen in the 10 patients evaluated. None of the patients’ disease worsened in the year following surgery. In the medication ‘off’ state, 3 patients improved by ⬃30%. As a group the CAPIT scores improved by 19%. No significant change was seen in PET scans [20]. Another study with PD patients has been recently reported where bilateral ‘solid piece’ grafts were placed. Half the patients were given sham burr holes as a control. The patients that received xenografts improved 28% and the sham patients 23%. This study failed to show significant group improvement in CAPIT scores and showed a remarkable sham placebo effect [10]. Phase I

Schumacher

150

safety trials also have been performed using porcine fetal neural cells in Huntington’s disease. On evaluation after one year following transplantation, no adverse events were seen. No significant deterioration and no improvement in functional capacity were seen [9]. A small group of patients with frontal lobe epilepsy had GABAergic porcine grafts transplanted in the seizure focus, and this study is still ongoing.

Future Direction in Xeno-Neurotransplantation

Studies in both animal models and in humans have shown that transplanted neurons across species barriers can survive and establish functional axonal and synaptic contact with the immunosuppressed host. Neurotransmitters can be replenished and neuronal circuitry re-established. Preclinical studies in animals have clearly shown that pathological and behavioral deficits can be reversed using xenografts. The value of having an unlimited supply of selected neuronal cells for transplantation cannot be underestimated. Xenografts can be carefully screened for the disease and selected for the function and precise embryonic age. The greatest obstacle in xenotransplantation is still graft rejection. This obstacle is somewhat offset by modern methods of immunosuppression, but it is not yet optimized. Novel strategies are underway to improve cell survival. Transgenic pigs have been genetically engineered to express human cell surface markers. These cells are less immunogenic and promote graft survival [3, 4]. Critics of neurotransplantation have cited problems or adverse events in recent controlled human studies using human fetal material. As a group, these studies failed to meet their endpoint of statistical significance in improvement. Of particular concern, dyskinesia was observed during defined ‘off’ periods in some patients that were transplanted [6]. This phenomenon may represent unregulated dopamine production by the graft or may reflect more ‘on’ time with dyskinesia. Those patients did have ‘on’ dyskinesia prior to transplantation and were successfully treated with medication and in 2 patients with globus pallidus stimulation. Although xenografts are not as viable as human allografts, we believe that studies should continue, in particular to engineer hybrid human-porcine cell lines that may show less immunogenicity and greater in vivo activity. The utility of the unlimited supply and possibility of tissue screening make xenografts an important resource. Our current level of understanding of the immune system limits the use of xenografts as a treatment in human disease. These studies also provide the information that will make transplantation with allografts or embryonic stem cells more successful.


151

Ultimately, the answer to curing neurodegenerative disease is not only in the protection of native cells but in reconstruction of damaged and lost neuronal circuitry. Pharmacological therapy cannot provide the neurotransmitter and signal regulation needed at the cellular level. This regulation can only be provided by cell configurations at the synaptic level. Until we can determine how to promote native regeneration and regrowth of neural elements, our best strategy for neurodegeneration will incorporate aspects of cellular replacement through transplantation.

References 1 2 3 4

5 6 7 8 9 10 11 12 13

14 15 16

17 18

Bjorklund A, Stenevi U, et al: Cross species neural grafting in the rat model of Parkinson’s disease. Nature (Lond) 1982;298:652–654. Cicchetti F, Fodor W, et al: Immune parameters relevant to neuroxenograft survival in the primate brain. Xenotransplantation 2003;10:41–49. Deacon P, Schumacher J, et al: Histological evidence of fetal pig neural cell survival after transplantation into a patient with Parkinson’s disease. Nat Med 1997;3:350–352. Deacon T, Fodor W, et al: Xenotransplantation of transgenic fetal pig dopamine neurons to rats and systemic prevention of host compliment-mediated cell lysis. Abstr Soc Neurosci 1998; 24:1056. Duan W: Immunological and inflammatory responses against intrastriatal neural grafts in the rat. Thesis, Lund University, 1997. Fahn S, Freed C, Breeze W: Transplantation of embryonic dopamine neurons for severe Parkinson’s disease. New Eng J Med 2001;334:710–719. Fahn S, Green P, et al: Double blind controlled trial of human embryonic dopaminergic tissue transplants in advanced Parkinson’s disease: Clinical outcomes. Neurology 1999;52:A405. Faustman D, Coe C: Prevention of xenograft rejection by masking donor HLA class 1 antigens. Science 1991;252:1700–1702. Fink J, Schumacher J, Elias S, Isacson O: Porcine xenographs in Parkinson’s disease and Huntington’s disease patients: Preliminary results. Cell Transplant 2000;9:273–278. Freeman T: Porcine xenografts in patients with Parkinson’s disease. Abstract: American Association of Neurological Surgeons 2003. Freeman T, Wojak J, et al: Cross-species intracerebral grafting of embryonic swine dopaminergic neuron. Prog Brain Res 1998;78:473–477. Isacson O, Bjorklund L: Parkinson’s disease. Interpretations of transplantation study are erroneous. Nature Neurosci 2001;4. Isacson O, Deacon P, Pakzaban P: Transplanted xenogenic neural cells in neurodegenerative disease models exhibit remarkable axonal targets specificity and distinct growth patterns of glial and axonal fibres. Nat Med 1995;1:1189–1194. Isacson O, Deacon T, Schumacher J: Immunobiology and neuroscience of xenotransplantation and neurological disease. San Diego, Academic Press, 1998, pp 365–387. Isacson O, Dunnett S, Bjorklund A: Graft-induced recovery in an animal model of Huntington’s disease. Proc Natl Acad Sci USA 1986;83:2728–2732. Kordower J, Freeman T, et al: Neuropathological evidence of graft survival and striatal reinnervation after the transplantation of embryonic mesencephalic tissue in a patient with Parkinson’s disease. New Eng J Med 1993;332:1118–1124. Lindvall O, Widner H, et al: Transplant of fetal dopamine neurons in Parkinson’s disease. Ann Neurol 1992;31:155–165. Pakzaban P, Deacon T, Isacson O: A novel mode of immunoprotection of neuroxenotransplants: Masking of donor major histocompatibility complex class 1 enhances transplant survival in the CNS. Neuroscience 1995;65:983–986.

Schumacher

152

19 20 21 22

Pakzaban P, Isacson O: Neuroxenotransplantation: Reconstruction of neuronal circuitry across species barriers. Neuroscience 1994;62:989–1001. Schumacher J, Ellias P, et al: Transplantation of embryonic porcine mesencephalic tissue in patients with PD. Neurology 2000;54:1042–1050. Shirai Y: Transplanting rat sarcoma in adult heterogenous animals. Jap Med World 1921;1:14–15. Widner H, Lindhval O (eds): Basic and Clinical Aspects of Neuroscience, vol 5. Heidelberg, Springer, 1993, pp 63–74.

James M. Schumacher, MD Center for Movement Disorders University of Miami School of Medicine, Miami, FL 33136 (USA) Tel. ⫹1 305 243 4675, Fax ⫹1 305 243 3337, E-Mail [email protected]


153


Adeno-Associated Viral Vectors for Clinical Gene Therapy in the Brain R. Jude Samulski, Jennifer Giles Gene Therapy Center, Department of Pharmacology, University of North Carolina at Chapel Hill, N.C., USA

Introduction

A number of recombinant viral vectors have been engineered for gene transfer to the brain. The vectors in widest use for neuroscience applications include herpes virus vectors, adenovirus vectors, and lentiviral vectors, in addition to the parvovirus vector, adeno-associated virus (AAV). Despite its relatively small capacity of ⬍5 kb, AAV vectors have gained wide acceptance as a preferred vector for gene transfer to the central nervous system (CNS), due to its advantages of neurotrophism, nonpathogenicity, and stable in vivo gene expression. AAV is one of the few viral vectors that have already proven itself as a gene transfer vector for functional genomics as well as clinical applications in gene transfer to the human brain. AAV is a nonpathogenic virus that is not associated with any human viral syndrome or disease. It depends on the presence of a helper virus, such as adenovirus or herpes virus, for replication. The wild-type AAV (wtAAV) has a 4.68-kb single-stranded DNA genome comprised of capsid (cap) and replication (rep) open reading frames flanked by inverted terminal repeats. Three structural proteins, VP1, VP2, and VP3, are encoded by the single cap gene using alternative splicing and alternative start codons. The AAV virion is composed of VP1, VP2, and VP3 at a ratio of 1:1:10, respectively. The rep gene codes for four overlapping proteins involved in AAV DNA replication and the control of AAV gene expression. The two larger rep proteins, Rep78 and Rep68, are controlled by the p5 promoter and are needed for viral DNA replication,

while the smaller Rep52 and Rep40 proteins are transcribed from the p19 promoter and serve to facilitate the accumulation of single-stranded virus. The inverted terminal repeats are the only cis-acting elements required for AAV replication, packaging, integration, and rescue [1].

Production of Clinical-Grade AAV Vector

Recombinant AAV (rAAV) is an increasingly important gene therapy vector. Perhaps most beneficially, wtAAV is innocuous and has a known integration site at chromosome 19qter13.4 [2]. Long-term transgene expression is facilitated by the ability of rAAV to persist in vivo episomally and possibly also by integrating into the host genome. This long-term persistence is further enhanced by the fact that AAV does not induce a cell-mediated immune response in the host [3]. Adding to its appeal as a therapeutic vector, rAAV has been shown to infect both dividing and nondividing cells in a broad range of tissues, including muscle, liver, brain, and retina [4]. The rAAV plasmid is constructed by replacing the entire AAV coding genome with a transgene expression cassette flanked by the viral inverted terminal repeats. The rAAV plasmid is then used to transfect cells concurrently with a helper virus infection and an AAV helper plasmid that contains the rep and cap genes needed to supply the Rep and Cap proteins in trans. This cotransfection procedure allows efficient rescue and encapsidation of the rAAV genome from the recombinant vector plasmid [5] and production of rAAV vectors that can then be used for gene therapy delivery. It is important to note that the same protocol can be used for production of any of the AAV serotypes or modified vectors. An Ad-free AAV production system, using cotransfection of plasmid encoding the Ad helper genes, is a recent development that allows quick, easy generation of rAAV vectors for typical lab-scale use [6]. In order to generate the high quantities of virus that will be needed for clinical applications, cell lines engineered to produce AAV vectors or alternative methods of transfection must be developed. Inducible cell lines for AAV production are a current focus of virology research. These cell lines, that contain integrated copies of some or all of the AAV genes needed for packaging, utilize a variety of approaches to provide the helper genes and a vector in the host genome [7, 8]. While there are many advances being made in the development of gene delivery systems targeting the CNS, production of clinical grade viral vectors continues to be a bottleneck in the progression of these therapies to the clinic. Production of viral vectors under conditions that satisfy FDA Good Laboratory Practices and Good Manufacturing Practice guidelines introduces a wide range of testing and facility modifications not needed for research-grade vectors. All

Adeno-Associated Viral Vectors for Gene Therapy

155

steps in the process must be thoroughly documented, from the plasmid and mammalian cells used to produce the virus to the final viral vector itself. Quality control and release assays for vectors to be used in the clinic are extensive, and include, to name but a few, ELISA assays to determine the physical titer of the virus, quantitative PCR to determine the genomic titer, determination of the infectious titer, assays for contamination by the mammalian cell line used to generate the virus, silver stain/Coomassie blue gel staining to assay for protein contamination, and Western blot to assess the ratio of capsid proteins VP1, VP2, and VP3. The unprocessed ‘bulk harvest,’ which includes the mammalian cells, media, and unpurified virus, undergoes a series of quality control assays as well. Those include tests for sterility, mycoplasma, and replicationcompetent AAV and adenovirus. After a series of purification steps, the final vector preparation is assayed for sterility, bacteria, fungi, endotoxin, and residual DNA. One of the unfortunate consequences of the need for adherence to the above regulations is a shortage in facilities that are able to produce clinicalgrade reagents for gene therapy trials. The National Institutes of Health (NIH) have designated two U.S. national vector labs for the production of viral vectors for the clinic, one at Indiana University that specializes in retroviral vectors and one at Baylor University specializing in adenoviral vectors. These facilities have succeeded in establishing the technical expertise required to generate viral vectors and the resources to pay for extensive testing of the final product. Unfortunately, a number of novel vectors that do not fall under the mainstream production procedure require expertise typically located in the labs that derive these new systems. In the case of AAV, a Human Applications Lab (HAL) at the University of North Carolina at Chapel Hill (UNC) is an academic facility which has succeeded in producing AAV for a clinical trial. The successment implementation of the recent Canavan’s disease clinical trial highlights the need for such a facility to produce reagents for Phase I clinical trials. With fewer than 1,000 children in the USA affected by the disease, it is not an attractive target for private industry, while most academic institutions do not have Good Laboratory Practices facilities. Facilities like the HAL at UNC fills a critical gap between clinicians interested in gene therapy applications for rare genetic disorders and the patients who have so much to gain through these pioneering clinical trials. The amount of vector necessary to treat 21 human patients, on the order of milliliters of the final product, has been scaled up from the quantities that are more typical of experiments in animal models, on the order of microliters, and it is expected that further scale-up for other clinical trials will be possible in the future, especially as technical advances in large-scale AAV production at our center are achieved.

Samulski/Giles

156

AAV Serotypes for Gene Transfer

AAV exists naturally as a variety of serotypes that have immunologically unique properties. To date, a total of eight mammalian serotypes have been discovered and tested as viral vectors [9–15]. AAV serotype 1 or AAV-1 was the first to be isolated and characterized. Although isolated from the rhesus monkey, epidemiological data indicates that it frequently infects humans as well; however, it has yet to be recovered from a human sample. AAV serotypes 2, 3, and 5 are human parvoviruses, and have a high level of infection in the general population, as shown in epidemiological studies. The most unique AAV serotype on the nucleic acid level, AAV serotype 4, was isolated from the African green monkey and is rarely found in humans. It has, however, demonstrated the ability to infect human cells in vitro [16]. AAV serotype 6 is not serologically unique, and is more than 99% homologous to AAV-1 in its capsid proteins at the amino acid level. An analysis of AAV-1 and AAV-2 nucleic acid sequences suggests that AAV-6 is the product of a recombination event between these two serotypes [10]. Most recently, AAV serotypes 7 and 8 were isolated from nonhuman primates. As would be suggested by the serological uniqueness of the AAV serotypes, comparison of their capsids shows that they are indeed heterogeneous. Initial studies to evaluate the different AAV serotypes as gene delivery vectors indicated that serotypes have unique tropisms and differing transduction efficiencies depending on the cell type transduced, when compared to AAV-2 or against each other [17]. Until recently, the majority of the research conducted using AAV-based vectors implored AAV-2. This serotype historically has been used to study critical steps in AAV DNA replication, site-specific recombination, and AAV viral gene expression [18]. For this reason, it was only natural to extend upon this base of knowledge in the early development of AAV vectors. Only after extensive use of AAV-2 vectors in vivo and the identification of limitations in efficient transduction did attention turn to the other serotypes. Each can be distinguished by the efficiency of transduction for specific target tissue when compared to traditional AAV-2 vectors. For example, AAV serotype 5, but not AAV-2, binds to the apical surface of airway epithelia and facilities gene transfer [12, 13]. AAV-1 appears more robust in muscle cells while AAV-4 has been suggested to infect primarily ependymal cells when introduced into the mouse brain [12, 13]. AAV serotypes 7 and 8 have been shown to have muscle (AAV-7) and liver-specific tropism (AAV-8). Surprisingly, only ninety amino acid differences exist between these two serotypes, strongly suggesting that the capsid domain responsible for tissue tropism can be narrowed down and eventually identified. Understanding these differences and the capsid regions required for tropism for cells of the CNS will be critical for developing effective therapies for neurometabolic disorders.


157

Fig. 1. Ribbon structure of AAV-2 VP3.

The above evidence clearly points to the benefits of exploiting the natural and unique tropisms of AAV serotypes other than AAV-2 to increase AAVmediated gene transfer efficiency in different cell types. In order to achieve the most efficient gene delivery to the CNS, an approach that takes advantage of different serotypes will be needed.

Crystal Structure of AAV and the Future of Specific Targeting with AAV Vectors

The structure of parvovirus capsid proteins is now known. The structure of six autonomous parvoviruses and one dependo-virus have been solved: B19 [19], canine parvovirus (CPV) [20], feline panleukopenia virus (FPV) [21], Galleria mellonella densovirus (an insect parvovirus) [22], Aleutian Mink Disease parvovirus [23], minute virus of mice [24], and AAV-2 [25]. Sequence alignment of the capsid genes of B19 and CPV/FPV shows only 23% amino acid identity [26], yet these capsid proteins share extensive basic structural motifs. The virions of these viruses are made up of sixty subunits, with the smallest capsid protein making up the majority of the virion. They all share the eight B-barrel motif with looped out regions between barrels (fig. 1) [20, 21, 26]. It is possible to change parvovirus tropism by swapping key capsid amino acids. CPV and FPV share 98% sequence similarity within the capsid-coding region [27]. However, these viruses have different host range infectivity [28]. Using recombinants of CPV and FPV capsid sequences, Parrish et al. [28] were

Samulski/Giles

158

BC1/C0 Loop

HI Loop

GH12/13 Loop

GH10/11 Loop

GH2/3 Loop

I Loop

Fig. 2. Ribbon diagram of AAV-2 VP3. Regions in bold from AAV-1, -2, -3, -7, and -8 alignment are highlighted in blue, many are surface-displayed and may reflect muscle versus liver tropism differences. They are labeled according to Xie et al. [25].

able to map specific functions to epitopes on the capsid including determinants of host range infectivity. Refinement of this work defined the determinants of host range to amino acids 93 and 323 of VP2. Coding sequence of these amino acids were introduced into FPV, which could then replicate in canine cells [29]. These data support the ability to interchange epitopes and tropism between parvoviruses. A comparison of parvovirus capsid structures indicates that they are quite similar. AAV-2 VP3 has eight B-barrel motifs that are separated by looped out regions (fig. 2) [26]. The known differences between AAV-2 and the other serotypes may provide the information essential for understanding receptor binding and entry step of AAV vectors. Similarly to the differences found between CPV and FPV, amino acid sequences in loops 3 and 4 may explain the differences in cellular receptors used by AAV2 and the other serotypes. The functional domains can be identified by regions of viral capsid homology. All serotypes of AAV have unique tropisms based on epitopes present on the virion shell. However, the epitopes responsible for those tropisms are not well understood. In the future it will be beneficial to understand which domains on the virion are responsible for each serotype’s unique tropism, to improve targeting of specific cell types. The virions of all serotypes of AAV are assembled from a homologous set of precursor capsid proteins. The alignment of amino acid sequence of the capsid proteins illustrates the degree of homology between each serotype [12], and


159

the domains within the capsid sequence with high or low levels of homology have been identified [30]. Areas of low homology are of special interest since they may hold the key to the determinants of the serotypes. Alignment of AAV serotypes elucidates those domains that share the lowest degree of amino acid homology. Prior to the availability of the crystal structure of AAV-2, we exchanged domains between serotypes 1, 3, 4, and 5 from amino acid 440–603 (AAV-2 numbering) and AAV-2 (fig. 2). This domain occupies almost all of the GH looped out domain, the largest and most variable domain between all serotypes, including the recently discovered AAV-7 and AAV-8 as well as the most autonomous parvoviruses. Using pair-wise comparisons this domain has the highest level of homology between AAV-2 and AAV-3 (70%) and the least homology between AAV-4 and AAV-5 (10%). Additionally, comparisons of the autonomous CPV and the FPV shows that amino acid substitutions that resulted in species-jumping from feline to canine are located in the homologous GH loop. This leads us to make the assumption that those domains that are surfacelocalized and have low homology between the AAV serotypes may be responsible for tropism differences between AAV serotypes. This information will also direct future chimeric vector design in order to incorporate phenotypes specific to vector application. Determining which amino acids are surface-displayed is essential for understanding tropism. To approach this problem in a rational way, the newly resolved AAV-2 crystal structure will be essential. The usefulness of the crystal structure has been demonstrated with the positioning of targeting insertions into the adenovirus knob HI loop. For the autonomous parvoviruses a wealth of information has been revealed through comparisons of the amino acid sequences and the crystal structure with respect to surface-display and tropism. The availability of the crystal structure now provides a specific road map to rational structural/functional analysis.

Delivery of Gene Therapy to the Brain

In spite of the tremendous growth in the field of gene therapy and the numerous clinical trials currently underway, relatively few applications target disorders of the CNS. This is due in large part to the complexity of the brain and its circuitry, which is intolerant to even mild inflammation or toxicity. Limited access to the brain itself makes direct injection difficult. Additionally, the protection afforded to the CNS by the blood-brain barrier hinders global delivery of viral vectors through venous injection or cerebrospinal fluid. One of the first experiments in rodents to demonstrate the utility of rAAV vectors in vivo was aimed at transduction of brain tissue into rats [31]. Many of

Samulski/Giles

160

Step 1 Receptor binding FGFR

HSPG

␣v␤5

H+

␣v␤5

H+

Step 2 Nuclear entry

Step 3 DNA template

Fig. 3. Diagram of potential rate-limiting steps in efficient AAV transduction.

the recent advances in the understanding of rAAV vectors have come about through the need to better understand in vitro and in vivo transduction. Although several recent studies have shown great promise in terms of duration of transgene expression in vivo, there has been a shortfall in transduction efficiency, which was unexpected, based on previous results in vitro [32]. High transduction efficiency is of particular importance in the treatment of global neurometabolic disorders which require gene delivery to every affected cell in order to be therapeutically useful. Regardless of the serotype, all of the AAV vectors follow three basic steps for productive infection (fig. 3). First, receptor binding to the cell membrane is required; second, internalization and nuclear entry; and third, DNA template formation. After receptor binding, internalization, and nuclear entry, AAV virions uncoat and release a single-stranded DNA template, which must convert to a duplex intermediate before transcription can ensue. The efficiency of forming the complementary strand can significantly impact vector transduction [33, 34].


161

a

wtAAV (ssDNA)

b

Duplex AAV (scDNA)

Fig. 4. Self-complementary AAV packages both strands.

There are two possible mechanisms by which single-stranded AAV genomes can be converted to duplex templates. The first mechanism relies on the reannealing of two single-stranded genomes of different polarity (⫹ and –). Since AAV packages both strands with equal efficiency, this polarity may offer a viable mechanism for solving the duplex template requirement. The second (and usual) mechanism by which single-stranded AAV genomes can be converted to duplex templates involves DNA replication. AAV productive infection relies on the 145-bp hairpin terminal repeat and a self-priming mechanism for viral DNA replication [35]. The terminal repeat exists as double-stranded DNA duplex ‘T’ shaped structure and serves as an origin of replication for the singlestranded viral template [35]. The single-stranded viral template and the terminal repeat hairpin structures are required to form a duplex intermediate [36]. Second-strand synthesis is a rate-limiting step for rAAV transduction. Evidence supporting this conclusion was found in experiments correlating the induction of transgene expression with the conversion of the single-stranded virion DNA to the duplex. Generation of a duplex DNA template is required before transcription can ensue. Careful analysis of this process has now determined unique proviral intermediates (monomer, dimer, concatemeric structures, and circular molecules), all of which are derived from input single-stranded viral DNA [3, 37–48]. The dimer length of vector molecules originally characterized comprise duplex monomers, which are covalently linked at one end and are identical to substrates characterized for wtAAV [34, 36]. The characteristic lag of vector gene expression after infection in nondividing cells correlates with the formation of these duplex DNA intermediates [34, 36, 40, 41, 47, 49]. Double-stranded vectors, (fig. 4) rather than the naturally packaged singled-stranded molecule, could bypass the rate limiting step of second-strand synthesis. Recently, we have generated a novel double-stranded AAV (dsAAV) vector and demonstrated that steps which influence traditional single-stranded AAV

Samulski/Giles

162

transduction (i.e., availability of host DNA pol) were not required for dsAAV vectors, supporting the importance of the duplex intermediate as a rate-limiting step in AAV transduction [50]. The use of dsAAV may be particularly useful in the treatment of global CNS disorders that necessitate very high transduction rates. At present, AAV vectors in vivo appear to have very little toxicity or immune consequences after vector transduction [51]. Increased understanding of the biology of AAV has led to the generation of preclinical data for the treatment of neurological disorders and in the case of Canavan’s disease and Parkinson’s disease, Phase I clinical trials.

Treatment of Focal Brain Disorders with rAAV Vectors

Huntington’s Disease Like many other neurological disorders, Huntington’s disease (HD) is caused by the degeneration of specific cell groups within the CNS. By directly targeting these cell groups with gene transfer to express neurotrophic factors, this degeneration may be reversed or avoided altogether. HD is manifested as an array of motor, cognitive, and psychiatric disturbances caused by the degeneration of medium-sized spiny neurons in the striatum and cerebral cortex. Glial cell-line-derived neurotrophic factor has been shown to prevent the loss of striatal neurons in animal models of HD [52–56]. In an animal model of HD, bilateral injections of a rAAV viral vector containing the glial cell-line-derived neurotrophic factor transgene into the striatum showed marked protection of striatal neurons and prevention of behavioral disturbances [57]. Because HD is passed on through an autosomally dominant gene, persons who have not yet sustained any neurological damage could be identified as having the disease and receive gene therapy, making it possible to avoid the debilitating effects of the disease altogether. Seizure Disorders Gene therapy with AAV also has been tested for the treatment of focal seizure disorders. A study by Haberman et al. [58] shows that the delivery of N-methyl D-aspartic acid receptor antisense using an AAV-derived vector can modulate seizure disorder in vivo. However, it was also shown that using different promoters to drive the transgene expression resulted in completely opposite physiological effects, increasing seizure sensitivity as opposed to reducing it. In order to work around this effect, the gene for the inhibitory neuroactive peptide, galanin, was delivered to cells using an AAV-derived vector, and constitutively secreted through the use of the fibronectin signal sequence. Importantly, the choice of promoter to drive the transgene did not impact the


163

decrease in seizure sensitivity [59]. This approach provides a new option for the long-term control of focal seizure disorders. Parkinson’s Disease Another devastating neurological disorder, Parkinson’s disease, is characterized by the degeneration of the substantia nigra pars compacta and consequently, reduced dopamine in the striatum. The resulting lack of inhibition of the subthalamic nucleus (STN) contributes to the motor abnormalities typical of the disease. Deep brain electrical stimulation of the STN has been shown to effectively reduce symptoms of Parkinson’s disease. One alternative approach for the treatment of Parkinson’s using gene therapy seeks to achieve the same effect biochemically. GABA, the brain’s major inhibitory transmitter, can be generated by two isoforms of glutamic acid decarboxylase. Stereotactic injection of rAAV carrying the glutamic acid decarboxylase transgenes into the STN of adult parkinsonian rats resulted in neuroprotection of the STN and a reduction in the excitatory phenotype of the disease. Robust expression of the transgene was seen up to 5 months after injection, with no significant immune response [60]. One of the two clinical trials currently underway that utilize rAAV vectors to treat neurological disorders uses this vector in human patients.

Global Gene Delivery with rAAV Vectors

Canavan’s Disease The original clinical trial in which rAAV was first used as a vector for gene delivery in humans is for the treatment of Canavan’s disease. This study, led by Dr. Paola Leone at UMDNJ-Robert Wood Johnson Medical School, was the first gene transfer clinical trial to use viral vectors to treat a neurodegenerative disorder. Canavan’s disease is an inherited disease, with autosomal recessive inheritance, that shortens life expectancy to a few years. Symptoms are generally first seen within the first 6 months of life and include megalocephaly and developmental delays. As the disease progresses, mental retardation, spasticity and cortical blindness develop, culminating in seizures and childhood death. A lack of the enzyme aspartoacylase (ASPA) that hydrolyzes N-acetylaspartic acid (NAA) into L-aspartate and acetate causes the damage associated with Canavan’s [61, 62]. It is thought that the accumulation of metabolic precursors such as NAA, the function of which remains undetermined, is toxic and leads to neurological damage [63]. It also appears that high levels of NAA affect the phenotype of developing myelinating cells through a complex set of

Samulski/Giles

164

gene expression effects which are currently being worked out [Leone, pers. commun.]. This clinical trial is intended to retard the damage of elevated NAA levels by injecting rAAV carrying the aspartoacylase gene into the brain. Preliminary data suggest beneficial effects of treatment on the biochemical and clinical level, with an absence of adverse events. dsAAV and Improved Global CNS Delivery? Because global neurometabolic disorders require long-term transgene expression, and in cases of intrinsically expressed gene products, transduction of nearly every affected cell, the development of less intrusive delivery methods is a key step in getting these treatments into the clinic. A recent study by Fu et al. demonstrated the utility of the dsAAV vector for global CNS distribution in a mouse model. An IV injection of 4 ⫻ 1011 particles of dsAAV2-expressing green fluorescent protein preceded by an injection of 12.5% mannitol showed a global distribution of the transgene in the brain and spinal cord 4–8 weeks after injection. No green fluorescent protein expression was seen using dsAAV2 in 12.5% mannitol or with the viral vector without mannitol. Additionally, no green fluorescent protein expression was seen when ssAAV was injected intravenously preceded by the injection of mannitol. Through the use of mannitol to transiently open the blood-brain barrier in conjunction with dsAAV to increase transduction efficiency, minimally invasive intravenous injections may provide another effective route to global gene delivery to the CNS, in addition to intraparenchymal injection protocols.

Conclusion

Safe and efficient delivery of corrective gene therapy to the CNS using rAAV vectors has great promise for the treatment of neurological disorders. An array of recent developments including advances in production, the elucidation of the rAAV crystal structure, and newly discovered serotypes will facilitate further clinical applications. While the CNS presents unique challenges to the development of effective therapies, the devastating nature of the diseases being targeted should serve as an impetus to overcome the inherent hurdles.

Acknowledgements Thanks to the members of the University of North Carolina, Chapel Hill (UNC) Gene Therapy Core.


165

References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

18 19 20 21 22 23 24 25 26

Berns KI: Parvorviridae and their replication; in Fields BN, Knipe DM (eds): Virology. New York, Raven, 1990, pp 1743–1763. Samulski RJ: Adeno-associated virus: Integration at a specific chromosomal locus. Curr Opin Genet Dev 1993;3:74–80. Xiao X, Li J, Samulski RJ: Efficient long-term gene transfer into muscle tissue of immunocompetent mice by adeno-associated virus vector. J Virol 1996;70:8098–8108. Monahan PE, Samulski RJ: AAV vectors: Is clinical success on the horizon? Gene Ther 2000;7:24–30. Samulski RJ, Chang LS, Shenk T: Helper-free stocks of recombinant adeno-associated viruses: Normal integration does not require viral gene expression. J Virol 1989;63:3822–3828. Xiao X, Li J, Samulski RJ: Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus. J Virol 1998;72:2224–2232. Qiao C, et al: Feasibility of generating adeno-associated virus packaging cell lines containing inducible adenovirus helper genes. J Virol 2002;76:1904–1913. Qiao C, et al: A novel gene expression control system and its use in stable, high-titer 293 cellbased adeno-associated virus packaging cell lines. J Virol 2002;76:13015–13027. Walters RW, et al: Incorporation of adeno-associated virus in a calcium phosphate coprecipitate improves gene transfer to airway epithelia in vitro and in vivo. J Virol 2000;74:535–540. Xiao W, et al: Gene therapy vectors based on adeno-associated virus type 1. J Virol 1999;73: 3994–4003. Muramatsu S, et al: Nucleotide sequencing and generation of an infectious clone of adenoaccociated virus 3. Virology 1996;221:208–217. Chiorini JA, et al: Cloning of adeno-associated virus type 4 (AAV4) and generation of recombinant AAV4 particles. J Virol 1997;71:6823–6833. Chiorini JA, et al: Cloning and characterization of adeno-associated virus type 5. J Virol 1999;73:1309–1319. Rutledge EA, Halbert CL, Russell DW: Infectious clones and vectors derived from adeno-associated virus (AAV) serotypes other than AAV type 2. J Virol 1998;72:309–319. Gao GP, et al: Novel adeno-associated viruses from rhesus monkeys as vectors for human gene therapy. Proc Natl Acad Sci USA 2002;99:11854–11859. Muzyczka N: Use of adeno-associated virus as a general transduction vector for mammalian cells. Curr Top Microbiol Immunol 1992;158:97–129. Rabinowitz JE, et al: Cross-packaging of a single adeno-associated virus (AAV) type 2 vector genome into multiple AAV serotypes enables transduction with broad specificity. J Virol 2002;76: 791–801. Berns KI, Hauswirth WW: Review: Adeno-associated viruses. Adv Virus Res 1979;25:407–449. Agbandje-McKenna M, et al: The structure of human parvovirus B19 at 8 Å resolution. Virology 1994;203:106–115. Tsao J, et al: The three-dimensional structure of canine parvovirus and its functional implications. Science 1991;25:1456–1464. Agbandje M, et al: Structure determination of feline panleukopenia virus empty particles. Proteins 1993;16:155–171. Simpson AA, et al: The structure of an insect parvovirus (Galleria mellonella densovirus) at 3.7 Å resolution. Structure 1998;6:1355–1367. McKenna R, et al: Three-dimensional structure of Aleutian mink disease parvovirus: Implications for disease pathogenicity. J Virol 1999;73:6882–6891. Agbandje-McKenna M, et al: Functional implications of the structure of the murine parvovirus, minute virus of mice. Structure 1998;6:1369–1381. Xie Q, et al: The atomic structure of adeno-associated virus (AAV-2), a vector for human gene therapy. Proc Natl Acad Sci USA 2002;99:10405–10410. Chapman M, Rossmann M: Structure, sequence, and function correlations among parvoviruses. Virology 1993;194:419–508.

Samulski/Giles

166

27 28

29

30 31

32 33 34 35 36 37

38 39 40 41 42 43 44 45

46

47 48 49 50

Martyn J, Davidson B, Studdert M: Nucleotide sequence of feline panleukopenia virus: Comparison with canine parvovirus identifies host-specific differences. J Gen Virol 1990; 71: 2747–2753. Parrish C, Aquadro C, Carmichael L: Canine host range and a specific epitope map along with variant sequences in the capsid protein gene of canine parvovirus and related feline, mink, and raccoon parvoviruses. Virology 1988;166:293–307. Chang S, Sgro J, Parrish C: Multiple amino acids in the capsid stucture of canine parvovirus coordinately determine the canine host range and specific antigenic and hemagglutination properties. J Virol 1992;66:6858–6867. Rabinowitz JE, Samulski J, Adeno-associated virus expression systems for gene transfer. Curr Opin Biotechnol 1998;9:470–475. Kaplitt MG, et al: Preproenkephalin promoter yields region-specific and long-term expression in adult brain after direct in vivo gene transfer via a defective herpes simplex viral vector. Proc Natl Acad Sci USA 1994;91:8979–8983. Flotte TR: Prospects for virus-based gene therapy for cystic fibrosis. J Bioenerg Biomembr 1993; 25:37–42. Fisher KJ, et al: Transduction with recombinant adeno-associated virus for gene therapy is limited by leading-strand synthesis. J Virol 1996;70:520–532. Ferrari FK, et al: Second-strand synthesis is a rate-limiting step for efficient transduction by recombinant adeno-associated virus vectors. J Virol 1996;70:3227–3234. Berns K: Parvoviridae: The viruses and their replication; in Fields BN, Knipe PM, Howley EA (eds): Virology. Philadelphia, Lippincott-Raven Publishers, 1996, pp 2173–2198. Straus SE, Sebring ED, Rose JA: Concatemers of alternating plus and minus strands are intermediates in adenovirus-associated virus DNA synthesis. Proc Natl Acad Sci USA, 1976;73:742–746. Duan D, et al: Lef1 transcription factor expression defines airway progenitor cell targets for in utero gene therapy of submucosal gland in cystic fibrosis. Am J Respir Cell Mol Biol 1998; 18: 750–758. Duan D, et al: Formation of adeno-associated virus circular genomes is differentially regulated by adenovirus E4 ORF6 and E2a gene expression. J Virol 1999;73:161–169. Duan D, et al: Structural analysis of adeno-associated virus transduction circular intermediates. Virology 1999;261:8–14. Snyder RO, Im DS, Muzyczka N: Evidence for covalent attachment of the adeno-associated virus Rep protein to the ends of the AAV genome. J Virol 1990;64:6204–6213. Snyder RO, Samulski RJ, Muzyczka N: In vitro resolution of covalently joined AAV chromosome ends. Cell 1990;60:105–133. Vincent-Lacaze N, et al: Structure of adeno-associated virus vector DNA following transduction of the skeletal muscle. J Virol 1999;73:1949–1955. McLaughlin SK, et al: Adeno-associated virus general transduction vectors: Analysis of proviral structures. J Virol 1988;62:1963–1973. Hong G, Ward P, Berns KI: Intermediates of adeno-associated virus DNA replication in vitro. J Virol 1994;68:2011–2015. Hoggan MD, et al: Continuous carriage of adenovirus associated virus genome in cell culture in the absence of helper adenovirus. Proceedings of the Fourth Lepetite Colloquium, Mexico, Cocoyac, 1972. Flotte TR, Afione SA, Zeitlin PL: Adeno-associated virus vector gene expression occurs in nondividing cells in the absence of vector DNA integration. Am J Respir Cell Mol Biol 1994; 11: 517–521. Clark KR, Sferra TJ, Johnson PR: Recombinant adeno-associated viral vectors mediate long-term transgene expression in muscle. Hum Gene Ther 1997;8:659–669. Berns KI, Linden RM: The cryptic life style of adeno-associated virus. Bioessays 1995;17:237–245. Miao CH, et al: The kinetics of rAAV integration in the liver [letter]. Nat Genet 1998;19: 13–15. McCarty DM, Monahan PE, Samulski RJ: Self-complementary recombinant adeno-associated virus (scAAV) vectors promote efficient transduction independently of DNA synthesis. Gene Ther 2001;8:1248–1254.


167

51

52 53 54 55

56

57 58 59 60 61 62

63

Samulski RJ, Sally M, Muzyczka N (eds): Adeno-associated viral vectors; in Friedman T (ed): The development of human gene therapy. Cold Spring Harbor, Cold Spring Harbor Laboratory Press, 1999, pp 131–172. Perez-Navarro E, et al: Glial cell line-derived neurotrophic factor protects striatal calbindinimmunoreactive neurons from excitotoxic damage. Neuroscience 1996;75:345–352. Perez-Navarro E, et al: Intrastriatal grafting of a GDNF-producing cell line protects striatonigral neurons from quinolinic acid excitotoxicity in vivo. Eur J Neurosci 1999;11:241–249. Gratacos E, et al: Neuroprotection of striatal neurons against kainate excitotoxicity by neurotrophins and GDNF family members. J Neurochem 2001;78:1287–1296. Araujo DM, Hilt DC: Glial cell line-derived neurotrophic factor attenuates the excitotoxininduced behavioral and neurochemical deficits in a rodent model of Huntington’s disease. Neuroscience 1997;81:1099–1110. Araujo DM, Hilt DC: Glial cell line-derived neurotrophic factor attenuates the locomotor hypofunction and striatonigral neurochemical deficits induced by chronic systemic administration of the mitochondrial toxin 3-nitropropionic acid. Neuroscience 1998;82:117–127. McBride JL, et al: Structural and functional neuroprotection in a rat model of Huntington’s disease by viral gene transfer of GDNF. Exp Neurol 2003;181:213–223. Haberman R, et al: Therapeutic liabilities of in vivo viral vector tropism: Adeno-associated virus vectors, NMDAR1 antisense, and focal seizure sensitivity. Mol Ther 2002;6:495–500. Haberman RP, Samulski RJ, McCown TJ: Attenuation of seizures and neuronal death by adenoassociated virus vector galanin expression and secretion. Nat Med 2003;9:1076–1080. Luo J, et al: Subthalamic GAD gene therapy in a Parkinson’s disease rat model. Science 2002; 298:425–429. Matalon R, et al: Aspartoacylase deficiency and N-acetylaspartic aciduria in patients with Canavan disease. Am J Med Genet 1988;29:463–471. Divry P, et al: N-acetylaspartic aciduria: Report of three new cases in children with a neurological syndrome associating macrocephaly and leukodystrophy. J Inherit Metab Dis 1988;11: 307–308. Hsich G, Sena-Esteves M, Breakefield XO: Critical issues in gene therapy for neurologic disease. Hum Gene Ther 2002;13:579–604.

R. Jude Samulski, PhD Gene Therapy Center, Department of Pharmacology CB# 7352, 7119 Thurston Bowles, University of North Carolina at Chapel Hill Chapel Hill, NC 27599–7352 (USA) Tel. ⫹1 919 962 3285, Fax ⫹1 919 966 0907, E-Mail [email protected]

Samulski/Giles

168


Molecular Mechanisms of Epilepsy and Gene Therapy Albert Telfeiana, Juanita Celixb, Marc Dichterc a

Division of Neurosurgery, Texas Tech University Medical Center, Lubbock, Tex., Department of Neurosurgery, University of Washington, Seattle, Wash., c Department of Neurology, University of Pennsylvania, Philadelphia, Pa., USA b

Introduction ‘Extreme remedies are appropriate for extreme diseases…’ Hippocrates (460–370 B.C.)

Epilepsy is a neurological disorder that afflicts 1–2% of the general population and encompasses a variety of disorders with seizures [1]. To best understand where we must go in the treatment of epilepsy, it is necessary to understand first where we have failed. The prognosis for seizure control in epilepsy with medication is good in ⬃60% of patients, and up to 40% of individuals suffer from intractable pharmacoresistant epilepsy. There are over twenty different anti-epileptic drugs available to the neurologist or neurosurgeon to treat seizures, but patients not controlled on monotherapy with the first anti-epileptic agent have only a 10% chance of being controlled with additional anti-epileptic agents, even when using medications that work by diverse mechanisms [2]. Newer is not necessarily better in terms of drug regimens. Only a small minority (⬍5%) of patients refractory to traditional drug therapy has been reported to become seizure free with a new generation anti-epileptic drug [2]. The more we understand about the genetic basis of this disease, the more naïve it appears that a single drug tailored to a specific channel or neurotransmitter receptor will effectively cure epilepsy. In fact, it appears that the better drugs have a wider basis in their mechanisms of action. Most clinically efficacious anti-epileptic drugs possess a combination of various properties. While there have been advances in the drugs available to control the seizures associated

with epilepsy, to date there is no effective therapy for the prevention of epilepsy. Generally, one third of all cases of epilepsy have a potential cause, the most obvious being trauma, tumors, and stroke, but treatment seems to have done little to prevent the process of increased intrinsic excitability, synchronization, or synaptic connectivity that may be responsible for the development of seizures. Here, we concentrate on how a better understanding of the molecular mechanisms of epilepsy may guide future therapies, such as gene therapy, for this disease.

Review of Pharmacotherapy in Epilepsy

Anti-epileptic drug therapy is the first-line treatment for epilepsy. Until the 1990s, there were only a handful of drugs available to treat the various seizure disorders. The earliest drugs used to control seizures (e.g., bromides, phenobarbital, valproic acid) were identified inadvertently. Later, a more scientific approach utilized animal models of epilepsy against which potential antiepileptic drugs were tested. The maximal electroshock (MES) model is used to evaluate agents for the ability to decrease seizure severity and to identify those drugs with efficacy in treating generalized tonic-clonic or partial seizures. The pentylenetetrazole (PTZ) model is utilized to test for agents that increase the seizure threshold and exert possible anti-absence seizure properties. While the early anti-epileptic drugs identified by these models were understood to function via action at the neural membrane or synapse, the animal models could not elucidate the important mechanisms of action. It was not until the introduction of more modern electrophysiological and pharmacological research techniques that the effects of anti-epileptic drugs at the neural membrane and synapse were determined. The diligent study of anti-epileptic therapies lead to the discovery of the principle mechanisms of action of the clinically efficacious drugs used to treat seizure disorders. In general, anti-epileptic agents control the initiation, maintenance, or propagation of epileptiform discharges through augmented inhibition, suppressed excitation, or modulation of action potential ion current, with effects at the level of localized neurons to entire neural networks. Typically, anti-epileptic drugs act on one of four classes of neuron ion channels: ␥-aminobutryic acid (GABAA) ligand-gated chloride channels, glutamate ligand-gated sodium and calcium channels (NMDA, N-methylD-aspartate; AMPA, ␣-amino-3-hydroxy-5-methyl-4-isoxazoleproprionic acid; and kainate), voltage-gated sodium channels, and voltage-gated T-type (low threshold) calcium channels. Most of the agents with activity in the MES model function through use-dependent inactivation of voltage-gated sodium

Telfeian/Celix/Dichter

170

channels. Drugs with efficacy in the PTZ model function via blockade of T-type calcium channels or augmentation of GABAA receptor-mediated chloride channels. Several of the recently developed anti-epileptic drugs that have been introduced during the past decade have new or additional mechanisms of action, targeting pre- and postsynaptic membrane-bound receptors and enzymes that function in the metabolism of neurotransmitters. Still, there are a few anti-epileptic agents whose mechanisms of action remain largely unknown (e.g., topiramate, felbamate). As we continue to make advances in cellular and molecular biology and gain a greater understanding of the excitability and synchronization of neurons in circuit, a fuller appreciation for the mechanisms of action of the efficacious anti-epileptic drugs will undoubtedly become more apparent.

Traditional Pharmacotherapy

Phenytoin, carbamazepine, and valproic acid are traditional anti-epileptic drugs with primary actions at the voltage-gated sodium channel. Phenytoin acts at the sodium channel in a voltage- and frequency-dependent manner. It preferentially binds to and stabilizes the channel in the open inactive state. Selective blockade of the inactive sodium channel prevents release of excitatory amino acid neurotransmitters, particularly glutamate and aspartate, delays channel recovery time, and slows propagation of electrical discharges. There is evidence that sustained, high-frequency repetitive firing plays a major role in neuron excitability, and that phenytoin functions to limit this repetitive firing [3]. As sodium channels may be more susceptible to blockade when they are in the open inactive state, preferential binding of phenytoin to the inactivated voltagegated sodium channel of the depolarized neuron during seizure activity is the likely mechanism through which phenytoin acts to limit sustained repetitive firing and terminate seizure activity. At high drug doses, phenytoin has additional clinically significant actions, including voltage-gated calcium channel blockade and decreased presynaptic glutamate release. The anti-epileptic properties of carbamazepine, initially developed as a tricyclic antidepressant, and valproic acid were fortuitously discovered. Similar to phenytoin, carbamazepine and valproic acid both function in a voltage- and frequency-dependent manner to bind the inactive sodium channel and delay recovery. Valproic acid also indirectly facilitates the inhibitory activity of GABA via increased synaptic neurotransmitter levels, while carbamazepine’s additional mechanisms of action include inhibition of norepinephrine reuptake, decrease in intracellular cAMP levels through interaction with adenosine receptors, and possible potentiation of GABA inhibition through interaction with the

Epilepsy and Gene Therapy

171

GABAB receptor. The clinical importance of these additional mechanisms of action remains to be determined. The voltage-gated calcium channels consist of at least four types, L-type, T-type, N-type, and P-type, with differing levels of voltage activation and inactivation. Ethosuximide is an anti-epileptic drug with activity at the T-type calcium channel, a low threshold voltage-gated channel. Ethosuximide functions primarily at the T-type channels of thalamic neurons, reducing the calcium current and interrupting the 3 Hz spike/wave thalamocortical action potentials typical of absence seizures. Activity at GABA ligand-gated chloride channels is the primary mechanism of action of the most well known anti-epileptic drugs, barbiturates and benzodiazepines. The GABAA receptor complex contains multiple modulatory sites for site-specific interaction with various agents. The GABA receptor subtypes have differential developmental and regional expression and varying sensitivities to the diverse ligands. Phenobarbital is one of the oldest antiepileptic drugs. At the GABAA receptor chloride channel, it potentiates the activity of GABA through enhanced duration of channel opening. At high drug doses, the barbiturates also interact with the N-type voltage-gated calcium channels to block the calcium influx and prevent release of excitatory amino acid neurotransmitters. Benzodiazepines also function at the GABAA receptor, where they increase the frequency of chloride channel opening without affecting the duration of opening.

Advances in Pharmacotherapy

The recent introduction of new generation anti-epileptic drugs illustrates our increased understanding of the multiple interrelated mechanisms by which synchronous excitatory electrical activity can be initiated, maintained or propagated, and enables a more rational approach to pharmacotherapy in epilepsy. While some of the newer agents were discovered using the MES and PTZ animal models of epilepsy, others were rationally designed to produce a specific molecular effect. Many newer agents have multiple mechanisms of action and predominantly function via metabotropic or ionotropic channels, similar to the traditional anti-epileptic drugs. Lamotrigine is a new generation drug that functions at the presynaptic open inactive voltage-gated sodium channel to delay channel recovery and decrease excitatory amino acid release. It may also interact with the GABA receptor, and it has efficacy in treating certain seizure types that cannot be explained by its primary mechanism of action. Unexplained therapeutic effects are common with many of the new generation agents. Gabapentin, which was developed based on the theoretical role of


172

GABA in epileptogenesis, is a novel drug formed by the linkage of a cyclohexyl group to GABA. This modification allows it to easily penetrate the blood-brain barrier (BBB). As a structural analog of the endogenous inhibitory neurotransmitter, gabapentin was designed to function via GABA-mimetic activity, but the drug does not interact with GABA receptors in the CNS. Neither does it interact with voltage-gated sodium or calcium channels, nor with NMDA ligandgated ion channels. It is unclear through which mechanism or combination of mechanisms gabapentin exerts its anti-epileptic effect, but it does indeed have anti-epileptic properties. A novel class of second generation anti-epileptic drugs is active in modulating local levels of GABA. Altered intracellular metabolism of endogenous neurotransmitter and inhibition of reuptake from the synaptic cleft are the mechanisms by which these new agents augment levels of inhibitory GABA within the brain and ultimately alter the balance of excitation and inhibition. Vigabatrin is an irreversible inhibitor of intracellular GABA transaminase, the protein responsible for the metabolic degradation of GABA. Inhibition of GABA degradation allows an increase in the local concentration of GABA and potentiation of its physiological role in limiting excitatory neural activity in the brain. Tiagabine increases synaptic GABA through blockade of neuronal and glial GABA reuptake from the synaptic cleft. Two of the new generation drugs were designed based on the hypothesized role of glutamate neurotransmission in seizure generation. These agents employ a novel mechanism of action with activity at the NMDA ligand-gated calcium channel. The NMDA receptor is believed to play a fundamental role in the initiation and propagation of epileptiform activity. NMDA receptormediated blockade of the calcium channel results in inhibition of neuronal hyperexcitability, with the potential to not only control seizure activity, but to modulate the underlying epileptogenic defect. In practice, the NMDA receptor antagonists have not proven efficacious in controlling seizure activity, and the incidence of adverse effects with this class of drugs limits their clinical usefulness. The elucidation of the basic mechanisms of action of the anti-epileptic drugs and the development of novel agents has afforded us a greater appreciation for the complex molecular mechanisms underlying neural excitability and synchronization in epileptogenesis. While these drug discoveries allow clinicians to move from the empirical treatment of epilepsy to a more rational approach to seizure control, they may not have a significant impact on the effective treatment of intractable seizures. An approach to epilepsy treatment based on innovative understandings of epileptogenic activity and novel avenues of investigation holds the greatest promise for the future of epilepsy research and the ultimate development of a cure.


173

The Future of Pharmacotherapy in Epilepsy

The future of pharmacotherapy in epilepsy will be based on an appreciation of the molecular mechanisms that result in abnormal synchronized excitatory electrical activity in the brain. Understanding the basic principles of initiation, maintenance, and propagation of seizure activity and their relationship to the factors that influence susceptibility to spontaneous recurrent seizures offers the potential for fundamental shifts in the paradigms of therapy in epilepsy. Integration of our knowledge of pharmacogenetics, drug resistance mechanisms, drug delivery systems, and cellular and gene therapy will allow us to develop a more powerful approach to drug therapy in epilepsy. The most promising strategies in pharmacotherapy should aim to propel us beyond treatment of the symptoms of epilepsy, namely suppression of seizures, to prevention or cure of this devastating neurological disease.

Pharmacogenetics: A New Approach to Drug Therapy

An increased understanding of both the molecular mechanisms of epilepsy and the mechanisms of action of the anti-epileptic drugs has made it increasingly possible to move away from the empirical treatment of epilepsy and employ a more rational approach to pharmacotherapy. Yet our considerable knowledge of the etiology of epilepsy, including the underlying neuropathological processes, electrophysiology and biochemistry of a seizure, and specific therapeutic and adverse effects of each anti-epileptic drug has not enabled clinicians to anticipate an individual patient’s response to a particular anti-epileptic agent. The individual response to a specific drug is still empirical. As a result, patients with refractory epilepsy undergo multiple trials of single or combination therapy with significant risk of CNS or systemic toxicity and severe idiosyncratic reactions. A novel approach to enhanced drug therapy in epilepsy is based upon our understanding of genetic polymorphism in drug metabolism. Genetic variations in drug pharmacokinetics are a likely factor in refractory seizures due to either lack of anti-epileptic drug efficacy or intolerable side effects. Genetic differences may influence both the pharmacokinetics and pharmacodynamic effects of a particular anti-epileptic drug. Abnormal drug metabolism can be due to genetic polymorphisms that result in altered metabolic enzymes. Rapid metabolizers may have low plasma concentrations of a drug despite appropriate dosing and good adherence to therapy, and essentially appear to be refractory to medical management. Slow metabolizers can experience high plasma concentrations at normal drug dosages with resultant CNS and systemic toxicity that limits the use of an anti-epileptic drug.


174

The effects of genetic polymorphisms of cytochrome P450 isoenzymes on phenytoin metabolism have been studied extensively. In adult patients with epilepsy, impaired CYP2C9 or CYP2C19 isoenzymes were associated with altered phenytoin hydroxylation, which resulted in dramatically increased plasma drug concentrations even at normal or low phenytoin doses [4]. Molecular studies revealed two G to A point mutations in the CYP2C19 gene as the cause of the defect in the functional CYP2C19 protein (de Morias et al., 1994a,b), and multiple amino acid variants in the CYP2C9 protein have been described (Kaminsky et al., 1992). Identification of the genetic mutations in CYP2C19 allows for widespread genotyping by simple polymerase chain reaction restriction fragment length polymorphism for the defective isoenzyme. Identification of genetic polymorphisms in anti-epileptic drug metabolism has broad implications for the effective drug treatment of epilepsy. Knowledge of an individual’s genotype has the potential to allow for tailored drug therapy and effective control of seizures in someone who was once considered medically refractory due to aberrant drug metabolism. Idiosyncratic hypersensitivity reactions may also be influenced by alterations in metabolic enzymes. It has been proposed that CYP450 anti-epileptic drug bioactivation results in reactive metabolites that mediate the chemical modification of detoxifying enzymes. The aberrantly modified enzyme leads to deficient detoxification of anti-epileptic drug and a subsequent increase in the availability of bioactivated drug. A host-dependent immune response is believed to have a role in the complex series of events that lead to a hypersensitivity reaction. Further characterization of the pathways involved may allow for genotyping studies that will identify patients at risk for adverse effects with anti-epileptic drug therapy, and permit clinicians to take a truly rational approach to pharmacotherapy in epilepsy.

Pharmacoresistance in Epilepsy

Of patients who are refractory to first-line conventional anti-epileptic drugs, less than 5% will attain good seizure control with use of the newer agents, and a patient whose seizures cannot be adequately controlled with one anti-epileptic drug has only a 5–10% chance of controlling their seizures with multiple anti-epileptic agents [2]. In this population, multiple trials of drugs with differing mechanisms of action do not improve the chance of a positive response. Given the character of global resistance to different anti-epileptic drugs with varying mechanisms of action, it is unlikely that acquired alterations in the multiple receptors upon which anti-epileptic drugs act can adequately explain


175

pharmacoresistance. It is instead hypothesized that pharmacoresistant epilepsy may be due to altered permeability of anti-epileptic drugs across the BBB and that the mechanisms whereby drug access to the brain is limited are most likely nonspecific [5]. It is further proposed that nonspecific drug resistance mechanisms may represent adaptive changes in the epileptic brain [5]. As we continue in search of an effective drug treatment for pharmacoresistant epilepsy, it is crucial that we appreciate the full nature of drug resistance mechanisms and the specific character of those mechanisms in the epileptogenic brain. Our understanding of epilepsy continues to evolve and the pursuit of an effective drug treatment for epilepsy continues to advance. But if we are to truly make breakthroughs in the pharmacotherapy of epilepsy we must move beyond the traditional line of investigation. An appreciation of the mechanisms of pharmacoresistance in epilepsy is essential to the development of better epilepsy therapy. The discovery of cellular membrane proteins that function in chemotherapy-resistant neoplasms has advanced our understanding of drug resistance in epilepsy. Extensive biochemical study of the molecular character of these proteins has elucidated the mechanisms by which drugs are denied access to or extruded from the intracellular space, and overexpression of multidrug transporters in neoplastic cells has been shown to correlate with resistance to chemotherapeutic drugs. Multidrug transporters in the BBB have been described and alterations in these drug transporters are one proposed mechanism of drug resistance in epilepsy. P-glycoprotein and members of the multidrug resistance-associated protein (MRP) family are the primary drug transporters identified to have a role in drug-resistant epilepsy. P-glycoprotein is a transporter expressed in endothelial cells of the BBB that functions in the active transport of lipophilic molecules from the intracellular space into the vascular space. Studies of epileptogenic tissue surgically removed from patients with intractable seizures show an overexpression of P-glycoprotein at the epileptic focus [6] and experiments in animals provide evidence for the role of P-glycoprotein in regulating the brain concentrations of multiple anti-epileptic drugs [7, 8]. The family of proteins known as MRP also functions in the transport of lipophilic molecules at the BBB. These proteins are normally expressed in various tissue types throughout the body. In the brain, MRP expression is found normally in capillary endothelial cells. Studies in human brain tissue have shown the abnormal expression of MRP in neurons and glia. In surgically resected lesional tissue from patients with refractory epilepsy due to hippocampal sclerosis, focal cortical dysplasia, and dysembryoplastic neuroepithelial tumors, abnormal MRP expression was demonstrated in reactive astrocytes, dysplastic neurons, and capillary endothelium [9], suggesting a physiological basis for drug resistance.


176

While the overexpression of multidrug transporters in the brain has implications for the effective control of seizure activity with an anti-epileptic drug, the inability to adequately control seizure activity may have etiological implications for drug-resistant epilepsy. A recent study of seizure-induced expression of P-glycoprotein in rodents demonstrates that both acute and chronic epileptic activity significantly increases the level of P-glycoprotein mRNA [10]. These findings suggest that uncontrolled seizures may contribute to pharmacoresistant epilepsy. For the subpopulation of patients with a genetic predisposition to drugresistant epilepsy, the lack of effective pharmacotherapy to adequately control seizure activity may be especially detrimental. The overexpression of drug transporters in an epileptic focus appears to promote pharmacoresistant epilepsy by limiting local access of anti-epileptic drugs. In the future, novel pharmacotherapeutic approaches to epilepsy may include the development of anti-epileptic drugs that are not substrates for membrane permeability proteins and the adjunctive use of multidrug transport inhibitors with traditional anti-epileptic drugs. The lipophilic nature of most anti-epileptic drugs enhances their ability to penetrate the brain, but may inadvertently promote drug resistance in those patients that overexpress multidrug transporters. The development of new drug delivery systems that effectively deliver anti-epileptic agents to the brain, but avoid the potential to promote multidrug transporter-mediated pharmacoresistant epilepsy, is an area that deserves attention.

Epilepsy: An Autoimmune Disorder?

As the scientific community continues to search for the mechanisms that underlie intractable epilepsy, our understanding of the complexity of factors that influence the development of seizures continues to expand. In the late 1970s, based on empirical evidence from children with epilepsy who received immunoglobulin for recurrent upper respiratory tract infections, the theory of an immunological component of epilepsy was revisited. A decrease in the frequency and severity of seizures was observed in this population. This prompted a flourish of research in the area of immunological mechanisms in the central nervous system. Research has revealed new evidence that certain refractory seizure disorders may be autoimmune mediated. The presence of specific antibodies has been identified in Rasmussen’s encephalitis, noninflammatory focal epilepsy, and pediatric ‘catastrophic’ epilepsy and research has shown these antibodies to be directed towards the GluR3 subtype of the AMPA glutamate receptor [11]. Characterization of the antibodies indicates that some interact with the GluR3


177

receptor at a site distant from the glutamate binding site and function in channel opening and neuronal activation [12]. This provided the first evidence of autoimmune-mediated activation of a metabotropic ion channel receptor in the central nervous system. Further studies demonstrated the influence of IgG immunoreactivity and complement activation on neuronal death [13]. Cortical dysfunction via immune-mediated mechanisms may be due to the combination of excitotoxic overactivation of the neuronal glutamate receptor and activation of complement leading to neuron injury. It is hypothesized that Rasmussen’s encephalitis is an autoimmune disease that results in neuronal damage and subsequent development of seizures. Animal studies revealed that rabbits immunized with a portion of the GluR3 protein demonstrated production of anti-GluR3 antibodies and development of a Rasmussen’s type disorder characterized by inflammatory cerebral lesions and recurrent seizures. In a study of pediatric patients with Rasmussen’s encephalitis, serum GluR3 autoantibodies were identified and titers were correlated with seizure frequency [14]. In both studies, a response to plasma exchange to remove circulating anti-GluR3 antibodies was demonstrated. In subsequent studies, a positive response to the use of intravenous immunoglobulin in patients with the adult-onset variant of Rasmussen’s encephalitis was observed [15, 16]. In separate animal studies, though, mice immunized with the GluR3 peptide displayed significant pathological brain abnormalities, but no seizure activity [17, 18], lending support to the hypothesis that production of anti-GluR3 autoantibodies may be necessary for the development of excitotoxic and complement-mediated neuronal damage in Rasmussen’s encephalitis, but is not sufficient to play a primary role in development of epileptic seizures. A causative autoimmune mechanism has also been hypothesized for the intractable epilepsy associated with West’s syndrome, Lennox-Gastaut syndrome, and Landau-Kleffner syndrome. While most of the support comes from noncontrolled studies, small case series, or single case reports, there is evidence that immunotherapy with intravenous immunoglobulin can completely control seizures in a portion of these patients. There are multiple factors that may contribute to the development of epilepsy as a result of autoimmune-mediated neuronal activity. The presence of autoantibodies could have a role in modulating epileptic activity independent of glutamate receptor activation. Immunoreactivity-associated epilepsy is seen in Hashimoto’s and viral encephalitis where it is mediated by autoantibodies against voltage-gated potassium channels (Lang and Vincent, 1996). Animal studies have shown that mice immunized with GluR3B peptide demonstrate production of anti-ssDNA antibodies at levels similar to those seen in the mouse model of systemic lupus erythematosus [17, 18]. It has been documented that systemic lupus erythematosus patients exhibiting high levels of anti-ssDNA


178

antibodies can experience seizures, but it is not known through what mechanism anti-DNA reactivity influences epileptogenesis. As would be anticipated in a complex disease process that results in epileptic activity, genetic factors are also likely to influence susceptibility to autoimmune seizure development. Many autoimmune disorders, such as systemic lupus erythematosus and juvenile onset diabetes mellitus, are associated with genetic variations in MHC gene expression. There is evidence of an increased incidence of Rasmussen’s encephalitis in patients with specific human leukocyte antigen haplotypes. While the precise mechanisms whereby autoantibodies alter neural circuits and influence seizure activity are being elucidated, there is evidence that treating appropriately selected patients with immunotherapy is beneficial in reducing seizure frequency and improving function. In select patients with Rasmussen’s encephalitis, treatment with intravenous human immunoglobulin or protein A immunoadsorption resulted in decreased seizure frequency, improved cognitive function, and improvement on SPECT imaging [15]. Plasma exchange and intravenous IgG are beneficial in patients with antibodyassociated Hashimoto’s or viral encephalitis and epilepsy. Despite the lack of definitive data as to the precise role of autoantibodies in seizure development, empirical evidence does support a possible role for immunotherapy in select types of intractable epilepsy. The influence of immunological mechanisms in certain epilepsy syndromes provides support for the multifactorial nature of seizure etiology and encourages researchers to consider novel approaches to pharmacotherapy in epilepsy.

Neuroprotection: Preventing Epilepsy?

Traditionally, pharmacological neuroprotection is a concept primarily associated with acute neurodegeneration due to cerebral ischemia or traumatic brain injury. The goal of drug therapy in these settings is to restore the normal biochemical environment and protect neurons from the cytotoxic effects of inflammation and hyperexcitability. Anti-thrombotic, thrombolytic, and anti-inflammatory agents are used in stroke and head injury to prevent the sequelae of such insults to the brain. Recent studies of anticoagulation with unfractionated heparin following ischemia and brain trauma in animal models show a decrease in lesion size and improvement in motor and cognitive deficits [19]. Unfractionated heparin is believed to have not only anti-coagulant, but anti-oxidant, anti-inflammatory, anti-excitatory, and neurotrophic effects that may act synergistically to provide neuroprotection in acute brain insult. After traumatic brain injury, neuroprotective therapy generally focuses on prevention of secondary brain injury. Many pharmacotherapeutic agents are being


179

investigated and a variety of promising therapeutic options have emerged, including glutamate receptor antagonists, calcium channel antagonists, and free radical scavengers. A common theme in neuroprotection is prevention of seizures in patients who are at increased risk for developing epilepsy. Potential causes of epilepsy include cerebrovascular disease, perinatal hypoxia or ischemia, infection, febrile seizures, tumors, congenital malformations, trauma, and status epilepticus. Underlying genetic factors may also have a role in determining seizure susceptibility after an insult. The ability of the anti-epileptic drugs to prevent epileptogenesis in at-risk patients is largely unknown. Clinical studies of phenytoin, phenobarbital, carbamazepine, and valproate have failed to show a protective effect in the development of epilepsy after head injury. Studies in animal models, though, have shown that valproate may be effective in preventing epilepsy. These findings emphasize the complex multifactorial nature of seizure development and emphasize the need for multifaceted treatment strategies in epilepsy. Identifying the fundamental mechanisms of epileptogenesis may allow us to develop therapies that target the underlying disease process and effectively alter the development or progression of epilepsy. There is undoubtedly a cascade of disparate events that occurs in the development of epilepsy. A primary insult in the setting of genetic susceptibility may lead to fundamental structural or biochemical changes that result in spontaneous seizures and epilepsy. Pharmacotherapeutic strategies that can alter the sequelae of brain injury, influence the process of epileptogenesis, prevent or terminate seizure activity, modify underlying pathology, or interfere with multidrug resistance mechanisms do have an essential role in the successful treatment of patients with epilepsy.

Defective Ion Channels: From Pathogenesis to Therapy

Further development of effective therapeutic strategies will depend upon elucidation of genetic factors that influence epilepsy and novel approaches to modulating genetic defects. Altered gene expression and mutated gene products are known to play an etiological role in epilepsy. Genetic mutations have been identified in some of the rare familial epilepsies as well as some types of idiopathic epilepsy. Many of the identified genes encode for ligand- or voltagegated channels. A variety of paroxysmal disorders are due to defective ion channels, or ‘channelopathies,’ including some of the myotonias and other neuromuscular disorders, and long QT-syndrome. Defects in sodium, potassium, calcium, chloride, or glycine-mediated channels have been identified in these disorders. Epilepsy models employing sporadic mutant animal strains, genetically


180

engineered animals, and knockout animals, coupled with human studies have helped elucidate the role of ion channel defects in epilepsy. Ion channels have a recognized role in generating electrical currents and a channel defect could easily alter the balance of excitation and inhibition in neural networks, resulting in neuronal hyperexcitability and progression of epileptogenesis. All of the genes identified in the human familial epilepsies encode ion channels or auxiliary subunits. Genetic defects in ion channels not only contribute to idiopathic epilepsy, but may also participate in the development of post-traumatic epilepsy. Neuronal injury induces the expression of ion channels, and overexpression of abnormal ion channels may lead to epilepsy after a cerebral insult [20]. Recognition of some of the multiple genetic defects associated with epilepsy illustrates the incredible heterogeneity of genes that function in producing the epileptic phenotype. Just as the scientific community has adjusted its view of cancer etiology to emphasize a multifactorial basis, the approach to epilepsy research must consider the varied genetic and environmental factors that underlie the mechanisms of epileptogenesis. It is unlikely that a single mechanism is responsible for the development of seizures and progression to epilepsy. A singular approach to epilepsy therapy based on the premise that a single agent operating on a single mechanism can potentiate effective treatment is fundamentally limited. An appreciation for the diversity of genetic influences in seizure disorders will guide future therapeutic strategies.

Advances in Drug Delivery

The disappointing failure of the new anti-epileptic drugs to significantly improve outcomes in the pharmacotherapeutic treatment of epilepsy has prompted researchers to re-evaluate the current approach to seizure control and pursue innovative avenues of investigation. Traditional approaches to anti-epileptic drug development have focused on formulating new agents or improving existing agents. A novel avenue of drug development focuses on advanced delivery systems. From engineered drug reservoirs to implantable drug pumps and synthetic polymers, these new methods of drug delivery may prove successful in improving the efficacy while decreasing the systemic effects of anti-epileptic drugs. Moreover, new drug delivery systems may have applications beyond simple delivery of anti-epileptic drugs to delivery of cell and gene therapy agents. Special oral formulations of traditional anti-epileptic drugs have been developed and are in widespread clinical use. Some of these new forms of medication were designed for use in certain seizure settings or specific patient


181

populations, while others were developed to achieve the elusive goal of pharmacotherapy with enhanced efficacy and decreased systemic toxicity. For emergency situations that require quick drug delivery and acute seizure exacerbations, such as occur in febrile seizures, traditional oral therapy is not an effective mode of delivery. Transmucosal administration of anti-epileptic drugs theoretically allows for faster drug delivery. A rectal gel formulation of diazepam is available and commonly used in terminating acute seizure activity. Nasal or buccal mucosa administration of anti-epileptic medication may also achieve rapid delivery. While there are no formulations currently approved for administration through these routes, the off-label use of liquid benzodiazepine formulations via nasal and buccal routes is common practice. Drug delivery through the respiratory mucosa is an attractive alternative mode of therapy. The administration of inhaled drugs is a daily practice in the induction of anesthesia and the treatment of acute asthma exacerbations. Recently, inhaled chemotherapeutic agents have been used to treat lung cancer. The engineering of micro- and nanoparticulate systems with an adhesive coating may allow for better delivery to the alveoli and greater uptake across the blood-air barrier, and holds promise for the development of an improved anti-epileptic drug delivery system. Drug administration in the pediatric population has been made easier with the introduction of syrups, sprinkles, and chewable formulations of certain anti-epileptic drugs. Depot forms of anti-epileptic drugs, such as transdermal patches or subcutaneous implants, may also prove effective in improving pharmacotherapy in the pediatric and other populations, but these drug formulations are still only theoretical. While new anti-epileptic drug formulations increase the treatment options available to patients with epilepsy and may improve adherence to therapy, they have not had a significant impact on the proportion of patients that achieve seizure control. Some researchers argue that the focal delivery of anti-epileptic agents holds the most promise for significantly improving seizure control in patients with epilepsy. This is an area of drug development that has gained much attention in recent years. The administration of a prodrug that is systemically inert and becomes activated at the seizure focus holds the potential to inhibit seizure activity with minimal systemic toxicity. The development of the prodrug DP-VPA is already underway [21]. This engineered drug is composed of valproic acid coupled to a phospholipid moiety that serves to inactivate the valproic acid. At the seizure focus, abnormal neuronal activity results in elevated activity of phospholipases, which function to cleave the phospholipid moiety and release the activated drug. Direct administration of anti-epileptic drugs into the CSF space provides for the possibility of better prevention of seizure activity and may have a role


182

in termination of seizures. The use of an intraventricular or intrathecal catheter to deposit drug directly into the CSF bypasses one of the major problems in CNS drug delivery, penetration of the BBB, with the additional benefit of decreased systemic toxicity. Subcutaneous infusion pump implants are already in clinical use for the long-term intrathecal administration of analgesics for chronic pain control and anti-spasticity medications following brain injury. While the application of an intrathecal infusion pump system in the delivery of anti-epileptic drugs is still speculative, a collaborative effort to investigate the infusion of an NMDA antagonist via the intrathecal route is underway. Direct infusion of drugs into the CNS has proven successful in certain settings, but there are limitations to this therapy. The delivery of anti-epileptic drugs directly into a seizure focus is an innovative approach with immense potential to improve pharmacotherapy in intractable epilepsy. Microinjection and microinfusion systems have been used to investigate the effects of focal application of anti-seizure agents on seizure activity in animals. Local perfusion of an anti-epileptic drug directly into the seizure focus in an animal model of epilepsy was effective in attenuating ictal and interictal events. Pioneering work in the development of a computer-controlled drug delivery system coupled with a seizure detection device is being conducted (Stein et al., 2000). The automated system employs a seizure-prediction algorithm that activates a programmable infusion pump to deliver a predetermined amount of anti-epileptic drug directly into the seizure focus. Animal studies show the ability of such a system to shorten seizure duration and prevent subsequent seizures. The ability to prevent progression of partial to generalized seizures would prove invaluable to many patients who suffer from intractable epilepsy. More powerful detection algorithms hold the potential to deliver anti-epileptic agents into the seizure focus before a seizure is clinically evident. Engineered drug reservoirs such as liposomes and nanoparticles have broad applications as delivery systems and are currently under investigation for use in delivering anti-epileptic drugs to the CNS and directly to the seizure focus. These inert carrier vehicles direct agents to a target tissue via ligand-receptor binding, theoretically increasing potency at the target site while decreasing toxicity in other tissues. The specific delivery and penetration parameters of unmodified carriers vary depending upon the lipid composition of the vehicle. In contrast, modified carriers are tagged with a ligand or receptor that functions to increase vehicle affinity for a specific cell type, thus enhancing drug delivery to the desired site of action. These tags include antibodies, hormones, cytokines, toxins and engineered ligands. A major obstacle in the development of a system of drug delivery to the CNS is penetration of the BBB. The cerebral capillary endothelium and astrocyte foot processes that comprise this protective barrier play an important role


183

in maintaining a constant chemical environment and shielding neurons from toxic agents. Hydrophilic and large molecules are normally excluded from the CNS by this impermeable barrier. Liposomes are generally unable to penetrate normal endothelial pores (2 nm). The tight junctions and lack of pores that characterize the cerebral endothelium present a significant impediment to the vehicle-mediated delivery of drugs to the CNS. Deliberate chemical disruption of the BBB is already used to gain direct access to the CNS for the delivery of chemotherapy agents, and researchers are investigating the use of electrical disruption as a similar means of gaining direct CNS access [Orr, 2000]. It is known that seizures disrupt the BBB, which holds interesting implications for the focal delivery of an anti-epileptic drug immediately preceding or at the onset of a seizure. Moreover, the methods of direct CNS delivery of anti-epileptic drugs may also apply to the delivery of cell and gene therapy agents. An alternative approach to disruptive penetration of the CNS proposes the use of active targeting of transport vectors to circumvent the protective nature of the BBB. Endogenous peptides, modified proteins, and monoclonal antibodies can be used to transport large, water-soluble molecules across the BBB and deliver therapeutic agents directly to targets in the CNS. A recent study demonstrated the CNS delivery of systemically administered vasoactive intestinal peptide coupled to a monoclonal antibody [Bikel et al., 2001]. The use of this approach to gain access across the BBB will depend upon the identification of appropriate targets within the CNS. Its value as an effective method of drug delivery in epilepsy will require the characterization of the seizure focus and the discovery of common targets. As researchers continue to elucidate the cellular and molecular nature of epileptogenic brain the prospects for development of effective drug delivery strategies to treat epilepsy will continue to improve. The use of polymers to deliver anti-epileptic drugs directly to the seizure focus is another method that has potential for seizure control in intractable epilepsy. A polymer is a complex of drug in a dissolvable matrix. As the matrix dissolves, drug is released into the immediate area. Polymers can be engineered to vary dissolution rates in response to changes in the chemical environment. The use of polymers as a strategy in drug delivery is being employed in the treatment of recurrent glioblastoma multiforme. A polymer composed of an alkylating agent is implanted in the tumor bed following resection and slowly releases the chemotherapy agent directly into the malignant tissue. The use of polymers in anti-epileptic drug delivery to the brain has been investigated in animal models and was shown to decrease or attenuate seizure activity [22]. While this promising delivery strategy is still in development, the potential requirement of multiple craniotomies may make its use impractical.


184

As the cellular and molecular mechanisms underlying epilepsy are elucidated, it has become apparent that traditional strategies of pharmacotherapy will continue to fail until we fully appreciate the complex interplay of multiple factors that influence the development of treatment-resistant epilepsy. Innovative methods of enhanced drug delivery do indeed hold promise for effective treatment of intractable epilepsy, but even direct delivery of antiepileptic agents into a seizure focus does not attempt to modulate the fundamental physiological and structural changes that characterize epileptogenic brain. While the development of novel drug delivery systems is primarily directed towards improving the pharmacological treatment of epilepsy, these strategies have been extended to the delivery of cell and gene therapy agents in the hope of ultimately curing this multifaceted disease.

Current Alternatives to Pharmacotherapy

When pharmacotherapy fails to adequately control seizures there are a variety of adjunctive or alternative therapies available. While some may be considered extreme, in the patient with intractable epilepsy these nonpharmacological therapies offer the only possibility of a life free from seizures. Noninvasive adjunctive therapy is limited to the ketogenic diet. The options for invasive treatment of epilepsy are varied and include conventional surgical techniques, multiple subpial transection, implantation of a vagus nerve stimulator, gamma knife radiosurgery, and implantation of depth electrodes for deep brain stimulation. Dietary Treatment of Epilepsy The ketogenic diet has been utilized in the adjunctive treatment of refractory epilepsy for over 75 years. Its pattern of use in treating intractable seizures has varied and we are currently experiencing resurgence in the use of the ketogenic diet to control seizures that are resistant to pharmacotherapeutic strategies. The high-fat, low-protein, and low-carbohydrate diet produces a ketotic state similar to that seen in starvation. The diet has been evaluated using the MES and PET infusion models in a manner similar to anti-epileptic drugs. It has been shown that ketosis functions to increase the seizure threshold, but there is no evidence that it lessens the severity of seizures. The exact mechanisms whereby a ketotic state produces an anti-epileptic effect are unclear. Ketosis causes changes in overall brain energy metabolism, cell membrane lipid composition, localized cerebral pH, and brain water content, though the extent to which any one or a combination of these alterations influences seizures is not certain. The results of recent work in the biochemical basis of the ketotic effect suggest that alterations in brain excitatory amino acid metabolism


185

may be primarily responsible for the anti-seizure property of the ketogenic diet. Altered metabolism has the overall effect of shifting neuronal amino acid equilibrium to favor glutamate production at the expense of aspartate production, enhance synthesis of GABA from the increased pool of glutamate, increase removal of glutamate from the synaptic cleft, and enhance release of glutamate in the form of glutamine into the vascular space. The ultimate result of these biochemical alterations is the ability to attenuate and possibly terminate the development of seizures. Some reports indicate the ketogenic diet to be at least as effective as the anti-epileptic drugs in controlling certain seizure types, and even exceeds pharmacotherapeutic efficacy in some cases. Studies of efficacy vary and indicate that 14–46% of children on the diet experience a greater than 50% reduction in seizure frequency and 7–54% of children achieve complete seizure control [23]. While this can represent a significant improvement in seizure activity, in practice it seldom offers freedom from the burden of daily debilitating seizures. What the ketogenic diet does offer is a new paradigm to think about the molecular mechanisms of epilepsy. A change in the primary metabolic substrate in the brain from glucose to ketones appears to alter the seizure threshold in one area of brain without affecting the normal global function throughout. The ketogenic diet has utility in treating seizures of multiple types and various etiologies. This lends to the theory that there are biochemical pathways common to the different seizure types and etiologies that make diverse seizure disorders responsive to a single therapy. Elucidation of the fundamental changes in cerebral energy metabolism that underlie the development of seizures holds major implications for the development of successful therapy for epilepsy. Surgery: A Cure for Epilepsy? For over a century, the surgical resection of epileptogenic lesions has proven curative in certain types of epilepsy. At present, it remains the only true cure for certain subtypes of this devastating disorder, but remains significantly underutilized. The remarkable advances in diagnostic tools available to evaluate the epileptogenic brain and localize seizure foci have enabled surgical intervention to play an increasingly important role in epilepsy therapy. The use of magnetic resonance imaging, PET and SPECT functional imaging, electrocorticography, and implanted depth electrodes allow better localization of seizure foci. Coupled with refined surgical techniques, localization-related refractory epilepsy may be particularly amenable to surgical therapy. While the standard of medical management of epilepsy defines intractable epilepsy as seizures that continue after 2 years of therapy with at least two first-line anti-epileptic drugs, it is not uncommon for the patient with refractory epilepsy to have been tried on numerous traditional and new


186

generation anti-epileptic drugs in various combinations for more than 2 years. In the pediatric population there is an even greater reluctance to consider early surgical intervention, as the natural history of seizures in this population may be self-limited. The natural history of epilepsy is dependent upon several factors, including the clinical presentation, epilepsy syndrome, and etiology of epilepsy. The importance of recognizing self-limited disorders from those that are truly refractory becomes evident when one considers the cognitive and psychosocial consequences of intractable epilepsy. The ability to reduce the high morbidity that accompanies both temporal- and extra-temporal localization-related epilepsy provides a strong argument for early surgical intervention in children with intractable seizures of these etiologies. There are also nonlocalized epilepsy syndromes with severe seizures in which surgical intervention can have a significant impact on the progression of the devastating neurodevelopmental consequences. The capacity to actually affect the course of progression of a seizure syndrome is of considerable importance in the surgical therapy of intractable epilepsy. Anti-epileptic drugs may be of great benefit in controlling the seizures associated with a particular disorder, but they cannot offer any advantage in altering the natural history of the epilepsy. A 30–35% recurrence rate of seizures after the discontinuation of an anti-epileptic drug is not uncommon, while surgical therapy for certain seizure disorders can reduce or completely eliminate seizures in 70–90% of treated patients [24]. Surgical treatment of epilepsy is often considered radical and carries known risks. A rigorous presurgical evaluation is essential to identify those patients that will benefit from surgical intervention. Anatomical or functional hemispherectomy is generally reserved for patients with progressive devastating nonlocalizing epilepsies who are not candidates for a localized resection. If accomplished early enough, neuronal plasticity will allow for some recovery of function in selected areas. As previously indicated, this intervention can have a significant affect on the progression of neurodevelopmental consequences. Corpus callosotomy has benefit in treating frequent intractable drop attacks from both tonic and atonic seizures. Surgical intervention in this setting cannot confer freedom from seizures and is instead aimed at improving quality of life. Focal temporal cortical resection is effectively curative in appropriately selected patients with intractable complex partial seizures originating in the temporal lobe. Multiple surgical techniques have been employed, from amygdalohippocampectomy to en bloc resection. Focal extratemporal cortical resection is the most commonly employed surgical intervention in children. Presurgical evaluation is complicated in this population as the seizure focus is generally more difficult to localize in extratemporal as compared to temporal lobe epilepsy, and the focus may be located in a silent region of the brain and


187

clinically apparent only after spread to adjacent areas. In patients with a seizure focus either completely or partially localized to eloquent cortex multiple subpial transection allows for the surgical manipulation of epileptogenic cortex without significant neurological deficit. The procedure is designed to disrupt horizontal neuronal connections without interrupting efferent cortical fibers. The successful surgical treatment of epilepsy depends on the identification of the appropriate surgical candidate. In all cases, early identification of suitable syndromes and timely surgical intervention offers the best possibility for a normal life. In the future, perhaps, surgical intervention in epilepsy will not be limited to removal of epileptogenic brain or mechanical disruption of neural connections, but will instead be utilized in the cellular and genetic therapy of epilepsy to fundamentally alter the pathological processes underlying this debilitating disorder. Vagus Nerve Stimulation In 1997, a new treatment modality for intractable epilepsy was introduced. Vagus nerve stimulation is FDA-approved for the adjunctive therapy of refractory partial seizures. Patients who are not candidates for epilepsy surgery or elect not to undergo intracranial surgery may benefit from the subpectoral implantation of a programmable pulse generator and left mid-cervical vagus nerve electrodes for continuous cyclical stimulation. Studies evaluating the mechanism of action of vagus nerve stimulation indicate it functions via immediate and long-term effects. Short-term changes in the nucleus of the solitary tract and its connections cause synchronization and desynchronization of electrical activity in the brain [Magnes et al., 1961; Peñaloza, 1964; Chase et al., 1967]. The solitary tract nucleus has projections to the parabranchial nucleus, hypothalamus, amygdala, infralimbic cortex, ventroposterior, intralaminar, and midline thalamic nuclei, insular cortex, dorsal raphe, and nucleus ambiguous. Long-term changes in cerebral blood flow and neurotransmitter concentrations have been documented [25, 26]. Increased noradrenergic and serotoninergic activity, which functions to increase the seizure threshold, has also been hypothesized to play a role [27]. Blinded randomized controlled trials of efficacy of vagus nerve stimulation are methodologically difficult, but longitudinal studies demonstrate a mean reduction in seizure frequency of 22–48%, which varies with intensity of stimulation, with some patients experiencing >75% reduction in seizures [28]. While this may represent a significant improvement in seizure status for an individual, vagus nerve stimulation does not confer complete seizure control. A novel approach to vagus nerve stimulation employs a transcutaneous stimulator for noninvasive therapy of refractory partial seizures [29].


188

Gamma Knife Radiosurgery Advances in our understanding of the molecular mechanisms of epilepsy coupled with an increasing appreciation for the role of surgical treatment of intractable seizures has lead to the application of a variety of novel surgical techniques for control of seizures. Gamma knife radiosurgery allows for the precise irradiation of a specific target with minimal radiation effects on surrounding tissue. It is commonly used in the ablation of arteriovenous malformations and neoplastic lesions. In the early 1980s animal experiments established the role of ionizing radiation in restricting the spread of discharges in the epileptic brain. A role for radiosurgery in epilepsy therapy was first noted in a series of patients who underwent gamma knife surgery for the treatment of cerebral arteriovenous malformations and showed a concomitant improvement of seizures. Complete seizure control can be achieved after radiosurgery treatment of a lesion with seizures at presentation, and a significant decrease in seizure frequency can be seen in adults with low-grade astrocytoma and intractable epilepsy following conformal radiotherapy. Brachytherapy and conventional radiotherapy for low grade tumors with refractory seizures can also have a significant impact on seizure frequency. The success of gamma knife treatment of seizures associated with mass lesions essentially introduced the possibility of radiosurgery as effective therapy for focal epilepsy. Less than a decade ago the first patient with intractable mesial temporal lobe epilepsy (MTLE) was treated with gamma knife entorhinoamygdalohippocampectomy. Since then, studies evaluating radiosurgery instead of microsurgery for MTLE indicate it is an efficacious and safe treatment option that can reduce the morbidity associated with invasive surgical intervention [30]. The mechanism of action of gamma knife surgery is largely unknown. Computed tomography and MRI imaging show radiation-induced structural changes in the mesial temporal lobe, but the significance of these findings is unclear. Clinical studies of gamma knife surgery suggest improvement of seizures may represent an actual anti-epileptic effect independent of structural alterations. Using animal models it has been demonstrated that non-necrotizing doses of irradiation can improve seizures [31] and the anti-epileptic effect increases with increasing doses [32]. Biochemically, it is theorized that radiation inhibits protein synthesis, thus preventing maintenance of spontaneous bursting in neurons [33], and has a differential effect on the inhibitory GABA system and the excitatory amino acid system [31]. Developments in the techniques for noninvasive surgery provide novel treatment options for intractable epilepsy. Gamma knife radiosurgery offers the possibility of effective treatment for certain types of refractory seizures with minimal impact on normal brain function.


189

Deep Brain Stimulation Brain stimulation is a novel therapeutic strategy for intractable epilepsy with the potential to effectively control seizures associated with certain epilepsy types in patients who are not candidates for surgical resection. Neuromodulation of brain structures by electrical stimulation has been used to treat neurological disorders of movement and chronic pain. In the treatment of Parkinson’s disease, high frequency stimulation of the thalamus, pallidum, and subthalamic nucleus produces the same effect as ablative neurosurgery. Previous attempts to influence seizure activity by electrical stimulation of deep brain structures have targeted the caudate nucleus, anterior thalamus, centromedian thalamic nucleus, posterior hypothalamus, and hippocampus. Results have been varied and limited by study design. Recent attention to deep brain stimulation in epilepsy therapy is focusing on the subthalamic nucleus as a target for stimulation. While the inhibitory effect of high-frequency stimulation of the subthalamic nucleus on the substantia nigra was initially based on the theory that electrical stimulation inhibits function, several studies provide evidence for the molecular mechanisms that underlie inhibition via electrical stimulation. High-frequency stimulation of subthalamic neurons has been shown to produce a long-lasting blockade of depolarization of voltage-gated channels [34]. Both spontaneous and induced epileptiform activity has been reduced or terminated by high-frequency cortical, subthalamic, and hippocampal stimulation, and this inhibition of activity occurs when neurons are depolarized [35]. There is also evidence that high-frequency stimulation may activate inhibitory GABAergic circuits in the basal ganglia and inhibit postsynaptic activity in the subthalamic nucleus [36, 37]. Direct inhibition of deep brain structures may not be the only effect of electrical stimulation. An excitatory effect of high-frequency stimulation has also been supported. Functional imaging provides evidence for activation of stimulated structures. Neurophysiological studies indicate the findings from microelectrode recordings of stimulated structures are inconsistent with the hypothesis that high-frequency stimulation inhibits the target structure [38, 39]. Stimulation of deep brain structures may also affect cortical activity through anti-dromic connections. Using animal models, stimulation of the subthalamic nucleus has demonstrated retrograde activation of the corticosubthalamic pathway, a major afferent projection to the subthalamic nucleus, evidenced by measurable cortical potentials [40, 41]. Preliminary studies in patients with epilepsy demonstrated evoked cortical potentials after subthalamic nucleus stimulation [38]. It is unclear precisely how retrograde cortical activation could suppress seizure activity. Anti-dromic activation of cortical interneurons may be the mechanism whereby cortical excitability is inhibited. Computer models of high-frequency stimulation suggest simultaneous neuronal excitation and


190

inhibition may mediate the therapeutic effects of subthalamic nucleus stimulation in epilepsy [42]. Stimulation of the subthalamic nucleus in the treatment of certain intractable epilepsy types in humans has been successful in reducing seizures. Modulation of glutamatergic subthalamic output may influence cortical excitability by inactivation of the nigral control system as well as activation of cortical GABAergic inhibition, resulting in suppression of seizure activity. A greater understanding of the basic principles by which stimulation of deep brain structures can influence the endogenous control systems in the brain and modulate cortical excitability will promote the application of deep brain stimulation to certain intractable epilepsies.

From Drug Delivery to Gene Delivery

Novel drug delivery strategies are being applied to the cellular and genetic treatment of epilepsy. Cellular transplantation and delivery of genetic material hold the potential to not only effectively treat intractable seizures, but also offer the possibility of altering the fundamental defects that result in epilepsy. Cellular Therapy for Epilepsy Cell transplantation or replacement theoretically offers the possibility of providing a continuous endogenous supply of a deficient neuromodulator to a localized area of brain, essentially enabling the restoration of functional neuronal connections. The feasibility of this innovative approach to treating pharmacoresistant epilepsy has already been demonstrated. The intraventricular grafting of an adenosine-releasing synthetic polymer in an animal model of epilepsy was shown to significantly decrease seizure activity [43]. As an endogenous inhibitory neuromodulator, adenosine has the potential to influence the development of synaptic connections and alter the balance of excitation and inhibition in the epileptogenic brain. This pioneering work has important implications for cellular therapy of epilepsy. The grafting of stem cells or free cells from another species has the potential to establish operative neural circuits in areas of the brain that are intrinsically defective. Several studies have investigated cell transplantation in animal models of epilepsy. Neuronal grafts from fetal rats have been transplanted into rats modeling temporal lobe epilepsy and amygdala-kindled rats and demonstrated the ability to restore GABAergic interneuron connections and reduce epileptiform after-discharges and clinical seizures, respectively [44–46]. More recently, fibroblasts engineered to release adenosine and encapsulated into polymers were grafted into the ventricles of kindled rats. The grafted rats demonstrated a


191

significant reduction in kindled seizure activity and significant suppression of after-discharges [47]. The transplantation of neuronal grafts as a therapeutic strategy in epilepsy holds tremendous promise for effectively treating this disorder. Currently, one of the major limitations of such a therapy is immunological rejection. As the brain is an immunologically privileged site, a host-versus-graft immunoreaction to transplanted cells can occur. The ability to utilize immature cells that are immunologically neutral is one possibility in circumventing this problem. Continued research in the areas of transplantation and neural stem cells is crucial and will undoubtedly lead to advances in the application of this treatment option. Strategies in Gene Therapy for Epilepsy Many concepts in drug delivery and cell therapy can be applied to gene therapy for epilepsy. What gene therapy offers as an epilepsy treatment strategy is the possibility of correcting the defect that underlies epileptogenic brain and essentially curing epilepsy. Just as we have adjusted our view of the pathogenesis of epilepsy to encompass a complex multifactorial etiology with genetic influence, we must also revise our approach to the treatment of epilepsy to include novel concepts in genetic therapy. The use of animal models in the study of the molecular mechanisms that underlie hyperexcitability and epileptogenesis has contributed significantly to our understanding of the genetic basis of epilepsy. Spontaneous epileptic mutants involving both mono- and polygenic inheritance allow researchers to progress from phenotype to genotype and identify many of the genes involved in the development of cortical hyperexcitability. The use of engineered transgenic mouse models permits a genotype to phenotype approach that can enable the elucidation of the critical steps in epileptogenesis and function in the systematic testing of pharmacological therapies in epilepsy. Through the use of genetic animal models we have gained tremendous insight into the genetic abnormalities that can influence the intrinsic excitability of epileptogenic brain, the mechanisms of altered synaptic transmission, and disruptions in neural networks. An estimated 40–50% of epilepsy and epilepsy syndromes are considered idiopathic and presumed to have a genetic basis [1]. Most of the epilepsy syndromes are not likely to be the result of a single genetic defect, but the outcome of multiple factors, including a genetic abnormality. Researchers have identified a number of genes with known causative roles in the pathogenesis of certain epilepsy syndromes. Many of these syndromes are associated with metabolic derangements or neurodegenerative disorders. The genetic defects in several of the progressive myoclonus epilepsy syndromes have already been elucidated. Unverricht-Lundoborg disease, also known as progressive myoclonic epilepsy type I, is due to truncation or unstable


192

insertion in the gene encoding the protease cystatin B and MERRF, or myoclonic epilepsy associated with ragged-red fibers, results from defects in mitochondrial DNA. The genetic mechanisms underlying some of the familial epilepsies have also been identified, including autosomal dominant nocturnal frontal lobe epilepsy, benign familial neonatal convulsions, generalized epilepsy with febrile seizures, and episodic ataxia with partial epilepsy. In all of the familial epilepsies, the identified genes encode for entire cation channels or their subunits, further evidence that channelopathies may have an essential role in the development of seizure activity and progression to epilepsy. Continued research in the molecular basis of the epilepsies will undoubtedly elucidate additional mechanisms whereby genetic defects lend to the development of epilepsy. Elucidating the genetic mechanism underlying an epileptic disorder provides for the possibility of effective treatment by replacing either the defective DNA or abnormal protein product. Autosomal dominant nocturnal frontal lobe epilepsy is due to a genetic defect that results in a dysfunctional neuronal nicotinic acetylcholine receptor ␣4 subunit. Knowledge of this defect offers the possibility of a genetic therapy that could restore the CHRNA4 gene or replace the abnormal receptor subunit. In the most prevalent epilepsies where a single genetic defect is not the known etiology, gene therapy theoretically remains a reasonable treatment option. Similar to cellular therapy, genetic therapy in the setting of an unknown gene defect may prove therapeutic if utilized as a modality of drug treatment. The introduction of genetic material into the brain in this setting cannot correct underlying pathology, but it can provide a continuous source of neurotransmitter to normalize the balance between excitation and inhibition. Strategies in gene therapy have been applied to the treatment of cerebral neoplasia, neurodegenerative disorders, lysosomal storage disease, Parkinson’s disease, and stroke with varying success. As a therapeutic approach in epilepsy, gene therapy is an area of investigation still in its infancy. The successful delivery of genetic material into the brain is based on strategies borrowed from novel approaches used in drug delivery and cellular therapy. Penetration of the BBB remains a formidable challenge, but many of the strategies utilized in drug delivery, such as the coupling of modified proteins or monoclonal antibodies to transport vectors, as previously discussed, can be applied to the delivery of DNA. There are several methods whereby DNA can be delivered to cells in the brain. Both ex vivo and in vivo approaches to gene therapy are utilized in the delivery of genetic material to the CNS. Ex vivo methods are characterized by the introduction of transgenes into cells that are then grafted into the brain, while in vivo methods employ vectors to introduce transgenes directly


193

into cells in the brain. The use of viral vectors and the development of engineered vectors, such as plasmids and peptide vehicles, represent the variety of techniques whereby genetic material is delivered to the CNS. Herpes simplex virus (HSV), adenovirus, adeno-associated virus (AAV), retrovirus, and lentivirus are the commonly used viral vectors in gene therapy. The neurotrophic nature of HSV makes it a good candidate for delivery of genetic material to the brain, but its long latency period and resultant transient expression of gene products presents the major obstacle to its effective use as a genetic vector. The therapeutic use of HSV as a genetic vector has been demonstrated in rodent models of neuroprotection in focal cerebral ischemia. The use of an engineered gene that encodes a herpes simplex enzyme designed to activate a prodrug, such as herpes simplex thymidine kinase, is another strategy in gene therapy utilizing the HSV. Adenovirus vectors are also in use in experimental models of gene therapy. Adenovirus is known to infect both dividing and nondividing cells, which permits its use in both rapidly dividing malignant brain tumors and neurological disorders of the postmitotic CNS, such as Parkinson’s disease and epileptic disorders that generally do not exhibit cell division. AAV shows great promise for effective use in the delivery of genetic material to the brain. In vivo studies utilizing recombinant AAV vectors have demonstrated long-term gene expression without evidence of infection or immune response, and in a mouse model of traumatic brain injury recombinant AAV was successfully delivered to the hippocampus. The feasibility of vector-mediated gene transfer into the epileptogenic brain has been demonstrated in both rats and humans. For example, the tremor rat is a genetic mutant that exhibits absence-like seizures and is used as a model of inherited epilepsy. This rat is now known to be an animal model for Canavan disease. A deletion of the aspartoacylase gene has been discovered in these animals and the resultant high levels of N-acetyl-aspartate are understood to be responsible for the epileptic seizures in tremor rats. Recent studies utilizing the intraventricular administration of a recombinant adenovirus carrying the rat aspartoacylase gene demonstrated significant inhibition of the generation of absence-like seizures in experimental animals [48]. In human studies, the effective use of viral vectors to mediate the transfer of genes into human epileptogenic brain slices has been shown [49]. Nonviral methods of delivery of genetic material to the brain offer an advantage over viral vectors, which can be limited by inadequate brain penetration and ineffective cell transfection. The use of liposome-packaged plasmids conjugated to a monoclonal antibody has been shown to successfully cross the BBB and access the microvasculature and parenchyma of the brain. Viral genetic material can also be packaged into liposomes, termed virosomes,


194

for delivery to the brain. Virosomes containing AAV plasmids have been delivered into the ventricles in both primates and humans. Cationic vectors, such as lipospermines and polyethylenimine, can also be used to facilitate the transport of genetic material into the brain. The current body of research in genetic therapy demonstrates the potential for effective treatment of certain epilepsy disorders using a gene therapy approach. The transfer of genetic material into neurons in a seizure focus and the expression of inhibitory neurotransmitters or neuropeptides, membrane transporters, postsynaptic receptor subunits, or antisense sequences provide for the possibility of altering the path of signal transduction and inhibiting the initiation, propagation, or maintenance of seizure activity. Gene therapy has the potential to influence epileptogenesis due to a variety of causes, including neurological injury, defective ion channels, or altered levels of neurotransmitter or receptors. Further development of viral vectors that can be used to transfer therapeutic genes offers the hope of a cure for certain epilepsy disorders. Other avenues of investigation must focus on delivery of transgenes to the target tissue. The stereotaxic procedures of molecular neurosurgery provide a powerful method of delivery of genetic material to cerebral tissue. Currently, stereotaxic techniques are applied to neuroablative, neuroaugmentative, and neuroendoscopic procedures, as well as radiation dosing, anatomical-physiological correlation in neuroimaging, tumor biopsy and resective therapy, and restorative surgical therapy. Both point-in-space and volumetric techniques are utilized in stereotaxic procedures, but volumetric stereotaxis provides many advantages in molecular neurosurgery including localization of a target structure, conceptualization of the three-dimensional shape of a target structure, preoperative planning of surgical approach and trajectory, and positional differentiation of target and adjacent tissue. A stereotaxic neurosurgical approach to genetic therapy in epilepsy can be utilized to deliver a transgene-containing viral vector or genetically engineered cells to a seizure focus. When a focal seizure origin cannot be identified the global delivery of genetic material via the endovascular system can be accomplished using stereotaxic intraventricular or intraparenchymal injection, interstitial infusion, or catheter-mediated delivery of transgenic vectors. As the molecular mechanisms of the epilepsies are elucidated, it is becoming apparent that an aggressive approach to gene therapy in epilepsy must be pursued. The delivery of genetic material to cerebral tissue offers a therapeutic strategy that can alter the pathological basis underlying epileptogenesis in the human brain. The possibility of a cure for certain epilepsy disorders is on the horizon and we must continue to pursue innovative and novel approaches to gene therapy for this devastating neurological disorder.


195

RNA Expression Profiling: Pharmacogenomics and Disease ‘Fingerprinting’

The analysis of mRNA expression in individual cells provides a strategy to compare the transcriptional profile of individual phenotypically characterized neurons. This approach has been implemented in human and experimental epilepsy models and in live as well as fixed cell types. The use of an oligo-dT primer and T7 RNA polymerase permits amplification of a broad population of expressed genes across many gene families. The size range and complexity of the amplified mRNA provides a comprehensive view of differential gene expression in single cells. Individual cell differences in gene expression could be used to develop new targets for epilepsy pharmacotherapy, ‘personalize’ treatment with existing drugs, or ‘fingerprint’ individuals for disease diagnostics. The concept of ‘personalized medicine’ is now within reach due to the landmark innovation of the biochip and the wealth of information created by the Human Genome Project. The massive amount of genomic information generated by sequencing efforts could only be exploited by using complex bioinformatics tools to comprehensively analyze systems at the DNA, RNA, and protein level. These bioinformatics tools together with the data available from RNA expression profiling using DNA chips has led to the comprehensive analyses of individual clinical samples in an attempt to describe disease and disease risk at the molecular level and integrate data to facilitate clinical decision making. Pharmacogenomics aims to optimize patient management by customizing and synthesizing drugs based on genetic variations in drug response. Its thrust is based on genome-based rational therapeutics that addresses interindividual variations or polymorphisms affecting metabolism, receptors, and absorption that can influence drug sensitivity, toxicity, and dosing. Potential benefits of pharmacogenomics include increasing efficacy and preventing adverse drug reactions, thus improving patient care and decreasing costs. All of these technologies are not yet in current clinical use and it is also too early to decide whether molecular ‘fingerprints’ or genomic profiles will have the diagnostic and prognostic power currently predicted. One approach that we have taken to investigate new targets for epilepsy pharmacotherapy has been to profile dentate gyrus granule cells from human epilepsy tissue. Mesial temporal lobe sclerosis is a common pathological finding in patients with medically intractable temporal lobe epilepsy. This disease is characterized by extensive cell loss in the hilus and the hippocampal CA1 and CA3 cell fields in addition to synaptic reorganization throughout the dentate gyrus. The dentate granule cells from hippocampal slices of patients diagnosed with medial temporal lobe sclerosis exhibit reduced synaptic inhibition with concomitant hyperexcitability. These physiological changes have been studied


196

Control human dentate gyrus

a Epileptic human dentate gyrus

b GABA␣1 GABA␤1 NMDA1a GAD65 GluR1 GLT-1 HES S49

GABA␣2 GABA␤2 NMDA2a GAD67 GluR2 Netrin-1 INX EAAC1

GABA␣3 GABA␤3 NMDA2b GFAP GluR4 Netrin-2 CREB c-fos

GABA␣4 GABA␥2 NMDA2c pBS GluR5 Nestin OTX-1 c-jun

GABA␣5 GABA␥3 NMDA2d ␤-Actin GluR6 BF1 ARC AChE

GABA␣6 NOS VGAT GABA-T GluR7 BF2 4B2 NCAM

c Fig. 1. Radiolabeled amplified aRNAs are used as probes of small scale cDNA arrays containing candidate genes of interest. For each 32P-labelled aRNA from a cell, duplicated slot blots were used for each hybridization reaction. An mRNA expression profile could then be obtained. a Expression profile for autopsy control dentate gyrus granule cell. b Expression profile for medial temporal lobe epilepsy dentate gyrus granule cell. c Candidate genes and controls used.

relative to the hippocampi of patients with temporal lobe tumors in which the cell loss and synaptic reorganization are not seen. The synaptic reorganization of both excitatory and inhibitory systems in the dentate gyrus of the hippocampus may be an important mechanism that contributes to chronic limbic seizures. Of interest is the role of neurotransmitter receptors and their uptake sites in the generation of seizures in MTLE. Differences in gene expression in temporal lobe epilepsy have been reported from investigations on surgically removed hippocampi implicating an up-regulation in the expression of excitatory neurotransmitter receptor genes in a role in epileptogenesis. Increases in mGluR1 (Mathern 97, 98; Lynd 96; Blumcke 00), mGluR2 (Blumcke 96), NMDAR2a (Mathern 99), NMDAR2b


197

Fig. 2. Harvesting single cells. Human hippocampi fixed in paraformaldehyde were stained with TA51, an antibody for neurofilament. First round synthesis of cDNA was done on the histologically sectioned tissue. Single cells were identified and removed from the dentate gyrus granule cell layer. A glass microelectrode is shown moving in to harvest a single granule cell. Conversion to double-stranded cDNA, then amplification of this cDNA as radiolabeled antisense mRNA was then performed.

(Mathern 99, 98) mRNA expression have all been reported in the dentate gyrus from hippocampal surgical specimens. These findings support the hypothesis that changes in hippocampal circuitry alter the postsynaptic gene expression in a way that contributes to chronic seizure. Our strategy has been to remove individual dentate gyrus granule cells from fixed specimens (fig. 1) of surgically removed hippocampi from patients with MTLE and autopsy hippocampi, stained with TA51, an antibody for neurofilament. Radiolabeled aRNA from these cells was used to probe cDNA arrays containing the GABAA ␣1–6, and ␤1–3 receptor subunits, mGluR1–6, NMDAR 1A-B, NMDAR2A-D receptor subunits, GAD65, GAD67, and VGAT. The relative intensity of each mRNA-cDNA hybrid is then quantified (fig. 2). Selective differences can be found at the level of gene expression in hippocampal dentate gyrus granule cells from MTLE patients compared to nonseizure autopsy controls. Reduced transcription of select receptors and increased expression of other subunits in MTLE may contribute to epileptogenesis. Although select differences in mRNA expression can be found in human epilepsy tissue, it is the level of functional receptor protein, and any associated regulatory component, which will determine the functional significance of these findings. Routine biochemical analysis (e.g., Western blots) cannot be


198

performed at the single cell level in order to determine how protein expression correlates with mRNA expression. Immunohistochemistry is a more qualitative technique although it can be resolved at the single cell level. However, it cannot be used for more than two simultaneous antigens and is still not very specific for receptor subunits. It is likely, thus, that physiological or pharmacological analyses will be required to determine the functional significance of the expression differences. Despite these difficulties, ultimately, in order to understand epilepsy and develop highly targeted therapies, molecular characterization of individual neuronal cell types in critical areas of the involved CNS is likely to be necessary. References 1 2 3 4

5 6 7

8 9 10 11

12 13 14 15 16 17

Hauser WA, Annegers JF, Kurland LT: Incidence of epilepsy and unprovoked seizures in Rochester, Minnesota: 1935–1984. Epilepsia 1993;34:453–468. Regesta G, Tanganelli P: Clinical aspects and biological bases of drug-resistant epilepsies. Epilepsy Res 1999;34:109–122. Macdonald RL, McLean MJ: Anticonvulsant drugs: Mechanisms of action. Adv Neurol 1986;44: 713–736. Mamiya K, Ieiri I, Shimamoto J, et al: The effects of genetic polymorphisms of CYP2C9 and CYP2C19 on phenytoin metabolism in Japanese adult patients with epilepsy: Studies in stereoselective hydroxylation and population pharmacokinetics. Epilepsia 1998;39:1317–1323. Loscher W, Potschka H: Role of multidrug transporters in pharmacoresistance to antiepileptic drugs. J Pharmacol Exp Ther 2002;301:7–14. Abbott NJ, Khan EU, Rollinson CM, et al: Drug resistance in epilepsy: The role of the blood-brain barrier. Novartis Found Symp 2002;243:38–47. Potschka H, Fedrowitz M, Loscher W: P-Glycoprotein-mediated efflux of phenobarbital, lamotrigine, and felbamate at the blood-brain barrier: Evidence from microdialysis experiments in rats. Neurosci Lett 2002;327:173–176. Sills GJ, Brodie MJ: Update on the mechanisms of action of antiepileptic drugs. Epileptic Disord 2001;3:165–172. Sisodiya SM, Lin WR, Harding BN, Squier MV, Thom M: Drug resistance in epilepsy: Expression of drug resistance proteins in common causes of refractory epilepsy. Brain 2002;125:22–31. Rizzi M, Caccia S, Guiso G, et al: Limbic seizures induce P-glycoprotein in rodent brain: Functional implications for pharmacoresistance. J Neurosci 2002;22:5833–5839. Mantegazza R, Bernasconi P, Baggi F, et al: Antibodies against GluR3 peptides are not specific for Rasmussen’s encephalitis but are also present in epilepsy patients with severe, early onset disease and intractable seizures. J Neuroimmunol 2002;131:179–185. Twyman RE, Gahring LC, Spiess J, Rogers SW: Glutamate receptor antibodies activate a subset of receptors and reveal an agonist binding site. Neuron 1995;14:755–762. He XP, Patel M, Whitney KD, Janumpalli S, Tenner A, McNamara JO: Glutamate receptor GluR3 antibodies and death of cortical cells. Neuron 1998;20:153–163. Andrews PI, McNamara JO: Rasmussen’s encephalitis: An autoimmune disorder? Curr Opin Neurobiol 1996;6:673–678. Leach JP, Chadwick DW, Miles JB, Hart IK: Improvement in adult-onset Rasmussen’s encephalitis with long-term immunomodulatory therapy. Neurology 1999;52:738–742. Villani F, Spreafico R, Farina L, et al: Positive response to immunomodulatory therapy in an adult patient with Rasmussen’s encephalitis. Neurology 2001;56:248–250. Levite M, Hermelin A: Autoimmunity to the glutamate receptor in mice – A model for Rasmussen’s encephalitis? J Autoimmun 1999;13:73–82.


199

18 19

20

21

22

23 24 25 26

27 28 29 30 31 32 33 34 35

36 37 38 39 40

41

Levite M, Fleidervish IA, Schwarz A, Pelled D, Futerman AH: Autoantibodies to the glutamate receptor kill neurons via activation of the receptor ion channel. J Autoimmun 1999;13:61–72. Stutzmann JM, Mary V, Wahl F, Grosjean-Piot O, Uzan A, Pratt J: Neuroprotective profile of enoxaparin, a low molecular weight heparin, in in vivo models of cerebral ischemia or traumatic brain injury in rats: A review. CNS Drug Rev 2002;8:1–30. Westenbroek RE, Bausch SB, Lin RC, Franck JE, Noebels JL, Catterall WA: Upregulation of L-type Ca2⫹ channels in reactive astrocytes after brain injury, hypomyelination, and ischemia. J Neurosci 1998;18:2321–2334. Bialer M, Johannessen SI, Kupferberg HJ, Levy RH, Loiseau P, Perucca E: Progress report on new anti-epileptic drugs: A summary of the Fifth Eilat Conference (EILAT V). Epilepsy Res 2001;43: 11–58. Tamargo RJ, Rossell LA, Kossoff EH, Tyler BM, Ewend MG, Aryanpur JJ: The intracerebral administration of phenytoin using controlled-release polymers reduces experimental seizures in rats. Epilepsy Res 2002;48:145–155. DiMario FJ Jr, Holland J: The ketogenic diet: A review of the experience at Connecticut Children’s Medical Center. Pediatr Neurol 2002;26:288–292. Snead OC 3rd: Surgical treatment of medically refractory epilepsy in childhood. Brain Dev 2001;23:199–207. Henry TR, Votaw JR, Pennell PB, et al: Acute blood flow changes and efficacy of vagus nerve stimulation in partial epilepsy. Neurology 1999;52:1166–1173. Ben-Menachem E, Hamberger A, Hedner T, et al: Effects of vagus nerve stimulation on amino acids and other metabolites in the CSF of patients with partial seizures. Epilepsy Res 1995;20: 221–227. Rafael H, Moromizato P: Vagus stimulator for seizures. J Neurosurg 1993;79:636–637. Valencia I, Holder DL, Helmers SL, Madsen JR, Riviello JJ Jr: Vagus nerve stimulation in pediatric epilepsy: A review. Pediatr Neurol 2001;25:368–376. Ventureyra EC: Transcutaneous vagus nerve stimulation for partial onset seizure therapy. A new concept. Childs Nerv Syst 2000;16:101–102. Regis J, Bartolomei F, de Toffol B, et al: Gamma knife surgery for epilepsy related to hypothalamic hamartomas. Neurosurgery 2000;47:1343–1351; discussion 1351–1352. Regis J, Kerkerian-Legoff L, Rey M, et al: First biochemical evidence of differential functional effects following Gamma Knife surgery. Stereotact Funct Neurosurg 1996;66(suppl 1):29–38. Mori Y, Kondziolka D, Balzer J, et al: Effects of stereotactic radiosurgery on an animal model of hippocampal epilepsy. Neurosurgery 2000;46:157–165; discussion 165–168. Chalifoux R, Elisevich K: Effect of ionizing radiation on partial seizures attributable to malignant cerebral tumors. Stereotact Funct Neurosurg 1996;67:169–182. Beurrier C, Bioulac B, Audin J, Hammond C: High-frequency stimulation produces a transient blockade of voltage-gated currents in subthalamic neurons. J Neurophysiol 2001;85:1351–1356. Bikson M, Lian J, Hahn PJ, Stacey WC, Sciortino C, Durand DM: Suppression of epileptiform activity by high frequency sinusoidal fields in rat hippocampal slices. J Physiol 2001;531: 181–191. Iribe Y, Moore K, Pang KC, Tepper JM: Subthalamic stimulation-induced synaptic responses in substantia nigra pars compacta dopaminergic neurons in vitro. J Neurophysiol 1999;82:925–933. Kayyali H, Durand D: Effects of applied currents on epileptiform bursts in vitro. Exp Neurol 1991;113:249–254. Montgomery EB Jr, Baker KB: Mechanisms of deep brain stimulation and future technical developments. Neurol Res 2000;22:259–266. Dostrovsky JO: Immediate and long-term plasticity in human somatosensory thalamus and its involvement in phantom limbs. Pain 1999;suppl 6:S37–S43. Maurice N, Deniau JM, Glowinski J, Thierry AM: Relationships between the prefrontal cortex and the basal ganglia in the rat: Physiology of the corticosubthalamic circuits. J Neurosci 1998;18: 9539–9546. Maurice N, Deniau JM, Menetrey A, Glowinski J, Thierry AM: Prefrontal cortex-basal ganglia circuits in the rat: Involvement of ventral pallidum and subthalamic nucleus. Synapse 1998;29: 363–370.


200

42 43

44

45

46

47 48

49

McIntyre CC, Grill WM: Selective microstimulation of central nervous system neurons. Ann Biomed Eng 2000;28:219–233. Boison D, Scheurer L, Tseng JL, Aebischer P, Mohler H: Seizure suppression in kindled rats by intraventricular grafting of an adenosine releasing synthetic polymer. Exp Neurol 1999;160: 164–174. Loscher W, Ebert U, Lehmann H, Rosenthal C, Nikkhah G: Seizure suppression in kindling epilepsy by grafts of fetal GABAergic neurons in rat substantia nigra. J Neurosci Res 1998;51: 196–209. Shetty AK, Turner DA: Fetal hippocampal grafts containing CA3 cells restore host hippocampal glutamate decarboxylase-positive interneuron numbers in a rat model of temporal lobe epilepsy. J Neurosci 2000;20:8788–8801. Shetty AK, Zaman V, Turner DA: Pattern of long-distance projections from fetal hippocampal field CA3 and CA1 cell grafts in lesioned CA3 of adult hippocampus follows intrinsic character of respective donor cells. Neuroscience 2000;99:243–255. Huber A, Padrun V, Deglon N, Aebischer P, Mohler H, Boison D: Grafts of adenosine-releasing cells suppress seizures in kindling epilepsy. Proc Natl Acad Sci USA 2001;98:7611–7616. Seki T, Matsubayashi H, Amano T, et al: Adenoviral gene transfer of aspartoacylase into the tremor rat, a genetic model of epilepsy, as a trial of gene therapy for inherited epileptic disorder. Neurosci Lett 2002;328:249–252. O’Connor WM, Davidson BL, Kaplitt MG, et al: Adenovirus vector-mediated gene transfer into human epileptogenic brain slices: Prospects for gene therapy in epilepsy. Exp Neurol 1997;148: 167–178.

Albert Telfeian, MD, PhD Neurosurgical Associates, LLP 3601 21st Street, Lubbock, TX 79410 (USA) Tel. ⫹1 806 797 2222, E-Mail [email protected]


201


Emerging Treatment of Neurometabolic Disorders Roscoe O. Brady, Roscoe O. Brady, Jr. Developmental and Metabolic Neurology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Md., USA

Introduction

Metabolic disorders are caused by mutations in genes that result in harmful reductions of catalytic activity of any of the enzymes required for the normal ‘housekeeping’ functions of cells. Many of these conditions are characterized by the accumulation of deleterious amounts of nondegraded materials within lysosomes, and are commonly known as ‘lysosomal storage disorders.’ Lysosomes are subcellular organelles which contain a plethora of enzymes that are necessary for the biodegradation of subcellular materials. These enzymes are preferentially active under acidic conditions that are characteristic of the intraluminal milieu of lysosomes. Substances that undergo lysosomal biodegradation include glycogen, mucopolysaccharides, and the major class of lipids called sphingolipids. Because approaches to enzyme replacement therapy (ERT) and gene therapy are particularly advanced in the sphingolipid storage disorders, we shall limit this chapter primarily to considerations of these conditions, although the basic techniques of gene and ERT have wide applicability to many metabolic disorders of the brain. We shall indicate briefly what has been accomplished to date, and then describe approaches that we believe will be required for the effective treatment of patients in which neurological involvement is a prominent cause of morbidity and mortality.

Background

Sphingolipids have a portion of their common structure comprised of the long chain amino alcohol sphingosine (fig. 1a). In these lipids, a long-chain fatty acid

a

Sphingosine CH3-(CH2)12-CH⫽CH-CH-CH-CH2OH OH NH2 D-erythro-trans-2-amino-4-octadecene-1,3-diol

b

Ceramide Sphingosine CH3-(CH2)12-CH⫽CH-CH-CH-CH2OH OH NH CH3-(CH2)22-C⫽O Fatty acid

c

Glucocerebroside Sphingosine-Glucose Fatty acid

d

Sphingomyelin Sphingosine-Phosphocholine Fatty acid

e

Globotriaosylceramide [Ceramidetrihexoside (GB3)] Sphingosine-Glucose-Galactose-Galactose Fatty acid

f

Ganglioside GM2 Sphingosine-Glucose-Galactose-N-acetylgalactosamine Fatty acid

N-acetylneuraminic acid

Ganglioside GM1

g

Sphingosine-Glucose-Galactose-N-acetylgalactosamine-Galactose Fatty acid

N-acetylneuraminic acid

Fig. 1. Structures of pertinent sphingolipids. a Sphingosine; b Ceramide; c Glucocerebroside; d Sphingomyelin; e Globotriaosylceramide; f Ganglioside GM2; g Ganglioside GM1.

Neurogenetic Diseases

203

is linked to the nitrogen atom covalently bound to carbon atom two of sphingosine, forming a moiety known as ceramide (fig. 1b). In all of the sphingolipid storage disorders (except Farber’s disease) in which ceramide itself is the principal accumulating substance, additional components are linked to the oxygen on carbon one of the sphingosine moiety of ceramide. The most prevalent metabolic storage disorder of humans is Gaucher’s disease. Here, the principal accumulating substance is glucocerebroside, comprised of ceramide to which a single molecule of glucose is linked by a ␤-glycosidic bond (fig. 1c). Another common sphingolipid storage disorder is Niemann-Pick’s disease. Here, the accumulating material is sphingomyelin (fig. 1d). Still another prominent condition is Fabry’s disease in which globotriaosylceramide (ceramidetrihexoside)(Gb3) accumulates in many organs (fig. 1e). Of particular significance to neuroscientists are Tay-Sach’s disease in which the ganglioside GM2 accumulates (fig. 1f ) and generalized gangliosidosis in which ganglioside GM1 is the major accumulating metabolite (fig. 1g). The nature of the metabolic abnormalities in the sphingolipid storage disorders was established 38 years ago by Brady et al. [1, 2] with the demonstration that the enzymatic defect in Gaucher’s disease was the insufficient activity of glucocerebrosidase, the enzyme that catalyzes the hydrolytic cleavage of glucose from glucocerebroside. There are three principal clinical phenotypes of Gaucher’s disease. The first is Type 1 (nonneuronopathic) Gaucher’s disease in which the CNS is not involved. The second is Type 2 (acute neuronopathic) Gaucher’s disease that is characterized by early and extensive CNS damage. The term neuronophagia is used in the context of Type 2 Gaucher, because of the widespread destruction of neurons by monocytes that are attracted into the brain from the circulation by cytokines that are elaborated by damaged neurons. The third is Type 3 (chronic neuronopathic) Gaucher’s disease in which signs of CNS involvement occur later than in Type 2 Gaucher patients. Neurological manifestations in Type 3 patients may be confined to horizontal, or less frequently, vertical gaze paresis. Some Type 3 patients also have progressive myoclonic epilepsy that is notoriously difficult to control. The identification of the enzymatic defects in Niemann-Pick’s disease [3], Fabry’s disease [4], generalized gangliosidosis [5], Tay-Sach’s disease [6], and Krabbe’s disease [7] followed soon after the elucidation of the enzymatic defect in Gaucher’s disease. This information was used to develop widely used enzymatic assays for the diagnosis [8], carrier detection [9] and prenatal identification of fetuses at risk for these conditions [10–12].

Development of ERT

A long period of time elapsed before effective treatment for any of these debilitating conditions became available. Once again, investigations into the

Brady/Brady

204

nature of Gaucher’s disease were of paramount importance. Since glucocerebrosidase activity was less than normal in patients with this disorder, it appeared to be a fairly rudimentary task to purify this enzyme and determine whether its administration to Gaucher patients would be beneficial [13]. In order to minimize the possibility of sensitizing recipients to the exogenous protein, human placental tissue was initially used as the source of glucocerebrosidase. This treatment strategy, whereby exogenous enzyme is administered to a patient in whom a deficiency caused disease, was later termed ‘enzyme replacement therapy.’ These initial studies, using protein extracted from human tissue, were performed before the advent of recombinant DNA techniques made it possible to clone a gene and express the resultant protein without the need for biological source material. The first major impediment in establishing the effectiveness of this therapeutic approach was the difficulty in obtaining sufficient quantities of purified glucocerebrosidase to undertake clinical trials. Eventually, a limited amount of the enzyme was isolated in a sufficiently pure form that was believed not to be harmful if injected into patients with Gaucher’s disease. Small quantities of placental glucocerebrosidase were injected intravenously into 2 splenectomized patients with Gaucher’s disease. Percutaneous liver biopsy was performed before administering the enzyme and another the day after the injection. In both patients, the quantity of glucocerebroside in the postinfusion biopsy specimens was 26% less than that in the preinfusion biopsy samples [26]. Moreover, there was a long-lasting reduction of glucocerebroside in the blood [14]. An additional year was required to isolate enough enzyme to examine its effect in a third Gaucher patient. In this patient, only an 8% reduction of glucocerebroside occurred in the liver following the enzyme delivery, and there was no change in the blood level. Based on the amount of glucocerebroside in the biopsy samples from the third recipient, it was deduced that insufficient glucocerebrosidase had been administered to obtain a significant clearance of the accumulated glucocerebroside. Because of the difficulty in obtaining large quantities of the enzyme from the placenta with the original methods of purification, a new isolation procedure was developed by Furbish et al. [15] in the mid-1970s. It was found that glucocerebrosidase obtained by this procedure was inconsistently delivered to lipidstoring macrophages such as Kupffer cells in the liver. Macrophages have a lectin on their surface that has a particularly high affinity for mannose-terminal glycoconjugates [16]. In order to target glucocerebrosidase to these cells, the oligosaccharide side chains of glucocerebrosidase were trimmed with three exoglycosidases to produce mannose as the terminal sugar [17]. In experimental animals, this modification of glucocerebrosidase resulted in a 50-fold increase in the uptake of enzyme by Kupffer cells. A clinical trial was conducted in which


205

190 IU of mannose-terminal glucocerebrosidase were injected weekly over a period of 6 months into 7 adults and one child with Type 1 Gaucher’s disease. Only the child showed any beneficial clinical response [18]. A dose-response study was carried out by performing percutaneous liver biopsies before and after injecting the enzyme over a wide dosing range. The quantity of enzyme that consistently produced a reduction of hepatic glucocerebroside was 60 IU/kg of body weight. When this amount of mannose-terminal placental enzyme was given to 12 adults with Type 1 Gaucher’s disease every 2 weeks for a period of 6 months, striking beneficial effects occurred in all of the recipients [19]. Based on these findings, mannose-terminal glucocerebrosidase was approved for the treatment of patients with Type 1 Gaucher’s disease by the U.S. Food and Drug Administration on April 5, 1991. Recombinant glucocerebrosidase was subsequently produced in Chinese hamster ovary cells. This product is biologically equivalent to placental glucocerebrosidase [20] and was approved for the treatment of Gaucher patients in the USA in 1994. ERT for Gaucher patients was later approved in 55 countries. At this time, more than 4,000 patients with Gaucher’s disease throughout the world are being treated by ERT, based on work originating back in the 1960s. Extension of this treatment to patients with neuronopathic forms of Gaucher’s disease is of great importance. Reduction of hepatosplenomegaly and skeletal improvement was universal in clinical trials in patients with Type 3 Gaucher’s disease, but improvement of the supranuclear gaze palsy manifested by these patients has been inconsistent [21]. ERT also has been examined in patients with Type 2 Gaucher’s disease. Again, systemic improvement occurred in infants, but no amelioration of the CNS impairment was evident [22]. This finding is not surprising since it has been known for many years that intravenously injected enzymes do not reach the brain because of the blood-brain barrier [23]. Alternative delivery strategies have, therefore, been explored for the treatment of patients with neuronopathic Gaucher’s disease.

Substrate Depletion

Inhibition of the formation of glucocerebroside (substrate depletion) was proposed a number of years ago as a therapeutic strategy for the treatment of metabolic storage disorders [24]. The effect of blocking glucocerebroside synthesis with N-butyl-deoxynojirimycin (NB-DNJ) has recently been examined in patients with Type 1 Gaucher’s disease, and some apparently salutary effects have been reported [25]. N-butyl-deoxynojirimycin is a small-molecular-weight compound that has been shown to reach the brain of experimental animals when given orally in large doses. A trial of N-butyl-deoxynojirimycin has been

Brady/Brady

206

undertaken in patients with Type 3 Gaucher’s disease in whom the systemic manifestations are controlled by intravenous administration of mannose-terminal glucocerebrosidase. Abnormal saccadic eye movements of the recipients will be monitored as a critical clinical endpoint. This general approach to substrate inhibition is also relevant to other important lysosomal storage diseases such as Tay Sach’s and Sandhoff’s diseases, and successful studies in animal models have led to studies in human subjects. Advances in neurosurgical delivery may help to increase the effectiveness of the approach and offset the current limitations (e.g., toxicity, nontargeted delivery, insufficient penetration of target tissue in the CNS) of systemic dosing of substrate inhibitors.

Intracerebral Injection of Mannose-Terminal Glucocerebrosidase

It is likely that much, if not all, of the glucocerebroside that accumulates in neurons in the brain of patients with Type 2 Gaucher’s disease originates from the catabolism of larger sphingoglycolipids such as gangliosides (fig. 1f, g). Ganglioside turnover is most active during the neonatal period of life. Thereafter, it decreases to a constant level that is approximately 5% of the maximum velocity. Glucocerebrosidase activity in patients with Type 2 Gaucher’s disease is very low, usually in the range of 1–2% of normal [2]. It is conceivable that these patients might improve if glucocerebrosidase could be supplied to the brain during the neonatal period to catabolize glucocerebroside in neurons during this critical period of development. Investigators in the Surgical Neurology Branch of the National Institute of Neurological Disorders and Stroke (NINDS) have developed a technique called convection-enhanced delivery to deliver proteins in solution directly to the brain [27]. It was desirable to determine whether glucocerebrosidase could be delivered to the brain by this technique. An investigation was carried out to examine the safety of the procedure and the distribution of intracerebrally injected glucocerebrosidase in normal rats [28]. The procedure was found feasible and innocuous to the recipient animals. Glucocerebrosidase was carried by convective flow along white matter fiber tracts from the site of administration in the striatum to the cerebral cortex (fig. 2). The half-life of injected glucocerebrosidase in the brain was ⬃9 h, which is comparable to that in other major organs such as the liver following intravenous administration. The enzyme was specifically taken up by neurons, precisely the cells that appear to require it to prevent neuronophagia that is a hallmark of this condition (fig. 3). The reason for the selective delivery to neurons is believed to be due to a mannose lectin on their surface [29, 30] that is present in a lesser amount but is qualitatively similar to that on macrophages.


207

Glucocerebrosidase diffusion in rat brain Enzymatic activity mU

aft

fore

Striatum Left hemisphere Control side

12 11 10 9 8 7 6 5 4 3 2 1

Injection site

fore

aft

Striatum Right hemisphere Infused side

Fig. 2. Distribution of human glucocerebrosidase in the brain of normal rats following intracerebral injection of the enzyme. [Reproduced from 28 with permission of Wiley-Liss, a subsidiary of John Wiley & Sons].

Fig. 3. Immunohistochemical staining of human glucocerebrosidase in neurons of normal rats following intracerebral injection of the enzyme. [Reproduced from 28 by permission of Wiley-Liss, a subsidiary of John Wiley & Sons].

The safety and intracerebral distribution of glucocerebrosidase will be examined in nonhuman primates. If the results of these investigations are favorable, it may be worthwhile exploring this approach for the treatment of patients with Type 2 Gaucher’s disease during the neonatal period. It is likely that a combination of intracerebral and intravenous administration of the enzyme will be necessary. Moreover, it may be useful to include a substrate-depleting agent in the treatment regimen to reduce glucocerebroside formation. If intracerebral administration of glucocerebrosidase proves beneficial in patients with

Brady/Brady

208

neuronopathic Gaucher’s disease, it is likely that this approach will be extended to a number of other human metabolic disorders in which the CNS is involved. Again, intracerebral delivery will require neurosurgical expertise for clinical trials and will require the development of additional instrumentation and technologies to uniformly disperse the enzyme in the brain. Some areas for future development from a neurosurgical perspective might involve improved colloidal formulations of the enzyme solution as well as pumps and dispersion (injection) devices in the brain.

Gene Therapy

Gene therapy has been examined in patients with Type 1 Gaucher’s disease using retroviral transduction of bone marrow stem and progenitor cells. In the first recipient, no expression of the transgene was detected. In the second subject, expression of the transgene was detected over a period of several months [31]. Several steps are necessary before gene therapy for patients with Type 1 Gaucher’s disease becomes realistic. Among the initial goals that must be reached are: (1) more effective transduction of stem and progenitor cells with the gene of interest; (2) selective enrichment of transduced cells before their reintroduction into patients; (3) development of procedure(s) for the delivery and implantation of the transduced cells into the patient’s bone marrow, and (4) elimination of harmful effects of retroviruses including various forms of leukemia [32, 33] which have been attributed to gene therapy in the context of X-linked severe combined immunodeficiency. The use of self-inactivating lentiviral vectors has recently come under investigation. The principal advantages of lentivirus vectors are: (1) efficient integration into the genomes of target cells; (2) sustained long-term gene expression; (3) no apparent immune response, and (4) ability to infect nondividing cells such as neurons. Delivery of genes into nondividing cells with pseudotyped, high-titer, replication-defective HIV-1 vector has already been achieved [34]. This strategy was improved by the construction of a herpes simplex virus VP22 fusion protein that greatly increased the intercellular delivery of the test protein [35]. A logical extension of this investigation was the use of such a construct to deliver genes and their protein products to nondividing cells in the CNS [36]. Use of herpes simplex virus VP22 greatly enhanced the delivery of proteins between cells. Incorporation of the neuron-specific enolase promoter resulted primarily in the transduction of neurons within the CNS. These encouraging findings have led to the development of a lentivirus gene construct containing the human glucocerebrosidase gene and VP22. It is expected that the fusion product will assist in the intercellular transport of the


209

therapeutic protein to a major portion of the brain. If this hypothesis is substantiated, it is expected to provide a significant impetus for serious consideration of gene therapy for patients with neuronopathic Gaucher’s disease. Whether the results of such investigations can be translated to other metabolic storage disorders remains to be established. The development and exploration of gene therapy in authentic animal analogs of human enzyme deficiency conditions should significantly accelerate our sense of the potential of this approach and hopefully reveal any unanticipated difficulties prior to their application to patients. In addition to lentivirus, a number of other viral vectors and nonviral gene delivery systems must be considered. As time goes by, the limitations of effective gene therapy are more related to technical obstacles that are gradually being overcome, rather than fundamental problems with the gene therapy approach. One can envision a future time in which the promise of ERT and gene transfer are fully realized through advances in neurosurgical delivery and improvements in vector design.

References 1 2 3

4 5 6 7 8 9 10 11 12 13 14

Brady RO, Kanfer JN, Shapiro D: Metabolism of glucocerebrosides. II. Evidence of an enzymatic deficiency in Gaucher’s disease. Biochem Biophys Res Commun 1965;18:221–225. Brady RO, Kanfer JN, Bradley RM, Shapiro D: Demonstration of a deficiency of glucocerebrosidecleaving enzyme in Gaucher’s disease. J Clin Invest 1966;45:1112–1115. Brady RO, Kanfer JN, Mock MB, Fredrickson DS: The metabolism of sphingomyelin. II. Evidence of an enzymatic deficiency in Niemann-Pick disease. Proc Natl Acad Sci USA 1966;55: 366–369. Brady RO, Gal AE, Bradley RM, Martensson E, Warshaw AL, Laster L: Enzymatic defect in Fabry’s disease. Ceramidetrihexosidase deficency. N Engl J Med 1967;276:1163–1167. Okada S, O’Brien JS: Generalized gangliosidosis: Beta-galactosidase deficiency. Science 1968; 160:1002–1004. Kolodny EH, Brady RO, Volk BW: Demonstration of an alteration of ganglioside metabolism in Tay-Sach’s disease. Biochem Biophys Res Commun 1969;37:526–531. Suzuki K, Suzuki Y: Globoid cell leucodystrophy (Krabbe’s disease): Deficiency of galactocerebroside beta-galactosidase. Proc Natl Acad Sci USA 1970;66:302–309. Kampine JP, Brady RO, Kanfer JN, Feld M, Shapiro D: The diagnosis of Gaucher’s disease and Niemann-Pick disease using small samples of venous blood. Science 1967;155:86–88. Brady RO, Johnson WG, Uhlendorf BW: Identification of heterozygous carriers of lipid storage diseases. Am J Med 1971;51:423–431. Brady RO, Uhlendorf BW, Jacobson CB: Fabry’s disease: Antenatal detection. Science 1971;172: 174–175. Epstein CJ, Brady RO, Schneider EL, Bradley RM, Shapiro D: In utero diagnosis of NiemannPick disease. Am J Hum Genet 1971;23:533–535. Schneider EL, Ellis WG, Brady RO, McCulloch JR, Epstein CJ: Infantile (Type II) Gaucher’s disease: In utero diagnosis and fetal pathology. J Pediatr 1972;81:1134–1139. Brady RO: Sphingolipidoses. N Engl J Med 1966;275:312–318. Brady RO, Pentchev PG, Gal AE, Hibbert SR, Dekaban AS: Replacement therapy for inherited enzyme deficiency: Use of purified glucocerebrosidase in Gaucher’s disease. N Engl J Med 1974;291:989–993.

Brady/Brady

210

15

16

17

18

19

20

21

22

23 24

25 26

27 28

29 30

31

32

Pentchev PG, Brady RO, Gal AE, Hibbert SR: Replacement therapy for inherited enzyme deficiency: Sustained clearance of accumulated glucocerebroside in Gaucher’s disease following infusion of purified glucocerebrosidase. J Mol Med 1975;1:73–78. Furbish FS, Blair HE, Shiloach J, Pentchev PG, Brady RO: Enzyme replacement therapy in Gaucher’s disease: Large-scale purification of glucocerebrosidase suitable for human administration. Proc Natl Acad Sci USA, 1977;74:3560–3563. Stahl PD, Rodman JS, Miller MJ, Schlesinger PH: Evidence for receptor-mediated binding of glycoproteins, glycoconjugates, and lysosomal glycosidases by alveolar macrophages. Proc Natl Acad Sci USA 1978;75:1399–1403. Brady RO, Furbish FS: Enzyme replacement therapy: Specific targeting of exogenous enzymes to storage cells; in Martonosi AT (ed): Membranes and Transport. New York, Plenum, 1982, vol 2, pp 587–592. Barton NW, Furbish FS, Murray GJ, Garfield M, Brady RO: Therapeutic response to intravenous infusions of glucocerebrosidase in a patient with Gaucher disease. Proc Natl Acad Sci USA 1990;87:1913–1916. Barton NW, Brady RO, Dambrosia JM, DiBisceglie AM, Doppelt SH, Hill SC, Mankin HJ, Murray GJ, Parker RI, Argoff CE, Grewal RP, Yu K-T: Replacement therapy for inherited enzyme deficiency – Macrophage-targeted glucocerebrosidase for Gaucher’s disease. N Engl J Med 1991;324:1464–1470. Grabowski GA, Barton NW, Pastores G, Dambrosia JM, Banerjee TK, McKee MA, Parker C, Schiffmann R, Hill SC, Brady RO: Enzyme therapy in Gaucher disease Type 1: Comparative efficacy of mannose-terminated glucocerebrosidase from natural and recombinant sources. Ann Intern Med 1995;122:33–39. Altarescu G, Hill S, Wiggs E, Jeffries N, Kreps C, Parker CC, Brady RO, Barton NW, Schiffmann R: The efficacy of enzyme replacement therapy in patients with chronic neuronopathic Gaucher’s disease. J Pediatr 2001;138:539–547. Prows CA, Sanchez N, Daugherty C, Grabowski GA: Gaucher disease: Enzyme therapy in the acute neuronopathic variant. Am J Med Genet 1997;71:16–21. Johnson WG, Desnick RJ, Long DM, Sharp HL, Krivit W, Brady B, Brady RO: Intravenous injection of purified hexosaminidase A into a patient with Tay-Sach’s disease. Birth Defects Orig Artic Ser 1973;IX:120–124. Radin NS: Inhibitors and stimulators of glucocerebroside metabolism. Prog Clin Biol Res 1982; 95:357–370. Cox T, Lachmann R, Hollak C, Aerts J, van Weekly S, Hrebicek M, Platt F, Butters T, Dwek R, Moyses C, Gow I, Elstein D, Zimran A: Novel oral treatment of Gaucher’s disease N-butyldeoxynojirimycin (OGT 918) to decrease substrate biosynthesis. Lancet 2000;355: 1481– 1485. Bobo RH, Laske DW, Akbasak A, Morrison PF, Dedrick RL, Oldfield EH: Convection-enhanced delivery of macromolecules in the brain. Proc Natl Acad Sci USA 1994;91:2076–2080. Zirzow GC, Sanchez OA, Murray GJ, Brady RO, Oldfield EH: Delivery, distribution and neuronal uptake of exogenous mannose-terminal glucocerebrosidase in the intact rat brain. Neurochem Res 1999;24:301–305. Burudi EM, Regnier-Vigouroux A: Regional and cellular expression of the mannose receptor in the post-natal developing mouse brain. Cell Tissue Res 2001;303:334–339. Schueler U, Kaneski C, Murray G, Sandhoff K, Brady RO: Uptake of mannose-terminal glucocerebrosidase in cultured human cholinergic and dopaminergic neuron cell lines. Neurochem Res 2002;27:325–330. Dunbar CE, Kohn DB, Schiffmann R, Barton NW, Nolta J, Esplin J, Pensiero M, Long Z, Lockey C, Emmons RVB, Cski S, Leitman S, Kreps CB, Carter C, Brady RO, Karlsson S: Retroviral transfer of the glucocerebrosidase gene into CD34⫹ cells from patients with Gaucher disease: In vivo detection of transduced cells without myeloablation. Hum Gen Ther 1998;9:2629–2640. Hacein-Bey-Abina S, von Kalle C, Schmidt M, Le Deist F, Wulffraat N, McIntyre E, Radford I, Villeval JL, Fraser CC, Cavazzana-Calvo M, Fischer A: A serious adverse event after successful gene therapy for X-linked severe combined immunodeficiency. N Engl J Med 2003;348: 255–256.


211

33 34

35

36

Verma IM: A voluntary moratorium? Mol Ther 2003;7:141. Reiser J, Harmison G, Kluepfel-Stahl S, Brady RO, Karlsson S, Schubert M: Transduction of non-dividing cells using pseudotyped defective high-titer human immunodeficiency virus type 1 particles. Proc Natl Acad Sci USA 1996;93:15266–15271. Lai Z, Han I, Zirzow GC, Brady RO, Reiser J: Intercellular delivery of a herpes simplex virus VP22 fusion protein from cells infected with lentiviral vectors. Proc Natl Acad Sci USA 2000;97: 11297–11302. Lai Z, Brady RO: Gene transfer into the central nervous system in vivo using a recombinant lentivirus vector. J Neurosci Res 2002;67:363–371.

Roscoe O. Brady, MD Building 10 Room 3D04, National Institutes of Health 9000 Rockville Pike, Bethesda, MD 20892–1260 (USA) Tel. ⫹1 301 496 3285, Fax ⫹1 301 496 9480, E-Mail [email protected]

Brady/Brady

212


Gene Therapy for Parkinson’s Disease Piotr Hadaczek, Marcel Daadi, Krystof Bankiewicz Molecular Therapy Laboratory, Department of Neurological Surgery, University of California, San Francisco, Calif., USA

Why Gene Therapy for Parkinson’s Disease?

The main existing pharmacological therapy for Parkinson’s disease (PD) centers on replacement of dopamine (DA) by administration of the DA precursor L-dopa. In many cases, agents that prolong the action of DA by preventing its breakdown are also used to potentiate L-dopa effects [1]. Current problems associated with L-dopa treatment include motor fluctuations and choreic or dystonic involuntary movements (dyskinesias), which are superimposed on underlying breakthrough symptoms of bradykinesia, rigidity, and postural instability [2]. With the inevitable progression of the disease, L-dopa loses its initial effects of symptom relief. The major limitations of L-dopa treatment are 3-fold: inability to achieve site-specific delivery, which results in unwanted side effects and limits the amount of drug which can be given [3]; nonsustained drug levels within the central nervous system (CNS), thought to contribute to unpredictable ‘on-off’ effects [4, 5]; and progressive degeneration of DA-secreting nerve cells during treatment [6]. Development of new therapeutic approaches to PD must address these inadequacies of L-dopa. Most importantly, L-dopa addresses the biochemical sequelae of PD but does not address the underlying causes. Therefore, a primary goal for therapy of PD is the development of neuroprotective therapy which will slow down and prevent the death of neurons in substantia nigra (SN). Gene therapy encompasses any technique whereby an absent or faulty gene is replaced by a working one, so that a cell can make the correct enzyme or protein and consequently eliminate the cause of the disease. As a result, gene transfer may serve as a compensation for missing or defective protein expression. Several features of PD make it particularly suited for a gene therapy-based approach to treatment: (1) The pathology of the disease has been well characterized (loss of dopaminergic neurons and degeneration of the nigrostriatal

circuitry); (2) the initial pathology is confined to a discrete location within the brain where stereotactic targeting is possible (i.e., global gene transfer is not required in early stages); (3) disease processes such as apoptosis occurring within the SN may be prevented with a gene transfer approach, and (4) established animal models are available for testing clinical efficacy, safety and prognostic assessments. Gene therapy models for PD have focused on two treatment strategies. One is the replacement of biosynthetic enzymes for DA synthesis. It has been hypothesized that the transfer of genes involved in DA production would help to ameliorate the direct motor symptoms of the disease by the sustained local delivery of this neurotransmitter. The biochemistry of the DA synthesis involves several enzymes and cofactors. The rate-limiting enzyme in DA production is tyrosine hydroxylase (TH), which converts the amino acid tyrosine to L-dopa. L-dopa is then converted to DA by the aromatic amino acid decarboxylase (AADC) [7]. Another cofactor that is essential for DA metabolism is 6-tetra-hydrobiopterin, the level of which is limited by the availability of the enzyme GTP-cyclohydrolase 1 (GTPCH-1) [8, 9]. In theory, these enzymes could be genetically manipulated to produce increased DA levels. Some studies have reported a benefit from such enzyme replacement therapy, but others have challenged the relevance of providing biosynthetic enzymes to a milieu in which the cells are dying and incapable of properly using DA in any case. Another treatment strategy is providing neurotrophic factors for protection and restoration of dopaminergic neurons, thereby preventing them from further degeneration. Within each of these separate strategies, both in vivo (direct transfer of the gene into brain) and ex vivo (transplantation of genetically engineered cells) approaches have been considered. In vivo Approach Neurodegenerative diseases like PD are chronic; therefore, treatment needs to be longlasting. This situation makes PD particularly suited to treatment with viral vectors, where a single application of a vector can result in prolonged, stable transgene expression of relevant enzymes or growth factors. In vivo gene therapy involves direct gene transfer into the host somatic cell via viral vectors or a liposome vehicle [10]. There are several advantages of direct gene transfer over cellular transplantation: (1) vector delivery may be less invasive for the synaptic circuitry and brain parenchyma; (2) there is limited risk of unregulated cellular proliferation; (3) tissue-specific delivery systems for regulatable transgene expression can be designed, and (4) multiple genes can be administered at the same time. If gene therapy is to become a truly practical mode of treatment of PD, the therapeutic gene will need to be expressed for a sustained length of time and it

Hadaczek/Daadi/Bankiewicz

214

will require stable transduction and maintenance of gene expression in the targeted region of the brain and limited immune response to the vector and the target product. Various vector systems represent different features (table 1) that have to be carefully considered before clinical application. Both viral and nonviral vectors have become an important and powerful tool for gene delivery to the human nervous system. Neurodegenerative diseases like PD are suitable candidates for ‘molecular neurosurgery’ approaches because they are localized and specific regions of the brain are responsible for their development. Researchers have developed many different virus-based systems to manipulate subtle neuronal cell biochemistry and physiology. Crucial issues that need to be taken into consideration include transgene transduction efficiency, adverse tissue responses, targeting specificity, and regulation of transgene expression. Issues related to vector toxicity, long-term expression, gene regulation, vector production, CNS administration, and axonal transport will need to be addressed to develop an optimal gene delivery system for PD. Herpes Simplex Virus (HSV) – Based Vectors Herpes Simplex Virus 1 (HSV-1) has some features, which make it attractive as a vehicle for the delivery of therapeutic genes to the nervous system. HSV-1 is neurotropic and viral genomes persist as extrachromosomal elements. A neuronal-specific HSV promoter is capable of remaining active during viral latency, making HSV-based vector systems less susceptible to promoter silencing. Additional modifications to the viral genome (e.g., removal of genes responsible for the lytic cycle) reduce the cytotoxicity of the vector [11]. One technical advantage of the HSV genome for vector construction is that viral genes are almost entirely found as contiguous transcribable units, which makes their genetic manipulation relatively straightforward. The large-size viral genome of 152 kb permits insertion of a large size transgene and the ability to deliver multiple therapeutic genes via a single vector source. In general, HSV-based vector systems can be assigned to one of two major categories, either recombinant viral vectors or defective viral vectors. The recombinant HSV vectors carry a foreign gene in the native viral genome. They lack essential viral genes crucial for replication, but retain their ability to enter into the latency state within neuronal cells. The other type of HSV vector, the plasmid-based amplicon (‘defective’ HSV vector) contains approximately 1% of the HSV-1 genome and its backbone includes an eukaryotic plasmid modified by the addition of an HSV origin of replication (ori) and packaging sequence. This system requires a helper virus such as the wild-type HSV for high-level transgene expression and packaging [12]. Titers of amplicon stocks are typically lower than those of recombinant vectors (⬃106–107 units). Since defective HSV vector stock preparations may contain helper virus, the use of such vectors

Gene Therapy for Parkinson’s Disease

215

Table 1. Review of viral vectors used in gene therapy systems Hadaczek/Daadi/Bankiewicz

Vector

HSV

HSV amplicon

Ad

Minimal Ad ‘gutless’

Lentivirus (HIV)

AAV

Diameter

200 nm

200 nm

70–100 nm

70–100 nm

80 nm

20 nm

Size of viral genome

152 kb

minimal (only HSV replication and packaging origins)

30–40 kb

500 bp

7 kb

4.7 kb

Insert capacity

⬍30 kb

⬎50 kb

⬍8 kb

36 kb

7 kb

4.8 kb

Occurrence in cell

episomal

episomal

episomal

episomal

integrated in genome

episomal/ integrated in genome

Vector type

recombinant

defective

recombinant

defective

defective

defective

Viral contamination during production

yes

yes

yes

minimal

no

minimal

216

Immunogenicity

high

high

high

low

low

low

Titers (TU/ml)

1011

107

1012

107

107

1012

Major advantage

large capacity of transgene; high titers

large capacity of transgene; low immunogenicity/ toxicity

high titers

low immunogenicity/ toxicity; large capacity of transgene

ability to transduce dividing and nondividing cells

low immunogenicity/toxicity

Major limitation

transient expression/triggers immune response

low titers

triggers immune response/transient expression

tedious production

possible conversion to HIV-1; random genomic integration

small size of insert

in vivo may result in the expression of cytotoxic gene products from the helper virus, leading to neuropathological effects. Progress in reducing cytotoxicity includes improvements in the packaging system such as increasing the ratio between defective viral vector and helper; usage of helper virus with a larger deletion in IE3 (immediate early gene); using helper-free packaging systems; and improving purification of amplicon from helper virus. Studies using HSV vectors for gene therapy in PD have had mixed results. In a rat 6-hydroxydopamine (6-OHDA) PD model, HSV-based vectors containing the TH gene were delivered to the rat striatum and the animals appeared to demonstrate behavioral and biochemical recoveries for one year [13, 14]. Using the same animal model, neuroprotective effects on dopaminergic neurons have been demonstrated using glial derived neurotrophic factor (GDNF) and the apoptosis inhibitor Bcl2 with HSV-derived vectors. It was also shown that cotransfection of HSV-GDNF and HSV-Bcl2 had additive neuroprotective properties [15]. Both vector systems have shown reduction of amphetamine-induced rotations in the 6-OHDA rat model of PD. As mentioned, the main limitations of HSV systems include CNS cytotoxic effects and poor long-term gene expression, with limited number of cells expressing the transgene [13]. For these reasons, clinical use of HSV-1 appears to be impractical unless changes in vector design are implemented. Adenovirus Vectors Adenoviruses (Ad) have been a popular vehicle for gene transfer. Their attractive features include the capacity to accommodate large transgene inserts up to 36 kb and the ability to infect a wide variety of cell types and species (including postmitotic cells). The four main cell types in the brain which can be transduced by Ad vectors are neurons, astrocytes, oligodendrocytes, and ependymal cells [16, 17]. Recombinant Ad vectors have focused on deletions of E1, E2, E3, and E4 genes to reduce immunogenicity [11]. In humans, perhaps the most important quality is that Ad is not associated with any neoplastic disease and causes relatively mild, self-limiting illness in immunocompetent individuals (respiratory infection, keratoconjunctivitis, gastroenteritis). The development of Ad vectors of first, second and third generations are all based on deletions of one or more of these genes. Replication-deficient Ad vectors are propagated on special cell lines that provide functions of the early transcription units. The first-generation vectors (lacking E1) can still induce substantial inflammation, despite being replicantdeficient. Both viral proteins and therapeutic proteins were found to be targets for immune attack. Despite the lack of E1, viral proteins are expressed on firstgeneration vectors at levels sufficient to elicit a T-cell response [18]. Unfortunately, in addition, even the therapeutic gene will often be recognized as foreign by the


217

host. To bypass such an immune response, use of tissue-specific promoters has been proposed [19]. To improve the utility of Ad vectors for gene therapy, investigators have further modified the virus by mutating or deleting regions E2-E4 (second- and third-generation Ad vectors) [20–23]. Ads have not gained widespread use because of their inconsistent performance, probably due to the instability of the deleted vector genome [24]. Deletion of all Ad protein-coding sequences is possible with fully deleted Ad vectors (minimal or ‘gutless’ Ad vectors). The only Ad sequences that need be retained are ⬃500 bp of cis-acting DNA elements, including the viral inverted terminal repeats located at each end of the genome and the viral packaging signal. Current methods for producing gutless Ad involve its coreplication in the presence of a second helper virus that provides replicative functions. The advantages of this system are increased cloning capacity (⬃37 kb), increased safety, and potentially reduced immune responses due to elimination of viral sequences. The episomal nature of Ad often means that ultimately the transgene will be expelled from the cytosol during cell division. However, the genome may persist as an episome in nondividing cells (neurons) with sustained transgenic expression for longer than one year [25]. Nevertheless, repeat vector administration is probably required in order to boost transgene expression levels to initial levels. Unfortunately, in most cases administration to immunocompetent individuals results in the formation of anti-Ad neutralizing antibodies (directed at the vector capsid) which presents a significant barrier to vector readministration [26–28]. In vivo use of Ad vectors in PD animal models has focused on delivery of either Ad-TH or Ad-GDNF. When Ad vector encoding the TH gene was introduced into the striatum of 6-OHDA-lesioned rats, a reduced frequency of apomorphine-induced rotational behavior was observed [29]. TH expression, being confined predominantly to astrocytes, was demonstrated only for 1–2 weeks following gene transfer and an inflammatory response with gliosis was detected. More recent experiments with Ad vector encoding TH under the control of the repressible tetracycline regulatory system (‘tet off’) also showed that this vector mediates synthesis of TH in striatal cells. Transgene expression was observed in a large proportion of cells for at least 17 weeks, resulting in a significant overall reduction of apomorphine-induced rotation for at least 30 days. However, after 6 weeks, the pre- and postinjection outcomes were comparable [30]. In studies with multisite partitioned delivery of Ad-TH, Leone et al. [31] showed a correlation between the numbers of TH-immunoreactive cells and the loss of apomorphine-induced rotation, with a near-linear relationship between TH expression and phenotypical recovery. Those data suggested that only a fraction of striatal cells need to be transduced in order to exert phenotypical effects. Neuroprotective effects on dopaminergic neurons have been demonstrated when Ad-GDNF vectors were delivered [32, 33], with increased survival of SN


218

dopaminergic neurons and preservation of dopaminergic innervation to the striatum [33]. Studies by Connor et al. [34] also demonstrated that GDNF expressing Ad injected into the striatum and SN of aged rats, one week prior to 6-OHDA lesioning, allowed the production of GDNF at DA nerve terminals. However, only striatal injections of Ad-GDNF protected against the development of behavioral deficits characteristic of unilateral DA depletion. These results show that increased levels of striatal, but not solely nigral, GDNF biosynthesis prevents DA neuronal loss and protects DA terminals from oxidative damage from 6-OHDA lesioning. The development of gutless Ad from first-generation vectors was inevitable because of the shortcomings of the latter. For now, gutless Ad appears to be a promising vector platform for genetic diseases where long-term gene expression is required. With the advancement of vectorology, Ad-based delivery systems may be amenable to clinical applications in the future, but many problems remain such as immunological sensitization. Lentivirus-Based Vectors Lentiviral (LV) vectors are derived from a group of pathogenic retroviruses, which include human immunodeficiency virus (HIV). The retroviral machinery requires the conversion of the RNA genome to double-stranded DNA, mediated by the reverse transcriptase enzyme that is present in the infectious virion. The last step of the replication cycle leads to the integration of the provirus into the host genome. Once integrated, the provirus is ready to be expressed. The first retroviral vectors used for gene transfer were murine leukemia virus. Their use in the CNS was largely limited to ex vivo gene therapy as they were not able to transduce nondividing neuronal cells [11]. Lentiviral-based vectors share the properties of commonly used retroviruses with additional advantages: they can infect both dividing and terminally differentiated cells such as neurons; they have a large cloning capacity (⬎9 kb); they can be stably integrated into the genome of the target cells; and they do not encode viral proteins that can trigger an immune response. In current versions of HIV-1-based LV vectors, up to 60% of the viral genome has been eliminated and only three or four of the nine genes of HIV-1 are retained [35, 36]. Viral particles are generated by transient transfection of 293T cells with a three-plasmid system consisting of packaging, envelope, and transfer vectors [37, 38]. Splitting of the viral genome limits the formation of replication-component particles [35]. Through integration, retroviral vectors offer the opportunity of long-lasting expression, a major advantage in the treatment of genetic diseases. The level of expression in the brain can be further increased by the introduction of postregulatory elements that stabilize nascent RNA transcripts [39, 40]. In most versions of LV vectors, the particles are pseudotyped with the G envelope protein of vesicular stomatitis virus,


219

which gives the vector the capacity to infect a broad range of tissues including nervous tissues, and is probably responsible for their high affinity for neurons [37, 38, 41–43]. Because retroviral vectors integrate into the host genomic DNA, there are some major biosafety issues. The first is that nonspecific integration represents a potential source of genetic mutation. Host genomic integration with lentivirus appears to be a random process and important host cellular genes may be disrupted or activated [44–46]; however, in ex vivo human gene therapy for severe combined immunodeficiency X-1, a retrovirus was found to have preferentially inserted into an oncogene sequence in 2 separate patients. Other risks may result from the vector preparation itself (i.e., toxicity of viral proteins or compounds derived from the production system). Generating replication-competent retrovirus also is a major concern; primate studies had highlighted this potential risk of [47]. To minimize the risk of such recombinants, a self-inactivating version of LV has been developed [40, 43]. The self-inactivating design results in the removal of the major part of the viral transcriptional elements prior to integration, which also minimizes the chance that genes adjacent to the vector integration site will become activated. As with other gene transfer vectors, immunogenicity of retroviral vectors needs to be studied further before widespread clinical applications are possible. LV gene transfer into the monkey nigrostriatal system has been shown to induce minor perivascular cuffing, but without an apparent inflammatory response [48]. In Fisher rats, after intraportal infusion into the liver of more than 8 ⫻ 108 transducing units of an LV, a mortality rate of 74% was observed [49], clearly unacceptable for clinical implementation. Despite these caveats, significant advances in defective LV systems have provided a new perspective on gene delivery to the brain. Use of LVs in PD animal models has permitted delivery of GDNF to the striatum or SN. For example, Deglon et al. [39] were the first to examine lenti-GDNF delivery in a rodent model of PD. Their study indicated a significant sparing of nigral neurons after unilateral injection of lenti-GDNF over the SN. Georgievska et al. [50] similarly demonstrated structural and functional protection in 6-OHDA-lesioned rats with a LV. Similar results were observed in a mouse model of PD [51] by a different group. In a MPTP monkey model of PD, Kordower et al. [52] tested LV for intracerebral GDNF delivery. In treated animals, severe motor deficits were partially corrected, loss of dopaminergic neurons in SN was partially spared, and striatal dopamine innervation was preserved up to 70–80%. Consistent with these results, striatal 18F-fluorodopa uptake [Positron Emission Tomography (PET) prior to euthanasia] was increased by 300% in the lenti-GDNF-treated striatum. This work certainly supports the eventual use of lentivirus-GDNF treatment in the gene therapy of PD, though issues of long-term efficacy and


220

toxicity must first be addressed. In another in vivo approach by Azzouz et al. [53], a self-inactivating minimal LV expressing TH, AADC, and GTPCH-1 in a single transcriptional unit was designed. After stereotactic delivery into the DA-denervated striatum of the 6-OHDA-lesioned rat, sustained expression of each enzyme and production of catecholamines was detected, resulting in significant reduction of apomorphine-induced motor asymmetry during testing. Expression of each enzyme in the striatum was observed for up to 5 months after injection. These data indicate that production of three catecholaminenergic enzymes by a single LV can achieve functional improvement in 6-OHDA-lesioned rats. These results are somewhat tempered by work from other groups suggesting that multiple enzymatic delivery in 6-OHDA PD rats with gene transfer did not appear to create any additive effect beyond TH gene transfer alone [Janson, pers. commun.]. HIV-2 derived LVs have been used extensively for gene delivery to human neuronal cells. HIV-2 appears to be slightly less pathogenic than HIV-1 and because of limited sequence homology; cross-packaging of HIV-2 vectors into HIV-1 cores will minimize recombination between sequences in the transfer and packaging vectors. Gene transfer of AADC gene using the above-mentioned system was first examined in vitro by D’Costa et al. [54]. SVG cells (human neuronal cells immortalized by SV40 transformation) were transduced by both HIV-1 and HIV-2 based vectors carrying a cassette containing the AADC gene. Subsequently, gene transfer was evaluated by determining the ability of the transduced cells to convert L-dopa into DA. This conversion was measured in the intracellular compartment as well as in the secreted form in the supernatant. The results showed that both HIV-1 and HIV-2 AADC vectors successfully imparted the ability on transduced cells to efficiently convert L-dopa into DA. It was noted that the observed higher transduction for HIV-1 cross-packaged vectors was partly due to the higher titer of the latter vector. This approach provides the ability to combine gene transfer and standard drug treatment. These outcomes suggest that efficient HIV-2 vectors with a therapeutic transgene selfpackaged in HIV-2 cores, or cross-packed in HIV-1 cores, can be generated for the future treatment of PD. Adeno-Associated Virus-Based Vectors Adeno-associated viral (AAV) vectors are favorable candidates as gene delivery vehicles. They have many advantageous properties for gene therapy applications. For example, the parental virus does not cause disease; no viral genes are included in AAV recombinants and therefore, host immune response is minimized, the vectors transduce dividing or nondividing cells and a wide range of cells and tissues; expression can persist, mediating impressive longterm gene expression. One main limitation of AAV vectors is their small transgene capacity of ⬍5 kb per particle.


221

Wild-type AAV is the smallest (⬃20 nm) and simplest of the DNA replication defective viruses (Parvoviruses). The nonenveloped wild-type AAV particle contains a linear single-stranded DNA genome of 4.7 kb encapsidated with a simple three-protein capsid. The conversion of the single-stranded DNA genome to a double-stranded molecule is an important event required for the efficient function of AAV as a delivery vehicle. Its rate largely depends on the physiological state of the cell and may be a limiting factor for the transduction efficiency. Wild-type AAV has been shown to stably integrate into a single specific site within chromosome 19q13.3-ter [55]. Latent persistence occurs when AAV infects cells in the absence of helper virus (Ad or HSV). When cells containing an AAV provirus are superinfected with a helper virus possessing trans-acting elements (necessary for replication and packaging), the integrated AAV genome is ‘rescued’ and replicated to yield progeny AAV particles. There is divergence in tropism among various AAV serotypes (types 1–5) [56]. For example, recombinant AAV-5 and AAV-2 preferentially transduce neurons. Viral receptors strictly define the specific tissue tropism of a particular viral serotype. The general principles of AAV vector construction are based on the substitution of the AAV coding sequence with foreign DNA (transgene) to generate a vector plasmid. Only the AAV inverted terminal repeats flanking the transgene cassette must be retained intact. Current methods to produce stocks of defective AAV often use a human cell line (typically 293 cells) that is cotransfected with an AAV vector and a helper plasmid containing the AAV coding sequences (rep and cap genes flanked by Ad). The transfected cells are subsequently superinfected with Ad plasmid, which serves as a helper virus. The result of this system is a mixture of AAV vector particles and Ad particles. The Ad can be inactivated by temperature (56oC for one hour) and separated by CsCl-density centrifugation. Other, more recent, methods for obtaining high titers of AAV with no contamination by helper Ad have been developed [57, 58]; these typically use ‘triple transfection’ with a rAAV vector and two helper plasmids that serve the replicative roles of Ad and the packaging role of the AAV wild-type sequence. Purification of AAV is critical for clinical trials. A variety of chromatographic methods (e.g., ion exchange, antibody and heparin affinity resins) have been used in both conventional chromatography and HPLC systems. Vectors derived from recombinant AAV appear to exist as episomes and have not been shown to integrate to a significant degree. After AAV particles enter the nucleus, the vectors become circularized and ligated into larger concatameric molecules. Most of such molecules appear to persist for prolonged periods, perhaps even for the lifetime of the cell in the case of nondividing cells such as neurons. This phenomenon may help to explain long-term expression of transgenes delivered to the cell by recombinant AAV. Recombinant AAV vectors are considered one of the safest viral delivery systems, with minimal induction


222

of innate immune response. Nevertheless, there are reports describing generation of humoral neutralizing antibody responses to AAV capsid proteins following systemic delivery. This response may reduce the efficiency of the transduction, a consideration for systemic readministration [59, 60]. AAV vectors were first introduced into clinical trials for the treatment of cystic fibrosis [61, 62], using inhaled delivery for the treatment of the lungs. In the brain, two clinical protocols have been initiated thus far using AAV, for Canavan disease [Janson C, pers. commun.] and PD [During M, pers. commun.]. In preclinical work, long-term expression of AAV transgenes has been demonstrated in the CNS, including in the SN, globus pallidus, and striatum [63–71]. Mandel and colleagues [72, 73] examined the neuroprotective effects of intranigral AAV-GDNF injected 3 weeks before or immediately after intrastriatal 6-OHDA lesions. Significant neuroprotective effects were observed on the histological level in both versions of that experiment. However, no functional recovery was detected. More recently, Kirik et al. [67] examined the regional effects of AAV-GDNF delivery. A 6-month period of sustained expression was reported. Interestingly, GDNF expression and its protective effect were observed at both injection sites (nigra and striatum), but preservation of striatal dopaminergic fibers occurred only with striatal injection of the vector. Functional recovery also occurred only when AAV-GDNF was transduced into the striatum. It appears, therefore, that protection of dopaminergic terminals in the striatum is a critical feature in promoting functional recovery. Another approach using AAV vectors, replacement of DA biosynthetic enzymes, has been examined in various animal models of PD. Injection of an AAV vector containing the TH gene resulted in expression of TH enzyme in neurons as early as 24 h postinjection and persisted up to 7 months [66]. That study was among the first attempts to use enzyme replacement strategy with AAV as the delivery system. More recent studies have confirmed the performance of AAV both in terms of efficiency and the absence of cytotoxicity [74]. AAV-TH alone, however, was reported by one group to produce neither significantly elevated L-dopa levels nor significant behavioral improvements [63]. In addition to TH, that group found that gene transfer for other enzymes (AADC, GTPCH-1) was necessary for efficient DA production. Replacement of two or even three crucial enzymes in PD can be therapeutic. Indeed, behavioral recovery and effective dopamine production was achieved in combination with therapy with AAV-TH and AAV-AADC [75]. In turn, triple transduction with AAV-TH, AAVAADC, and AAV-GTPCH 1 showed improved rotational behavior lasting at least 12 months, and elevated DA production in rat striatum, compared with double transduction with AAV-TH and AAV-AADC [69]. This strategy extended the preclinical exploration to a primate model of PD and also showed some behavioral improvement with restoration of DA synthesis [76].


223

AAV/Lac-Z

AADC-IR

AAV/AADC

AADC-IR

Fig. 1. Convection-enhanced delivery (CED) of AAV2-AADC. Efficient technique for vector delivery is required for successful gene transfer. Convection-enhanced delivery can distribute AAV-2 vector in a nontraumatic and uniform fashion within monkey striatum [71]. Immunostaining of the monkey brain section for AADC shows the extent of transduction with AAV2-AADC (right). The section from the control monkey transduced with AAV2LacZ (left). Residual immunoreactivity (IR) for AADC is seen only in nucleus accumbens which is spared in PD.

An alternative approach of combined drug and gene transfer proposed by Bankiewicz and colleagues [70, 71] is based on the premise that a reduction in AADC might contribute to the loss of L-dopa therapeutic efficacy. Therefore, gene transfer to restore the decarboxylating capacity of L-dopa may result in a therapeutic gain with continued L-dopa dosage. In a MPTP monkey model [71], AAV2-AADC injected alone in the striatum was found to confer long-term (3.5 years) expression of the AADC gene (fig. 1, 2) with robust conversion of peripheral L-dopa to DA and some behavioral improvement. Modulating intrastriatal DA levels, by combination of AADC gene delivery and oral adjustments of L-dopa dosage, may provide a treatment strategy that could prolong L-dopa efficacy and reduce side effects seen from chronic high-dose oral drug therapy. Because DA levels are difficult to regulate after single or multiple gene transduction, the AADC and L-dopa approach is inherently more safe, though longterm efficacy is still unproven.


224

Fig. 2. Sustained AADC gene expression following AAV2-AADC gene delivery in PD monkey. AAV2-mediated gene expression can be long lasting (over 3.5 years) [147]. Picture was taken at 9 months post-transduction. Due to the neurotropic nature of AAV2, mostly medium spiny neurons are targeted.

An interesting approach to correct the physiological circuit affected by PD with AAV was recently proposed by Luo et al. [77]. The basis of their study design was the idea that marked improvement of the motor symptoms of PD occurs following subthalamic nucleus (STN) ablation or high-frequency stimulation. The projection axons from the STN end in excitatory synapses on target neurons in the SN pars reticulata, a major output pathway to the thalamus. Luo et al. generated AAV vectors containing two isoforms of glutamic acid decarboxylase (GAD65, GAD67), an enzyme which is responsible for conversion of glutamate to gamma amino butryic acid. Adult male rats were stereotactically injected into the STN with AAV-GAD vectors. Expression of transgenes was observed up to 5 months posttransduction. Transduced neurons, when driven by electrical stimulation, produced mixed inhibitory responses associated with the gamma amino butryic acid release. Three weeks after surgery, the ipsilateral medial forebrain bundle was lesioned with 6-OHDA, while control animals received AAV-GFP (green fluorescent protein) or PBS. These lesions led to impaired general locomotor activity and apomorphine-induced rotations contralateral to the denervated side in control animals. In the GAD65-treated rats, however, abnormal apomorphine-induced rotation was decreased by 65%. Immunohistochemical data revealed that 80% of dopaminergic neurons survived in the ventral tegmental area and 35% in the SN pars compacta. These results suggest that neurons generally considered excitatory and glutamatergic can express


225

GAD transcripts. Hence, AAV-GAD gene transfer into excitatory neurons may have clinical potential for the treatment of PD or other conditions associated with excessive excitation. One major concern regarding this study has been that AAV can spread from the site of administration via axonal transport, and thus expression could adversely influence neurons beyond the targeted site. Moreover, data in a primate model has not been made public and it will be important to confirm results from the rat model in a large-animal model that more accurately reflects the human physiology. In a Phase I study that was initiated at Long Island Jewish Hospital (USA), ablative surgery in the STN is proposed in case of any adverse effects involving uncontrolled expression of the GAD transcript. Hybrid Vectors As all vector systems have certain advantages and disadvantages, researchers have tried to combine elements from different viruses to create hybrid vectors with the most advantageous features for gene delivery into the CNS. Problems with current viral vectors include toxicity of viral proteins, difficulty in regulating transgene expression, and poor efficiency of transgene delivery and stability in host cells. New generations of chimeric viral vectors will be focused on targeting of specific tissues and cell types; achieving stable and regulated transgene expression through integration into the host genome or maintenance as episomal elements; accommodating large transgenes; retaining high-transduction efficiency; and minimizing adverse cytotoxic and/or immune responses. Different versions of chimeric delivery systems have already been proposed: Ad/EBV hybrid vectors, HSV/EBV/RV hybrid amplicon vectors, Ad/RV, Ad/AAV, HSV/ AAV, and others. Costantini et al. [78] used a HSV/AAV hybrid system and showed high-transduction efficiency and stability in culture.

Nonviral Approaches for Gene Therapy of PD

All the limitations of viral gene delivery systems (mentioned above) emphasize the need for alternative therapies with high effectiveness, specificity, and minimal side effects. Therefore, nonviral vectors have become an attractive option for gene delivery. Their low immunogenicity and easy large-scale production capability are among their most important characteristics. Naked DNA It was demonstrated by Wolff and coworkers [79] that simple administration of ‘naked’ or free DNA by intramuscular injection resulted in a fairly high level of expression in muscle. Later studies confirmed naked DNA gene transfer in other tissues (e.g., lung, heart, liver, kidney). The most likely mechanism


226

for cellular entry of such foreign DNA is based on receptor-mediated endocytosis [80]. While syncytial muscle fibers can readily uptake naked DNA, other tissues such as brain cells are not nearly as permissive, limiting this approach. Nevertheless, a number of routes and methods have been proposed for delivery of naked DNA into peripheral tissues, which may also apply to peripheral nerves. Examples include topical or intradermal; direct injections into deeper tissues, including intratumoral injections; and intravenous or intra-arterial. Most studies with naked DNA have focused on intratumoral injections as a possible anti-tumor strategy. For example, Coll et al. [81] showed that injection of naked DNA carrying Bax or p53 genes into a xenograft model of human lung non-small cell carcinoma could inhibit tumor growth. Naked DNA can be used as a DNA tumor vaccine; one such study showed anti-tumor immunity when naked DNA encoding the tumor antigen carcinoembryonic antigen or CEA was delivered by intrasplenic administration [82]. The transfer of naked DNA is gaining growing acceptance as a form of nonviral gene therapy; however, this technique is not sufficiently efficient in the brain. Lipid Vectors Liposomes have been used as drug carriers for many years. Several different liposomal formulations have been used in clinical trials. Cationic lipid is the most commonly used for such a purpose. To further stabilize liposomal structure, various polymers (commonly polyethyleneglycol or PEG) have been used, which may result in improved pharmacokinetics and biodistribution [83]. Cationic liposome-DNA complexes (plasmid DNA encapsulated in liposomal vesicle) are the most studied nonviral gene delivery systems in humans. After reaching the target cell, the DNA is carried across the plasma membrane, either by fusion or by endocytosis. Subsequently, DNA must be released from the endosome into the cytoplasm to avoid degradation in the lysosomes. Finally, the DNA must relocate from the cytoplasm into the nucleus to direct the expression of the gene products. All of these three steps (entry into the cell, escape into the cytoplasm, entry into the nucleus) are the main areas for chemists to design optimal formulations of lipids for gene delivery into cells. In the 6-OHDA rat model, Zhang et al. [84] demonstrated that it is possible to normalize brain TH enzyme activity by liposomal gene transfer via intravenous administration. The TH gene was encapsulated in 85-nm polyethylene glycol immunoliposomes and targeted to the brain with a monoclonal antibody to the rat transferrin receptor. In this manner, the gene was successfully delivered across the bloodbrain barrier and the plasma membrane. Three days after intravenous administration, striatal TH activity was normalized in association with a 70% reduction in apomorphine-induced rotation behavior, an approach that was repeated by others [10, 85].


227

Polymer-DNA Complexes (Polyplexes) Similar to cationic lipidic vectors, polycationic polymers can interact with the negative phosphate groups of DNA. These polyplexes protect DNA from degradation and enhance DNA uptake into the cell, resulting in efficient gene transfer. Cells take up condensed particles through a number of natural processes such as endocytosis, pinocytosis, or phagocytosis. Similar to lipid vectors, a polyplex has to pass the plasma and nuclear membranes. Different strategies have been proposed to improve transfection efficiency, improve specific targeting (e.g., conjugation with different ligands), prolong gene expression (e.g., insertion of regulatory sequences), and minimize toxicity. The most common polymers for DNA delivery include poly-L-lysine, protamine, polyethylenimine, and dendrimer. Several groups have already used polyplexes in animal models for cancer, an important application of gene therapy. Polyethylenimine/DNA complexes were also found to be efficient for in vivo gene transfer into neurons after stereotactic injection into the brain [86, 87]. A study by Wang et al. [88] demonstrated that polyethylenimine/DNA complexes migrate by retrograde axonal transport to neuronal cell bodies after being internalized by nerve terminals in the muscle, and confirmed the feasibility of nonviral gene delivery to the CNS via peripheral injection sites. This approach may have a number of clinical applications including PD, but specificity remains a problem.

Regulation of Gene Expression

Many proteins of therapeutic value posses a narrow window for optimal mode of action and have side effects and toxicities when overproduced. Therefore, gene therapy systems that introduce expression of an endogenous protein ideally should be regulated in vivo to achieve sustained transgene expression. For example, in the case of PD, too much DA production as a result of excessive DA biosynthetic enzyme expression can result in unmanageable dyskinesias and other serious side effects [89–91]. Early gene delivery systems generally relied on viral promoters to drive constitutive expression [11]. Disadvantages include loss of transgene expression over time and lack of well-regulated expression. Using promoters that are specific for particular cell types and tissues is one method of gene regulation, as their presence in a physiologically specific environment prevents gene silencing or shutdown of expression. The neuron-specific enolase, enkephalin, Purkinje cell-specific L7 protein, and myelin-basic protein promoters have been used as transcriptional activators in viral vectors to express transgenes in neurons, cerebellar Purkinje cells, and oligodendrocytes [92–95]. Xu et al. [96] compared a range of different mammalian CNS expression cassettes in AAV vectors using


228

different promoter sequences. The highest expression of reporter genes occurred when endogenous, nonviral promoters such as neuron-specific enolase and beta-actin were used in AAV-based vectors delivered into rat brain. The commonly used basal CMV promoter was found to be the weakest of those tested in vitro and in vivo. The choice of the proper promoter, therefore, is an important component of successful transgene expression. In a retroviral-mediated gene transfer system, Cortez et al. [97] used a glial fibrillary acidic protein promoter, whose activity is up-regulated in areas of gliosis often characteristic for PD. When astrocytes were transduced with the TH gene and implanted into the striatum of rats lesioned with 6-OHDA, a significant reduction in the turning behavior occurred for at least 4 weeks after grafting. The glial fibrillary acid protein promoter is of interest for gene therapy for neurodegenerative disorders, as it is active in the CNS throughout adult life and may serve as a disease-specific activator, since expression increases following many types of brain insults. It is important to investigate additional promoters to express transgenes in subpopulations of neurons most affected in neurodegenerative diseases such as dopaminergic neurons in PD. Gene expression can be manipulated by introducing a hybrid gene formed by linking a regulatory element upstream of the gene to be transcribed. One such strategy is to use a small-molecule drug that can cross the blood-brain barrier to act on drug-dependent promoters which directly activate or repress target gene transcription. Current drug-dependent gene-regulation systems use three general types of transcription factors: (1) drug-responsive elements (e.g., tetracycline, rapamycin); (2) nuclear hormone receptors (e.g., glucocorticoid-regulated systems), and (3) heterodimeric proteins (i.e., chemical-induced dimerization). At this time, the most commonly exploited transgene regulation systems use tetracycline as the activator or suppressor. The tetracycline-repressible system (‘tet-off’) works via negative control: the expression of the target gene is on in the absence of tetracycline and off in its presence [98]. The repressible system requires two gene sequences, the tetracycline transactivator (tTA) and the target gene that contains a promoter with tetracycline-binding sites (tetracycline operator, TetO). In the absence of an antibiotic, tTA has affinity for the TetO sites and stimulates transcription of the transgene (mode ON). When tetracycline is present, tTA protein changes its conformation and reduces its affinity for TetO sites so that the transcription is shut down [99]. The magnitude of the transgene repression in vivo can be as high as ⬃100-fold. The tetracycline-inducible system (‘tet-on’) uses positive control and works in the opposite manner [99]. The rtTA (reverse tTA) gene encodes a protein that has a very low affinity to TetO sites; however, when an antibiotic is added, the rtTA protein is converted to an active form which gains the ability to bind TetO sites and activates transgene expression. A similar principle is applied in anti-progestin or other hormone-inducible


229

systems. In the presence of an inducer, transgene expression is ‘on,’ while in the absence of a hormone, the promoter is not activated and the expression is effectively blocked. An advantage of drug-controlled gene transfer systems is that genes can be delivered in a relatively dose-dependent manner with more consistent and predictable expression. As these are only the first steps in controlling expression of transgenes, it is important to understand the limitations of regulated gene therapy systems before applying them in clinical trials. More studies in animal models should address issues of safety. The ideal solution would be to develop a system that would place a transgene under the control of both a tissue-specific promoter and a disease-specific promoter. The first reports of such advanced systems have already been published [100] and may soon be used in human studies.

Targeting of Viral Vectors

Viral surface proteins that bind to the specific cell receptors work as the primary means of initiating cellular attachment. The expression of specific surface molecules produces the tissue and cellular tropism for particular viral vectors. Most viral vectors transduce a relatively broad spectrum of host cells. The main goal of targeted gene therapy is to specifically infect a single cell type or group of cells and the choice of the right vector is crucial for specific and targeted gene delivery. For example, there are eight natural AAV serotypes which have been studied for gene transfer (AAV-1–8). In the majority of studies for PD, the neurotropic AAV-2 vector was used. AAV-5 has much broader tropism and also drives efficient gene expression in astrocytes and epithelial cells [56]. Other serotypes demonstrate preference for skeletal muscle (AAV-1), neurons (AAV-3), or ependymal cells (AAV-4). It is possible to modify viral surface structure by attaching or conjugating receptor ligands or antibodies. Restricting the vector’s ability to infect unwanted tissues decreases nonspecific infectivity, which is, of course, an unwelcome result of every in vivo gene therapy strategy. Incorporation of the vesicular stomatitis viral glycoprotein has been shown to increase infectivity of retrovirus [101–103], but decreases with HSV virions [104]. It is possible to generate an Ad vector expressing a chimeric fiber protein which alters the recognition profile of the virus. In the CNS, one could design a strategy with a fiber protein conjugated to a neurotrophic factor. This would preferentially target the vector to neurons expressing the receptor for the conjugated neurotrophic factor. The enhancement of the affinity of the virus for a particular cell type by modification of the viral coat could result in lowering the number of viral vector particles to


230

be used in vivo, an added advantage especially because it may have an important effect on the immune response against immunogenic vectors like Ad. Nonviral vectors (e.g., DNA-polyplexes, liposomes) have practically no selectivity at the level of their incorporation into cells; therefore, introducing specific ligands has been the major solution in designing targeted gene delivery. Attachment or incorporation of antibodies is a commonly used strategy. Receptor targeting increases transduction efficiency of disease-affected cells, while decreasing gene delivery to nontarget cells. This is perhaps of importance in PD where a very specific and isolated subset of neurons in the nigra and striatum is primarily affected. Of course, the use of selected promoters that are active only within subsets of cells or the use of cell-type-specific drug-inducible promoters, are solutions to the problem of nonspecificity in this context.

The ex vivo Approach

Ex vivo gene transfer offers the potential for persistent and regulated local and widespread delivery of therapeutic agents into the CNS. Several strategies utilizing genetically engineered cells for treating PD are currently under investigation. These strategies consist of introducing therapeutic genes to cells and grafting the modified cells into the diseased brain region. In PD models, genetically modified cells may be aimed at DA replacement in the denervated striatum, whereby the therapeutic cells are transduced with multiple genes that encode for the enzymes and cofactors involved in the biosynthesis of DA; or protecting the remaining midbrain dopaminergic cells that are still functional from degeneration. The therapeutic effect also may be aimed at rescuing DA cells that have begun the process of degeneration through the production of local trophic factors. Among growth factors that have been described to support the survival and/or regeneration of the midbrain DA neurons are brain-derived neurotrophic factor, basic fibroblastic growth factor (bFGF), insulin-derived growth factor, and glial cell line-derived neurotrophic factors (i.e., GDNF, neurturin, persephin, artemin). GDNF has the most prominent and selective effect in rescuing midbrain DA neurons, increasing DA activity and improving behavioral deficits of both rodent and primate models of PD.

Source of Cells for ex vivo Gene Therapy

There are multiple potential sources of cells for ex vivo gene therapy. The cells used as delivery vehicles must meet at least three basic criteria: (1) they should not form tumors in vivo; (2) they should graft at the site of the diseased


231

brain area, and (3) they should not elicit a strong host immune response. Autologous cells isolated from patient biopsies meet these criteria and neural cells derived from the brain in particular represent an attractive source. However, this source is unpractical because of the difficulties in isolating and maintaining neuronal cells in culture and obtaining adequate numbers for clinical applications. Myoblasts, skin fibroblasts, and bone marrow stem cells all have been considered for ex vivo gene therapy for PD. While myoblasts demonstrated some limited success, bone marrow stem cells transduced with either L-dopa or TH demonstrated limited neuronal differentiation and functional integration within host tissue. Skin fibroblasts have been a popular source for autologous cells; they demonstrated good survival in a primate model of PD, with expression of transgene that lasted for several months [74]. However, long-term gene expression by grafted fibroblasts has not been shown to be successful in rat models of PD. The reason for this failure may be due to inflammatory cytokine reaction to traumatic changes in the host tissue. It is important to note that these cells do not process the cellular machinery to store and release DA; to be competent for such a function within the host striatum, fibroblasts would need to be transduced with DA transporters and other genes involved in DA storage and release mechanisms [74, 105, 106]. Indeed, Lee et al. [106] cotransduced rat fibroblasts with both vesicular monoamine transporter-2 and AADC genes and demonstrated that these cells were then capable of converting L-dopa to DA and of storing DA. Transplantation of these engineered fibroblasts into a rat model of PD resulted in efficient DA production and storage. Other cell lines have been explored as a vehicle for gene transfer in PD preclinical studies, which include immortalized fibroblasts, immortalized fetal astrocytes, schwanoma and glioma cell lines, and endothelial cells [107–111]. The cells were engineered to produce enzymes or trophic factors and then grafted into the striatum of PD model. While these cell lines survived and expressed the transgene after implantation, most of them also gave rise to tumors, initiated immunological rejection, or did not integrate and died. Neural precursor cells are capable of giving rise to neurons, astrocytes and oligodendrocytes, and migrating and integrating into the local circuitry. These cells are preferred for grafting applications, as they approximate the normal physiological activity of neural cells. One approach is to isolate purified midbrain dopaminergic neurons by using a cell type-specific live monitoring technique. To achieve this selective isolation, Sawamoto et al. [112] generated transgenic mice and rats expressing GFP under the control of a 9-kb rat TH promoter. The authors demonstrated that expression of GFP protein was specific to DA neurons in the mesencephalon. The rat fetal midbrain was dissected out and dissociated cells were sorted using the fluorescent activated cell sorting. This purification step yielded an enriched population of TH-GFP positive neurons


232

(⬎60%). This sorted and enriched population of DA neurons improved the behavioral deficits of 6-OHDA lesioned rats. A similar approach was applied by the same group to isolate midbrain precursor cells [113]. The GFP reporter gene was under control of the neural-specific second intronic enhancer of the nestin gene. Nestin GFP mice showed strong fluorescence in the ventricular zone where precursors are known to reside. Neurospheres generated from the fluorescent activated cell-sorted nestin-GFP precursors were implanted into the 6-OHDA rat model of PD and 5 weeks after transplantation the rats demonstrated a significant behavioral improvement. Other studies have used nonmidbrain-derived stem cells transduced with the transcription factor Nurr1 (which promotes a dopaminergic neuronal phenotype) in order to induce stable midbrain DA lineage. For instance, to induce a dopaminergic cell lineage, Wagner et al. [114] transduced C17.2 cells [115] with Nurr1 and incubated the cultured cells with soluble factors derived from ventral mesencephalic type 1 astrocytes. This treatment resulted in the induction of dopaminergic fate in 80% of the total cells [114]. The genetic modification of stem cells with Nurr1 was also required to efficiently convert embryonic stem (ES) cells to DA neurons. Kim et al. [116] demonstrated that overexpression of Nurr1 alone promoted a 10-fold increase in the number of TH-expressing neurons. The administration of Shh (Sonic hedgehog) and FGF8 resulted in an additional 5.6-fold increase in the proportion of TH-positive neurons. The ability of these newly induced TH-positive neurons to synthesize and release DA was demonstrated using HPLC. After implantation into the 6-OHDA lesioned rats, TH-positive neurons survived and extended processes within the host parenchyma. These cells were postmitotic as demonstrated by the absence of the cell proliferation marker expression Ki-67. In other studies, Nurr1 ES cells grafted in parkinsonian rats improved their rotational behavioral and motor skills as tested in the adjusting step, cylinder, and paw-reaching tests. Fetal Mesencepahlic DA Neurons In both rodent and monkey models of PD, midbrain-derived fetal tissue implants are able to survive, extend neurites, make functional synaptic contact with host neurons, and secrete DA leading to a dramatic improvement in behavioral deficits [117–120]. Based on these studies, clinical trials of primary fetal nigral cell transplantation for medically intractable PD were initiated in the 1980s. Open-label clinical trials with mesencephalic DA neurons obtained from 6- to 9-week-old aborted human fetuses demonstrated graft survival, DA storage and release (assayed with PET), and significant and persistent improvement as measured with Unified Parkinson Disease Rating Scales (UPDRS) relative to the baseline. Histopathological demonstration of striatal DA reinnervation was obtained for a period extending up to 10 years [121–125]. These encouraging


233

findings have been recently called into question, however, by the first double-blind placebo-controlled clinical trial [89]. This large, placebo-controlled study involving sham surgeries enrolled 40 patients, 20 transplant subjects and 20 controls who underwent sham surgery. The graft survival rate in transplanted patient was 85%, without the use of immunosuppression. Young patients (⬍60 years) demonstrated 28% and 34% improvement in UPDRS total and motor ‘off’ components, respectively. However, despite cell survival and DA production demonstrated with PET, clinical improvement was not as dramatic in patients in the typical age range for PD (⬎60 years). Of greatest concern were 5 patients who developed severe dyskinesias and/or dystonia in the absence of levodopa. This graft-induced dyskinesia was termed ‘runaway’ dyskinesia because it was an uncontrolled, off-medication side effect. The authors’ interpretation was that the side effects were due to graft overgrowth and an excess of DA production. Subsequently, it has been argued that this conclusion was somewhat simplistic and the negative outcome of this study is partially attributable to extended culture of donor tissue, unconventional neurosurgical procedures, and an absence of immunosuppression [126, 127]. Nevertheless, together with a study by Fahn and coworkers [89] using standardized cell transplantation procedures and assessment protocols, Hagell et al. [128] observed a pronounced ‘runaway’ dyskinesias due to fetal mesencephalic DA cell grafts. As with the trial by Fahn, these dyskinesias persisted after withdrawal of L-dopa and DA agonists [128]. In the latter study, the authors argued that the ‘runaway’ dyskinesias are not caused by overgrowth of grafted cells, but could be due to micrograft DA spillover which overstimulated supersensitive receptors outside the graft-innervated area. These authors also speculated that the development of ‘runaway’ dyskinesias could be due to the extended storage or culture of donor tissue before grafting or to transplantation-evoked changes in host striatum or nondopaminergic components of the grafts. More recently, a third report on the second double-blind, placebo-controlled trial of fetal nigral transplants [91] demonstrated a failure to induce significant motor improvement relative to placebo. Moreover, patients who received grafts also developed severe ‘runaway’ dyskinesias that tended to appear 6–12 months after transplantation. Patients with a higher dose (4 donor grafts) showed improvement at 6 and 9 months and deteriorated thereafter, coincident with the termination of cyclosporine intake, suggesting a possible immune reaction against the graft. Indeed, activated microglia immunostaining with CD45 antibody demonstrated an increased immune reaction particularly surrounding the graft deposits compared to the placebo. Significant improvement was noted in patients with milder disease at baseline, UPDRS ⱕ 49. The development of runaway dyskinesias did not depend on differences in fluorodopa uptake, nor on the dose of cells implanted.


234

A study by Ma et al. [129] had suggested that higher fluorodopa uptake coincides with dyskinesia side effects. Contrary to Hagell’s [128] interpretation, the use of fresh donor tissue (less than 48 h storage) did not prevent the development of debilitating side effects [91]. In view of the new risks of transplantation that were exposed, Fahn et al. recommended against using fetal nigral cells as a cell-based therapy for PD at this time, yet suggest that patients with milder disease may benefit from such a strategy if they receive grafts with a higher number of surviving cells and a more prolonged immunosuppressant. It appears likely that overproduction and focal pulsatile delivery of DA within the dennervated striatum can lead to the development of uncontrolled dyskinesias, and future use of DA tissue grafts will need to address this issue of DA regulation and immune regulation. Stem Cells: ES and Brain-Derived The use of stem cells for ‘cell therapy’ of PD requires directing the cells toward dopaminergic lineages before transplantation or at early stages of grafting. This preconditioning of stem cells has been tested in vitro using growth factors, cytokines, and conditioned media for forebrain-derived neural precursors [130, 131] and for midbrain cells as well [132–135]. Transplantation of nonimmortalized stem cells into parkinsonian animal models has led to survival, integration, and expression of TH, the rate-limiting enzyme of DA synthesis, and to behavioral improvement of the lesioned animals [133,135–138]. Clonally derived neural stem cells [115] were shown to spontaneously differentiate into TH-expressing cells in a rat model of PD, a characteristic that is dependant on the culture confluency of the clone and the host’s microenvironment [139]. DA neurons can be derived form ES cells using a multistep protocol. In a study by Kawasaki et al. [140], ES cells were maintained in an undifferentiated state in media supplemented with serum and leukemia inhibitory factor. To direct differentiation toward the DA lineage, ES cells were cocultured with PA6 stromal cells for 8 days and then ascorbate was added to the media for 6–12 days. With this treatment, 16% of the total cell population was converted into DA neurons expressing TH- and DA lineage-specific transcription factors Nurr1 and Ptx3. After implantation into the 6-OHDA rat model of PD, the ES-derived neurons maintained the DA phenotype and did not form tumors. However, behavioral analysis of these animals was not reported. In another study by Lee et al. [141], ES cells were maintained in serum and media containing leukemia inhibitory factor. Removing leukemia inhibitory factor for 4 days and then growing the cells on adhesive substrate for 24 h induced the formation of embryoid bodies. Stem cells were subsequently expanded in serum-free media for 6–10 days before inducing DA neurons with bFGF, sonic hedgehog (Shh-N), and FGF8 for an additional 6 days. Maturation of DA neurons was established by


235

culturing the cells in media supplemented with cAMP and ascorbate. Under these conditions 23% of the total cells were dopaminergic. These DA neurons expressed transcription factors characteristic of midbrain DA neurons; however, they were not assessed for their functional ability to reverse behavioral deficits in PD animal model. Naïve ES cells also have been implanted into 6-OHDA denervated rat striatum [142]. In this study, ES cells gave rise to TH-immunoreactive neurons that also expressed DA transporter and AADC. Using PET and functional MRI, the grafted ES cells demonstrated the appropriate dopaminergic neuronal properties which paralleled behavioral recovery demonstrated with the apomorhine rotational test [142]. This study also reported, however, that ES cell transplantation led to the generation of lethal teratomas in 20% of implanted animals. While these findings are encouraging, further in vitro manipulation of ES cells and long-term posttransplant survival studies are required to provide assurance that tumor formation does not occur, an unacceptable outcome in a disease with existing drug and surgical therapies. One possible technology to control both growth rate and lineage of the cells before transplantation is genetic modification, such as providing a repressible regulatory unit or suicide gene that could be induced as required.

Detection of Gene Expression

PET Neuroimaging techniques and behavioral analyses make it possible to assess in vivo the state of the DA system in patient and animal models. Functional studies provide valuable information about the structure and function of DA neurons and the effects of therapeutic approaches. These techniques permit quantitative measurement of changes in DA terminals, receptors, and release of DA in vivo. PET and the use of specific radiolabeled ligands can noninvasively quantify pre- and postsynaptic markers of the DA system. Many of these tracers bind selectively to specific transporters, such as DA transporter or the vesicular monoamine transporter 2. The type of ligand utilized will determine the information we can obtain about a particular system [143]. Such noninvasive techniques are ideal for longitudinal studies in experimental models of PD. Conceivably, they could be used to monitor the progress in effectiveness of gene therapy approaches by monitoring the transgene expression. Studies from our lab demonstrated that PET was successfully applied to monitor AADC expression introduced by AAV-based vector [71]. Another study from our laboratory in the MPTP monkey model of PD demonstrated the ability to distinguish between dopaminergic changes in the putamen and the SN


236

compacta using PET scanning [144]. The previously cited study by Kordower et al. [48] with AAV-GDNF used PET to monitor therapeutic effects of gene transfer in a MPTP model. The choice of imaging ligands is especially important because the use of growth factors and other therapeutic interventions often specifically target either the terminals or cell bodies. In vivo detection of gene expression, as seen in studies after AADC gene transfer in MPTP-treated monkeys, is important because it provides quantitative assessment of gene transfer. This approach applies for both in vivo and in vitro gene transfer where duration, levels, and location of AADC gene expression can be detected. Microdialysis In vivo intracerebral microdialysis has been used in rats, nonhuman primates, and humans to monitor the extracellular level of neurotransmitters. This method can be very useful to evaluate alterations in brain metabolism. Microdialysis probes, connected to microinjection pumps, are stereotactically inserted into targeted points in the brain. Artificial cerebrospinal fluid is administered at a slow rate and dialysates are collected to microtubes for chemical analysis. The amount of neurotransmitters in each fraction is determined by HPLC. Alternatively, by directly sampling cerebrospinal fluid from the lateral ventricles, the levels of amino acids can be measured. Microdialysis was used during stereotaxic thalamic surgery for PD tremor for neurochemical characterization of the target area [145]. Studies by Fedele et al. [146] confirmed that this method might be used in PD to measure amino acid release in human basal ganglia. Using this approach one can assess the level of DA restoration via gene therapy. The very same procedure was successfully used by Pernaute et al. [70] to evaluate the functional effect of AAV-mediated gene transfer of aromatic L-amino acid decarboxylase into the striatum of 6 OHDA-lesioned rats.

Conclusions

PD is characterized primarily by the degeneration of a specific population of neurons in SN and a decline in local neurotransmitter synthesis. Replacement therapy with the DA precursor L-dopa has been a mainstay of therapy for PD. However, L-dopa addresses only the biochemical consequences of the disease and leads to long-term side effects such as dyskinesias. Prevention of further loss of dopaminergic neurons by neuroprotection, perhaps using gene transfer, is one alternative approach to treatment. Advances in cellular and genetic engineering also will permit stem cell transplants to replace neurons once they have been lost. Gene therapy for PD based on in vivo or ex vivo strategies is realistic, but will depend on the progress that is made over the years to come.


237

References 1 2 3 4 5 6 7

8

9 10

11 12 13 14

15

16

17 18

19

20 21

22

Fahn S: Is levodopa toxic? Neurology 1996;47:S184–S195. Hurtig HI: Problems with current pharmacologic treatment of Parkinson’s disease. Exp Neurol 1997;144:10–16. Harder S, Baas H, Rietbrock S: Concentration-effect relationship of levodopa in patients with Parkinson’s disease. Clin Pharmacokinet 1995;29:243–256. Chase TN, Juncos J, Serrati C, Fabbrini G, Bruno G: Fluctuation in response to chronic levodopa therapy: Pathogenetic and therapeutic considerations. Adv Neurol 1987;45:477–480. Obeso JA, Grandas F, Herrero MT, Horowski R: The role of pulsatile versus continuous dopamine receptor stimulation for functional recovery in Parkinson’s disease. Eur J Neurosci 1994;6:889–897. Melamed E, Hefti F: Mechanism of action of short- and long-term L-DOPA treatment in parkinsonism: Role of the surviving nigrostriatal dopaminergic neurons. Adv Neurol 1984;40:149–157. Nagatsu T, Levitt M, Udenfriend S: Conversion of L-tyrosine to 3,4-dihydroxyphenylalanine by cell-free preparations of brain and sympathetically innervated tissues. Biochem Biophys Res Commun 1964;14:543–549. Nagatsu T, Yamaguchi T, Rahman MK, Trocewicz J, Oka K, Hirata Y, Nagatsu I, Narabayashi H, Kondo T, Iizuka R: Catecholamine-related enzymes and the biopterin cofactor in Parkinson’s disease and related extrapyramidal diseases. Adv Neurol 1984;40:467–473. Nagatsu T, Ichinose H: Molecular biology of catecholamine-related enzymes in relation to Parkinson’s disease. Cell Mol Neurobiol 1999;19:57–66. Imaoka T, Date I, Ohmoto T, Nagatsu T: Significant behavioral recovery in Parkinson’s disease model by direct intracerebral gene transfer using continuous injection of a plasmid DNA-liposome complex. Hum Gene Ther 1998;9:1093–1102. Hermens WT, Verhaagen J: Viral vectors, tools for gene transfer in the nervous system. Prog Neurobiol 1998;55:399–432. Spaete RR, Frenkel N: The herpes simplex virus amplicon: A new eucaryotic defective-virus cloning-amplifying vector. Cell 1982;30:295–304. During MJ, Naegele JR, O’Malley KL, Geller AI: Long-term behavioral recovery in parkinsonian rats by an HSV vector expressing tyrosine hydroxylase. Science 1994;266:1399–1403. Geller AI, Yu L, Wang Y, Fraefel C: Helper virus-free herpes simplex virus-1 plasmid vectors for gene therapy of Parkinson’s disease and other neurological disorders. Exp Neurol 1997;144: 98–102. Natsume A, Mata M, Goss J, Huang S, Wolfe D, Oligino T, Glorioso J, Fink DJ: Bcl-2 and GDNF delivered by HSV-mediated gene transfer act additively to protect dopaminergic neurons from 6-OHDA-induced degeneration. Exp Neurol 2001;169:231–238. Akli S, Caillaud C, Vigne E, Stratford-Perricaudet LD, Poenaru L, Perricaudet M, Kahn A, Peschanski MR: Transfer of a foreign gene into the brain using adenovirus vectors. Nat Genet 1993;3:224–228. Davidson BL, Allen ED, Kozarsky KF, Wilson JM, Roessler BJ: A model system for in vivo gene transfer into the central nervous system using an adenoviral vector. Nat Genet 1993;3:219–223. Yang CC, Xiao X, Zhu X, Ansardi DC, Epstein ND, Frey MR, Matera AG, Samulski RJ: Cellular recombination pathways and viral terminal repeat hairpin structures are sufficient for adenoassociated virus integration in vivo and in vitro. J Virol 1997;71:9231–9247. Pastore L, Morral N, Zhou H, Garcia R, Parks RJ, Kochanek S, Graham FL, Lee B, Beaudet AL: Use of a liver-specific promoter reduces immune response to the transgene in adenoviral vectors. Hum Gene Ther 1999;10:1773–1781. Zhou H, O’Neal W, Morral N, Beaudet AL: Development of a complementing cell line and a system for construction of adenovirus vectors with E1 and E2a deleted. J Virol 1996;70:7030–7038. Gorziglia MI, Kadan MJ, Yei S, Lim J, Lee GM, Luthra R, Trapnell BC: Elimination of both E1 and E2 from adenovirus vectors further improves prospects for in vivo human gene therapy. J Virol 1996;70:4173–4178. Krougliak V, Graham FL: Development of cell lines capable of complementing E1, E4, and protein IX defective adenovirus type 5 mutants. Hum Gene Ther 1995;6:1575–1586.


238

23

24

25

26

27

28

29

30

31

32

33

34

35 36 37

38 39

40 41

Yeh P, Dedieu JF, Orsini C, Vigne E, Denefle P, Perricaudet M: Efficient dual transcomplementation of adenovirus E1 and E4 regions from a 293-derived cell line expressing a minimal E4 functional unit. J Virol 1996;70:559–565. Lieber A, He CY, Kirillova I, Kay MA: Recombinant adenoviruses with large deletions generated by Cre-mediated excision exhibit different biological properties compared with first-generation vectors in vitro and in vivo. J Virol 1996;70:8944–8960. Zermansky AJ, Bolognani F, Stone D, Cowsill CM, Morrissey G, Castro MG, Lowenstein PR: Towards global and long-term neurological gene therapy: Unexpected transgene dependent, highlevel, and widespread distribution of HSV-1 thymidine kinase throughout the CNS. Mol Ther 2001;4:490–498. Dong JY, Wang D, Van Ginkel FW, Pascual DW, Frizzell RA: Systematic analysis of repeated gene delivery into animal lungs with a recombinant adenovirus vector. Hum Gene Ther 1996;7: 319–331. Kaplan JM, St George JA, Pennington SE, Keyes LD, Johnson RP, Wadsworth SC, Smith AE: Humoral and cellular immune responses of nonhuman primates to long-term repeated lung exposure to Ad2/CFTR-2. Gene Ther 1996;3:117–127. St George JA, Pennington SE, Kaplan JM, Peterson PA, Kleine LJ, Smith AE, Wadsworth SC: Biological response of nonhuman primates to long-term repeated lung exposure to Ad2/CFTR-2. Gene Ther 1996;3:103–116. Horellou P, Vigne E, Castel MN, Barneoud P, Colin P, Perricaudet M, Delaere P, Mallet J: Direct intracerebral gene transfer of an adenoviral vector expressing tyrosine hydroxylase in a rat model of Parkinson’s disease. Neuroreport 1994;6:49–53. Corti O, Sanchez-Capelo A, Colin P, Hanoun N, Hamon M, Mallet J: Long-term doxycyclinecontrolled expression of human tyrosine hydroxylase after direct adenovirus-mediated gene transfer to a rat model of Parkinson’s disease. Proc Natl Acad Sci USA 1999;96:12120–12125. Leone P, McPhee SW, Janson CG, Davidson BL, Freese A, During MJ: Multi-site partitioned delivery of human tyrosine hydroxylase gene with phenotypic recovery in Parkinsonian rats. Neuroreport 2000;11:1145–1151. Choi-Lundberg DL, Lin Q, Chang YN, Chiang YL, Hay CM, Mohajeri H, Davidson BL, Bohn MC: Dopaminergic neurons protected from degeneration by GDNF gene therapy. Science 1997;275: 838–841. Bilang-Bleuel A, Revah F, Colin P, Locquet I, Robert JJ, Mallet J, Horellou P: Intrastriatal injection of an adenoviral vector expressing glial-cell-line-derived neurotrophic factor prevents dopaminergic neuron degeneration and behavioral impairment in a rat model of Parkinson disease. Proc Natl Acad Sci USA 1997;94:8818–8823. Connor B, Kozlowski DA, Schallert T, Tillerson JL, Davidson BL, Bohn MC: Differential effects of glial cell line-derived neurotrophic factor (GDNF) in the striatum and substantia nigra of the aged Parkinsonian rat. Gene Ther 1999;6:1936–1951. Dull T, Zufferey R, Kelly M, Mandel RJ, Nguyen M, Trono D, Naldini L: A third-generation lentivirus vector with a conditional packaging system. J Virol 1998;72:8463–8471. Zufferey R, Nagy D, Mandel RJ, Naldini L, Trono D: Multiply attenuated lentiviral vector achieves efficient gene delivery in vivo. Nat Biotechnol 1997;15:871–875. Naldini L, Blomer U, Gage FH, Trono D, Verma IM: Efficient transfer, integration, and sustained long-term expression of the transgene in adult rat brains injected with a lentiviral vector. Proc Natl Acad Sci USA 1996;93:11382–11388. Naldini L, Blomer U, Gallay P, Ory D, Mulligan R, Gage FH, Verma IM, Trono D: In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science 1996;272:263–267. Deglon N, Tseng JL, Bensadoun JC, Zurn AD, Arsenijevic Y, Pereira de Almeida L, Zufferey R, Trono D, Aebischer P: Self-inactivating lentiviral vectors with enhanced transgene expression as potential gene transfer system in Parkinson’s disease. Hum Gene Ther 2000;11:179–190. Zufferey R, Dull T, Mandel RJ, Bukovsky A, Quiroz D, Naldini L, Trono D: Self-inactivating lentivirus vector for safe and efficient in vivo gene delivery. J Virol 1998;72:9873–9880. Blomer U, Naldini L, Kafri T, Trono D, Verma IM, Gage FH: Highly efficient and sustained gene transfer in adult neurons with a lentivirus vector. J Virol 1997;71:6641–6649.


239

42

43 44

45 46 47

48 49 50

51

52

53

54

55 56

57 58 59

60

61

Cisterni C, Henderson CE, Aebischer P, Pettmann B, Deglon N: Efficient gene transfer and expression of biologically active glial cell line-derived neurotrophic factor in rat motoneurons transduced with lentiviral vectors. J Neurochem 2000;74:1820–1828. Miyoshi H, Blomer U, Takahashi M, Gage FH, Verma IM: Development of a self-inactivating lentivirus vector. J Virol 1998;72:8150–8157. Li Z, Dullmann J, Schiedlmeier B, Schmidt M, von Kalle C, Meyer J, Forster M, Stocking C, Wahlers A, Frank O, Ostertag W, Kuhlcke K, Eckert HG, Fehse B, Baum C: Murine leukemia induced by retroviral gene marking. Science 2002;296:497. Marshall E: Clinical research. Gene therapy a suspect in leukemia-like disease. Science 2002; 298:34–35. Murre C: Intertwining proteins in thymocyte development and cancer. Nat Immunol 2000;1: 97–98. Donahue RE, Kessler SW, Bodine D, McDonagh K, Dunbar C, Goodman S, Agricola B, Byrne E, Raffeld M, Moen R: Helper virus induced T cell lymphoma in nonhuman primates after retroviral mediated gene transfer. J Exp Med 1992;176:1125–1135. Kordower JH, Bloch J, Ma SY, Chu Y, Palfi S, Roitberg BZ, Emborg M, Hantraye P, Deglon N, Aebischer P: Lentiviral gene transfer to the nonhuman primate brain. Exp Neurol 1999;160:1–16. Park F, Ohashi K, Chiu W, Naldini L, Kay MA: Efficient lentiviral transduction of liver requires cell cycling in vivo. Nat Genet 2000;24:49–52. Georgievska B, Kirik D, Rosenblad C, Lundberg C, Bjorklund A: Neuroprotection in the rat Parkinson model by intrastriatal GDNF gene transfer using a lentiviral vector. Neuroreport 2002; 13:75–82. Bensadoun JC, Deglon N, Tseng JL, Ridet JL, Zurn AD, Aebischer P: Lentiviral vectors as a gene delivery system in the mouse midbrain: Cellular and behavioral improvements in a 6-OHDA model of Parkinson’s disease using GDNF. Exp Neurol 2000;164:15–24. Kordower JH, Emborg ME, Bloch J, Ma SY, Chu Y, Leventhal L, McBride J, Chen EY, Palfi S, Roitberg BZ, Brown WD, Holden JE, Pyzalski R, Taylor MD, Carvey P, Ling Z, Trono D, Hantraye P, Deglon N, Aebischer P: Neurodegeneration prevented by lentiviral vector delivery of GDNF in primate models of Parkinson’s disease. Science 2000;290:767–773. Azzouz M, Martin-Rendon E, Barber RD, Mitrophanous KA, Carter EE, Rohll JB, Kingsman SM, Kingsman AJ, Mazarakis ND: Multicistronic lentiviral vector-mediated striatal gene transfer of aromatic L-amino acid decarboxylase, tyrosine hydroxylase, and GTP cyclohydrolase I induces sustained transgene expression, dopamine production, and functional improvement in a rat model of Parkinson’s disease. J Neurosci 2002;22:10302–10312. D’Costa J, Harvey-White J, Qasba P, Limaye A, Kaneski CR, Davis-Warren A, Brady RO, Bankiewicz KS, Major EO, Arya SK: HIV-2 derived lentiviral vectors: Gene transfer in Parkinson’s and Fabry disease models in vitro. J Med Virol 2003;71:173–182. Kotin RM, Menninger JC, Ward DC, Berns KI: Mapping and direct visualization of a region-specific viral DNA integration site on chromosome 19q13-qter. Genomics 1991;10:831–834. Okada T, Nomoto T, Shimazaki K, Lijun W, Lu Y, Matsushita T, Mizukami H, Urabe M, Hanazono Y, Kume A, Muramatsu S, Nakano I, Ozawa K: Adeno-associated virus vectors for gene transfer to the brain. Methods 2002;28:237–247. Xiao X, Li J, Samulski RJ: Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus. J Virol 1998;72:2224–2232. Urabe M, Ding C, Kotin RM: Insect cells as a factory to produce adeno-associated virus type 2 vectors. Hum Gene Ther 2002;13:1935–1943. Chirmule N, Xiao W, Truneh A, Schnell MA, Hughes JV, Zoltick P, Wilson JM: Humoral immunity to adeno-associated virus type 2 vectors following administration to murine and nonhuman primate muscle. J Virol 2000;74:2420–2425. Halbert CL, Standaert TA, Wilson CB, Miller AD: Successful readministration of adeno-associated virus vectors to the mouse lung requires transient immunosuppression during the initial exposure. J Virol 1998;72:9795–9805. Flotte T, Carter B, Conrad C, Guggino W, Reynolds T, Rosenstein B, Taylor G, Walden S, Wetzel R: A phase I study of an adeno-associated virus-CFTR gene vector in adult CF patients with mild lung disease. Hum Gene Ther 1996;7:1145–1159.


240

62

63

64 65

66

67

68 69

70

71

72

73

74

75

76

77

Wagner JA, Messner AH, Moran ML, Daifuku R, Kouyama K, Desch JK, Manley S, Norbash AM, Conrad CK, Friborg S, Reynolds T, Guggino WB, Moss RB, Carter BJ, Wine JJ, Flotte TR, Gardner P: Safety and biological efficacy of an adeno-associated virus vector-cystic fibrosis transmembrane regulator (AAV-CFTR) in the cystic fibrosis maxillary sinus. Laryngoscope 1999; 109:266–274. Mandel RJ, Rendahl KG, Spratt SK, Snyder RO, Cohen LK, Leff SE: Characterization of intrastriatal recombinant adeno-associated virus-mediated gene transfer of human tyrosine hydroxylase and human GTP-cyclohydrolase I in a rat model of Parkinson’s disease. J Neurosci 1998;18:4271–4284. McCown TJ, Xiao X, Li J, Breese GR, Samulski RJ: Differential and persistent expression patterns of CNS gene transfer by an adeno-associated virus (AAV) vector. Brain Res 1996;713:99–107. Leff SE, Spratt SK, Snyder RO, Mandel RJ: Long-term restoration of striatal L-aromatic amino acid decarboxylase activity using recombinant adeno-associated viral vector gene transfer in a rodent model of Parkinson’s disease. Neuroscience 1999;92:185–196. Kaplitt MG, Leone P, Samulski RJ, Xiao X, Pfaff DW, O’Malley KL, During MJ: Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat Genet 1994;8:148–154. Kirik D, Rosenblad C, Bjorklund A, Mandel RJ: Long-term rAAV-mediated gene transfer of GDNF in the rat Parkinson’s model: Intrastriatal but not intranigral transduction promotes functional regeneration in the lesioned nigrostriatal system. J Neurosci 2000;20:4686–4700. Lo WD, Qu G, Sferra TJ, Clark R, Chen R, Johnson PR: Adeno-associated virus-mediated gene transfer to the brain: Duration and modulation of expression. Hum Gene Ther 1999;10:201–213. Shen Y, Muramatsu SI, Ikeguchi K, Fujimoto KI, Fan DS, Ogawa M, Mizukami H, Urabe M, Kume A, Nagatsu I, Urano F, Suzuki T, Ichinose H, Nagatsu T, Monahan J, Nakano I, Ozawa K: Triple transduction with adeno-associated virus vectors expressing tyrosine hydroxylase, aromatic-L-amino-acid decarboxylase, and GTP cyclohydrolase I for gene therapy of Parkinson’s disease. Hum Gene Ther 2000;11:1509–1519. Sanchez-Pernaute R, Harvey-White J, Cunningham J, Bankiewicz KS: Functional effect of adenoassociated virus mediated gene transfer of aromatic L-amino acid decarboxylase into the striatum of 6-OHDA-lesioned rats. Mol Ther 2001;4:324–330. Bankiewicz KS, Eberling JL, Kohutnicka M, Jagust W, Pivirotto P, Bringas J, Cunningham J, Budinger TF, Harvey-White J: Convection-enhanced delivery of AAV vector in parkinsonian monkeys; in vivo detection of gene expression and restoration of dopaminergic function using pro-drug approach. Exp Neurol 2000;164:2–14. Mandel RJ, Spratt SK, Snyder RO, Leff SE: Midbrain injection of recombinant adeno-associated virus encoding rat glial cell line-derived neurotrophic factor protects nigral neurons in a progressive 6-hydroxydopamine-induced degeneration model of Parkinson’s disease in rats. Proc Natl Acad Sci USA 1997;94:14083–14088. Mandel RJ, Snyder RO, Leff SE: Recombinant adeno-associated viral vector-mediated glial cell line-derived neurotrophic factor gene transfer protects nigral dopamine neurons after onset of progressive degeneration in a rat model of Parkinson’s disease. Exp Neurol 1999;160:205–214. Bankiewicz KS, Leff SE, Nagy D, Jungles S, Rokovich J, Spratt K, Cohen L, Libonati M, Snyder RO, Mandel RJ: Practical aspects of the development of ex vivo and in vivo gene therapy for Parkinson’s disease. Exp Neurol 1997;144:147–156. Fan DS, Ogawa M, Fujimoto KI, Ikeguchi K, Ogasawara Y, Urabe M, Nishizawa M, Nakano I, Yoshida M, Nagatsu I, Ichinose H, Nagatsu T, Kurtzman GJ, Ozawa K: Behavioral recovery in 6-hydroxydopamine-lesioned rats by cotransduction of striatum with tyrosine hydroxylase and aromatic L-amino acid decarboxylase genes using two separate adeno-associated virus vectors. Hum Gene Ther 1998;9:2527–2535. Muramatsu S, Fujimoto K, Ikeguchi K, Shizuma N, Kawasaki K, Ono F, Shen Y, Wang L, Mizukami H, Kume A, Matsumura M, Nagatsu I, Urano F, Ichinose H, Nagatsu T, Terao K, Nakano I, Ozawa K: Behavioral recovery in a primate model of Parkinson’s disease by triple transduction of striatal cells with adeno-associated viral vectors expressing dopamine-synthesizing enzymes. Hum Gene Ther 2002;13:345–354. Luo J, Kaplitt MG, Fitzsimons HL, Zuzga DS, Liu Y, Oshinsky ML, During MJ: Subthalamic GAD gene therapy in a Parkinson’s disease rat model. Science 2002;298:425–429.


241

78

79 80 81

82

83

84

85

86

87

88 89

90 91

92

93

94 95

96

97

Costantini LC, Jacoby DR, Wang S, Fraefel C, Breakefield XO, Isacson O: Gene transfer to the nigrostriatal system by hybrid herpes simplex virus/adeno-associated virus amplicon vectors. Hum Gene Ther 1999;10:2481–2494. Danko I, Williams P, Herweijer H, Zhang G, Latendresse JS, Bock I, Wolff JA: High expression of naked plasmid DNA in muscles of young rodents. Hum Mol Genet 1997;6:1435–1443. Budker V, Budker T, Zhang G, Subbotin V, Loomis A, Wolff JA: Hypothesis: Naked plasmid DNA is taken up by cells in vivo by a receptor-mediated process. J Gene Med 2000;2:76–88. Coll JL, Negoescu A, Louis N, Sachs L, Tenaud C, Girardot V, Demeinex B, Brambilla E, Brambilla C, Favrot M: Antitumor activity of bax and p53 naked gene transfer in lung cancer: In vitro and in vivo analysis. Hum Gene Ther 1998;9:2063–2074. White SA, LoBuglio AF, Arani RB, Pike MJ, Moore SE, Barlow DL, Conry RM: Induction of anti-tumor immunity by intrasplenic administration of a carcinoembryonic antigen DNA vaccine. J Gene Med 2000;2:135–140. Papahadjopoulos D, Allen TM, Gabizon A, Mayhew E, Matthay K, Huang SK, Lee KD, Woodle MC, Lasic DD, Redemann C, Martin FJ: Sterically stabilized liposomes: Improvements in pharmacokinetics and antitumor therapeutic efficacy. Proc Natl Acad Sci USA 1991;88:11460–11464. Zhang Y, Calon F, Zhu C, Boado RJ, Pardridge WM: Intravenous nonviral gene therapy causes normalization of striatal tyrosine hydroxylase and reversal of motor impairment in experimental parkinsonism. Hum Gene Ther 2003;14:1–12. Segovia J, Vergara P, Brenner M: Astrocyte-specific expression of tyrosine hydroxylase after intracerebral gene transfer induces behavioral recovery in experimental parkinsonism. Gene Ther 1998;5:1650–1655. Boussif O, Lezoualc’h F, Zanta MA, Mergny MD, Scherman D, Demeneix B, Behr JP: A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: Polyethylenimine. Proc Natl Acad Sci USA 1995;92:7297–7301. Abdallah B, Hassan A, Benoist C, Goula D, Behr JP, Demeneix BA: A powerful nonviral vector for in vivo gene transfer into the adult mammalian brain: Polyethylenimine. Hum Gene Ther 1996;7:1947–1954. Wang S, Ma N, Gao SJ, Yu H, Leong KW: Transgene expression in the brain stem effected by intramuscular injection of polyethylenimine/DNA complexes. Mol Ther 2001;3:658–664. Freed CR, Greene PE, Breeze RE, Tsai WY, DuMouchel W, Kao R, Dillon S, Winfield H, Culver S, Trojanowski JQ, Eidelberg D, Fahn S: Transplantation of embryonic dopamine neurons for severe Parkinson’s disease. N Engl J Med 2001;344:710–719. Koh JY, Gwag BJ, Lobner D, Choi DW: Potentiated necrosis of cultured cortical neurons by neurotrophins. Science 1995;268:573–575. Olanow CW, Goetz CG, Kordower JH, Stoessl AJ, Sossi V, Brin MF, Shannon KM, Nauert GM, Perl DP, Godbold J, Freeman TB: A double-blind controlled trial of bilateral fetal nigral transplantation in Parkinson’s disease. Ann Neurol 2003;54:403–414. Andersen JK, Frim DM, Isacson O, Breakefield XO: Herpesvirus-mediated gene delivery into the rat brain: Specificity and efficiency of the neuron-specific enolase promoter. Cell Mol Neurobiol 1993;13:503–515. Kaplitt MG, Kwong AD, Kleopoulos SP, Mobbs CV, Rabkin SD, Pfaff DW: Preproenkephalin promoter yields region-specific and long-term expression in adult brain after direct in vivo gene transfer via a defective herpes simplex viral vector. Proc Natl Acad Sci USA 1994;91:8979–8983. Hashimoto M, Aruga J, Hosoya Y, Kanegae Y, Saito I, Mikoshiba K: A neural cell-type-specific expression system using recombinant adenovirus vectors. Hum Gene Ther 1996;7:149–158. Peel AL, Zolotukhin S, Schrimsher GW, Muzyczka N, Reier PJ: Efficient transduction of green fluorescent protein in spinal cord neurons using adeno-associated virus vectors containing cell type-specific promoters. Gene Ther 1997;4:16–24. Xu R, Janson CG, Mastakov M, Lawlor P, Young D, Mouravlev A, Fitzsimons H, Choi KL, Ma H, Dragunow M, Leone P, Chen Q, Dicker B, During MJ: Quantitative comparison of expression with adeno-associated virus (AAV-2) brain-specific gene cassettes. Gene Ther 2001;8:1323–1332. Cortez N, Trejo F, Vergara P, Segovia J: Primary astrocytes retrovirally transduced with a tyrosine hydroxylase transgene driven by a glial-specific promoter elicit behavioral recovery in experimental parkinsonism. J Neurosci Res 2000;59:39–46.


242

98 Gossen M, Bujard H: Tight control of gene expression in mammalian cells by tetracycline-responsive promoters. Proc Natl Acad Sci USA 1992;89:5547–5551. 99 Hinrichs W, Kisker C, Duvel M, Muller A, Tovar K, Hillen W, Saenger W: Structure of the Tet repressor-tetracycline complex and regulation of antibiotic resistance. Science 1994;264:418–420. 100 Chyung YH, Peng PD, Kay MA: System for simultaneous tissue-specific and disease-specific regulation of therapeutic gene expression. Hum Gene Ther 2003;14:1255–1264. 101 Aiken C: Pseudotyping human immunodeficiency virus type 1 (HIV-1) by the glycoprotein of vesicular stomatitis virus targets HIV-1 entry to an endocytic pathway and suppresses both the requirement for Nef and the sensitivity to cyclosporin A. J Virol 1997;71:5871–5877. 102 Bartz SR, Vodicka MA: Production of high-titer human immunodeficiency virus type 1 pseudotyped with vesicular stomatitis virus glycoprotein. Methods 1997;12:337–342. 103 Arai T, Takada M, Ui M, Iba H: Dose-dependent transduction of vesicular stomatitis virus G proteinpseudotyped retrovirus vector into human solid tumor cell lines and murine fibroblasts. Virology 1999;260:109–115. 104 Anderson DB, Laquerre S, Ghosh K, Ghosh HP, Goins WF, Cohen JB, Glorioso JC: Pseudotyping of glycoprotein D-deficient herpes simplex virus type 1 with vesicular stomatitis virus glycoprotein G enables mutant virus attachment and entry. J Virol 2000;74:2481–2487. 105 Kang UJ, Fisher LJ, Joh TH, O’Malley KL, Gage FH: Regulation of dopamine production by genetically modified primary fibroblasts. J Neurosci 1993;13:5203–5211. 106 Lee WY, Chang JW, Nemeth NL, Kang UJ: Vesicular monoamine transporter-2 and aromatic L-amino acid decarboxylase enhance dopamine delivery after L-3, 4-dihydroxyphenylalanine administration in Parkinsonian rats. J Neurosci 1999;19:3266–3274. 107 Shimohama S, Rosenberg MB, Fagan AM, Wolff JA, Short MP, Breakefield XO, Friedmann T, Gage FH: Grafting genetically modified cells into the rat brain: Characteristics of E. coli betagalactosidase as a reporter gene. Brain Res Mol Brain Res 1989;5:271–278. 108 Wolff JA, Fisher LJ, Xu L, Jinnah HA, Langlais PJ, Iuvone PM, O’Malley KL, Rosenberg MB, Shimohama S, Friedmann T, Gage FH: Grafting fibroblasts genetically modified to produce L-dopa in a rat model of Parkinson disease. Proc Natl Acad Sci USA 1989;86:9011–9014. 109 Tornatore C, Baker-Cairns B, Yadid G, Hamilton R, Meyers K, Atwood W, Cummins A, Tanner V, Major E: Expression of tyrosine hydroxylase in an immortalized human fetal astrocyte cell line; In vitro characterization and engraftment into the rodent striatum. Cell Transplant 1996;5: 145–163. 110 Keir SD, Miller J, Yu G, Hamilton R, Samulski RJ, Xiao X, Tornatore C: Efficient gene transfer into primary and immortalized human fetal glial cells using adeno-associated virus vectors: Establishment of a glial cell line with a functional CD4 receptor. J Neurovirol 1997;3:322–330. 111 Horellou P, Brundin P, Kalen P, Mallet J, Bjorklund A: In vivo release of dopa and dopamine from genetically engineered cells grafted to the denervated rat striatum. Neuron 1990;5:393–402. 112 Sawamoto K, Nakao N, Kakishita K, Ogawa Y, Toyama Y, Yamamoto A, Yamaguchi M, Mori K, Goldman SA, Itakura T, Okano H: Generation of dopaminergic neurons in the adult brain from mesencephalic precursor cells labeled with a nestin-GFP transgene. J Neurosci 2001;21: 3895–3903. 113 Sawamoto K, Nakao N, Kobayashi K, Matsushita N, Takahashi H, Kakishita K, Yamamoto A, Yoshizaki T, Terashima T, Murakami F, Itakura T, Okano H: Visualization, direct isolation, and transplantation of midbrain dopaminergic neurons. Proc Natl Acad Sci USA 2001;98:6423–6428. 114 Wagner J, Akerud P, Castro DS, Holm PC, Canals JM, Snyder EY, Perlmann T, Arenas E: Induction of a midbrain dopaminergic phenotype in Nurr1-overexpressing neural stem cells by type 1 astrocytes. Nat Biotechnol 1999;17:653–659. 115 Snyder EY, Deitcher DL, Walsh C, Arnold-Aldea S, Hartwieg EA, Cepko CL: Multipotent neural cell lines can engraft and participate in development of mouse cerebellum. Cell 1992;68:33–51. 116 Kim JH, Auerbach JM, Rodriguez-Gomez JA, Velasco I, Gavin D, Lumelsky N, Lee SH, Nguyen J, Sanchez-Pernaute R, Bankiewicz K, McKay R: Dopamine neurons derived from embryonic stem cells function in an animal model of Parkinson’s disease. Nature 2002;418:50–56. 117 Bjorklund A, Stenevi U: Reconstruction of the nigrostriatal dopamine pathway by intracerebral nigral transplants. Brain Res 1979;177:555–560. 118 Perlow MJ, Freed WJ, Hoffer BJ, Seiger A, Olson L, Wyatt RJ: Brain grafts reduce motor abnormalities produced by destruction of nigrostriatal dopamine system. Science 1979;204:643–647.


243

119 Redmond DE, Sladek JR Jr, Roth RH, Collier TJ, Elsworth JD, Deutch AY, Haber S: Fetal neuronal grafts in monkeys given methylphenyltetrahydropyridine. Lancet 1986;1:1125–1127. 120 Bankiewicz KS, Plunkett RJ, Jacobowitz DM, Porrino L, di Porzio U, London WT, Kopin IJ, Oldfield EH: The effect of fetal mesencephalon implants on primate MPTP-induced parkinsonism. Histochemical and behavioral studies. J Neurosurg 1990;72:231–244. 121 Freeman TB, Olanow CW, Hauser RA, Nauert GM, Smith DA, Borlongan CV, Sanberg PR, Holt DA, Kordower JH, Vingerhoets FJ: Bilateral fetal nigral transplantation into the postcommissural putamen in Parkinson’s disease. Ann Neurol 1995;38:379–388. 122 Hauser RA, Freeman TB, Snow BJ, Nauert M, Gauger L, Kordower JH, Olanow CW: Long-term evaluation of bilateral fetal nigral transplantation in Parkinson disease. Arch Neurol 1999;56: 179–187. 123 Kordower JH, Freeman TB, Snow BJ, Vingerhoets FJ, Mufson EJ, Sanberg PR, Hauser RA, Smith DA, Nauert GM, Perl DP, Olanow CW: Neuropathological evidence of graft survival and striatal reinnervation after the transplantation of fetal mesencephalic tissue in a patient with Parkinson’s disease. N Engl J Med 1995;332:1118–1124. 124 Kordower JH, Freeman TB, Chen EY, Mufson EJ, Sanberg PR, Hauser RA, Snow B, Olanow CW: Fetal nigral grafts survive and mediate clinical benefit in a patient with Parkinson’s disease. Mov Disord 1998;13:383–393. 125 Piccini P, Brooks DJ, Bjorklund A, Gunn RN, Grasby PM, Rimoldi O, Brundin P, Hagell P, Rehncrona S, Widner H, Lindvall O: Dopamine release from nigral transplants visualized in vivo in a Parkinson’s patient. Nat Neurosci 1999;2:1137–1140. 126 Brundin P, Dunnett S, Bjorklund A, Nikkhah G: Transplanted dopaminergic neurons: More or less? Nat Med 2001;7:512–513. 127 Nikkhah G: Neural transplantation therapy for Parkinson’s disease: Potential and pitfalls. Brain Res Bull 2001;56:509. 128 Hagell P, Piccini P, Bjorklund A, Brundin P, Rehncrona S, Widner H, Crabb L, Pavese N, Oertel WH, Quinn N, Brooks DJ, Lindvall O: Dyskinesias following neural transplantation in Parkinson’s disease. Nat Neurosci 2002;5:627–628. 129 Ma Y, Feigin A, Dhawan V, Fukuda M, Shi Q, Greene P, Breeze R, Fahn S, Freed C, Eidelberg D: Dyskinesia after fetal cell transplantation for parkinsonism: A PET study. Ann Neurol 2002;52: 628–634. 130 Carpenter MK, Cui X, Hu ZY, Jackson J, Sherman S, Seiger A, Wahlberg LU: In vitro expansion of a multipotent population of human neural progenitor cells. Exp Neurol 1999;158:265–278. 131 Daadi MM, Weiss S: Generation of tyrosine hydroxylase-producing neurons from precursors of the embryonic and adult forebrain. J Neurosci 1999;19:4484–4497. 132 Ling ZD, Potter ED, Lipton JW, Carvey PM: Differentiation of mesencephalic progenitor cells into dopaminergic neurons by cytokines. Exp Neurol 1998;149:411–423. 133 Studer L, Tabar V, McKay RD: Transplantation of expanded mesencephalic precursors leads to recovery in parkinsonian rats. Nat Neurosci 1998;1:290–295. 134 Yan H, Bunge MB, Wood PM, Plant GW: Mitogenic response of adult rat olfactory ensheathing glia to four growth factors. Glia 2001;33:334–342. 135 Sanchez-Pernaute R, Studer L, Bankiewicz KS, Major EO, McKay RD: In vitro generation and transplantation of precursor-derived human dopamine neurons. J Neurosci Res 2001;65: 284–288. 136 Arenas E: Stem cells in the treatment of Parkinson’s disease. Brain Res Bull 2002;57:795–808. 137 Svendsen CN, Caldwell MA, Shen J, ter Borg MG, Rosser AE, Tyers P, Karmiol S, Dunnett SB: Long-term survival of human central nervous system progenitor cells transplanted into a rat model of Parkinson’s disease. Exp Neurol 1997;148:135–146. 138 Carvey PM, Ling ZD, Sortwell CE, Pitzer MR, McGuire SO, Storch A, Collier TJ: A clonal line of mesencephalic progenitor cells converted to dopamine neurons by hematopoietic cytokines: A source of cells for transplantation in Parkinson’s disease. Exp Neurol 2001;171:98–108. 139 Yang M, Stull ND, Berk MA, Snyder EY, Iacovitti L: Neural stem cells spontaneously express dopaminergic traits after transplantation into the intact or 6-hydroxydopamine-lesioned rat. Exp Neurol 2002;177:50–60.


244

140 Kawasaki H, Mizuseki K, Nishikawa S, Kaneko S, Kuwana Y, Nakanishi S, Nishikawa SI, Sasai Y: Induction of midbrain dopaminergic neurons from ES cells by stromal cell-derived inducing activity. Neuron 2000;28:31–40. 141 Lee SH, Lumelsky N, Studer L, Auerbach JM, McKay RD: Efficient generation of midbrain and hindbrain neurons from mouse embryonic stem cells. Nat Biotechnol 2000;18:675–679. 142 Bjorklund LM, Sanchez-Pernaute R, Chung S, Andersson T, Chen IY, McNaught KS, Brownell AL, Jenkins BG, Wahlestedt C, Kim KS, Isacson O: Embryonic stem cells develop into functional dopaminergic neurons after transplantation in a Parkinson rat model. Proc Natl Acad Sci USA 2002;99:2344–2349. 143 Sanchez-Pernaute R, Brownell AL, Isacson O: Functional imaging of the dopamine system: In vivo evaluation of dopamine deficiency and restoration. Neurotoxicology 2002;23:469–478. 144 Eberling JL, Bankiewicz KS, Jordan S, VanBrocklin HF, Jagust WJ: PET studies of functional compensation in a primate model of Parkinson’s disease. Neuroreport 1997;8:2727–2733. 145 Meyerson BA, Linderoth B, Karlsson H, Ungerstedt U: Microdialysis in the human brain: Extracellular measurements in the thalamus of parkinsonian patients. Life Sci 1990;46:301–308. 146 Fedele E, Mazzone P, Stefani A, Bassi A, Ansaldo MA, Raiteri M, Altibrandi MG, Pierantozzi M, Giacomini P, Bernardi G, Stanzione P: Microdialysis in Parkinsonian patient basal ganglia: Acute apomorphine-induced clinical and electrophysiological effects not paralleled by changes in the release of neuroactive amino acids. Exp Neurol 2001;167:356–365. 147 Bankiewicz KS, Daadi MM, Pivirotto P, Bringas J, Sanchez-Pernaute R, Herscovitch P, Carson R, Eckelman W, Cunningham J, Reutter B, VanBrocklin HF, Eberling JL. Long term evaluation of AAV/AADC gene transfer in parkinsonian monkeys. 33rd Annual Meeting of Society for Neuroscience, 2003, Washington DC. Neuroscience 2003, p. Program No. 299.214.

Dr. Krystof Bankiewicz Department of Neurosurgery, University of California San Francisco MCB, 1855 Folsom Street, Room 225, San Francisco, CA 94103 (USA) Tel. ⫹1 415 502 3132, Fax ⫹1 415 514 2777, E-Mail [email protected]


245


Simplifying Complex Neurodegenerative Diseases by Gene Chip Analysis Clemens R. Scherzera, Steven R. Gullansa, Roderick V. Jensenb a

Laboratory for Functional Genomics, Center for Neurologic Diseases, Harvard Medical School, Brigham and Women’s Hospital, Cambridge, Mass., and b Department of Physics, Wesleyan University, Middletown, Conn., USA

Genes for many autosomal dominant or recessive neurodegenerative diseases have been already identified. However, little is known about the complex genetics behind the vast majority of sporadic or ‘idiopathic’ neurodegenerative diseases. These diseases are likely to be caused by the combinatorial effect of several susceptibility genes acting in concert with environmental risk factors. Identifying the relevant genes, elucidating their molecular function, and defining targets for neuroprotective drugs pose great challenges and will require novel scientific methodologies. These genetic strategies will help to bring the benefits of the recent genomic revolution to the clinic and the operating room, by developing treatment strategies for neurodegenerative diseases. Traditional scientific approaches have always focused on serial studies of one gene at a time. For complex diseases that are caused by a multiplicity of susceptibility genes, high-throughput analysis of many genes in parallel is a more efficient and informative approach, though cost considerations have been a major problem in the past. Gene chips or ‘microarrays’ attach probes for transcripts of tens of thousands of genes onto a rigid support such as a glass slide and permit a comprehensive genome-wide analysis of transcript changes. This chapter will discuss how gene chip technology can be applied to the investigation of neurodegenerative diseases. We will address how gene chips can identify candidate disease-modifying genes and prioritize susceptibility genes for genotyping in complex neurodegenerative diseases. We envision that in the future, gene chip analysis will efficiently detect the molecular fingerprints associated with distinct clinical states and will define unique gene activity profiles

or ‘mRNA barcodes’ for specific clinical traits. In clinical practice, these tools may assist in the diagnosis and prognosis of neurodegenerative diseases, more accurately predict individual treatment responses, and be used as markers of disease risk in presymptomatic subjects.

Current Best Practices of Microarray Technique: Refining Modifier Candidates

Primary Screen In our experience, many investigators would like to use a combination of microarrays, bioinformatics, and simple validation experiments to define a short list of one to ten high-priority candidate genes. A stepwise filtering process is generally applied to the initial microarray datasets. We typically start with error models tailored to the specific microarray platforms, to optimize quantification of the gene expression levels. We also recommend a stringent three-step statistical analysis to minimize false positives due to biological or technical variation and to correct for multiple testing. First, a selective intensity filter is applied to exclude genes with low hybridization signal intensities, because false-positive results are particularly high for low-intensity genes. With Affymetrix gene arrays, we generally require that the gene ‘Average Difference’ or ‘Signal’ be greater that the ‘Target Intensity’ (defined as the trimmed-mean expression level on the array) for at least one sample in the study. This will focus further analysis on the 30–40% most abundant transcripts. Second, a ratio threshold (generally fold changes of >1.5–2.0) is applied to eliminate small changes in expression that are of unclear technical and biological significance. Although smaller fold-changes may be statistically significant they are very difficult to verify by other means (e.g., quantitative polymerase chain reaction with reverse transcription; RT–PCR). Finally, a t-like test statistic is used to identify genes that are expressed differentially on the basis of confidence values or P values [1]. Permutation tests (e.g., Significance Analysis of Microarrays [2]) are performed to estimate the significance of the test statistic and to correct for multiple testing. The number of false positives expected by chance alone is determined by repeatedly permuting the samples’ class labels and computing t statistics for all genes in the scrambled data. Secondary Screen To qualify each gene further after the primary microarray assessment, a secondary screen may be required to independently confirm the observed changes in gene expression. If the primary screen results in a relatively short

Simplifying Complex Neurodegenerative Diseases

247

list (less than fifty genes), quantitative RT–PCR can be performed (on samples) for technical validation. Investigators may further prioritize genes as candidate targets on the basis of their organismal roles; for example, hormones may be favored as potential therapeutic proteins, or receptors or enzymes that are amenable to modulation by small-molecule drugs may be chosen (for further study). For genes with unknown or unclear functions, prioritizing those of greatest physiological relevance requires further analysis such as quantitative RT–PCR or protein expression analysis. Western blot or immunohistochemistry are preferred for protein analysis, but an antibody is not always readily available. The secondary screening process may obtain a more detailed dissection of the biological process using time series, more diverse biological samples, and anatomical specificity. Shotgun Microarrays Secondary screens become labor intensive, time consuming, and expensive if a large list of genes need be confirmed. Therefore, we have begun to use multiple microarray platforms for efficient technical validation of large numbers of differentially expressed genes. Different high-density oligonucleotide platforms (e.g., Affymetrix, Amersham, Agilent) spot distinct probes for the genes interrogated and have distinct technical advantages and weaknesses. Our results suggest that for the more highly expressed transcripts, 70–80% of the >2-fold gene expression changes are concordant when the same RNA sample is run on Affymetrix and Amersham arrays. In our opinion, the current optimal secondary screen takes advantage of two independent high-density oligonucleotide platforms in a cross-validation strategy that we term ‘shotgun’ or sequential microarray analysis. Error Minimization When using microarrays to identify differentially expressed genes, it is important to recognize the inherent error caused by technical and biological variations. Reproducibility and sensitivity problems can generate both falsenegative and false-positive results. But these issues can be addressed readily through robust experimental design, rigorous statistical analysis, the use of biological and technical replicates, and independent verification by quantitative RT–PCR or other microarray platforms. Although microarrays represent a powerful tool for forming initial hypotheses, it is essential to consider the limitations of interpreting biological responses through measurements of mRNA abundance alone. Measurements of mRNA do not directly reflect protein quantities, enzyme activities, or extranuclear signal transduction. Microarray experiments also may fail to resolve true ‘modifier genes’ from homeostatic responses that attempt to restore the original state of the system. Generally, microarray

Scherzer/Gullans/Jensen

248

measurements fail to resolve cause from effect. Thus, successful use of microarray technology requires that sources of error be controlled carefully in the design and execution of experiments. Biological Validation The primary microarray screen will identify a shortlist of high-priority modifier candidates. Each type of selective profiling identifies differentially expressed genes characteristic for a particular RNA source. The choice of (tissue) source and controls will modify the biases flowing into the results of the screen. Invariably, validation experiments will be indicated to distinguish microarray-derived candidates that are strong modifiers of the disease process and to overcome the limitations of each RNA source. Several approaches can be taken to validate and to prioritize candidate modifiers once a shortlist has been identified. Among the most important are gene knockout and knock-in strategies in cells and model organisms, because these can replicate more closely the actions of potential modifiers and identify phenotypic changes and mechanisms. For a high-throughput genetic validation of microarray candidates, simple model organisms such as yeast, flies, and worms are most frequently used. An elegant application of this strategy resulted in the discovery of a new modifier candidate for multiple sclerosis (MS). Microarray analysis of MS lesions yielded new modifiers of MS that were validated in autoimmune encephalomyelitis [3]. In a landmark study [3], Lawrence Steinman’s group at Stanford defined microarray-derived modifiers of human MS. By combining expression analysis and high-throughput sequencing of expressed sequence tags in a rat model of MS and human MS plaque tissue, they found an increase in osteopontin mRNA abundance in both human and rat tissues. The biological role of osteopontin in the progression of MS was then further validated in knockout mice: osteopontin-deficient mice were resistant to the progressive MS subtype and had significantly more remissions compared to wild-type mice. Using microarrays as a screening tool, osteopontin is now a promising novel drug target for blocking progressive MS in humans.

Prioritizing Candidate Suppressors or Enhancers of Neurodegeneration through Gene Chip Analysis

Selective Vulnerability Profiles When using microarrays to discover modifier genes in neurodegenerative diseases, genome-wide mRNA expression profile is determined in postmortem brain tissue from patients. The investigator applies a series of noise filters and


249

significance statistics to identify candidate modifier genes that are differentially expressed in patient tissue out of the tens of thousands of genes interrogated by human genome arrays. Neurodegenerative processes are highly selective for specific neuronal populations and brain regions and are often associated with characteristic histological lesions. Each neurodegenerative disease preferentially affects distinct neuronal populations and distinct brain regions and is associated with hallmark histopathological lesions. This vulnerable neuronal population is often distributed in distinct brain regions. For example, in Parkinson’s disease (PD) dopaminergic neurons localized to the substantia nigra pars compacta are predominantly affected, while dopaminergic cells in other brain regions are less vulnerable. Regional and cellular profiling techniques have been developed that are tailored to investigate the selective regional and cellular vulnerability of neurodegenerative diseases. Expression analysis of vulnerable brain regions (regional profiling), vulnerable neuronal or glial populations (cellular profiling), or characteristic histological lesions such as MS plaques [3] (lesion profiling) has lead to intriguing results reflecting the strengths and weaknesses of each approach. Regional Profiling Nonspecific gene expression changes related to neuronal loss or reactive glial proliferation must be considered in the interpretation of gene expression in affected brain regions. Hauser et al. [in preparation] have used disease controls with dopaminergic cell loss such as progressive supranuclear paralysis to control cell loss not specific to PD pathogenesis. Alternatively, expression changes of neuronal markers such as neurofilaments or of neuronal specific subpopulations such as tyrosine hydroxylase and other dopamine biosynthesis enzymes, and glial markers such as glial fibrillary acidic protein, may be used to estimate the range of gene expression changes accounted for by unspecific cell loss and gliosis alone. Validation of regional expression changes in vulnerable neuronal populations by double-labeling immunohistochemistry or doublelabeling in situ hybridization can address this concern. Analysis of gene expression in patients ‘at risk’ or at presymptomatic disease stages could reduce some of these biases but tissue availability and diagnostic uncertainty limit this approach. Cellular Profiling Laser-capture microdissection (LCM) of vulnerable neuronal populations allows direct sampling of the neuronal population of interest under the microscope [4–6]. LCM controls for some biases associated with regional profiling such as reactive gliosis or nonspecific neuron loss. Distinct considerations


250

guide the interpretation of LCM expression profiles. During interpretation of results, one must take into account whether gene expression changes observed are specific to the disease in question or whether they may be generally found in dying neurons irrespective of the specific disease process. Comparison with cellular profiles in disease controls could help to estimate this bias. In addition, a selection bias might be introduced by LCM; cellular profiling might select for neurons less affected in the disease process. This is particularly a concern if advanced disease stages are profiled. For example, in PD, an estimated 70% of nigral neurons have died prior to the onset of clinical symptoms [7]. Dopaminergic neurons that survive the disease process and thus are found in postmortem tissue might reflect a particularly resistant subpopulation rather than reflecting the transcription profile of vulnerable dopaminergic cells. The cellular gene expression profile thus might identify transcripts of genes conferring enhanced resistance within the vulnerable cell population. Extraneuronal Profiling A novel approach to avoid some of these limitations has made use of altered gene expression in peripheral tissues of patients with neurodegenerative diseases. In this paradigm, neurodegenerative diseases are approached as a systemic disease with systemic changes in the expression of disease-modifying and susceptibility genes that act in a combinatorial fashion with localizing factors unique to vulnerable neuronal populations and lead to selective neurodegeneration. Biochemical and transcriptional alterations in peripheral tissues such as platelets [8], lymphocytes [9, 10], fibroblasts [11] and muscle of neurodegenerative patients have been extensively documented in Alzheimer’s disease (AD), PD, and other neurodegenerative diseases. Indeed, most genes implicated in familial AD [8, 11–13] and familial PD [14, 15] are ubiquitously expressed. To gain insight into the molecular basis of these alterations, we [23] screened differential gene expression in lymphoblasts of controls and two independent groups of AD patients using cDNA microarrays. This genomic screen identified six differentially expressed genes. One of the six genes (LR11) is a novel neuronal ApoE receptor and thus an excellent candidate modifier. Subsequent validation experiments in the brain indicated that LR11 was enriched in vulnerable cortical and hippocampal pyramidal neurons in human control brains, and that it was concentrated in neuronal endosomal-lysosomal compartments. In striking contrast to normal tissue, LR11 was diminished in AD brains with dramatic reductions in surviving neurons. In cultured cells, LR11 overexpression markedly reduced extracellular A␤ levels, providing a mechanistic link between LR11 and A␤ clearance [Levey, unpubl. observations].


251

Thus, changes in LR11 expression in AD lymphoblasts and brain, and its effects on extracellular A␤, suggested an important role for this apoE receptor in AD pathogenesis. Detection of Neuroprotective Targets: Gene Chip Analysis of Early Disease Stages Candidate Modifier Screen in Animal Models Toxic and genetic animal models of neurodegenerative diseases faithfully replicate key features of human neurodegenerative diseases. Microarray analysis of tissue from animal models, which is generally more available than human tissue samples, allows for dissection of the molecular machinery involved in progressive neurodegeneration. In extension of the ‘static’ gene expression snapshot detectable in human postmortem tissue representative of the disease endpoint, transgenic animal models allow for detection of the ‘dynamic’ range of gene expression changes during the disease progression, at any selected timepoints when the animals are sacrificed. This approach is particularly valuable in the analysis of chronic progressive neurodegenerative diseases. Pathology may begin several years prior to the onset of clinical symptoms and progresses from early disease stages associated with low morbidity and good response to medications to clinically debilitating end stages associated with the depletion of select neuronal populations. Specimens from animal models can capture these changes over the entire course of a disease, in statistically meaningful numbers. For example, in PD, tremor and bradykinesia develop only after an estimated 70% of vulnerable dopaminergic neurons in the substantia nigra have already died during the presymptomatic stage, spanning a period of years [7]. It is a fundamental goal for the neurologist to develop medications that stop or slow disease progression at presymptomatic or early disease stages. Modeling changes in presymptomatic or early symptomatic stages is especially crucial for understanding molecular pathogenesis and, perhaps even more importantly, for identifying therapeutic targets that might help to slow the disease process before it reaches the threshold for clinical symptoms. In one model of PD, Drosophila expressing human ␣-synuclein (␣S) carrying the disease-linked A30P mutation in a panneural pattern faithfully replicate age-dependent onset and chronic progression of human PD. Transgenic ␣S Drosophila develop adult-onset, progressive degeneration of dopaminergic cells, with widespread Lewy body inclusions and impaired locomotor function as monitored by progressive loss of climbing ability [16]. Loss of dopaminergic neurons and inclusion formation are first detected at 10 days of age, while at day 1 post-eclosion, the A30P-␣S Drosophila are still histologically and behaviorally normal.


252

To identify gene expression changes at presymptomatic, early and advanced disease stages, our group hybridized RNA extracted from fly heads to highdensity oligonucleotide arrays spotted with probes representing the entire Drosophila genome. In presymptomatic ␣S transgenics, microarray analysis was more sensitive than conventional neuropathological techniques in elucidating disease-associated changes [17]. It was interesting that despite a ‘normal’ phenotype at this stage, in the one-day-old ␣S transgenics, transcription of thirty six genes was significantly and reproducibly dysregulated. These abnormalities presaged neuronal loss, Lewy body-like inclusion formation, and locomotor impairment at later stages. We found that the ␣S signature genes are dysregulated independent of disease stage in both presymptomatic and symptomatic animals (fig. 1). This suggests that parts of the molecular machinery dysregulated during symptomatic disease stages is already altered in presymptomatic transgenics prior to the onset of neurodegeneration (fig. 1). Thus, temporal profiling of progressive gene expression changes in neurodegenerative disease models provides unbiased starting points for defining disease mechanisms and for identifying potential targets for neuroprotective drugs at preclinical stages. Discovery of Susceptibility Genes by Converging Arraying and Mapping Genes controlling a certain clinical trait may cause variation in the trait through differential transcription due to DNA polymorphisms [18] that regulate transcription. Microarray analysis can assist traditional linkage analysis by identifying polymorphic transcription and in prioritizing candidate susceptibility genes. The correlation structure between transcript abundance and classical genetic linkage has been used to identify susceptibility loci for complex diseases such as diabetes [19] and asthma [20]. Most recently, convergence of gene expression and linkage analysis implicated a novel gene, glutathione S-transferase omega in the control of age-at-onset of AD [24]. A genetic linkage screen for age-at-onset in AD and PD has identified several chromosomal regions that may harbor novel age-at-onset genes [21]. The most interesting finding was a ⬃15 cM linkage region on chromosome 10q. This linkage peak was large, spanning over fifteen megabases and several hundred genes. Gene expression analysis probing for 22,000 human genes on RNA from 6 AD patient hippocampus and matched normal controls was performed to identify genes with polymorphic transcription in AD versus control brain. Fifty-two genes were identified that demonstrated significant differences in gene expression levels between AD and controls. Four of these fifty-two genes were physically located in the chromosome 10q linkage region. Genotyping fourteen single nucleotide polymorphisms in 1773 AD and 635 PD patients spanning these 4 candidates, and one functionally related


253

a

b

c

d

f e

1


254

isoform, identified allelic association (p ⫽ 0.001) for age-at-onset with glutathione S-transferase omega, one of the 4 microarray-defined candidates. Thus, integration of the independent genetic linkage, gene expression, and allelic association evidence implicated a novel gene as a significant biological factor in the control of age-at-onset in AD. The Practice of Genomic Neurosurgery: Diagnosis, Subtype-Classification, and Treatment Personalization through Microarray-Derived Biomarkers The prospect of neuroprotective therapy has highlighted the crucial need for disease-specific biomarkers that identify patients at early stages and allow monitoring of disease progression. In addition, biomarkers for disease subtypes are needed to efficiently design clinical trials for neuroprotective drugs. In our opinion, transcription levels of susceptibility, disease-modifying, and treatment-modifying genes will result in defined gene expression ‘barcodes’ based on haplotypes or single-nucleotide polymorphisms. These gene expression barcodes will serve to diagnose patients with neurodegenerative diseases, to classify disease subtypes, and to predict treatment responses. Finally, using bioinformatics techniques such as the ‘gene ratios’ [22], a small number of genes will be extracted from the gene expression patterns that best define a clinical state. This small subset of genes will then be assayed by simple and widely available standard laboratory techniques such as quantitative real-time PCR. Fig. 1. Gene expression changes presage neurodegeneration in a Drosophila model of PD (from [17]). a 51 signature genes tightly associated with A30P-␣-synuclein expression independent of disease stage are clustered by hierarchical average-linkage analysis and visualized in a colorgram. The branches of the dendrogram comprising the cluster of four independent samples of presymptomatic 1-day-old transgenics are highlighted in pink. Expression levels higher than the mean are displayed in red, lower than the mean in blue. b–d While histology and behavior are normal in presymptomatic 1-day-old ␣S-transgenics, microarray profiles reveal a PD-specific expression signature. Graphs show the average fold change of select genes in different functional classes at day 1, 10, and 30 for ␣S transgenics (left panels) and tau transgenics (right panels). In R406W-tau transgenics, expression of the ␣S-signature genes is generally unchanged (changes not significant by SAM). Time points representing symptomatic stages of PD pathology are shaded gray. Signature expression of down-regulated lipid genes (b, and green font in a), up-regulated membrane transporters (c, orange font), and defense response genes (d, blue font) is detectable at the presymptomatic stage. e Using the ␣S-associated signature genes as classifiers, blinded hierarchical average-linkage analysis correctly distinguishes the eight ␣S samples from tau transgenics and, as expected, from normal controls. f Progressive up-regulation of a set of energy genes also begins in presymptomatics. This increase may be a compensatory response different from the energy genes uniquely down-regulated at day 1 (fig. 1a). [Reproduced with permission from Hum Mol Genet 2003;12:2457–2466, Oxford University Press.]


255

Acknowledgements We thank Dr. Mel Feany at Brigham and Women’s Hospital, Harvard Medical School, for her helpful comments and critical review of the manuscript. The authors are supported by grants from the Harvard Center for Neurodegeneration and Repair, the Paul B. Beeson Career Development Award in Aging Research, the George C. Cotzias Memorial Fellowship from the American Parkinson Disease Association (to C.R.S.), and the Michael J. Fox Foundation (C.R.S, S.R.G., R.V.J).

References 1 2 3

4 5 6

7 8

9 10

11

12 13

14

15 16

Slonim DK: From patterns to pathways: Gene expression data analysis comes of age. Nat Genet 2002;32(suppl):502–508. Tusher VG, Tibshirani R, Chu G: Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 2001;98:5116–5121. Chabas D, Baranzini SE, Mitchell D, Bernard CC, Rittling SR, Denhardt DT, Sobel RA, Lock C, Karpuj M, Pedotti R, et al: The influence of the proinflammatory cytokine, osteopontin on autoimmune demyelinating disease. Science 2001;294:1731–1735. Bahn S, Augood SJ, Ryan M, Standaert DG, Starkey M, Emson PC: Gene expression profiling in the post-mortem human brain – No cause for dismay. J Chem Neuroanat 2001;22:79–94. Wittliff JL, Erlander MG: Laser capture microdissection and its applications in genomics and proteomics. Methods Enzymol 2002;356:12–25. Kamme F, Salunga R, Yu J, Tran DT, Zhu J, Luo L, Bittner A, Guo HQ, Miller N, Wan J, Erlander M: Single-cell microarray analysis in hippocampus CA1:Demonstration and validation of cellular heterogeneity. J Neurosci 2003;23:3607–3615. Fearnley JM, Lees AJ: Ageing and Parkinson’s disease: Substantia nigra regional selectivity. Brain 1991;114:2283–2301. Di Luca M, Colciaghi F, Pastorino L, Borroni B, Padovani A, Cattabeni F: Platelets as a peripheral district where to study pathogenetic mechanisms of Alzheimer disease: The case of amyloid precursor protein. Eur J Pharmacol 2000;405:277–283. Ibarreta D, Urcelay E, Parrilla R, Ayuso MS: Distinct pH homeostatic features in lymphoblasts from Alzheimer’s disease patients. Ann Neurol 1998;44:216–222. Caronti B, Tanda G, Colosimo C, Ruggieri S, Calderaro C, Palladini G, Pontieri FE, Di Chiara G: Reduced dopamine in peripheral blood lymphocytes in Parkinson’s disease. Neuroreport 1999; 10:2907–2910. Citron M, Vigo-Pelfrey C, Teplow DB, Miller C, Schenk D, Johnston J, Winblad B, Venizelos N, Lannfelt L, Selkoe DJ: Excessive production of amyloid beta-protein by peripheral cells of symptomatic and presymptomatic patients carrying the Swedish familial Alzheimer disease mutation. Proc Natl Acad Sci USA 1994;91:11993–11997. Li QX, Fuller SJ, Beyreuther K, Masters CL: The amyloid precursor protein of Alzheimer disease in human brain and blood. J Leukoc Biol 1999;66:567–574. Schlossmacher MG, Ostaszewski BL, Hecker LI, Celi A, Haass C, Chin D, Lieberburg I, Furie BC, Furie B, Selkoe DJ: Detection of distinct isoform patterns of the beta-amyloid precursor protein in human platelets and lymphocytes. Neurobiol Aging 1992;13:421–434. Shin EC, Cho SE, Lee DK, Hur MW, Paik SR, Park JH, Kim J: Expression patterns of alphasynuclein in human hematopoietic cells and in Drosophila at different developmental stages. Mol Cells 2000;10:65–70. Sunada Y, Saito F, Matsumura K, Shimizu T: Differential expression of the parkin gene in the human brain and peripheral leukocytes. Neurosci Lett 1998;254:180–182. Feany MB, Bender WW: A Drosophila model of Parkinson’s disease. Nature 2000;404: 394–398.


256

17 18

19

20

21

22

23

24

Scherzer CR, Jensen RV, Gullans SR, Feany MB: Gene expression changes presage neurodegeneration in a Drosophila model of Parkinson’s disease. Hum Mol Genet 2003;12:2457–2466. Schadt EE, Monks SA, Drake TA, Lusis AJ, Che N, Colinayo V, Ruff TG, Milligan SB, Lamb JR, Cavet G, et al: Genetics of gene expression surveyed in maize, mouse and man. Nature 2003; 422:297–302. Eaves IA, Wicker LS, Ghandour G, Lyons PA, Peterson LB, Todd JA, Glynne RJ: Combining mouse once genic strains and microarray gene expression analyses to study a complex trait: The NOD model of type 1 diabetes. Genome Res 2002;12:232–43. Karp CL, Grupe A, Schadt E, Ewart SL, Keane-Moore M, Cuomo PJ, Kohl J, Wahl L, Kuperman D, Germer S, et al: Identification of complement factor 5 as a susceptibility locus for experimental allergic asthma. Nat Immunol 2000;1:221–226. Li YJ, Scott WK, Hedges DJ, Zhang F, Gaskell PC, Nance MA, Watts RL, Hubble JP, Koller WC, Pahwa R, et al: Age at onset in two common neurodegenerative diseases is genetically controlled. Am J Hum Genet 2002;70:985–993. Gordon GJ, Jensen RV, Hsiao LL, Gullans SR, Blumenstock JE, Ramaswamy S, Richards WG, Sugarbaker DJ, Bueno R: Translation of microarray data into clinically relevant cancer diagnostic tests using gene expression ratios in lung cancer and mesothelioma. Cancer Res 2002;62: 4963–4967. Scherzer CR, Offe K, Gearing M, Rees HD, Fang G, Heilman CJ, Schaller C, Bujo H, Levey AI, Lah JJ: Loss of apolipoprotein E receptor LR11 in Alzheimer disease. Arch Neurol 2004;61: 1200–1205. Li YJ, Oliveira SA, Xu P, Martin ER, Stenger JE, Scherzer CR, Hauser MA, Scott WK, Small GW, Nance MA, Watts RL, Hubble JP, Koller WC, Pahwa R, Stern MB, Hiner BC, Jankovic J, Goetz CG, Mastaglia F, Middleton LT, Roses AD, Saunders AM, Schmechel DE, Gullans SR, Haines JL, Gilbert JR, Vance JM, Pericak-Vance MA: Glutathione S-transferase omega-1 modifies age-atonset of Alzheimer disease and Parkinson disease. Hum Mol Genet 2003;12:3259–3267.

Clemens R. Scherzer, MD Laboratory for Functional Genomics, Center for Neurologic Diseases Harvard Medical School, Brigham and Women’s Hospital 65 Landsdowne Street, Suite 327, Cambridge, MA 02319 (USA) Tel. ⫹1 617 768 8697, Fax ⫹1 617 768 8595, E-Mail [email protected]


257


Molecular Pathology of Dementia Emerging Treatment Strategies

Gunnar K. Gouras Department of Neurology and Neuroscience, Laboratory of Alzheimer’s Disease Neurobiology, Weill Medical College of Cornell University, New York, N.Y., USA

Introduction

Dementia is defined as a progressive and abnormal decline in cognition, typically over months to years. Delirium differs from dementia in that it is an acute or subacute impairment in cognitive abilities, often caused by a reversible toxic or metabolic insult. Neurodegenerative diseases of aging that cause dementia are a growing public health concern as life expectancy increases. Alzheimer’s disease (AD), the most common cause of dementia, currently afflicts about 4 million Americans. Annual costs associated with the care of patients with AD to our society (USA) have been estimated to exceed USD 100 billion annually, and will only increase unless new therapeutic approaches are devised. Major categories in the differential diagnosis of dementia include diverse neurodegenerative diseases, toxic-metabolic encephalopathies, vascular dementia, structural lesions and dementia of depression (table 1). Currently, causes of dementia warranting surgical interventions include brain masses, subdural hematoma and hydrocephalus. At times, brain tumors, such as a frontal glioma or meningioma, can present with mainly gradual cognitive impairment. An important and neurosurgically treatable cause of dementia, presenting with gradual memory impairment in the elderly, is normal pressure hydrocephalus, which is characterized by the triad of gait impairment, dementia, and urinary incontinence. Obstructive hydrocephalus presenting with dementia may be secondary to the obstruction of CSF flow, as may be caused by a colloid cyst of the third ventricle or aqueductal stenosis. It is interesting that recent work aimed at decreasing the symptoms or progression of Alzheimer’s have looked at shunting CSF fluid as one possible approach (e.g., Eunoe, COGNIShunt System), and clinical trials at Stanford

Table 1. Major categories of dementia Neurodegenerative diseases • Alzheimer’s disease • Diffuse Lewy body disease and Parkinson’s disease • Frontotemporal dementia (Pick’s disease; corticobasal ganglionic degeneration) • Huntington’s disease • Creutzfeldt-Jakob disease (prion diseases) • Progressive supranuclear palsy • Other neurodegenerative diseases (SCAs; lipid storage diseases; demyelinating diseases) Vascular dementia Toxic-metabolic encephalopathy (endocrine, infectious, nutritional, toxins) Normal pressure hydrocephalus Structural lesions • Subdural hematoma • Brain tumor • Obstructive hydrocephalus

University propose a role of CSF purification or clearance as a potential neurosurgical treatment option for selected patients, outside of those with normal pressure hydrocephalus. Other neurosurgical approaches for dementia include the delivery of in vivo or ex vivo gene therapy in the form of recombinant enzymes or growth factors, as well as stem cell transplants to regenerate lost neurons and axons. In this chapter, we will provide an overview of neurodegenerative diseases, the most common cause of dementia in the elderly, and discuss some emerging biological treatment strategies including various molecular neurosurgical approaches. Central Role of Amyloid Beta Peptide (‘Abeta’) in AD

Over the past two decades there has been tremendous progress in better understanding of the molecular biology, pathology, and genetics of AD [8, 23]. The discovery of the peptide sequence of -amyloid (Abeta), the principle component of senile plaques [7, 16] initiated the modern era of molecular biology research into AD. Despite these advances, current treatment for AD remains strictly palliative. There is only one class of medication that is currently F.D.A. approved for the treatment of AD, the cholinesterase inhibitors, and these drugs do not appear to be as effective in mid- to late-stage AD. The use of these drugs evolved from neurochemical studies conducted in the 1970–80s which demonstrated reductions of cholinergic neurotransmission in AD brain tissue, especially in the basal forebrain.

Molecular Pathology of Dementia

259

Of the two hallmarks of AD neuropathology, neurofibrillary tangles and senile plaques, the neurofibrillary tangles have been viewed as less specific to AD because they are found in a variety of other neurodegenerative diseases. For example, familial mutations in tau, the principle component of neurofibrillary tangles, are associated with genetic forms of fronto-temporal dementia. Nonetheless, indirect evidence suggests that tau may be essential to the loss of neurons which occurs in AD [28]. It is interesting that both Abeta and tau may be elevated following traumatic brain injury, particularly chronic brain injury, suggesting that they may constitute a response to injurious stimuli. The importance of Abeta to the AD disease process was strengthened by the discovery of mutations within the Abeta precursor protein (-APP) gene that segregate with autosomal dominant forms of early onset familial AD (FAD). Subsequently, transgenic mice harboring human APP with FAD mutations were developed that reproduce AD-like brain amyloidosis [6, 11]. The discovery of autosomal dominant FAD mutations in presenilin (PS) 1 (chromosome 14) and 2 (chromosome 1) also led to the findings that these forms of FAD invariably lead to increased generation of the longer Abeta42 form of Abeta [23]. Abeta peptides range in length up to 42 or 43 amino acids. Both the Abeta N-terminus and C-terminus reveal heterogeneity, but tend to be mainly referred to in the literature as Abeta 1–40 (Abeta40) and Abeta 1–42 (Abeta42) peptides. The slightly extended Abeta42 aggregates more readily and is the main constituent of senile plaques in AD [9]. The mechanism whereby Abeta is involved in AD pathogenesis remains controversial. Numerous investigators have demonstrated that Abeta isoforms are neurotoxic when added to cultured neurons in vitro and when injected into the brain of experimental animals in vivo [29]. Accordingly, neuritic plaques found in the extracellular space in AD brains are presumed to be toxic to surrounding neurons and their processes. Recent evidence suggests that AD, analogous to a growing number of diverse neurodegenerative diseases, is also characterized by the intracellular accumulation of its disease-linked Abeta peptide. Indeed, a recent immunoelectron microscopy study observed accumulation of Abeta within neurons, especially within distal neuronal processes of transgenic APP mice with aging prior to and with the onset of plaque pathology [30]. Currently, increases in Abeta oligomers (mainly soluble Abeta40 but also insoluble Abeta42) are viewed as important neurotoxic intermediates in the development of neuronal dysfunction [13, 23]. Increased soluble Abeta levels appear to be the best Abeta correlate of cognitive dysfunction from mild cognitive function through more severe stages of AD [18; fig. 1 for schema of AD pathogenesis]. A major focus of AD research continues to be to better

Gouras

260

FAD mutations (APP.PS1, PS2)

Apo E

High cholesterol head trauma

↑Soluble A

Aging

↑Insoluble A oligomers

Neuronal and synaptic dysfunction

MCI

A plaques and NFTs

Loss of nerve cells and synapses

Dementia

Fig. 1. Schema of AD pathogenesis.

understand the biochemical mechanisms by which the genetically diseaselinked APP/Abeta, PS, and apolipoprotein-E are involved in neurobiology and the AD disease process. Overall, there is growing optimism that with major advances in our understating of neurodegenerative diseases of aging, therapies for these incurable disorders are not far away [8; table 2; fig. 2]. Inhibitors of ␥- and ␤-Secretase

The increasing evidence for the importance of Abeta accumulation in AD has made reduction of Abeta a leading target for AD therapy. Strategies to accomplish Abeta reduction include inhibition of the proteases that lead to the cleavage of Abeta from its larger precursor, -APP (fig. 3). The -site APP cleaving enzyme (BACE) was discovered to be responsible for the initial cleavage of APP at the N-terminus of Abeta that first produces an APP C-terminal fragment or CTF [31]. PS is viewed as critical for the subsequent -secretase cleavage of the CTFs that generates the 40 or 42 C-terminal ends of Abeta [23]. Given that BACE knockout mice appear normal while PS1 knockout mice are embryonically lethal [25], and reports linking PS to cleavages of several other important proteins, BACE inhibition currently is the leading therapeutic target for Abeta inhibition [2]. In common with other therapies directed at the CNS, efficient delivery of such secretase inhibitors could be critical and may depend on placement of indwelling infusion devices into the CSF or brain parenchyma to bypass the blood brain barrier.


261

Amyloid precursor protein (APP) metabolism -secretase (BACE)

-secretase

N

C 1 11 17

40/42

-secretase

or

CTF

Amyloidogenic 40/42

1 4 kD A

11 40/42 3.4 kD A

-amyloid

Nonamyloidogenic CTF

sAPP 17 40/42 3 kD p3

p3

Fig. 2. Amyloid precursor protein (APP) metabolism.

Table 2. Outline of therapeutic treatment strategies for AD Cholinesterase inhibitors (i.e., Aricept, Exelon, Reminyl) Anti-Abeta approaches • BACE inhibition • gamma-secretase inhibition (inhibition of PS, Aph1, PEN2 and/or Nicastrin) • anti-Abeta vaccine • cholesterol-lowering agents • anti-aggregation compounds (e.g., beta-sheet breakers) • promoters of Abeta degrading enzymes (neprilysin, IDE, neutral endopeptidase) Anti-oxidants (e.g., Vitamin E) Anti-inflammatory medications (e.g., NSAIDs) Estrogen replacement therapy → NO LONGER RECOMMENDED! Anti-tau strategies Anti-apoptosis agents Neurotrophic factors (e.g., NGF) CSF shunting Stem cells Gene therapy

Gouras

262

Abeta Vaccine: Theory and Practice

Other Abeta-based therapeutic approaches that are under investigation include anti-Abeta vaccines, in which Abeta is deliberately administered with an adjuvant in order to stimulate the body’s natural immune response to clear the substance from the blood and the brain. Initial excitement demonstrating reductions of cerebral Abeta plaques in transgenic FAD APP mice in response to intravenous infusion of Abeta or anti-Abeta antibodies was seriously dampened by the discontinuation of a clinical trial of Abeta infusion into AD patients secondary to the occurrence of encephalitis in ⬃5% of patients. Despite this setback for an anti-Abeta vaccine, reports indicating cognitive stabilization in some patients with high anti-Abeta antibody titers provide some encouragement [10]. Overall, an immunological approach to AD has received great interest in the field, but remains a therapeutic direction in AD that is less well understood than others. Further research on the neuroimmunology of Abeta may help elucidate the mechanism whereby anti-Abeta antibodies may benefit patients with AD. Some investigators have postulated that anti-Abeta antibodies in the systemic circulation may act as a ‘peripheral Abeta sink’ that could draw Abeta out of the brain [5]. Theoretically, evidence for the sink hypothesis could also be viewed as a mechanism whereby neurosurgical CSF shunting [27] may have therapeutic potential for AD, since it would be expected to divert Abeta out of the brain.

Other Strategies Based on Reduction of Abeta

Retrospective studies suggest that intake of cholesterol-lowering 3-hydroxy-3-methyl glutaryl coenzyme A reductase inhibitors, commonly known as statins, are associated with protection against AD. In addition, treatment of APP transgenic mice with a high fat cholesterol diet increases, while treatment with statins reduces, Abeta plaque pathology [21, 22]. There is a compelling biological mechanism by which cholesterol influences Abeta generation, since the critical -secretase cleavage of Abeta occurs within the membrane lipid bilayer and could be modulated by higher or lower cholesterol; supportive evidence comes from drops in Abeta levels in cultured cells that parallels transgenic mice studies. Prospective clinical trials to assess the efficacy of cholesterol-lowering strategies for the prevention and/or treatment of AD are ongoing, and preliminary results suggest an effect of statin treatment. In the past, retrospective studies suggested that estrogen replacement for postmenopausal women might reduce the incidence of AD. Studies in tissue culture and on APP transgenic mice support the efficacy of both estrogen and


263

Therapeutic anti-A targets in Alzheimer’s disease Cholesterol modulators - and -secretase inhibitors Antioxidants

Gonadal hormones

Immunotherapy

-sheet breakers

?

Intracellular A

A plaques ?

Signal transduction modulators

Metal chelators Modulators of inflammation

Fig. 3. Sites for intervention vis-à-vis Abeta for Alzheimer’s disease: treatment strategies aimed at reducing Abeta are either directed at APP/Abeta metabolism within nerve cells or Abeta associated with extracellular plaques.

noninflammatory treatments. Unfortunately, prospective studies of patients with mild-to-moderate AD have failed to demonstrate benefits for estrogen. Indeed, patients treated with estrogen experienced more adverse side effects, especially vascular complications such as deep vein thrombosis and heart attack. In addition, estrogen intake increases the risk for the development of breast and uterine cancer. Moreover, the Women’s Health Initiative Memory Study (WHIMS) [32] showed that rather than helping as had been originally proposed, estrogen-progesterone hormone replacement therapy was harmful and actually increased the occurrence of Alzheimer’s dementia over the treatment period. The use of anti-inflammatory medications such as NSAIDs has also been proposed as a possible treatment option for incipient AD, given that experimental evidence supports the role of various inflammatory processes in AD, including the role of activated microglia and brain cytokine release. However, given a lack of clinical data to support this approach and bearing in mind the results of estrogen clinical trials which appear to contradict earlier in vitro and in vivo work, caution is warranted in recommending anti-inflammatory drugs at this time. It appears that inflammatory processes do affect the amount of Abeta produced in cell culture and mouse models of AD, yet, prospective, randomized clinical trials are required to validate this approach. Metal chelation has been proposed in the past as a treatment for AD, since certain divalent or trivalent metals appear to increase aggregation of Abeta in vitro and the antibiotic clioquinol was reported to reduce plaque pathology in transgenic APP mice [4]. However, metal accumulation is very likely to be

Gouras

264

an epiphenomenon of other changes and there are no convincing data to support the role of metals in the pathology of AD or the use of chelation therapy. Signal transduction modulators such as GSK3, CDK5, PI3K, or PKC modulators have been proposed as treatment strategies for AD, acting via mechanisms that may affect both Abeta and tau. Augmenting the various proteases involved in degrading Abeta in brain is another possible treatment approach. Reduced clearance and degradation of Abeta may be more important for most cases of AD than the increased generation of Abeta associated with the relatively rare early onset forms of AD. Some important enzymes in the brain that have been demonstrated to degrade Abeta in vitro or in vivo include neprilysin, insulin degrading enzyme, neutral endopeptidase, and other metalloproteinase enzymes. Protease augmentation in targeted brain areas would be particularly amenable to an in vivo or an ex vivo gene transfer approach. Excitotoxicity is a general phenomenon involved in neuronal cell damage, in which glutamatergic, excitatory transmission leads to excessive cell stimulation, pathological Ca influx, and associated processes such as apoptosis. Despite failure of other glutamatergic antagonists in the past to prevent neurodegeneration, the N-methyl-D-aspartate (NMDA) antagonist memantine was reported in a prospective clinical study to improve function in patients with moderate to severe AD [20], though the biological mechanism by which it may be beneficial for AD is not known. NMDA-mediated excitotoxicity is one potential mechanism for protective effects of NMDA antagonists, though the role of NMDA toxicity has not been precisely defined in the pathophysiology and treatment of many neurodegenerative diseases. A recent neurobiological study suggested that NMDA antagonists may provide therapeutic benefit in AD by other biological effects on Abeta and neuronal physiology [12]. Increasing evidence supports the hypothesis that oxidative stress is a mechanism whereby aging is the major risk factor for age-related neurodegenerative diseases [1]. While the biochemical pathways involved in the physiologically relevant oxidative stress associated with age-related diseases of the brain require more definition, effects on the progression of AD in a large double-blind clinical study following treatment with the anti-oxidant vitamin E suggests that this approach requires further study [24]. Because oxidative species may accentuate or accelerate the negative effects of Abeta deposition, it is possible that lowering oxidative stress may have synergistic effects on lowering Abeta. Gene Therapy for Alzheimer’s? The understanding that a significant component of AD, especially early onset forms of AD, are inherited or that almost half of the more typical late


265

onset cases of AD are associated with a single genetic risk factor, the apolipoprotein E4 allele, has led to a growing awareness that gene therapy may be an important direction for future therapy. Other neurodegenerative diseases are known to have, and are increasingly being found to have, complex genetic etiologies. Since a common theme in many neurodegenerative diseases (e.g., AD, Huntington’s disease, Parkinson’s disease) is the abnormal intracellular accumulation of insoluble peptides, the most direct method to reduce the expression of a disease-linked protein is reduced transcription of the protein, increased transcription of proteases involved in the degradation of a given protein, or reduction in proteases involved in the generation of toxic proteolytic fragments (e.g., generation of Abeta from APP). In addition to standard gene transfer using viral and nonviral vectors, siRNA has emerged as a powerful new method to reduce gene transcription and is being employed in experimental strategies to reduce protein accumulation associated with neurodegenerative diseases such as AD. Effective gene therapy will require improvements in vectors to reliably transduce cells in the CNS while minimizing local inflammation and/or tissue damage. Once technical obstacles for safe and effective gene transfer to the CNS can be accomplished, manipulation of genes may one day revolutionize the treatment of neurodegenerative diseases. Stem cells are increasingly being explored as a vector to deliver ex vivo gene transfer and may be useful for dementias to replenish neurons that are destroyed as a result of a neurodegenerative disease. Neurotrophic factors, important in the development of the nervous system, continue to be produced in adulthood and have been proposed as a therapeutic option for neurodegenerative diseases such as AD, particularly in combination with stem cell-based ex vivo gene therapy. Too Much Focus on Abeta? Despite this preponderance of anti-amyloid approaches for treating AD, there is not unanimous agreement in the field whether the focus on Abeta is justified. Arguments against the primary role of Abeta in AD include the following: only a small group of AD patients have mutations in APP (1%); Abeta plaques do not correlate with cognitive dysfunction as well as synaptic loss; and significant amounts of A plaques can be found in postmortem brains of people without clear clinical evidence of dementia. It should also be pointed out that despite its apparently deleterious role in AD, Abeta40 (i.e., the majority of soluble Abeta), Abeta42 (i.e., largely insoluble Abeta fraction), and other proteolytic fragments of Alzheimer’s precursor protein such as sAPPalpha or sAPPbeta may have important physiological functions. APP knockout mice were originally reported to have normal brains, but subsequent work has demonstrated abnormalities in synaptic markers in APP knockout

Gouras

266

mouse brain and decreased neurite extension in cultured APP knockout mouse neurons [19]. Indeed, recent work has indicated that neuronal activity increases Abeta secretion, which in turn may provide a negative feedback to modulate further neuronal activity [12]. It was reported that transection of the perforant path, the major outflow tract from the entorhinal cortex to hippocampus, reduced plaque pathology in the hippocampus of transgenic APP mice [15, 26]. Because APP can be transported down axons by fast axonal transport [14], normal transport of APP may be required to develop plaques in the terminal fields of axonal projections. Therapeutic implications include a concern that treatments which interfere with axonal transport may reduce plaques in transgenic mouse models of beta-amyloidosis, and therefore appear to have therapeutic potential for AD while in fact they may be detrimental to neurons and not beneficial in AD. Aggregation of Insoluble Proteins in Neurodegenerative Diseases There are a number of major similarities among diverse neurodegenerative diseases of aging, which include age-related development of aberrant cellular accumulation and aggregation of insoluble proteins in vulnerable brain regions; a role for oxidative stress; and the occurrence of inflammation. While it is still debated whether intra- or extracellular aggregates are directly toxic to the cell or are an attempt by the cell to defend itself, the prevailing view is that such aggregates within aging cells of the brain are likely to be detrimental. Mutations associated with Parkinson’s disease on two different proteins, parkin, a ubiquitin ligase, and ubiquitin C-terminal hydrolase L1, have pointed to a role for the ubiquitin-proteasome degradation pathway for intracellular proteins in neurodegenerative diseases [17]. Evidence also indicates that endosomallysosomal system abnormalities occur early in the development of AD [3]. These multiple lines of evidence indicate that the reduced ability of selective neuronal populations to efficiently degrade disease-linked protein aggregates are a final common pathway in neurodegenerative diseases of aging. Future neurosurgical interventions that may be critical for the treatment of neurodegenerative diseases such as AD could include placement of infusion devices for delivery of treatments as diverse as gene therapeutic agents (i.e., viral vectors), protease inhibitors, anti-Abeta antibodies to bypass the bloodbrain barrier, stem cells, and CSF shunting to reduce levels of soluble brain Abeta in patients with AD.

Acknowledgments The author thanks Dr. Michael T. Lin for his critical reading of the chapter and helpful discussions.


267

References 1 2 3

4

5

6

7 8 9

10

11 12 13 14

15

16

17 18

19

Beal MF: Mitochondria, free radicals, and neurodegeneration. Curr Opin Neurobiol 1996;6: 661–666. Cai H, Wang Y, McCarthy D, Wen H, Borchelt DR, Price DL, Wong PC: BACE1 is the major betasecretase for generation of Abeta peptides by neurons. Nat Neurosci 2001;4:233–234. Cataldo AM, Peterhoff CM, Troncoso JC, Gomez-Isla T, Hyman BT, Nixon RA: Endocytic pathway abnormalities precede amyloid beta deposition in sporadic Alzheimer’s disease and Down syndrome: Differential effects of APOE genotype and presenilin mutations. Am J Pathol 2000; 157:277–286. Cherny RA, Atwood CS, Xilinas ME, Gray DN, Jones WD, McLean CA, Barnham KJ, Volitakis I, Fraser FW, Kim Y, Huang X, Goldstein LE, Moir RD, Lim JT, Beyreuther K, Zheng H, Tanzi RE, Masters CL, Bush AI: Treatment with a copper-zinc chelator markedly and rapidly inhibits betaamyloid accumulation in Alzheimer’s disease transgenic mice. Neuron 2001;30:665–676. DeMattos RB, Bales KR, Cummins DJ, Dodart JC, Paul SM, Holtzman DM: Peripheral antiA beta antibody alters CNS and plasma A beta clearance and decreases brain A beta burden in a mouse model of Alzheimer’s disease. Proc Natl Acad Sci USA 2001;98:8850–8885. Games D, Adams D, Alessandrini R, Barbour R, Berthelette P, Blackwell C, Carr T, Clemens J, Donaldson T, Gillespie F, et al: Alzheimer-type neuropathology in transgenic mice overexpressing V717F beta-amyloid precursor protein. Nature 1995;373:523–527. Glenner GG, Wong CW: Alzheimer’s disease: Initial report of the purification and characterization of a novel cerebrovascular amyloid protein. Biochem Biophys Res Commun 1984;120:885–890. Golde TE: Alzheimer disease therapy: Can the amyloid cascade be halted? J Clin Invest 2003; 111:11–18. Iwatsubo T, Odaka A, Suzuki N, Mizusawa H, Nukina N, Ihara Y: Visualization of A beta 42(43) and A beta 40 in senile plaques with end-specific A beta monoclonals: Evidence that an initially deposited species is A beta 42(43). Neuron 1994;13:45–53. Hock C, Konietzko U, Streffer JR, Tracy J, Signorell A, Muller-Tillmanns B, Lemke U, Henke K, Moritz E, Garcia E, Wollmer MA, Umbricht D, de Quervain DJ, Hofmann M, Maddalena A, Papassotiropoulos A, Nitsch RM: Antibodies against beta-Amyloid slow cognitive decline in Alzheimer’s disease. Neuron 2003;38:547–554. Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, Younkin S, Yang F, Cole G: Correlative memory deficits, Abeta elevation, and amyloid plaques in transgenic mice. Science 1996;274:99–102. Kamenetz F, Tomita T, Hsieh H, Seabrook G, Borchelt D, Iwatsubo T, Sisodia S, Malinow R: APP processing and synaptic function. Neuron 2003;37:925–937. Klein WL, Krafft GA, Finch CE: Targeting small Abeta oligomers: The solution to an Alzheimer’s disease conundrum? Trends Neurosci 2001;244:219–224. Koo EH, Sisodia SS, Archer DR, Martin LJ, Weidemann A, Beyreuther K, Fischer P, Masters CL, Price DL: Precursor of amyloid protein in Alzheimer disease undergoes fast anterograde axonal transport. Proc Natl Acad Sci USA 1990;87:1561–1565. Lazarov O, Lee M, Peterson DA, Sisodia SS: Evidence that synaptically released beta-amyloid accumulates as extracellular deposits in the hippocampus of transgenic mice. J Neurosci 2002;22: 9785–9793. Masters CL, Simms G, Weinman NA, Multhaup G, McDonald BL, Beyreuther K: Amyloid plaque core protein in Alzheimer disease and Down syndrome. Proc Natl Acad Sci USA 1985;82: 4245–4249. McNaught KS, Olanow CW: Proteolytic stress: A unifying concept for the etiopathogenesis of Parkinson’s disease. Ann Neurol 2003;53(suppl 3):S73–S84. Naslund J, Haroutunian V, Mohs R, Davis KL, Davies P, Greengard P, Buxbaum JD: Correlation between elevated levels of amyloid beta-peptide in the brain and cognitive decline. JAMA 2000;283:1571–1577. Perez RG, Zheng H, Van der Ploeg LH, Koo EH: The beta-amyloid precursor protein of Alzheimer’s disease enhances neuron viability and modulates neuronal polarity. J Neurosci 1997; 17:9407–9414.

Gouras

268

20 21

22

23 24

25 26

27

28 29 30

31

32

Reisberg B, Doody R, Stoffler A, Schmitt F, Ferris S, Mobius HJ: Memantine Study Group. Memantine in moderate-to-severe Alzheimer’s disease. N Engl J Med 2003;348:1333–1341. Refolo LM, Pappolla MA, LaFrancois J, Malester B, Schmidt SD, Thomas-Bryant T, Tint GS, Wang R, Mercken M, Petanceska SS, Duff KE: A cholesterol-lowering drug reduces beta-amyloid pathology in a transgenic mouse model of Alzheimer’s disease. Neurobiol Dis 2001;8:890–899. Refolo LM, Malester B, LaFrancois J, Bryant-Thomas T, Wang R, Tint GS, Sambamurti K, Duff K, Pappolla MA: Hypercholesterolemia accelerates the Alzheimer’s amyloid pathology in a transgenic mouse model. Neurobiol Dis 2000;7:321–331. Selkoe DJ: Alzheimer’s disease: Genes, proteins, and therapy. Physiol Rev 81:741–766. Sano M, Ernesto C, Thomas RG, Klauber MR, Schafer K, Grundman M, Woodbury P, Growdon J, Cotman CW, Pfeiffer E, Schneider LS, Thal LJ: A controlled trial of selegiline, alpha-tocopherol, or both as treatment for Alzheimer’s disease. The Alzheimer’s Disease Cooperative Study. N Engl J Med 1997;336:1216–1222. Shen J, Bronson RT, Chen DF, Xia W, Selkoe DJ, Tonegawa S: Skeletal and CNS defects in Presenilin-1-deficient mice. Cell 1997;89:629–639. Sheng JG, Price DL, Koliatsos VE: Disruption of corticocortical connections ameliorates amyloid burden in terminal fields in a transgenic model of Abeta amyloidosis. J Neurosci 2002;22: 9794–9799. Silverberg GD, Levinthal E, Sullivan EV, Bloch DA, Chang SD, Leverenz J, Flitman S, Winn R, Marciano F, Saul T, Huhn S, Mayo M, McGuire D: Assessment of low-flow CSF drainage as a treatment for AD: Results of a randomized pilot study. Neurology 2002;59:1139–1145. Trojanowski JQ, Lee VM: The role of tau in Alzheimer’s disease. Med Clin North Am 2002;86: 615–627. Yankner BA: Mechanisms of neuronal degeneration in Alzheimer’s disease. Neuron 1996;16: 921–932. Takahashi RH, Milner TA, Li F, Nam EE, Edgar MA, Yamaguchi H, Beal MF, Xu H, Greengard P, Gouras GK: Intraneuronal Alzheimer abeta42 accumulates in multivesicular bodies and is associated with synaptic pathology. Am J Pathol 2002;161:1869–1879. Vassar R, Bennett BD, Babu-Khan S, Kahn S, Mendiaz EA, Denis P, Teplow DB, Ross S, Amarante P, Loeloff R, Luo Y, Fisher S, Fuller J, Edenson S, Lile J, Jarosinski MA, Biere AL, Curran E, Burgess T, Louis JC, Collins F, Treanor J, Rogers G, Citron M: Beta-secretase cleavage of Alzheimer’s amyloid precursor protein by the transmembrane aspartic protease BACE. Science 1999;286:735–741. Shumaker SA, Legault C, Rapp SR, Thal L, Wallace RB, Ockene JK, Hendrix SL, Jones BN, Assaf AR, Jackson RD, Kotchen JM, Wassertheil-Smoller S, Wactawski-Wende J (for the WHIMS Investigators): Estrogen plus progestin and the incidence of dementia and mild cognitive impairment in postmenopausal women. JAMA 2003;289:2651–2662.

Gunnar K. Gouras, MD Assistant Professor, Department of Neurology & Neuroscience Director, Laboratory of Alzheimer’s Disease Neurobiology Weill Medical College of Cornell University, 525 East 68th Street, New York, NY 10021 (USA) Tel. 1 212 746 6598, Fax 1 212 746 8741, E-Mail [email protected]


269


Expanding the Role of Deep Brain Stimulation from Movement Disorders to Other Neurological Diseases Massimo Leone, Angelo Franzini, Giovanni Broggi, Gennaro Bussone Istituto Nazionale Neurologico Carlo Besta, Milano, Italy

Introduction

Deep brain stimulation (DBS) is a relatively new treatment modality, first developed in the 1980s by Benabid and colleagues [3] in France, that until recently has been limited to the treatment of complex movement disorders such as Parkinson’s disease through stimulation of the thalamus, globus pallidus, or subthalamic nucleus. However, recent work suggests that techniques of intracranial stimulation may have much wider applications in diseases that are not well managed by any other treatments. Among the new targets for DBS are intractable epilepsy, intractable dystonia, and intractable headache [1–3]. Because the molecular effects of DBS are still largely unknown, it is possible that DBS may also have applications in neurodegenerative and psychiatric diseases; as the molecular basis of disorders such as Alzheimer’s disease or schizophrenia are elucidated, and effects of DBS on gene transcription and cellular activity are better understood, additional applications in different brain regions may become apparent. It is quite plausible that DBS will offer primary or adjunct treatment options for a variety of complex neurological disorders in the future, many of which the neurosurgeons are not accustomed to treating, but which nevertheless are prevalent and debilitating, perhaps offering new career opportunities for neurosurgeons specializing in functional and restorative neurosurgery. In the special case of hypothalamic DBS, the topic of this chapter, applications outside of intractable cluster headache (CH) may include complex neuroendocrine

Table 1. The cluster headache attack, diagnostic criteria A. At least five seperate attacks fulfilling criteria below B. Severe or very severe unilateral orbital, supraorbital, and/or temporal pain lasting 15 min to 3 h if untreated C. Headache is accompanied by at least one of the following: 1. Ipsilateral conjunctival injection and/or lacrimation 2. Ipsilateral nasal congestion and/or rhinorrhea 3. Ipsilateral eyelid edema 4. Ipsilateral forehead and facial sweating 5. Ipsilateral miosis and/or ptosis 6. Sense of restlessness or agitation

or pain disorders. The field of hypothalamic microinstrumentation is in its infancy, and these studies are a first step toward defining the roles and limitations of DBS outside of the basal ganglia. CH is an interesting clinical syndrome, the excruciating severity of which is not widely appreciated. It is a primary pain syndrome characterized by unilateral and incapacitating headache attacks and also associated with ipsilateral autonomic phenomena [4]. Its prevalence is less than one per 1000 and males are more affected than females, the ratio being about 3:1 [5, 6]. There are two main forms of the disease, episodic and the chronic. About 80% of CHs occur in the episodic form; in this case, the attacks are grouped in so-called ‘cluster periods’ usually lasting 1–2 months at a time. In the chronic form, attacks occur for more than one year without remission or with remissions lasting less than one month. The duration of single attacks ranges from 15 min to 3 h at a time, from once every other day to almost continually at eight times a day [1]. Often, attacks appear at the same hours of the day or night and (for unknown reasons) cluster periods often start in autumn and spring [7]. Circadian rhythmicity is a clinical landmark of the syndrome and for this reason it also has been nicknamed as ‘clock headache.’ The excruciating pain is orbital, supraorbital, and/or temporal in location. At least one of the following autonomic phenomena are present during the attack, ipsilateral to the pain: severe conjunctival injection and/or lacrimation, nasal congestion and/or rhinorrhea, eyelid edema, forehead and facial sweating, miosis and/or ptosis, and a sense of restlessness or agitation (table 1).

Pathophysiology of CH and Links to the Hypothalamus

The pain of CH is probably the most severe known to humans, similar in severity to subarachnoid hemorrhage. Accordingly, it has been termed the

Hypothalamic DBS: Therapeutic Approach for Intractable Cluster Headache

271

‘suicide headache.’ Until a few years ago, CH was still regarded as a vascular headache. In the last years much has been learned about its pathophysiology and as a consequence the vascular theory of the disease has been discarded. The observation of specific events during the headache attacks involving both cranial nerves and the brain suggests that the neural phenomena are much more relevant to the pathophysiology of CHs. Cluster pain is probably initiated by the activation of the first (ophthalmic) division of the trigeminal nerve, also associated with an increase of calcitonin-gene related peptide in the ipsilateral jugular blood during the attack. Autonomic symptoms are due to the activation of the cranial parasympathetic outflow from the VIIth cranial nerve, associated with an increase of vasoactive intestinal polypeptide in the ipsilateral jugular blood during the attack [8]. The relapsing-remitting course of cluster attacks, mainly in autumn and spring, and the clockwise regularity of single attacks initially suggested that the hypothalamus, site of circadian phenomena or the ‘biological clock,’ might play a crucial role in the pathophysiology of this form of headache [9]. A number of neuroendocrinological abnormalities have been reported in CH [10–12], lending further support to this hypothesis. The first direct evidence showing a pivotal role of the hypothalamus in CH came from positron emission tomography (PET) studies. An increased regional blood flow in the posterior inferior ipsilateral hypothalamic gray matter during the acute stage of cluster attack has been shown both in nitroglycerine-induced attacks [13] and in spontaneous cluster attacks [14]. In a voxel-based magnetic resonance imaging (MRI) study, an increased neuronal density was found in the same brain region that was known to be activated in the PET studies, again ipsilateral to the pain [15]. These structural changes were seen independent of the headache state, suggesting an inherent dysfunction of the hypothalamus in CH rather than an epiphenomenon. Although CH was previously described as a vascular headache, and vascular phenomena also appears to be involved, the striking circadian rhythmicity of this strictly unilateral pain syndrome cannot be explained by a simplistic vascular hypothesis. The case report we present in this chapter illustrates the relevance of the hypothalmus to the pathophysiology of CH. We report on a patient suffering from chronic intractable CHs on the left side, who had initially received complete surgical section of the left trigeminal sensory root to relieve the pain [16, 17]. After the operation, he was completely anesthetic over the entire left trigeminal distribution and the left corneal reflex was absent but he continued to have cluster attacks. Blink reflexes of the left supraorbital nerve produced neither ipsilateral nor contralateral blink reflex responses. With the complete section of the left trigeminal sensory root, the brain cannot perceive vasodilatation or a peripheral neural inflammatory process; as a consequence, none of these peripheral structures, be they neural or vascular, are necessary for

Leone/Franzini/Broggi/Bussone

272

an attack to happen [18]. It may be concluded that CH is generated primarily from within the brain.

Pharmacological Treatments

The pharmacological treatment of CH is aimed at aborting the ongoing attacks (i.e., acute therapy) and preventing recurrent attacks during the cluster period (i.e., prophylaxis) [19, 20]. The gold standard in the treatment of acute attacks of CH is the subcutaneous administration of the 5HT1B/D agonist sumatriptan [21], also used in migraine attacks. Alternatively inhalation of 100% oxygen can be used. Verapamil [22], lithium carbonate [23], methysergide [24] and cortisone are the most effective drugs to prevent the incidence of CH. Valproic acid, topiramate, gabapentine, naratriptan, melatonin and local application of civamide or anesthesia of the greater occipital nerve may also be of some help [25].

Surgical Treatment

Radiofrequency thermocoagulation of the trigeminal ganglion has been reported to be effective in about 75% of chronic drug-resistant CH patients [26–29]. Other procedures on the trigeminal nerve have been tempted with inconsistent results [30, 31]. Complications include recurrence of headache, in which case, a repeat procedure may be necessary [28]. Other surgical sequelae are corneal analgesia with resulting potential corneal infection or opacification, and anesthesia dolorosa.

DBS of the Hypothalamus

PET has shown the activation of the ipsilateral posterior inferior hypothalamic gray matter during CH attacks [13, 14], which is apparently specific for the condition [32], while voxel-based morphometric MR has documented alterations in the same area [15], strongly suggesting that the CH generator is located there. We reasoned that the stereotactic stimulation of this area might prevent the activation and hopefully relieve intractable forms of CH [16, 17]. The first hypothalamic implantation using DBS to relieve intractable chronic CH was done in July 2000 [16]. Due to the brilliant results, both in term of painfree state and absence of relevant adverse events, 13 new patients have been implanted so far. A summary of the first implanted patient follows.


273

The First Patient Implanted with Hypothalamic DBS

At the time of treatment, this patient was a 39-year-old right-handed man who had suffered daily CH attacks for 5 years [16]. The attacks lasted from 30 min up to 4 h at a time, occurred two to eight times a day, and were associated with striking oculo-facial phenomena. The majority of the attacks were on the right side and sometimes on the left, but they were never bilateral. Cerebral MR, MR angiography, and catheter angiography were unremarkable. He became completely drug resistant. He was operated on four times on the right trigeminal nerve; after the last thermal rhizotomy, the right side headache attacks ceased, but from that moment, the left-side attacks worsened, again with striking autonomic phenomena. This new onset of CH was completely drug refractory. In addition, the patient was blind on the right as a result of vitreous humor hemorrhage and left trigeminal surgery was highly contraindicated by the risk of corneal sequelae, which could have left the patient totally blind. Stereotactic electrode implantation targeting the posterior inferior ipsilateral hypothalamic gray matter was then proposed. After informed consent the operation was performed under local anesthesia using a CRW frame on July 14, 2000. The electrode (Medtronic 3089, Minneapolis, Minn., USA) was inserted at coordinates 6 mm posterior to the anterior commissure-posterior commissure midpoint, 2 mm left of the midline, and 8 mm below the commissural plane [15, 16]. Intraoperative electrical stimulation induced no side effects. The permanent generator (Soletra, Medtronic, Minneapolis, Minn., USA), embedded in a subclavicular pocket, was connected by subcutaneous tunnelization. Therapeutic stimulation was in continuous unipolar mode. The position of the permanent electrode was verified by postoperative MR. When stimulated at 180 Hz, 3 V, 60-␮s pulse width, the attacks disappeared after 48 h [16]. Twice, unknown to the patient, the stimulator was switched off and the left side attacks reappeared within 48 h later. When the stimulator was turned on again, the attacks disappeared 48 h later [16]. More than 4 years after the operation, the patient remains pain free [17].

Patient Selection for Hypothalamic DBS

First of all, it should be kept in mind how debilitating CH can be, when it has a chronic course and does not respond to any pharmacological treatments. All patients who received hypothalamic implantation suffered daily attacks in the years before the operation, notwithstanding all kinds of prophylactic drugs, including high dosage steroids. They did not tolerate the painful condition and true to the name ‘suicide headache,’ 2 of the patients had attempted suicide. Another patient had a myocardial infarction while waiting to be operated,


274

probably induced by her very high intake of daily triptans (up to 78 mg injectable sumatriptan/day); thus, she could not take triptans anymore and had to increase daily steroid consumption to control in some way the pain episodes. Three weeks after increased dosage of steroids, she died because of repeated gastrointestinal bleeding. Another patient also had myocardial infarction probably related to triptan overuse and had to increase steroid daily intake; in a few months he had developed, among other side effects, a myopathy in the legs and became unable to climb stairs or to stand from the sitting position. Steroids were tapered and strength in the leg gradually improved, but the CHs worsened. He was asked to restart steroids but the myopathy worsened (confirmed with EMG). Again, steroids were stopped, symptoms of myopathy improved, but CHs worsened. In light of these side effects of standard treatments, it is easy to understand how much chronic CHs may interfere with patients’ life. The hypothalamic DBS operation should be considered only in patients with chronic CH for at least 2 years, with daily attacks [33]. From a clinical point of view, 2 years is sufficient time to apply the full range of available CH prophylactic medications [19, 33] and to exclude that a remission period does not occur spontaneously. At present, DBS should be considered only in patients with strictly unilateral CH (no side shift), since contralateral attacks are well known to develop in CH patients after procedures on the trigeminal nerve [33]. Even though we have implanted 2 patients on both sides because they suffered from intractable bilateral chronic CH, bilateral DBS cannot be recommend at this time until more experience on unilateral DBS has been accumulated [33]. Before considering this operation, it is very important to closely monitor the patient over a period of time to verify the diagnosis and assess the attacks firsthand; hence, the patient has to be admitted in order to witness the attacks. Candidates for hypothalamic DBS must be psychologically stable, with a normal psychological profile. Neuropsychological profiling has to be tested before the operation and periodically afterwards. It is also important to exclude any intracranial abnormality that may contraindicate DBS or which could be a potential underlying cause of CH [33] by performing cranial MRI with gadolinium, craniocervical transition, MRI arterial and venous angiography, and CT of the skull base. The patient must be informed that continuous hypothalamic stimulation might have effects on his/her fertility and sexual behavior, although this aspect has not been well studied. At the present stage, the effects of hypothalamic stimulation on pregnancy are completely unknown and for this reason, we recommend that pregnant patients should not receive DBS. After electrode implantation, the stimulator is turned on only when typical spontaneous CH attack has occurred [33]. A list of primary contraindications to hypothalamic DBS also are listed in table 2. All surgical candidates are informed of the classic surgical procedures that are available for the treatment of intractable chronic CH (open microvascular


275

Table 2. Criteria for electrode implants in intractable chronic cluster headache 1. CCH diagnosed according to IHS criteria; in addition both of the following: a) CCH for at least 24 months b) Attacks should normally occur on a daily basis 2. Attacks must have always been strictly unilateral 3. Patients must be hospitalized to witness attacks and document their characteristics 4. All state of the art drugs for CH prophylaxis must have been tried in sufficient dosages (unless contraindicated or have unacceptable side effects, etc.) alone and in combination, where applicable. These comprise verapamil, lithium carbonate, methysergide, valproate, topiramate, gabapentin, melatonin (where available), pizotifen, indomethacin and steroids 5. Normal psychological profile 6. No medical/neurological conditions contraindicating DBS including: a) Recent myocardial infarction b) Cardiac arrhythmia c) Cardiac malformation d) Epilepsy e) Stroke f) DBS for other reasons g) Degenerative disorder of central nervous system h) Arterial hypertension or hypotension, not controlled by drugs i) Autonomic nervous system disorder j) Endocrinological illnesses k) Major disturbance in electrolyte balance (e.g., due to renal insufficiency or hyperaldosteronism) 7. Normal neurological examination except for symptoms characteristic of CH (e.g., persistent Horner’s syndrome) 8. Normal CT scan (base of the skull window). Normal cerebral MRI including cranio-cervical transition and MRI arterial and venous angiography 9. Neurosurgical team experienced at performing stereotactic implant of electrodes 10. Patient should not be pregnant 11. Ethics Committee/Institutional Review Board approval 12. Patient informed and gives written consent

decompression/lesion of cranial nerves in the cerebellopontine angle and percutaneous radiofrequency trigeminal rhizotomy). They can choose among the various surgical options once detailed informations on the procedures are given. The Surgical Protocol

Ipsilateral Posterior Hypothalamus Electrode Implantation Affixing a stereotactic apparatus (Leksell frame, Elekta, Sweden) is performed under local anesthesia. If sedation is required, low doses of midazolam


276

(0.05–0.1 mg/kg) or propofol (0.5–1 mg/kg) are used [34]. Antibiotic treatment was given to all patients during the perioperative period. A preoperative MRI (brain axial volumetric fast spin echo inversion recovery) is used to obtain highdefinition anatomical images that allow for the precise determination of the AC-PC line. MR images are fused with 2 mm thick CT slices obtained under stereotactic conditions by using an automated technique that is based on a mutual-information algorithm (Frame-link 4.0, Sofamor Danek Stealthstation, Medtronic, Memphis, Tenn., USA). The workstation also provides stereotactic coordinates of the target: 5 mm behind the mid-commissural point, 8 mm below this point and 2 mm lateral from the midline [34]. A rigid cannula is inserted through a precoronaric paramedian burr hole and positioned up to 10 mm from the target. This cannula is used both as a guide for microrecording (Lead Point Medtronic, Minneapolis, Minn., USA) and for the placement of the definitive electrode (DBS-3389, Medtronic, Minneapolis, Minn., USA). Macrostimulation (1–7 V, 60 ␮s, 180 Hz) is used to evaluate potential side effects. All patients subjected to stimulus intensities higher than 4 V have shown ocular deviation that was followed by verbal reports of extreme proportions (‘I feel near to death’; ‘I am at the edge of the end’, etc). When other side effects could be ruled out at standard parameters of stimulation, the guiding cannula was then removed and the electrode was secured to the skull with microplates. The extension was then connected to the electrode, tunneled, and brought out percutaneously for subsequent trial stimulations. On the day following surgery, an additional MR study is done for the purpose of checking the electrode position. After 3–15 days of trial stimulation, the electrodes are then connected to a pulse generator (Itrel II, Medtronic) positioned subcutaneously into the subclavicular area. The following parameters of chronic stimulation have been employed: frequency 180 Hz and a 60-␮s pulse width, with gradually increasing amplitude values [16, 17].

Clinical Results of Hypothalamic DBS

The results of this study are presented in table 3. All the patients have achieved near-complete or complete pain relief, as a result of the long-term high-frequency hypothalamic stimulation that was continued in the follow-up evaluations [35]. Eight out of the 13 implanted patients have remained pain free/almost pain free without any medication (table 3), while 5 needed low doses of methysergide or verapamil to be pain free/almost pain free. It should be noted that these same drugs had been completely ineffective prior to the operative procedure. We observed no noxious side effects from chronic high-frequency hypothalamic stimulation nor did we observe any acute complications from the


277

Table 3. Chronic intractable cluster headache sufferers who received hypothalamic implantation Pt. no and gender

Age Chronic Side CH since

Implant. date

Date of improvement

Amplitude (V)

Drugs

Pt. 1 M a)

39

July 14, 2000

None

Pt. 1 M b)

40

Pt. 2 M

50

Pt. 3 F

1997

Pain free 1 month later

3.0

Right May 31, 2001

Left

Immediately pain free

0.5

1997

Left

November 17, 2000

Pain free 2 months later. Occasional attacks in the last 9 months

1.1

Methysergide 3–4.5 mg/day

63

1994

Left

May 22, 2001

Pain free 2 months later

3.0

Verapamil 80–360 mg/day

Pt. 4 M

52

1997

Right October 11, 2001

Pain free 4 months later. July 2002 electrode replacement

3.1

Methysergide 4.5 mg/day

Pt. 5 M

30

2000

Left

March 22, 2002


1.4

None

Pt. 6 M

46

2000

Left

May 31, 2002


2.8

Verapamil 360 mg/day

Pt. 7 F a)

27

2001

Left

September 12, 2002


2.0

None

Pt. 7 F b)

27

2002

Right January 9, 2003


1.3

Pt. 8 M

25

2001

Right July 10, 2003

2 brief attacks in the last 30 days

2.1

None

Pt. 9 M

43

2000

Left

2 attacks/day, very brief duration and intensity: no sumatriptan need!

2.5

None

Pt. 10 M

46

2001

Right July 30, 2003

Pain free 3 weeks after

1.5

Verapamil 360 mg/day

Pt. 11 M

50

2001

Right August 26, 2003

3 attacks per week

2.0

None

Pt. 12 M

36

2000

Left

September 25, 2003

From 10 attacks/day to 2/day (decreasing frequency and pain intensity)

2.7

None

Pt. 13 M

24

2001

Left

October 15, 2003

From 10–12 attacks/day to 3/day (decreasing frequency and pain intensity)

2.1

None

July 29, 2003


278

implant procedure [35]. Patient 1 presented with some signs of mild hypersexual and hyperphagic behavior prior to the operation, which seemed to be resolved by stimulation [17]. In fact, this patient showed a 25-kg weight reduction at the 18 months follow-up.

Discussion

Less than 20% of CHs have a chronic course and about 20% of the chronic forms become drug resistant. In such cases, a surgical option has to be considered. Until a few years ago, surgical treatment of CH was not done, due to limited knowledge about the pathophysiology of the disease, and was essentially based on the interruption of the autonomic pathways [36–40] (i.e., greater superficial petrosal nerve, intermedius nerve section, sphenopalatine ganglion lesions) and/or on a partial or total trigeminal lesion [29–31, 41–46] (i.e., thermal rhizotomy, glycerolysis, direct nerve sectioning, peripheral avulsions). There appears, however, to be a direct relationship between sensory deficit and subsequent discomfort with facial numbness, keratitis, dysesthesias and sometimes anesthesia dolorosa and success rate using classical surgical approaches. In addition to these troubling side effects, the recurrence rate of CH remains high [29–31, 41–46] and even a complete trigeminal deafferentiation can be followed by the persistence of attacks of CH [18]. Microvascular decompression of the trigeminal nerve could obtain pain relief without lesioning nervous structures but unfortunately, the long-term results of these procedures is unsatisfying [26]. For many years CH was considered and treated as having a peripheral vasogenic origin. However, the striking circadian and circannual rhythmicity of the disease indicated that the hypothalamus was probably involved in the pathogenesis of CH. The recent functional PET [13, 14] and morphological, voxelbased morphometry MRI [15] studies shed light on a new pathophysiological process to explain CH in which the posterior inferior hypothalamic gray matter could be the cluster generator [13]. If a central dysfunction involving hypothalamic circuitry is linked to CH, it seems reasonable to question whether surgical strategies may be used to rebalance the unbalanced or disturbed circuits. According to the current models of basal ganglia circuitry, the akinetic and rigid symptoms of Parkinson’s disease result from the hyperactivity of the globus pallidus internus and substantia nigra pars reticulata, as a consequence of an increased glutamatergic drive from a disinhibited subthalamic nucleus. Although the precise mechanisms of high-frequency DBS remain unknown, the therapeutic effect found after long-term high-frequency DBS in Parkinson’s disease seems to be a result of the inhibitory effect of current delivery to subthalamic nucleus hyperactive neurons [47]. It is possible to suggest that a similar


279

mechanism may explain DBS effectiveness in relieving chronic CH. In fact, the observed increased blood flow at the hypothalamic level observed during CH attacks may originate from an increased neuronal activity at that level. Notwithstanding the hypothesis that hypothalamic DBS rebalances the hyperactive hypothalamus (accounting for the therapeutic effect in our patients), a more generic analgesic effect coming from an activation of pain-modulating pathway, such as the one involving the release of endogenous opiates, cannot be excluded at the present stage. Future work will look at the effects of DBS on the mechanisms of pain. Over 30 years ago, other authors targeted the hypothalamus to relieve painful conditions [48–50]. Surgical procedures on the posteromedial hypothalamus were published by Sano and colleagues [49, 50] in the 1970s to treat otherwise untreatable facial pain and behavioral disorders such as violence and aggression, in the days when so-called psychosurgery was in vogue. Intraoperative high-frequency stimulation of the posteromedial hypothalamus produced analgesic effects, autonomic responses such as hypertension, tachycardia, respiratory suppression, hyperpnea, tachypnea and mydriasis as well as somatomotor responses. No such effects were observed in our series of CH patients, probably because of differences in both the targeting and the stimulation parameters that have been used [16, 17, 34, 35]. Now that the physiological basis of disorders such as CH and other pain syndromes are able to be precisely measured with new techniques such as PET, fMRI, and intraoperative microdialysis monitoring, it is hoped that functional and restorative neurosurgery will re-emerge from the discredited shadows of early 1960s psychosurgeries, much in the same manner, that basal ganglia surgery has enjoyed a renaissance, since a decadeslong hiatus through the 1960s and 1970s until the pioneering work of Laitinen and colleagues in the 1980s [51].

Conclusions and Future Directions in Hypothalamic DBS

In this chapter, we report the first large series of successfully treated chronic CH sufferers using long-term high-frequency hypothalamic stimulation. These results provide clear evidence that the hypothalamic stimulation offers a safe and effective treatment for CH without any of the troublesome side effects associated with peripheral nerve lesioning procedures. The rationale underlying hypothalamic DBS in CH is based on more advanced morpho-functional studies pointing to the hypothalamus as the ‘CH generator.’ It is hypothesized that the prolonged hypothalamic stimulation rebalances the genetic and cellular mechanisms that leading to hyperfunctioning hypothalamic neurons. It should be underscored that this is the first direct therapeutic application of neuroimaging


280

functional data in a primary headache syndrome, and such that the surgical approach used is reversible (by turning off or altering current) in the event of serious complications. Other conditions in which hypothalamic DBS may provide a useful experimental model in the future include complex neuro-endocrine and pain disorders.

References 1 2 3

4 5 6 7 8 9 10

11 12

13 14 15

16 17 18 19 20

Vonck K, Boon P, Achten E, De Reuck J, Caemaert J: Long-term amygdalohippocampal stimulation for refractory temporal lobe epilepsy. Ann Neurol 2002;52:556–565. Hodaie M, Wennberg RA, Dostrovsky JO, Lozano AM: Chronic anterior thalamus stimulation for intractable epilepsy. Epilepsia 2002;43:603–608. Benabid AL, Koudsie A, Benazzouz A, Vercueil L, Fraix V, Chabardes S, Lebas JF, Pollak P: Deep brain stimulation of the corpus luysi (subthalamic nucleus) and other targets in Parkinson’s disease. Extension to new indications such as dystonia and epilepsy. J Neurol 2001;248(suppl 3):III37–47. Headache Classification Committee of The International Headache Society. The International Classification of Headache Disorders, ed 2. Cephalalgia 2004;24:1–195. D’Alessandro R, Gamberini G, Benassi G, Morganti G, Cortelli P, Lugaresi EI: Cluster headache in the Republic of San Marino. Cephalalgia 1986;6:159–162. Manzoni GC, Terzano MG, Bono G, Micieli G, Martucci N, Nappi G: Cluster headache – Clinical findings in 180 patients. Cephalalgia 1983;3:21–30. Kudrow L: The cyclic relationship of natural illumination to cluster period frequency. Cephalalgia 1987;7:76–78. Goadsby PJ, Edvinsson L: Human in vivo evidence for trigeminovascular activation in cluster headache: Neuropeptide changes and effects of acute attacks therapies. Brain 1994;117:427–434. Kudrow L: Cluster Headache, Mechanism and Management, ed 1. New York, Oxford University Press, 1980. Waldenlind E, Gustafsson SA, Ekbom K, Wetterberg L: Circadian secretion of cortisol and melatonin in cluster headache during active cluster periods and remission. J Neurol Neurosurg Psychiatry 1987;50:207–213. Leone M, Bussone G: A review of hormonal findings in cluster headache. Evidence for hypothalamic involvement. Cephalalgia 1993;13:309–317. Leone M, D’Amico D, Moschiano F, Fraschini F, Bussone G: Melatonin vs placebo in the prophylaxis of cluster headache: A double-blind pilot study with parallel group. Cephalalgia 1996; 16:494–496. May A, Bahra A, Büchel C, Frackowiak RS, Goadsby PJ: Hypothalamic activation in cluster headache attacks. Lancet 1998;352:275–278. Sprenger T, Boeker H, Toelle TR, Bussone G, May A, Leone M: Specific hypothalamic activation during a spontaneous cluster headache attack. Neurology 2003;in press. May A, Ashburner J, Buchel C, McGonigle DJ, Friston KJ, Frackowiak RS, Goadsby PJ: Correlation between structural and functional changes in brain in an idiopathic headache syndrome. Nat Med 1999;5:836–838. Leone M, Franzini A, Bussone G: Stereotactic stimulation of posterior hypothalamic gray matter in a patient with intractable cluster headache. N Engl J Med 2001;345:1428–1429. Leone M, Franzini A, Broggi G, May A, Bussone G: Long-term follow up of bilateral hypothalamic stimulation for intractable cluster headache. Brain 2004;in press. Matharu MS, Goadsby PJ: Persistence of attacks of cluster headache after trigeminal nerve root section. Brain 2002;125:976–984. Dodick D, Rozen T, Goadsby P, Silberstein S: Cluster headache. Cephalalgia 2000;20:787–803. Dodick DW, Capobianco DJ: Treatment and management of cluster headache. Curr Pain Headache Rep 2001;5:83–91.


281

21

22

23

24 25 26 27 28 29 30 31 32 33

34 35

36 37 38 39 40 41 42 43 44 45

46

Ekbom K, Monstad I, Prusinski A, Cole JA, Pilgrim AJ, Noronha D: Subcutaneous sumatriptan in the acute treatment of cluster headache: A dose comparison study. The Sumatriptan Cluster Headache Study Group. Acta Neurol Scand 1993;88:63–69. Leone M, D’Amico D, Frediani F, Moschiano F, Grazzi L, Attanasio A, Bussone G: Verapamil in the prophylaxis of episodic cluster headache: A double-blind study vs. placebo. Neurology 2000; 54:1382–1385. Bussone G, Leone M, Peccarisi C, Micieli G, Granella F, Magri M, Manzoni GC, Nappi G: Double blind comparison of lithium and verapamil in cluster headache prophylaxis. Headache 1990;30:411–417. Curran DA, Hinterburger H, Lance JW: Methysergide. Res Clin Stud Headache 1967;1:74–122. May A, Leone M: Up to date on cluster headache. Curr Opin Neurol 2003;16:333–340. Lovely TJ, Kotsiakis X, Jannetta PJ: The surgical management of chronic cluster headache. Headache 1998;38:590–594. Mathew NT, Hurt W: Percutaneous radiofrequency trigeminal gangliorhizolysis in intractable cluster headache. Headache 1988;28:328–331. Jarrar RG, Black DF, Dodick DW, Davis DH: Outcome of trigeminal nerve section in the treatment of chronic cluster headache. Neurology 2003;60:1360–1362. Taha Jm, Tew JM Jr: Long-term results of radiofrequency rhizotomy in the treatment of cluster headache. Headache 1995;35:193–196. Pieper DR, Dickerson J, Hassenbusch SJ: Percutaneous retrogasserian glycerol rhizolysis for treatment of chronic intractable cluster headaches: Long-term results. Neurosurgery 2000;46:363–368. Watson CP, Morley TP, Richardson JC, Schutz H, Tasker RR: The surgical treatment of chronic cluster headache. Headache 1983;23:289–295. May A, Bahra A, Büchel C, Frackowiak RS, Goadsby PJ: PET and MRA findings in cluster headache and MRA in experimental pain. Neurology 2000;55:1328–1335. Leone M, May A, Franzini A, Broggi G, Dodick D, Rapoport A, Goadsby PJ, Schoenen J, Bonavita V, Bussone G: Deep brain stimulation for intractable chronic cluster headache: Proposals for patient selection. Cephalalgia 2004;in press. Franzini A, Ferroli P, Leone M, Broggi G: Stimulation of the posterior hypothalamus for treatment of chronic intractable cluster headaches: First reported series. Neurosurgery 2003;52:1095–1099. Leone M, Franzini A, D’Amico D, Grazzi L, Rigamonti A, Mea E, Broggi G, Bussone G: Long-term follow-up of hypothalamic stimulation to relieve intractable chronic cluster headache. Neurology 2004;62(suppl 5):355. Gardner WJ, Stowell A, Dutlinger R: Resection of the greater superficial petrosal nerve in the treatment of unilateral headache. J Neurosurg 1947;4:105–114. Sachs E Jr: Further observations on surgery of the nervus intermedius. Headache 1969;9:159–161. Sachs E Jr: The Role of nervus intermedius in facial neuralgia: Report of four cases with observations on the pathways for taste, lacrimation and pain in the face. J Neurosurg 1968;28:54–60. Stowell A: Physiologic mechanisms and treatment of histaminic or petrosal neuralgia. Headache 1970;9:187–194. Sweet WH: Surgical treatment of chronic cluster headache. Headache 1988;28:669–670. White JC, Sweet WH: Periodic migrainous neuralgia; in Pain and the Neurosurgeon: A Forty-Year Experience. Springfield, Charles C Thomas, 1969, pp 345–434. Wilkins RH, Morgenlander JC: Results of surgical treatment of cluster headache: Initial relief followed by recurrence. Neurosurgery 1991;31:91–106. Maxwell RE: Surgical control of chronic migrainous neuralgia by trigeminal gangliorhizolysis. J Neurosurg 1982;57:459–466. Morgenlander JC, Wilkins RH: Surgical treatment of cluster headache. J Neurosurg 1990;72: 866–871. North RB, Kidd DH, Piantadosi S, Carson BS: Percutaneous retrogasserian glycerol rhizotomy: Predictors of success and failure in treatment of trigeminal neuralgia. J Neurosurg 1990;72: 851–856. O’Brien MD, MacCabe JJ: Trigeminal nerve section for unremitting migrainous neuralgia; in Rose FC, Zilkha KJ (eds): Progress in Migraine Research. London, Pitman, 1981, vol 1, pp 185–187.


282

47

48

49 50 51

Benazzouz A, Gross C, Feger J, Boraud T, Bioulac B: Reversal of rigidity and improvement in motor performance by subthalamic high-frequency stimulation in MPTP-treated monkeys. Eur J Neurosci 1993;5:382–389. Fairman D: Hypothalamotomy as a new perspective for alleviation of intractable pain and regression of metastatic malignant tumors; in Fusek I, Kunk Z (eds): Present Limits of Neurosurgery. Prague, Avicenum, 1972, pp 525–528. Sano K: Sedative neurosurgery with special reference to posteromedial hypothalamus. Neurol Med Chir (Tokyo) 1962;4:112–142. Sano K, Mayanagi Y, Sekino H, Ogashiwa M, Ishijima B: Results of stimulation and destruction of the posterior hypothalamus in man. J Neurosurg 1970;33:689–707. Laitinen LV: Brain targets in surgery for Parkinson’s disease. Results of a survey of neurosurgeons. J Neurosurg 1985;62:349–351.

Massimo Leone, MD Istituto Nazionale Neurologico Carlo Besta Via Celoria 11, IT–20133 Milano (Italy) E-Mail [email protected]


283


Molecular Mediators of Pain Priya Chaudhary, Kim Burchiel Department of Neurological Surgery, L472, Oregon Health & Science University, Portland, Oreg., USA

Introduction: What Is Pain?

Pain is an universal, subjective and unpleasant sensation. Acute and subchronic pain has a protective role as it warns of tissue damage. However, chronic and severe pain offers no survival advantage and causes suffering. Particularly in chronic pain, sensory processing from an affected region becomes abnormal and innocuous stimuli (e.g., thermal, touch/pressure) that would normally not cause pain may do so (i.e., allodynia) or noxious stimuli may elicit exaggerated perceptions of pain (i.e., hyperalgesia). In addition, sensations similar to electric tingling or shocks (i.e., paresthesias) and/or sensations having unpleasant qualities (i.e., dysesthesias) may be elicited by normal stimuli. Pain is initiated by specialized sensory nociceptors in the peripheral tissues in response to noxious stimuli [1, 2]. The dorsal root ganglion (DRG) neurons provide a site of communication between the periphery and the spinal cord. Peripheral nociceptors have the machinery for encoding noxious stimuli into action potentials. Central terminals mediate synaptic transmission as well as presynaptic modulation. At the spinal cord level, pain impulses undergo substantial modulation by local mechanisms and by projections from the supraspinal structures (inhibition and facilitation). The processed signal is transmitted to the brainstem and thalamic sites and finally to the cerebral cortex, where it elicits the sensation of pain [3, 4]. Nociceptors are sensory nerve endings which respond to stimuli, which threaten or are capable of causing tissue damage. In addition to activating centripetal discharge, the nociceptive stimuli cause primary afferent (sensory) fibers to release endogenous chemicals. Cutaneous primary afferent nociceptor fibers can be classified into three types, C, A␤ and A␦ based on their soma diameter, structure and conduction velocity of their axon. Multiple classes of

Pathological condition Nerve injury Chemical and inflammatory mediators

Transcription DNA Pre-mRNA

Splicing Alternative splicing

Translation mRNA Stabilization destabilization

Functional protein (pain mediators)

Protein

Post-translational modifications like glycosylation phosphorylation dephosphorylation

Fig. 1. Schematic demonstration of gene expression steps subjected to possible regulation during pain [redrawn from 177].

C and A␦ exist with differing sensory properties. There are two main categories of A␦ and C nociceptors: A␦ mechanical nociceptors and C-polymodal nociceptors. A␦ mechanical nociceptors are activated by mechanical stimuli that damage the tissue. C-polymodal nociceptors are capable of responding to mechanical, thermal and chemical stimuli. The other nociceptor types include A␦ mechanoheat nociceptors, A␦ and C cold nociceptors and C mechanical nociceptors. Acute pain corresponds to the activation of nociceptors with little intervention from higher modulatory mechanisms. However, injury by physical, chemical or immunological means also causes long-term alterations in the expression levels of excitatory mediators, neuropeptides, neurotransmitters, inhibitory neuromodulators, neurotrophic factors, peripheral terminal receptor functions, and signal transduction molecules (fig. 1). These substances exert a variety of actions on local tissue, vasculature, and the afferent fibers. Acute nociceptive, inflammatory and neuropathic pain to some degree, all depends on the activation of primary afferent neurons in the DRG and trigeminal ganglion. A variety of mediators are involved in the central transmission. These substances, which are capable of altering the properties of nociceptors, are broadly termed as modulators (fig. 2). In this chapter, we will describe various mediators like amines (e.g., histamine, serotonin), kinins (e.g., bradykinin; BK), prostanoids (e.g., prostaglandins; PGs), cytokines (e.g., interleukins, tumor necrosis factor; TNF), neuropeptides [e.g., substance P (SP), and calcitonin gene related peptide (CGRP)], energy sources [e.g., adenosine triphospate (ATP)],

Molecular Mediators of Pain

285

Inhibitory influences: Opioids (␮, ␦, ␬) ␣2-adrenoceptor (␣2c) Adenosine (A1) Cannabinoids (CB1, CB2)

GABA (GABA B) Orphanin (ORL1) Somatostatin Immune cells Mast cell

Platelets

Sensory neuron Tissue injury inflammation Dorsal root ganglion

Spinal cord

Blood

Sensory nerve vessel ending

Sympathetic nerves

Excitatory influences: Prostanoids (EP, IP) Bradykinin (B1, B2) Histamine (H1) Serotonin (5-HT1, 5-HT2, 5-HT3, 5-HT4) ATP (P2X3) TRPV1 Sodium channels

Adrenoceptor (␣2A) Glutamate (NMDA, AMPA, KA) Acetylcholine (N) Adenosine (A2a, A3) Tachykinins (NK1, NK2) Nerve growth factor (TrkA)

Fig. 2. Excitatory and inhibitory influences on peripheral nerve activity by mediators released by tissue injury and inflammation and by a variety of agents acting on neuroreceptors [178].

diffusible gas molecules (e.g., nitric oxide; NO), ions (e.g., H⫹, Na⫹, K⫹), and neurotrophins [NT; e.g., nerve growth factor (NGF)]. Modulators such as opioids, opioid receptors, cannabinoid receptors, and somatostatin are also described (fig. 3, table 1). Effective treatments for pain can be developed by understanding the cellular mechanisms, molecular mechanisms, and mediators that produce pain. This chapter highlights the importance of pain mediators and modulators in developing novel approaches for the treatment of pain. Since gene therapy promises to provide new, effective and innovative solutions for the treatment of pain, we describe some aspects of how gene therapy can be used in pain treatment. We have organized mediators and modulators into categories such as neuroreceptors, ion channels, excitatory receptors, inhibitory receptors, immune mediators, peptides, inflammatory mediators, growth factors and signal transduction molecules, while recognizing that they work in concert to create the sensation of pain.

Chaudhary/Burchiel

286

PG

ATP Histamine 5-HT

EP3

P2Y2

H1 5-HT4/7 CGRP½

BK B2 (B1)

CGRP

H2

Histamine

SP NK1 A2A

Adenosine

PLC

Dorsal horn

[Ca2⫹]i

AC

EP2/4

gK⫹

␣2 NK1

gCa2⫹ EAA

gNa⫹

PG

NAD SP

5-HT2A

NMDA

5-HT3

AMPA/Kainate

5-HT

P2X2/ ASIC Capsaicin VDCC P2X3 ATP Protons Vanilloids

Fig. 3. Roles of diverse receptors and intracellular signals in mediating pain at the polymodal C fiber terminal [4].

Table 1. Receptors localized on primary afferent fibers and their ligands from neuronal and non-neuronal origins [181] Receptors associated with nociceptors ATP, neurokinin-1, GABA, neuropeptide Y, acetylcholine, somatostatin, prostaglandin E, cholecystokinin, adrenergic, 5-HT, glutamate, bradyinin, noradrenaline, capsaicin, opioid, angiotensin II, adenosine (A1 and A2), cannabinoid and menthol receptors Ligands with non-neuronal sources Acetylcholine, ATP, prostaglandin E, opioids, adenosine, glutamate, bradykinin, noradrenaline, serotonin Ligands in nociceptors Substance P, opioid, ATP, adenosine, neuropeptide Y, glutamate, cholecystokinin, somatostatin


287

Neuronal Pain Receptors

Vanilloid (Capsaicin) Receptors: VR1/TRPV1 TRPV1 belongs to the transient receptor potential (TRP) family of ion channels [5–7]. It is a nonspecific cation channel expressed preferentially in small sensory neurons. TRPV1-expressing neurons are divided into two groups: peptidergic, expressing SP and CGRP; and nonpeptidergic, expressing P2X3 purinoceptor. TRPV1 detects painful stimuli activated by heat and tissue acidosis, or H⫹ ions [8]. Endogenous substances released from activated immune cells during inflammation also activate TRPV1 and lead to CGRP release. BK, PGE2, and NGF are compounds released during inflammation and are also known to modulate the activity of TRPV1 [9, 10]. Certain cannabinoids, which are products of lipoxygenase pathways of arachidonic acid metabolism (e.g., 12- or 15-hydroxyperoxy-eicosatetraenoic acid) and N-arachidonyl dopamine are endogenous ligands, which activate TRPV1. TRPV1 and its ligands have an important role in mediating inflammatory pain. TRPV1 knockout mice have demonstrated that TRPV1 is important for inflammation-induced hyperalgesia [11, 12]. The natural vanilloid capsaicin, an ingredient of hot pepper, activates TRPV1 and has been most commonly used to study the properties of TRPV1 receptors [5]. Most mechanistic studies of capsaicin-induced activation of nociceptive neurons have been made using cultured sensory neurons and isolated nerves in vitro. Capsaicin causes depolarization, during which there is an increase in membrane permeability to cations (Ca2⫹, Na⫹). Vanilloids have a biphasic response consisting of an initial excitatory response and a refractory phase or desensitization [13]. Capsazepine, a competitive antagonist, and ruthenium red, a noncompetitive antagonist, both block TRPV1. ATP, protein kinase C (PKC), and protein kinase A (PKA) can also modulate the properties of the TRPV1 channels. ATP acts as an allosteric factor and enhances the effect of capsaicin on rat TRPV1 channels [14]. PKC can lead to the activation of capsaicin receptor even in the absence of ligands such as H⫹ or heat [15]. PKA sensitizes the receptor to vanilloids and anandamide [16]. Analyses of mutations in TRPV1 have been the key to understanding the function of this receptor. Site-directed mutagenesis experiments identified a glutamic acid residue (Glu600) near the putative pore which is thought to serve as a key regulatory site, setting the sensitivity to noxious stimuli in response to changes in extracellular proton concentration [17]. Jordt et al. [17] also showed that protons, vanilloids, and heat promote channel opening through separate pathways, since mutations at E648 selectively abolish protonevoked channel activation without diminishing responses to other noxious stimuli.

Chaudhary/Burchiel

288

The example of anandamide, a cannabinoid that plays a dual role in the body, serves to indicate the complexity of the vanilloid system. It can act as a pro- or anti-inflammatory ligand depending on whether it activates TRPV1 or cannabinoid receptor B1 (CB1). Anandamide acts as a full agonist of human TRPV1 [18] and inhibits capsaicin-induced CGRP release in skin sensory afferents and from the dorsal horn, possibly through the activation of CB1 [19, 20]. Understanding the functioning of the receptor (open and closed states) and signal transduction pathways by which TRPV1 is activated could lead to the identification of novel pain-relieving targets. Alternatively, down-regulation of TRPV1 expression or factors promoting the closed state or desensitization of the channel represents a promising therapeutic strategy for novel analgesic drugs (table 1). Cannabinoid Receptors Cannabinoid and vanilloid receptors are colocalized in the primary sensory neurons and the dorsal horn of the spinal cord [7]. CB1, CB2 are G-protein coupled receptors (GPCRs). The CB1 receptors are expressed in areas involved in modulation of nociception such as periaqueductal grey, spinal cord dorsal horn, and the DRG. CB2 receptors are expressed in nonneuronal cells such as mast cells and other immune cells. Behavioral tests indicate that cannabinoids have anti-nociceptive effects in animal models of acute pain and in persistent pain following peripheral inflammation [21] or nerve injury [22]. It is now undisputable that cannabinoid receptor modulation has therapeutic value in anti-nociception, although concomitant modulatory activity of dopaminergic systems may have adverse psychotropic effects. Anandamide (an endo-cannabinoid) is formed from the hydrolysis of a phospholipid precursor catalyzed by a phospholipase D and is inactivated via reuptake by anandamide membrane transporter (AMT) and enzymatic hydrolysis by fatty acid amide hydrolase enzyme. Anandamide activates the CB1 receptor in the brain and also acts as a full agonist of TRPV1 [18, 23]. Since anandamide and capsaicin share the same TRPV1-binding site, compounds which influence the activity of AMT may facilitate the action of anandamide at the TRPV1. Activation of AMT thus enhances the activity of anandamide at the TRPV1, and AMT inhibitors block the anandamide activity. Since the anandamide-binding site on CB1 is extracellular, AMT could play an important role in distributing anandamide between the intra- and extracellular compartments and activating TRPV1 or CB1 [24]. Cold Receptors Mammals detect temperature effects with specialized neurons in the peripheral nervous system (PNS). Cold and menthol-sensitive receptor (TRPM8)


289

belongs to the TRP family of excitatory ion channels and functions as a transducer of cold stimuli [25]. About 10% of trigeminal ganglion neurons express this cold receptor, and thus it is still possible that cold excites the sensory neurons by activating the cold and menthol-sensitive receptor as well as by modulating other excitatory and inhibitory channels present on these neurons. TRPM8 was the first molecule to be identified that responds to cold temperatures and stimulation with menthol. ANKTM1 has recently been characterized as a cold-activated channel. This channel has a lower activation temperature compared to the cold and menthol receptor, TRPM8. ANKTM1 shares little amino acid similarity with TRPM8. ANKTM1 is found in a subset of nociceptive sensory neurons where it is coexpressed with TRPV1 (the capsaicin/heat receptor) but not TRPM8 [26]. Understanding the role of TRPM8 and ANKTM1 may help to identify a target for therapeutic applications using cold receptors. Cold treatment is already used as a method of relief from pain, due in part to its effect on inflammation. In some cases, hypersensitivity to cold can lead to cold allodynia in patients suffering from neuropathic pain, and this also constitutes a therapeutic rationale. Proteinase-Activated Receptors Proteinases like thrombin and trypsin not only act as degradative enzymes, but also act as signaling molecules that regulate proteinase-activated receptors (PAR) [27, 28]. Proteinases are released during inflammatory processes. Proteolytic cleavage of the extracellular amino terminus of PAR exposes a tethered ligand domain, which acts as a receptor-activating ligand. Synthetic peptides corresponding to this proteolytically revealed new N-terminal domain (PARactivating peptides) constitute selective agonists for these receptors. The PAR receptor family is known to have four members PAR1, PAR2, PAR3 and PAR4 which are all G-protein coupled [29]. PARs are expressed on endothelium, platelets, inflammatory cells, fibroblasts and nociceptive primary afferents [30]. Several studies suggest that these receptors might be mediators of neurogenic inflammation and may cause nociception. PAR agonists produce thermal and mechanical hyperalgesia, which is diminished in mice lacking the NK-1 receptor [27, 28]. In the DRG, more than half of the neurons expressing PAR2 also coexpress CGRP and SP, which play a role in vasodilatation and inflammatory responses. Thus, proteases and PARs may play a previously unknown novel role in pain. This may have a potential for developing therapeutic targets in inflammation and pain. Adrenoceptors Adrenoceptors (ARs) mediate some of the main actions of the natural catecholamines, epinephrine and norepinephrine. ARs include ␣1, ␣2, ␤1, ␤2 and ␤3. ARs are members of the much larger family of GPCRs, which include

Chaudhary/Burchiel

290

muscarinic cholinergic receptors, serotonin receptors, dopamine receptors, neurokinin receptors as well as the photoreceptor rhodopsin. ARs are distributed at sites, which are associated with the transfer of sensory, nociceptive information. For example ␣2 AR subtypes have been found in the DRG, superficial laminae of the spinal cord, and the thalamic nuclei. Different subtypes of ␣2 ARs play a role either in anti-nociception or have been implicated in causing hyperalgesia. Clonidine is an ␣2-adrenergic agonist that has been used as analgesic agent to control severe, acute and chronic pain conditions following epidural or spinal administration. Clonidine relieves hyperalgesia in patients with sympathetically maintained pain but has no effect on sympathetically independent pain. Clonidine not only produces significant analgesia on its own but also potentiates the analgesia produced by opiates [31]. More research is needed to understand the roles played by individual AR subtypes. Cholinergic Receptors Acetylcholine activates cholinergic receptors, both muscarinic acetylcholine receptors and nicotinic acetylcholine receptors (nAChR). At least five primary mACh receptor subtypes are known (M1–M5). They are GPCRs. M1, M3, and M5 mediate their effects through increases in intracellular calcium, whereas M2 and M4 mediate their effects through decreases in cAMP production. Nicotinic receptors on the other hand are ligand-gated channels and at least 11 nAChR subunits ␣2–␣9 and ␤2–␤4 have been identified. The activation of neuronal nAChR produces significant increases in intracellular Ca2⫹ and may play a role in cellular signaling. These receptors are known to produce spinal and supraspinal analgesia. The central anti-nociceptive effects of nicotine, a neuronal nAChR agonist, have been known for many years. Epibatidine is the most potent known agonist at several nicotinic receptor subtypes and mediates anti-nociceptive effects. However, it is too toxic for use in humans [32]. Only recently a potent nAChR agonist, ABT-594 was shown to have antinociceptive properties equal in efficacy to those of morphine [33].

Ion-Gated Channels

ATP-Gated Ion Channels Micromolar concentrations of extracellular ATP (nucleotide) activate sensory neurons via ATP-gated ion channels, cell surface receptors known as P2 receptors [34]. P2 receptors are classified into two categories: the P2X family consisting of ligand-gated cation channels and the P2Y family made up of the GPCRs [35]. Seven subtypes of P2X and eight subtypes of P2Y family have been identified [36]. P2X receptors are found as homomultimeric or heteromultimeric channels.


291

P2X3 is expressed selectively by nociceptors in the DRG, predominantly on the nonpeptidergic neurons [37]. DRG neurons respond to ATP by an increase in the free intracellular calcium or by depolarization. P2X3 immunoreactive terminals have also been detected in the lamina II of the dorsal horn. P2Y receptors are expressed on large DRGs and produce action potentials by light touch. These observations suggest that P2 receptors play a role in signal transduction of pain from the periphery to the spinal cord. P2X3 receptors are down-regulated following peripheral nerve injury (e.g., sciatic nerve cut) and their expression can be regulated by glial cell-derived neurotrophic factor (GDNF) [37]. In contrast to the sciatic nerve cut example, the P2X3 receptor is up-regulated in the trigeminal ganglion after nerve injury [38]. In human embryonic kidney cells expressing the P2X2 homomer or P2X2/P2X3 heteromer, acidification (pH ⬍ 6.3) increased the ATP-induced current [39]. Since inflammation causes a decrease in the tissue pH, these ATP receptors may play an important role in inflammatory pain. Suramin and PPADS are nonspecific blockers of the P2X1, P2X2, P2X3, P2X5 receptors at micromolar concentrations and other P2X receptors at higher concentrations [40]. The more specific inhibitor Trinitrophenyl-ATP, selectively inhibits P2X1, P2X3 receptors and heteromeric channels that contain one of these receptors subunits [41]. Thus, the P2X antagonist at the sensory terminal may help in reducing pain caused due to inflammation. Adenosine Receptors Adenosine and ATP influence pain transmission at peripheral and spinal sites. Four adenosine receptor types have been cloned: the A1, A2a, A2b and A3 receptors. The A2a receptor is found in the large neuronal cells of the rat DRG. At the peripheral nerve terminals adenosine A1 receptor activation causes antinociception [42] and adenosine A2 receptor activation produces pronociception [42]. Adenosine A3 receptor activation produces pain behaviors due to release of histamine and 5-HT from mast cells and subsequent actions on the sensory nerve terminal [43]. Acid-Sensing Ion Channels Acid-sensing (proton-gated) ion channels (ASICs) use protonation for the activation of ionic current suggesting the importance of pH regulation in the normal functioning of the nervous system. Severe tissue acidosis that accompanies inflammation is painful and sensory neurons respond to acidic tissue pH with increased firing. Proton-gated cation channels in sensory nerve endings are thought to be responsible for the activation of nociceptive afferents by acid. Members of the ASIC family include ASIC1a, splice variant ASIC1b (BNC2), ASIC2a (MDEG1, BNC1), ASIC2b (MDEG2), ASIC3 (DRASIC- dorsal root

Chaudhary/Burchiel

292

ASIC), and ASIC4 [44]. These ion channels can form homo- or heteromultimers with other ASICs and are expressed widely in the nervous system. These channels are sensitive to amiloride at high concentrations and are selective for Na⫹. No specific blockers for these individual channels have been suggested. Only recently, psalmotoxin 1 was used to inhibit currents through ASIC1a [45]. The discovery of blockers for these channels is important to evaluate the role played by these channels in nociception. Potassium Channels K⫹ channels form the largest family of ion channels. The common feature of all K⫹ channels is the presence of a conserved motif called the P domain. The 2P domain, leak/background K⫹ channels are non-voltage-gated channels. These background channels are widely distributed in the nervous system. The 2P K⫹ channels play an essential role in setting the neuronal membrane potential and in tuning the action potential duration. They are represented by TWIK-1, TWIK-2 (weak inward rectifiers), TREK-1, TREK-2 (Twik-related K⫹ channel), TRAAK (TWIK-related arachidonic acid-stimulated K⫹; lipid-sensitive mechano-gated K⫹ channels) and TASK-1, TASK-2, TASK-3 (TWIK-related acid sensing K⫹ channel; acid-sensitive outward rectifiers), [46, 47]. The TREK-1, TREK-2 and TRAAK channel activity is elicited by increasing mechanical pressure. These channels are also reversibly opened by polyunsaturated fatty acids including arachidonic acid. TREK-1 is opened by intracellular acidosis, membrane stretch, cell swelling, arachidonic acid and heat [48, 49]. PGE2 and cAMP can close the channel by a PKA-mediated phosphorylation of Ser333. Since TREK-1 is present in sensory neurons as well in the hypothalamus, it is a good candidate as a temperature sensor [50]. TASK-1, TASK-2 and TASK-3 are sensitive to variations of extracellular pH in the physiological range. TASK-3 operates in the pathophysiological range of pH, closes at pH 6.0 and cytosolic arachidonic acid (10 ␮M), which suggests that it may play a role in inflammation [51]. The recent demonstration that TASK-1, TREK-1 and TREK-2 channels are activated by inhalational general anesthetics, and that TRAAK is activated by the neuroprotective agent riluzole, indicates that this novel class of K⫹ channels are interesting targets for new therapeutic developments [52]. Sodium Channels Na⫹ channels are important in electrogenesis within primary sensory neurons. These channels are involved in multiple functions like transduction, signal amplification, and genesis of action potentials [53]. There is growing evidence that modulation of these currents/channels is an endogenous mechanism used to control neuronal excitability [54]. Voltage-gated Na⫹ channels, which produce the inward membrane current necessary for regenerative action potential production


293

have emerged as targets in the study of pathophysiology of pain and in the search for new pain therapies. Nine voltage gated Na⫹ channel ␣ subunits (Nav 1.1–1.9) and three subunits (␤1–3) have been cloned [55]. Based on the sensitivity to tetrodotoxin (TTX), these currents are divided into two types, TTX-sensitive and -resistant channels. The TTX-sensitive channel is present in all sensory neurons. TTX-resistant SNS/PN3/Nav1.8 and NaN/SNS2/Nav1.9 channels have been detected in small diameter, unmyelinated sensory afferent neurons [2]. The ratio of these two types of Na⫹ channels can have a profound effect on excitability. The slowly inactivating, rapidly repriming SNS/PN3/Nav1.8 channel is the most likely candidate for the repetitive firing of the injured peripheral nerve [56]. Several studies have demonstrated that TTX-resistant Na⫹ channels can be modulated by inflammatory molecules such as PGs and serotonin through the cAMP-PKA cascade. Down-regulation of TTX-resistant Na⫹ channels (Nav1.8 and Nav1.9) and up-regulation of TTX-sensitive Type III Na⫹ channels (Nav1.3) has been detected after nerve injury [57–60]. Local anesthetics such as lidocaine and mexiletine or anticonvulsants such as carbamazepine and phenytoin have been used in the treatment of neuropathic pain, although clinical performance has been hindered by a number of side effects. There is interest in delineating mechanisms underlying membrane excitability, action potential generation and transmission in nociceptive neurons [54, 58]. Targeting sodium channels (especially the Nav1.8 and Nav1.9) in the periphery could be a novel opportunity for producing analgesia without having major side effects in the central nervous system (CNS).

Excitatory Receptors

Glutamate Receptors Glutamate receptors play a key role in pain perception. Glutamate acts through ionotropic glutamate receptors (iGluRs, coupled to ion channels) and metabotropic glutamate receptors (mGluRs, coupled to intracellular secondary messengers). Animal studies indicate that glutamate in the periphery plays an important role in response to inflammatory agents such as intraplantar formalin [61] and causes pain-related behaviors [62]. The nociceptive-specific primary afferent fibers are a source of peripheral glutamate [63]. Ionotropic glutamate receptors include those activated by ␣-amino-3hydroxy-5-methyl-4-isoxzolepropionic acid, N-methyl-D-aspartate (NMDA) and kainate. The NMDA receptor (NMDA-R) is distinctive and unique. It acts as both ligand and voltage-gated, and is selectively permeable to Ca2⫹ ions. As a consequence, NMDA-R mediated alterations in intracellular Ca2⫹ levels regulate a variety of signaling pathways, ranging from localized, acute effects on receptor and

Chaudhary/Burchiel

294

channels activities to long-term effects on nuclear gene transcription. The involvement of peripheral NMDA-R in inflammatory nociception offers an attractive target for antagonists that do not cross the blood brain barrier. Such agents might be powerful anti-nociceptive agents without having the CNS side effects [64]. There are eight cloned mGluRs (mGluR1–8), divided into three groups based on sequence similarity, pharmacology and intracellular effector systems. Group I consists of mGluR1 and mGluR5, group II is made up of mGluR2 and mGluR3, and group III has mGluR4, mGluR6, mGluR7, mGluR8 [65]. The mGluRs are activated upon the release of glutamate in the dorsal horn subsequent to the activation of sensory neurons. mGluRs are also activated in the peripheral primary afferent terminals of sensory neurons in response to inflammatory stimulus and in experimental neuropathic pain elicited by ligation of L5/L6 spinal nerves [66, 67]. Group I mGluRs act through the activation of phospholipase C, which leads to the release of calcium from intracellular stores and activation of PKC. mGluRs also activate other kinases that can modulate the function of vanilloid receptors, TTX-resistant channels that have been implicated in the production of pain. The mGluRs contribute to nociceptive processes such as hyperalgesia since receptor antagonists attenuate pain. mGluR5 antagonists [SIB-1757, 2-methyl6-(phenylethynyl)-pyridine (MPEP)] and mGluR1 receptor antagonists [7-(hydroxyimino)cyclo-propa[b]chromen-1a-carboxylate ethyl ester (CPCCOEt), 2-methyl-4-carboxyphenylglycine (LY367385)], cause reversal of pain symptoms [66–68]. The reduction of hyperalgesia by mGluR antagonists is important in designing drugs that could target the painful neuropathies or inflammatory pain conditions. Elucidation of the underlying molecular mechanisms by which the glutamate receptors enhance pain sensitivity, may lead to designing inhibitors of glutamate release, selective glutamate receptor antagonists or the inhibitors of intracellular glutamate-activated pathways [69–71].

Inhibitory Receptors

g-Amino-Butyric Acid Receptors ␥-amino-butyric acid (GABA) is a major inhibitory neurotransmitter and acts via three receptor subtypes, GABAA, GABAB, and GABAC [72]. Endogenous peripheral GABA arises from primary afferent fibers (glutamate is converted to GABA by glutamate decarboxylase). GABAA receptors, present on some unmyelinated afferent axons [73] are, therefore, involved in modulating pain signaling. The GABAmimetic agents have a broad spectrum of pharmacological actions, including analgesia. Both directly acting (GABAA and GABAB


295

agonists) and indirectly acting GABAergic agents (GABA uptake inhibitors and GABA-transaminase inhibitors) produce analgesia [74]. Gabapentin, a synthetic structural analog of GABA, given systemically, is clinically effective in chronic neuropathic pain conditions [75, 76]. Pharmacological actions of gabapentin are unclear. It is known to interact with an auxiliary subunit of voltage-sensitive Ca2⫹ channel and modulation of GABA and glutamate synthesis [77].

Immune Mediators

Cytokines The role of neuroinflammation and neuroimmune activation in pain involves the infiltration of immune cells to the site of injury (CNS or PNS) and activation of endothelial cells, microglia, and astrocytes. Activation of these cells leads to the production of cytokines and chemokines [78]. Proinflammatory cytokines like interleukin (IL-1, IL-6) and TNF have been implicated in the genesis and maintenance of pain [79–83]. Proinflammatory cytokines can exaggerate pain responses by directly acting on the cytokine receptors found on neurons or by indirectly stimulating the release of other substances that could act on neurons. Cytokines can cause neuronal hyperexcitability, via alterations in ion channels. DRG neurons are known to express TNF receptor type I (TNFRI) and interleukin receptor I [84, 85]. Thus, DRG neuronal response during pain is affected by the surrounding inflammatory cytokines. In parallel, antiinflammatory cytokines such as IL-4, IL-10, IL-13 and IL-1ra are produced and reduce hyperalgesic effects of the proinflammatory cytokines that are initially produced. Inflammatory pain, therefore, is the result of interplay between hyperalgesic and analgesic mediators. Drugs such as immunosuppressants influencing this interplay may also impair endogenous hyperalgesic and analgesic mechanisms. TNF␣, IL-1␣ and IL-1␤ are the first cytokines involved in Wallerian degeneration. Indirectly, these cytokines further regulate macrophage recruitment, myelin removal, survival of PNS neurons, regeneration, and pain through the regulation of NGF production [86, 87]. Increased levels of spinal interleukins have been detected following spinal nerve transection, L5 nerve root injury [88], peripheral nerve injury, acute peripheral inflammation (formalin or zymosan subcutaneous injections in the hind paw) [89], experimental traumatic spinal cord injury in rats [90] and TNF␣ injection in the sciatic nerve [82]. The role of cytokines in neuropathic pain is further demonstrated by the ability of corticosteroids (immunosuppressants), thalidomide, and anti-inflammatory cytokine IL-10 to alleviate neuropathic pain [91, 92].

Chaudhary/Burchiel

296

Peptide Mediators: Opioids

Opioid Peptides Dynorphins, enkephalins and ␤-endorphins (␤-EP) are the main groups of opioid peptides. Recently, a novel group of endogenous opioid peptides have been discovered in the brain and named endomorphins, including endomorphin-1 and endomorphin-2 [93]. These opioid peptide-containing neurons have been found in the thalamus, periaqueductal grey, cortex and spinal cord, regions involved in nociceptive responses. Opioids act through opioid receptors. Opioid receptors are found on the primary afferent terminals, DRG and on immune cells. Three members of the opioid receptor family cloned in the early 1990s include ␦-opioid receptor, ␮-opioid receptor and ␬-opioid receptor. Endomorphins have been shown to induce analgesia via ␦-opioid receptors [94]. These three receptors belong to a family of seven transmembrane GPCRs and share considerable homologies. In addition to the well-established opioid receptors, an orphan opioid-like receptor 1 has been cloned. Nociceptin, a novel opioid-like heptadecapeptide, is believed to be the endogenous ligand for opioid-like receptor 1. The activation of these receptors causes reduction in excitability and decreased propagation of action potentials in the sensory neurons [95]. The role of opioids in the anti-nociceptive processes has been well documented for many centuries and opioids are arguably the earliest and most useful medicines known to man. However, using opioids for chronic and neuropathic pain remains somewhat controversial. Clinical evidence suggest that neuropathic pain is not opioid resistant but that only reduced sensitivity to systemic opioids is observed, i.e., an increase in opioid dose is needed to obtain significant analgesia. This reduced efficiency may be due to changes in spinal opioid receptors or signal transduction pathways [96]. The important problem in administering chronic opioids to control pain is the development of tolerance and dependence. Problems of tolerance are not observed, however, with peripherally applied opioids [97].

Nonopioid Peptides

SP SP is one of the most intensively studied sensory neuropeptides, an undecapeptide belonging to the tachykinin peptide family, which includes SP and neurokinin A/B. These peptides act through neurokinin receptors, NK1, NK2, and NK3. A subpopulation of DRG neurons synthesizes and transports SP to the spinal cord where it is released upon noxious stimulation. Released SP


297

interacts with neurons of lamina I, which express the SP receptor (SPR or neurokinin receptor). This receptor is expressed on spinothalamic and spinobrachial neurons located in the lamina I, suggesting that these neurons play a role in nociception. Increases in NK1 levels in the superficial laminae of the dorsal horn have been detected in the sciatic nerve cut and inflammatory animal models [98, 99]. Upon binding, SP and SPR are internalized [100]. This SP induced internalization of SPR has been exploited as a means of entry into spinal cord neurons in experimental models of pain treatment. SP was conjugated to the ribosome-inactivating protein saporin. This SP-conjugated neurotoxin resulted in the death of NK1 positive neurons, which led to the inhibition of hyperalgesia [101]. The data suggest that a small population of SPR-expressing neurons are important in the maintenance of hyperalgesia; however, the role of a variety of other non-SPR receptors present on these cells should not be underestimated [102]. The novel approach of receptor internalization and introduction of therapeutic compounds may be of future use in targeting of spinal neurons involved in transmitting chronic pain. CGRP Probably the most abundant neuropeptide in small sensory neurons, activation of sensory unmyelinated neurons by noxious stimuli evokes the release CGRP from peripheral nerve endings. CGRP exerts its effects through CGRP1 and CGRP2 receptors, both of which are coupled to adenylyl cyclase. Administration of neutralizing antibody to CGRP in the spinal cord produces analgesia [103]. Neurotensin Neurotensin (NT) is a brain-gut tridecapeptide with dual functionality. It acts as a neurotransmitter/neuromodulator in the nervous system, and as a paracrine and circulating hormone in the periphery. NT acts through three receptors, NTS1, NTS2, and NTS3. NTS1 and NTS2 belong to a family of GPCRs with seven transmembrane domains, whereas NTS3 is a single transmembrane domain protein. Most of the known peripheral and central effects of NT are mediated through NTS1. NT receptors have been demonstrated on small DRG neurons. Sciatic nerve transection causes a marked decrease in the number of NT receptor mRNA-positive small neurons in DRGs, NT mRNApositive neurons in the dorsal horn, and NT-immunocreactive cell bodies and fibers in laminae I-II. Thus, axotomy causes down regulation of several NT systems at the spinal level, suggesting that the possible effects of NT on primary sensory neurons is attenuated after peripheral axotomy [104, 105]. NT administration into the CSF produces dose-related anti-nociceptive responses [106], which may represent a possible NT-mediated approach to pain relief.

Chaudhary/Burchiel

298

Somatostatin Somatostatin or somatotrophin-release inhibiting factor was first isolated as a 14-amino acid peptide that reduced the release of growth hormone from the pituitary. Later, it was found to act as a neuromodulator in the mammalian CNS. It is now known to have an inhibitory effect on nociceptors. Five somatostatin receptor genes have been cloned, SST1–5 [107]. These receptors are G-protein coupled with seven transmembrane domains, which interact with a wide range of downstream signaling targets [108]. Experimental and clinical data suggest that the SST2 receptor might be involved in nociceptive transmission at the central and peripheral sites. Thus, SST2-selective drugs may prove to be important analgesics [109]. Neuropeptide Y Neuropeptide Y (NPY) is a 36-amino acid peptide having diverse biological activity. NPY was originally isolated from the mammalian brain tissue and its three receptors (Y1, Y2, Y3) are relatively abundant in the brain and spinal cord [110]. The anti-nociceptive property of NYP is due to the inhibition of SP release from the primary afferent fibers [111]. Elevated levels of NPY are detected in the spinal gray matter and the DRG after sciatic nerve transection. Galanin Galanin is a 29-amino acid peptide expressed in the DRG and spinal dorsal horn interneurons and regulated by nerve injury and peripheral inflammation. Three G-protein coupled galanin receptor subtypes have been identified: GAL1–3. The role of galanin in pain processing at the spinal level appears to be quite complex. It has been known to produce both facilitatory [112, 113] and inhibitory [114] effects on nociceptive behaviors. This peptide is overexpressed in sensory neurons following peripheral nerve damage. The precise role of the peptide was unclear until the generation of a galanin-knockout mouse. Galanin is now known to act as a neuromodulator and is also important in regeneration [115]. Galanin influences pain processing at the dorsal horn level, particularly via GAL1 (inhibitory) receptors on dorsal horn neurons in response to pain arising from nerve injury (neuropathic pain). GAL1 receptor agonists could perhaps be used to treat neuropathic pain [116]. A better understanding of the role of galanin and its receptors may lead to potential therapeutic treatment options. Cholecystokinin This peptide originally was isolated from mammalian gastrointestinal tract and was later detected in brain. It is also present in primary sensory neurons.


299

Cholecystokinin A and B (CCKA and CCKB) receptors are found in the CNS and PNS. CCKB receptor is predominant in the brain, spinal cord and DRG. The peptide and CCKB receptor levels increase after peripheral axotomy [117, 118].

Chemical and Inflammatory Mediators

BK Biologically active kinins, including bradykinin (BK) and kallidin, are short-lived peptide mediators predominantly generated by the enzymatic action of kallikreins on kininogen precursors [119]. Kinins are involved in neurogenic inflammation through the activation of A␦ and C fibers. A diverse spectrum of physiological and pathological actions attributed to local kinin production is a consequence of the activation of GPCRs. Kinins act through B1 and B2 receptors. B1 receptors are not normally expressed but are expressed in pathological conditions and by the proinflammatory agents such as lipopolysaccharides and cytokines. B2 receptors are expressed constitutively in the PNS and CNS. Kinins also act partly through nonreceptormediated release of histamine and 5-hydroxytryptamine (5-HT) from mast cells and sensitize primary afferents through interactions with inflammatory mediators such as cytokines and PGs leading tohyperalgesia and allodynia [120]. Various transduction pathways have been suggested for the effects of BK. BK-mediated increase in membrane excitation (depolarization) reflects an increase in cation membrane conductance due to Na⫹. This action involves the production of diacyl glycerol, increase in intracellular Ca⫹⫹ and activation of PKC. The increase in calcium levels could cause neuropeptide release, PG synthesis and stimulation of NO synthase [121]. A new pathway for BK effects is through the activation of capsaicin receptors via the production of 12-lipoxygenase metabolites [122]. Experiments have found B2 to have proalgesic actions on sensory neurons. B2 receptor antagonists are effective in their anti-nociceptive properties. However, given that B2 receptors play an important role in the control of physiological processes such as the cardiovascular system, the blockade of B2 receptors has many undesirable side effects. B1 receptors are activated under inflammatory conditions and B1 receptor antagonists are thus being developed to control inflammatory pain. Considering these facts and the widespread distribution of kinin receptors in many tissues, it is no surprise that the therapeutic potential of kinins and kinin receptor antagonists remains the focus of numerous investigations.

Chaudhary/Burchiel

300

PGs Prostanoids include PGD2, PGE2, PGF2␣, PGI2 and thromboxane A2. The enzyme PGH synthase (PGHS), also known as cyclooxygenase (COX) catalyzes the synthesis of PG from arachidonic acid. Arachidonic acid is kept esterified by enzymes until mobilized by phospholipases (PLA2). COX1 is the constitutively active isoform generating PG required for cellular function. COX2 generates huge amounts of PG under pathological conditions. High concentrations of PGs contribute to the excitation of neurons through the suppression of a K⫹ conductance and the increase in Na⫹ (and Ca2⫹) conductance, which leads to an increase in neuropeptide release from C fiber terminals. Prostanoids exert a variety of actions on various tissues and cells. PGs act at peripheral sensory neurons and at central sites within the spinal cord and brain to evoke hyperalgesia [123]. Of the many species of PG known, PGE2 and PGI2 are the major contributors to hyperalgesia. There are at least 8 types and subtypes of the prostanoid receptors in mouse and man: DP, EP1, EP2, EP3, EP4, FP, IP, TP [124]. The PG receptors belong to a GPCR superfamily of seven-transmembrane spanning proteins. Nonsteroidal anti-inflammatory drugs (NSAIDs; e.g., aspirin, indomethacin, ibuprofen), block PGH synthase-derived PG synthesis and are commonly used analgesics and anti-inflammatory agents [125]. Aspirin blocks substrate access and orientation at the COX active site by covalently acetylating a serine residue. The coxibs, e.g., celecoxib (Celebrex) and rofecoxib (Vioxx), are newer selective COX-2 inhibitors that have been used clinically for managing pain [126]. COX-1-derived ‘homeostatic’ PGs are not inhibited by the coxibs. Second-generation coxibs, e.g., valdecoxib and etoricoxib, are under development. Serotonin (5-HT) Serotonin or 5-HT is a neurotransmitter involved in various physiological processes. There exist at least 14 subtypes of 5-HT receptors known to be encoded by distinct genes. Splice variants of many of the subtypes have also been identified resulting in the discovery of at least thirty distinct protein products that recognize 5-HT as their physiological ligand [127]. The 5-HT receptors have been divided into seven subfamilies by convention. The 5-HT1, 5-HT2, 5-HT4, 5-HT5, 5-HT6, and 5-HT7 receptors couple to G-proteins, whereas the 5-HT3 receptors are 5-HT-gated ion channels [128]. Primary afferent fibers (C and A␦) are excited by 5-HT, which appears to involve the activation of 5-HT3 receptors directly gating ion channels permeable to Na⫹ (and K⫹). Like the B2 receptors, 5-HT3 receptors are coupled to PLC and initiate changes in the afferent fibers involving diacylglycerol-induced activation of PKC and IP3-induced increases in intracellular Ca2⫹. 5-HT3


301

receptor antagonists have anti-nociceptive properties in several inflammatory pain models. The understanding of the serotonergic analgesic system will help in the development of new nonopioid, nonaddictive analgesics. Histamine Histamine, a constituent of mast cells, can activate polymodal nociceptors and release pain-related neuropeptides. Histamine generally elicits an itchy sensation rather than pain; however, higher concentrations may induce pain. Pharmacological studies have suggested that a subgroup of primary sensory neurons is responsive to histamine via the H1 receptor, which is coupled to PLC [129].

Growth Factors

NTs NTs are molecules promoting the survival, growth and maintenance of neurons. NGF, brain-derived neurotrophic factor, NT-3, NT4/5 are important NTs (i.e., growth regulators) essential for the development and maintenance of sensory neurons. NT receptor p75 (i.e., low affinity receptor capable of binding to all NTs) and a family of tyrosine kinases TrkA (i.e., binds NGF), TrkB (i.e., binds brain-derived neurotrophic factor and NT4/5) and TrkC (i.e., binds NT3) are located on adult sensory neurons. All the three Trk receptors show discrete but partly overlapping distributions to subpopulations of primary sensory neurons. Peripheral nerve injury results in apoptosis of DRG neurons and downregulation of TrkA in DRG and spinal cord [130]. Administration of exogenous NGF counteracts the degenerative changes in the NGF-responsive axotomized neurons [130, 131]. Recent evidence suggests that NGF is a peripherally produced mediator of some persistent inflammatory pain states. It has also been demonstrated that administration of NGF produces thermal and mechanical hyperalgesia [132]. GDNF GDNF, a member of the transforming growth factor-␤ → (TGF-␤→) superfamily, is a trophic factor with important effects on the primary sensory neurons [133, 134]. GDNF mediates its actions through a multicomponent receptor system composed of a glycosyl-phosphatidylinositol-linked protein (designated GDNFR-␣ or GFR␣-1), a ligand-binding domain, and the transmembrane protein tyrosine kinase Ret, which acts as the signal-transducing domain. About a third of the primary sensory neurons express Ret mRNA [135].

Chaudhary/Burchiel

302

GDNF can both prevent and reverse signs of neuropathic pain. It reduces ectopic discharges in damaged sensory neurons by normalization of expression of sodium channel [57].

Vasoactive Intestinal Peptide (VIP) and Pituitary Adenylyl Cyclase Activating Polypeptide (PACAP) PACAP and VIP are members of the vasoactive intestinal peptide/ secretin/ glucagon family of peptides with neurotransmitter, neuroprotective, and neurotrophic functions. PACAP is widely expressed in many central and peripheral neurons [136], in trigeminal ganglion [137], in gastrointestinal tract, and adrenal glands. It is expressed in two alternatively processed forms PACAP-27 and PACAP-38 and exerts its effects through three different receptors: PAC1 (previously called Type I PACAP receptor), VPAC1 (Type II), and VPAC2 (Type III) [138]. These receptors belong to a family of seven-transmembrane GPCRs. PAC1 receptors are coupled to both adenylate cyclase and phospholipase C [139, 140], and VPAC receptors are mostly coupled to adenylyl cyclase. PAC1 and VPAC receptors play an important role in the transmission of sensory information [141]. PACAP, vasoactive intestinal peptide, and other neuropeptides like CCK and NPY and their receptors are up-regulated after nerve injury [142, 143].

Other Mediators

Nitric Oxide (NO) NO, a free radical gas acts as a messenger molecule and plays a role in synaptic transmission both in the CNS and PNS. Immunohistochemical data suggests that NO synthase, the enzyme that synthesizes NO from L-arginine, is present in the CNS and PNS. Recent studies have suggested a role of NO in nociceptive processing [144]. NO modulates spinal and sensory neuron excitability through multiple mechanisms [145]. The activation of excitatory amino acid receptors such as NMDA receptors causes intraneuronal elevation of calcium, which stimulates NO synthase and production of NO [145, 146]. This formation of NO due to the activation of NMDA receptor indicates that NO may act as a mediator of NMDA-induced nociceptive effects. NO has also been implicated in the development of hyperexcitability resulting in hyperalgesia by increasing nociceptive transmitters at the central terminals. NO biosynthesis inhibitors like NG-nitroarginine-L-methyl ester produce anti-nociceptive effects.


303

Signal Transduction

PKC Activation of PKC has been implicated in the induction and/or maintenance of neuropathic pain behaviors [147–149]. Spinal cord administration of GM1 ganglioside, an intracellular inhibitor of PKC translocation/activation, reverses both increased levels of membrane-bound PKC and painrelated behaviors [150]. Many different groups have used nonspecific PKC blockers, but these studies have not identified which of the ten isoforms of PKC are involved in maintaining hyperalgesia. Malmberg et al. [151] created a mouse lacking PKC gamma (PKC-␥). Mice that lacked PKC-␥ displayed normal responses to acute pain stimuli, but they failed to develop a neuropathic pain syndrome after partial sciatic nerve section. Thus, selective inhibitors of PKC-␥ may help to alleviate nerve injury-induced neuropathic pain states. Since acute pain responses in PKC-␥ null mice were not affected, the added advantage of using selective PKC-␥ inhibitors is that the acute pain responses, which have an important role in detecting injury, are left untouched. PKC-␧ has also been known to play a role in nociception. PKC-␧ is activated by NGF and regulates the responses to NGF including activation of extracellular signal-regulated kinases (ERK1, ERK2), isoforms of mitogenactivated protein (MAP) kinases, and neurite outgrowth [152]. Studies on PKC-␧ null mice indicate that PKC-␧ is required for the full expression of carrageenan-induced hyperalgesia [153], suggesting a role in pain due to inflammation. Inhibitors of PKC-␥ will help in reducing pain without affecting normal nociceptive responses.

Gene Therapy for Pain?

The explosion in research into understanding the neural mechanisms involved in pain has led to a search for more effective molecular treatment options, compared to traditional pharmacological approaches. Establishing and understanding various mediators and modulators involved in the pathophysiology of pain will help in designing novel therapeutic agents. Gene therapy offers a reasonable and physiological methodology to treat pain. Gene therapy can be attempted at various levels: transcription, mRNA stability, and/or translation. Gene transfer therapy to treat pain can focus on a combination of pharmacological, cellular and molecular approaches [154]. Pain therapy can be delivered in the CNS or PNS, or both sites simultaneously. Targets in the PNS offer advantages since pain information can be blocked before it reaches the spinal

Chaudhary/Burchiel

304

Table 2. Examples of potential targets for gene therapy [156] Target protein

Classification

NK-1 (SP) receptor Protein kinase C Vanilloid receptors PN3/Nav1.8 Cannabinoids Acetylcholine receptors NMDA subtype of glutamate receptors

G-protein coupled receptor kinase nonselective cation channels sodium channel G-protein coupled receptors G-protein coupled receptors ligand-gated ion channels

Table 3. Loss of function strategies [182]

Knock down or antisense technology Small molecule inhibitors Ribozymes Aptamers RNA interference (RNAi) Inhibitory peptides Antibodies

cord, and also because delivery in the PNS is not confounded by the problems of CNS administration [155, 156] (table 2). A molecule is a good target in the treatment of pain if it satisfies the following considerations: it should have a major role in pain sensation; interactions with other molecules (agonists and antagonists) should have been studied extensively; it should not disturb other normal physiological functions; and it should be expressed in specific neuronal types so that the shut-off of therapy is possible. Viral vectors, antisense oligonucleotide technology, and RNA interference (RNAi) might all be useful techniques in controlling pain by up-regulating anti-nociceptive and down-regulating pronociceptive targets (table 3). Viral Vectors Viral-derived vectors have a natural ability to penetrate cells and deliver a transgene into the host nucleus. The viral vector has the ability to attach, transfer its genome and the transgene into the host, but is incapable of replication (fig. 4). The most widely used viral vectors are derived from adenoviruses, adeno-associated virus, herpes simplex viruses (HSV), or retroviruses. The properties of various virus vector systems are described below (table 4). An


305

36kb

Sequence

Inverted terminal repeats

Packaging sequence

Viral packaging

Function

Locus control region

Ligand response element

cDNA

Poly A

Heterologous sequence

Inverted terminal repeats

Drug Therapeutic RNA Tissue Removal of gene processing viral genes specificity regulation

Fig. 4. Features of an optimized adenovirus gene therapy vector. Schematic diagram of a gutted adenoviral vector with an adenoviral packaging sequence and terminal repeats (ITR), containing a minimum of adenoviral genome sequences [179].

Table 4. Comparison of properties of various vector systems [183] Features

Retroviral

Lentiviral

Adenoviral

AAV

Naked/lipid DNA

Maximum insert size Concentrations (viral particles per ml) Route of gene delivery Integration Duration of expression in vivo Stability Ease of preparation (scale-up)

7–7.5 kb

7–7.5 kb

⬃30 kb

3.5–4.0 kb

Unlimited size

⬎108

⬎108

⬎1011

⬎1012

No limitation

Ex vivo

Ex/in vivo

Ex/in vivo

Ex/in vivo

Ex/in vivo

Yes Short

Yes Long

No Short

Yes/No Long

Very poor Short

Good Pilot scale-up, up to 20–50 liters Few

Not tested Not known

Good Easy to scale-up

Very good Easy to scale-up

Few

Extensive

Good Difficult to purify, difficult to scale-up Not known

Immunological problems Preexisting host immunity Safety problems

None

Unlikely

Unlikely, Yes Yes No except maybe AIDS patients Insertional Insertional Inflammatory Inflammatory None mutagenesis? mutagenesis? response, response, toxicity toxicity

Chaudhary/Burchiel

306

ITR

⫹

LoxP LoxP

E1

ITR

ITR

⫹

Helper virus

Foreign gene

ITR

Vector Cre-recombinase expressing 293 cells

Reamplify in Cre-recombinase expressing cells ⫹ Helper virus

ITR

⫺

E1

ITR

ITR

⫹

ITR

LoxP Not packaged

Packaged

Fig. 5. The replication defective but still infective virus is dependent on the use of Crerecombinase expressing 293 cell line and a helper virus containing loxP-flanked packaging sequence. The Cre-recombinase enzyme excises any segment of DNA flanked by a loxP sequence (30 bp). Infection of the 293 cells with the helper virus with its ␺ sequence flanked by loxP site results in excision of the viral packaging sequence, rendering the helper virus DNA unpackagable. The helper virus provides all the functions for the packaging of the gutless virus [180].

ideal viral vector for gene therapy should be stable, have tissue-specific gene expression, and should not elicit a host immune response. A recently described helper-dependent gutless adenovirus is devoid of all viral genome. This vector is designed by deletion of all of the viral genome except for the inverted terminal repeat and the packaging (␺) sequence essential for viral packaging. This virus can carry up to 30 kb of transgene (fig. 5), making it a useful size for gene transfer. HSV, an enveloped double-stranded DNA virus, is another commonly used gene transfer virus. The wild-type virus is responsible for the common cold sore. The wild-type virus replicates initially in the skin or mucous membranes and is then taken up by sensory nerve terminals. It can then establish a life long latent state in the nucleus of the sensory ganglion neurons. This unique feature of the HSV can be exploited for applications directed towards conditions that affect the PNS [157]. Peripheral inoculation of HSV vectors on


307

abraded skin [158] or scarified cornea [159] allows the introduction of a transgene to the sensory ganglia neurons without the need for other sophisticated methods of viral vector delivery. Three weeks after infection with HSV encoding rat preproenkephalin A via the hind footpads, strong expression of preproenkephalin A mRNA was detected in the rat lumbar DRG [158, 160]. Wilson and Yeomans and colleagues [161, 162] used a similar approach and demonstrated the anti-hyperalgesic effect of preproenkephalin overexpression in primary sensory neurons on sensitization of sensory afferents by dimethylsulfoxide or capsaicin application. Hao et al. [163] have used HSV vectormediated expression of proenkephalin in the DRG and demonstrated an anti-allodynic effect in neuropathic pain. They also showed that the enkephalin release enhances the effect of morphine, reducing ED50 of morphine 10-fold and the animals also did not develop tolerance to the continued production of vector-mediated enkephalin over a period of several weeks. Taken together, several of the above studies suggest that viral vector-mediated expression of proenkephalin may be a novel way to treat patients with neuropathic pain. Production of site-specific peptides is of great interest in the field of gene therapy, but these may require various modifications in order to facilitate secretion or activity in vivo. Addition of N-terminal signal peptide is not always sufficient to achieve this goal. To overcome this problem, addition of the preprosequence of mouse nerve growth factor to ␤-EP was tested [164]. Retrovirusmediated expression of a hybrid construct of the preprosequence of NGF and human ␤-EP in primary fibroblasts resulted in the secretion of ␤-EP. Transplantation of such ␤-EP-secreting cells into the brain or spinal cord could provide an ex vivo gene therapy approach for the treatment of chronic, opioidsensitive pain states [164]. Concerns about viral vector distribution in the CNS have limited current gene therapy efforts (table 5). The problem of CNS distribution might be overcome by transferring genes to the meninges surrounding the spinal cord. For example, a recombinant adenovirus encoding a secreted form of ␤-EP was delivered by intrathecal infusion and the resulting increase in ␤-EP secretion by the meningeal piamater cells attenuated inflammatory hyperalgesia in a carrageenan-injection model of persistent pain. This method can be adapted to treat pain in neurodegenerative disorders in which broad spatial distribution of therapeutic effect is critical [165]. Viral vectors have also been used to deliver antisense molecules to control the expression of specific genes in vivo [166, 167]. Thus, viral vector approaches can be used to treat chronic pain states in which plastic changes occur in the neuronal systems. However, in spite of attempts in carefully designing viral vectors, they have not been widely accepted for the use in the treatment of pain.

Chaudhary/Burchiel

308

Table 5. Comparison of viral vectors and antisense oligonucleotides [155] Advantages Viral vectors

Disadvantages

• effective delivery of large • immunogenicity prevents exogenous DNA

• longer duration of action • possibility of specific •

targeting effective in nondividing cells

• • • •

• • ODN

• specificity of gene • • • •

inhibition minimal immunogenicity effective in nondividing cells not integrated into host genome short duration of action

repeated administration of viral vector direct cellular toxicity inability to target specific subset of cells difficulty in penetrating blood brain barrier may cause insertional mutagenesis (due to integration of viral vector into host genome) possibility of viral replication possibility of creation of a new recombinant virus in vivo (for retrovirus)

• difficulty in generating • • •

and isolating an active oligomer difficulty in gaining intracellular access toxicity may be caused by nonspecific effects of the ODN difficulty in penetrating blood brain barrier

Antisense Oligonucleotide Technology Antisense oligonucleotides are complementary to a portion of target mRNA [168] and have advantages over viral vectors (table 5). Binding of the antisense molecule to target mRNA disturbs the ability of the mRNA to be read by the translational machinery, and thus blocks the synthesis of the encoded protein (fig. 6). Antisense oligonucleotides are modified to enhance entry into cells and are made to be resistant to nucleases within the cell. Three regions in a DNA sequence are considered best for standard antisense design; the 5⬘ cap region, the AUG initiation codon, and the 3⬘ untranslated region of mRNA. Most antisense molecules are 15–20 bases long. This length is sufficient to pick out an unique sequence from others. The oligonucleotides enter the cells by fluid phase pinocytosis, receptor-mediated endocytosis or both.


309

Stimuli Pain

mRNA

DNA Antisense oligo

Antisense oligo

Pronociceptive protein

mRNA

DNA

Nucleus

Nucleus

Normal gene activity

Block of gene activity

Fig. 6. Antisense oligonucleotide and potential sites of action [155].

Antisense technology might be used to knock down targets involved in nociception such as NMDA receptors, PKC, neurokinin 1 receptor, and sodium channels [155] (table 6). Although this seems to be a feasible technology, there are practical difficulties in designing a perfect oligomer with greatest specificity and determining that the significant effects observed are due to the antisense oligomer, and not due to other nonantisense effects. RNAi Post-transcriptional gene silencing and RNAi involve the specific suppression of genes by complementary dsRNA [169]. RNAi provides a powerful method of gene silencing in eukaryotic cells. Specific genetic interference by double-stranded RNA in Caenorhabditis elegans was first discovered by Fire et al. [170] in 1998. Double-stranded RNA rather than single-stranded antisense RNA is introduced within cells. Once inside the cell, the double-stranded RNA molecules are cleaved by ribonuclease III into twenty-one to twenty-two nucleotide short interfering RNAs which are replicated by an RNA-dependent RNA polymerase. The short interfering RNA duplexes bind to a nuclease complex to form the RNA-induced silencing complex, which then targets the homologous endogenous mRNA sequence, thus blocking further protein synthesis [171]. However, not many experiments have been performed to determine the utility of RNAi as a method of gene knockdown in postmitotic mammalian neurons. Only recently, Krichevsky and Kosik [172] have applied the RNAi to postmitotic primary neuronal cultures. Thus, in the future one would potentially be able to specifically block target molecules in DRG neurons to control pain.

Chaudhary/Burchiel

310

Table 6. Effects of antisense oligonucleotide sequence in animal pain models [177] mRNA

Species

Effects

References

c-fos

Rat

↓ c-fos protein immunoreactivity

[184]

DOR

Mice

↓ DOR protein but not mRNA and anti-nociception produced by DOR-selective agonist (D-Ala2) deltorphin II

[185]

GAL

Rat

↓ axotomy-induced upregulation of Gal protein, but no change in GAL mRNA

[186]

NK-1 R

Rat

↓ behavioral response to formalin and NK-1 receptor protein immunoreactivity in spinally SP-treated rats

[187]

Nav1.8 SNS/PN3

Rat

↓ SNS/PN3 protein immunoreactivity decreased in DRG and chronic nerve or tissue injury-induced hyperalgesia and allodynia ↓ TTX-R Na⫹ current density in cultured sensory neurons and PGE2-induced hyperalgesia

[188] [189] [190]

PKC ␣

Human (in vitro)

Inhibition of phorbol ester-induced reduction of bradykinin-evoked calcium mobilization

[191]

NMDAR1

Mouse

↓ pain behavior and decreased receptor binding

[192]

Rat

↓ immunoreactive staining for NMDA-R1 and ↓ formalin-evoked behaviors

[193]

Rat

Delayed onset of mechanical and thermal hyperalgesia in chronic neuropathic pain model

[194]

PSD95/SAP90

Other Gene Therapy Methods Ribozymes are RNA molecules, which also act as enzymes. These catalytic RNA molecules can be designed to recognize and bind to a specific mRNA and cause cleavage of mRNA, thus preventing its translation into protein. While they represent an alternative to RNAi, achieving specificity and delivery of these enzymes within the living tissue is difficult. Neural stem cells are self-renewing precursors of neurons and glia with numerous potential ex vivo gene therapy applications. The advantages of using these precursors include their theoretically limitless clonal expansion in tissue culture [173]. Neural progenitor cells could be genetically modified to express exogenous genes for neurotransmitters, neurotrophic factors, or various ion


311

channels. Other ex vivo transfected (iso- or xenogenic) cells can also be used to treat nervous system disorders. Several drawbacks using such methods include risk of toxicity, possible tumor formation, instability of transgenes, and lack of cell specificity. In spite of these challenges, Ishii et al. [174] have successfully used a combination of cell transplantation and gene transfer for the delivery of ␤-EP into the subarachanoid space in rats. The rats that received ␤-EP producing cells showed prominent analgesic effects for up to a month after transplantation. Another study used xenogenic tumor cells secreting ␤-EP and immunologically isolated in polymer capsules (microcapsules) to reduce pain when transplanted into the CSF of rats [175]. Synthetic DNA delivery systems like liposomes have become increasingly popular methods of gene transfer. Introduction of DNA into cells can now be safely achieved by complexing DNA with cationic lipids. These complexes are endocytosed into the cells, which involves binding, internalization, formation of endosomes, fusion with lysosomes and lysis. Finally, the DNA which survives endocytotic processing and degradation by nucleases reaches the nucleus [176]. Liposomes have been successfully used to inject DNA complexes into rodent brains, but gene expression is transient. These methods could be used in the treatment of various neurological diseases including the treatment of intractable pain, where transient expression of the transgene is needed.

Future Directions in Pain Research

Pain sensation is complex and involves integrative mechanisms at the PNS and CNS. Neuropathic pain is unresponsive to most conventional therapy. In recent years, much has been elucidated concerning neuroanatomical circuits, mediators of pain and transduction pathways involved in pain processing. This information has led to the development of new and unconventional therapeutic options for the treatment of pain. Ion channels (e.g., Nav1.8, Nav1.9), neurotransmitters, neuropeptides and their receptors present on pain-sensing neurons are important potential targets for therapy. PKC and other kinases also offer as targets for analgesic development. Among gene therapy options, the use of antisense oligonucleotides seems promising but delivery to specific cell types remains problematic. Viral vectors are attractive candidates but safety and targeting issues remain. Modulation of proteins involved in hyperalgesia by understanding the proteome will lead to more effective therapies for pain relief. Focal drug delivery via microinfusion systems may be necessary adjuncts to analgesic design. In the future, pain management will be a multidisciplinary approach that will include pharmacological intervention, minimally invasive procedures, and gene therapy targeted to

Chaudhary/Burchiel

312

specific cell types or specific neuron class with high efficacy and minimum side effects. It is difficult to predict which of these approaches may lead to a clinically applicable means of producing analgesia. What is certain is that each of these therapies must be precisely regulated for optimal clinical effects with optimal pharmacological specificity. However this field evolves, future analgesics depend on a growing knowledge of the nociceptive system and its aberrations.

References 1 2 3 4 5 6 7 8

9 10 11

12

13 14

15 16

17 18

Waldmann R, Lazdunski M: H(⫹)-gated cation channels: Neuronal acid sensors in the NaC/DEG family of ion channels. Curr Opin Neurobiol 1998;8:418–424. McCleskey EW, Gold MS: Ion channels of nociception. Annu Rev Physiol 1999;61:835–856. Basbaum AI, Fields HL: Endogenous pain control systems: Brainstem spinal pathways and endorphin circuitry. Annu Rev Neurosci 1984;7:309–338. Millan MJ: The induction of pain: An integrative review. Prog Neurobiol 1999;57:1–164. Caterina MJ, Schumacher MA, Tominaga M, Rosen TA, Levine JD, Julius D: The capsaicin receptor: A heat-activated ion channel in the pain pathway. Nature 1997;389:816–824. Gunthorpe MJ, Benham CD, Randall A, Davis JB: The diversity in the vanilloid (TRPV) receptor family of ion channels. Trends Pharmacol Sci 2002;23:183–191. Ahluwalia J, Urban L, Capogna M, Bevan S, Nagy I: Cannabinoid 1 receptors are expressed in nociceptive primary sensory neurons. Neuroscience 2000;100:685–688. Tominaga M, Caterina MJ, Malmberg AB, Rosen TA, Gilbert H, Skinner K, Raumann BE, Basbaum AI, Julius D: The cloned capsaicin receptor integrates multiple pain-producing stimuli. Neuron 1998;21:531–543. Reichling DB, Levine JD: Heat transduction in rat sensory neurons by calcium-dependent activation of a cation channel. Proc Natl Acad Sci USA 1997;94:7006–7011. Cesare P, McNaughton P: A novel heat-activated current in nociceptive neurons and its sensitization by bradykinin. Proc Natl Acad Sci USA 1996;93:15435–15439. Caterina MJ, Leffler A, Malmberg AB, Martin WJ, Trafton J, Petersen-Zeitz KR, Koltzenburg M, Basbaum AI, Julius D: Impaired nociception and pain sensation in mice lacking the capsaicin receptor. Science 2000;288:306–313. Davis JB, Gray J, Gunthorpe MJ, Hatcher JP, Davey PT, Overend P, Harries MH, Latcham J, Clapham C, Atkinson K, Hughes SA, Rance K, Grau E, Harper AJ, Pugh PL, Rogers DC, Bingham S, Randall A, Sheardown SA: Vanilloid receptor-1 is essential for inflammatory thermal hyperalgesia. Nature 2000;405:183–187. Szallasi A, Blumberg PM: Vanilloid (Capsaicin) receptors and mechanisms. Pharmacol Rev 1999;51:159–212. Kwak J, Wang MH, Hwang SW, Kim TY, Lee SY, Oh U: Intracellular ATP increases capsaicinactivated channel activity by interacting with nucleotide-binding domains. J Neurosci 2000;20: 8298–8304. Premkumar LS, Ahern GP: Induction of vanilloid receptor channel activity by protein kinase C. Nature 2000;408:985–990. De Petrocellis L, Harrison S, Bisogno T, Tognetto M, Brandi I, Smith GD, Creminon C, Davis JB, Geppetti P, Di Marzo V: The vanilloid receptor (VR1)-mediated effects of anandamide are potently enhanced by the cAMP-dependent protein kinase. J Neurochem 2001;77:1660–1663. Jordt SE, Tominaga M, Julius D: Acid potentiation of the capsaicin receptor determined by a key extracellular site. Proc Natl Acad Sci USA 2000;97:8134–8139. Smart D, Gunthorpe MJ, Jerman JC, Nasir S, Gray J, Muir AI, Chambers JK, Randall AD, Davis JB: The endogenous lipid anandamide is a full agonist at the human vanilloid receptor (hVR1). Br J Pharmacol 2000;129: 227–230.


313

19 20 21 22

23

24 25 26

27

28 29 30 31 32 33

34 35 36 37

38 39 40 41 42 43 44

Richardson JD, Kilo S, Hargreaves KM: Cannabinoids reduce hyperalgesia and inflammation via interaction with peripheral CB1 receptors. Pain 1998;75:111–119. Richardson JD, Aanonsen L, Hargreaves KM: Antihyperalgesic effects of spinal cannabinoids. Eur J Pharmacol 1998;345:145–153. Martin WJ, Loo CM, Basbaum AI: Spinal cannabinoids are anti-allodynic in rats with persistent inflammation. Pain 1999;82:199–205. Herzberg U, Eliav E, Bennett GJ, Kopin IJ: The analgesic effects of R(⫹)-WIN 55,212–2 mesylate, a high affinity cannabinoid agonist, in a rat model of neuropathic pain. Neurosci Lett 1997; 221:157–160. Zygmunt PM, Petersson J, Andersson DA, Chuang H, Sorgard M, Di Marzo V, Julius D, Hogestatt ED: Vanilloid receptors on sensory nerves mediate the vasodilator action of anandamide. Nature 1999;400: 452–457. Di Marzo V, Blumberg PM, Szallasi A: Endovanilloid signaling in pain. Curr Opin Neurobiol 2002;12:372–379. McKemy DD, Neuhausser WM, Julius D: Identification of a cold receptor reveals a general role for TRP channels in thermosensation. Nature 2002;416:52–58. Story GM, Peier AM, Reeve AJ, Eid SR, Mosbacher J, Hricik TR, Earley TJ, Hergarden AC, Andersson DA, Hwang SW, McIntyre P, Jegla T, Bevan S, Patapoutian A: ANKTM1, a TRP-like channel expressed in nociceptive neurons, is activated by cold temperatures. Cell 2003;112: 819–829. Vergnolle N, Bunnett NW, Sharkey KA, Brussee V, Compton SJ, Grady EF, Cirino G, Gerard N, Basbaum AI, Andrade-Gordon P, Hollenberg MD, Wallace JL: Proteinase-activated receptor-2 and hyperalgesia: A novel pain pathway. Nat Med 2001;7:821–826. Vergnolle N, Wallace JL, Bunnett NW, Hollenberg MD: Protease-activated receptors in inflammation, neuronal signaling and pain. Trends Pharmacol Sci 2001;22:146–152. Hollenberg MD: Protease-activated receptors: PAR4 and counting: How long is the course? Trends Pharmacol Sci 1999;20:271–273. Mantyh PW, Yaksh TL: Sensory neurons are PARtial to pain. Nat Med 2001;7:772–773. Furst S: Transmitters involved in antinociception in the spinal cord. Brain Res Bull 1999;48: 129–141. Flores CM: The promise and pitfalls of a nicotinic cholinergic approach to pain management. Pain 2000;88:1–6. Bannon AW, Decker MW, Holladay MW, Curzon P, Donnelly-Roberts D, Puttfarcken PS, Bitner RS, Diaz A, Dickenson AH, Porsolt RD, Williams M, Arneric SP: Broad-spectrum, non-opioid analgesic activity by selective modulation of neuronal nicotinic acetylcholine receptors. Science 1998;279:77–81. Burnstock G: A unifying purinergic hypothesis for the initiation of pain. Lancet 1996;347: 1604–1605. Burnstock G: P2X receptors in sensory neurones. Br J Anaesth 2000;84:476–488. North RA: Molecular physiology of P2X receptors. Physiol Rev 2002;82:1013–1067. Bradbury EJ, Burnstock G, McMahon SB: The expression of P2X3 purinoreceptors in sensory neurons: Effects of axotomy and glial-derived neurotrophic factor. Mol Cell Neurosci 1998;12:256–268. Eriksson J, Bongenhielm U, Kidd E, Matthews B, Fried K: Distribution of P2X3 receptors in the rat trigeminal ganglion after inferior alveolar nerve injury. Neurosci Lett 1998;254:37–40. Stoop R, Surprenant A, North RA: Different sensitivities to pH of ATP-induced currents at four cloned P2X receptors. J Neurophysiol 1997;78:1837–1840. North RA, Barnard EA: Nucleotide receptors. Curr Opin Neurobiol 1997;7:346–357. Thomas S, Virginio C, North RA, Surprenant A: The antagonist trinitrophenyl-ATP reveals co-existence of distinct P2X receptor channels in rat nodose neurones. J Physiol 1998;509:411–417. Taiwo YO, Levine JD: Direct cutaneous hyperalgesia induced by adenosine. Neuroscience 1990; 38:757–762. Sawynok J: Adenosine receptor activation and nociception. Eur J Pharmacol 1998;347:1–11. Waldmann R: Proton-gated cation channels – Neuronal acid sensors in the central and peripheral nervous system. Adv Exp Med Biol 2001;502:293–304.

Chaudhary/Burchiel

314

45

46 47 48 49 50 51 52 53 54 55 56 57 58

59 60

61 62 63 64 65 66

67

68 69

Escoubas P, De Weille JR, Lecoq A, Diochot S, Waldmann R, Champigny G, Moinier D, Menez A, Lazdunski M: Isolation of a tarantula toxin specific for a class of proton-gated Na⫹ channels. J Biol Chem 2000;275: 25116–25121. Patel AJ, Honore E: Properties and modulation of mammalian 2P domain K⫹ channels. Trends Neurosci 2001;24:339–346. Yost CS: Potassium channels: Basic aspects, functional roles, and medical significance. Anesthesiology 1999;90:1186–1203. Patel AJ, Honore E, Maingret F, Lesage F, Fink M, Duprat F, Lazdunski M: A mammalian two pore domain mechano-gated S-like K⫹ channel. Embo J 1998;17:4283–4290. Maingret F, Patel AJ, Lesage F, Lazdunski M, Honore E: Mechano- or acid stimulation, two interactive modes of activation of the TREK-1 potassium channel. J Biol Chem 1999;274:26691–26696. Maingret F, Lauritzen I, Patel AJ, Heurteaux C, Reyes R, Lesage F, Lazdunski M, Honore E: TREK-1 is a heat-activated background K(⫹) channel. Embo J 2000;19:2483–2491. Kim Y, Bang H, Kim D: TASK-3, a new member of the tandem pore K(⫹) channel family. J Biol Chem 2000;275:9340–9347. Lesage F, Lazdunski M: Molecular and functional properties of two-pore-domain potassium channels. Am J Physiol Renal Physiol 2000;279:F793–F801. Waxman SG, Dib-Hajj S, Cummins TR, Black JA: Sodium channels and pain. Proc Natl Acad Sci USA 1999;96:7635–7639. Gold MS, Reichling DB, Shuster MJ, Levine JD: Hyperalgesic agents increase a tetrodotoxinresistant Na⫹ current in nociceptors. Proc Natl Acad Sci USA 1996;93:1108–1112. Baker MD, Wood JN: Involvement of Na⫹ channels in pain pathways. Trends Pharmacol Sci 2001;22:27–31. Eglen RM, Hunter JC, Dray A: Ions in the fire: Recent ion-channel research and approaches to pain therapy. Trends Pharmacol Sci 1999;20:337–342. Boucher TJ, Okuse K, Bennett DL, Munson JB, Wood JN, McMahon SB: Potent analgesic effects of GDNF in neuropathic pain states. Science 2000;290:124–127. Cummins TR, Waxman SG: 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 1997;17:3503–3514. Dib-Hajj S, Black JA, Felts P, Waxman SG: Down-regulation of transcripts for Na channel alphaSNS in spinal sensory neurons following axotomy. Proc Natl Acad Sci USA 1996;93:14950–14954. Dib-Hajj SD, Tyrrell L, Black JA, Waxman SG: NaN, a novel voltage-gated Na channel, is expressed preferentially in peripheral sensory neurons and down-regulated after axotomy. Proc Natl Acad Sci USA 1998;95:8963–8968. Omote K, Kawamata T, Kawamata M, Namiki A: Formalin-induced release of excitatory amino acids in the skin of the rat hindpaw. Brain Res 1998;787:161–164. Carlton SM, Hargett GL, Coggeshall RE: Localization and activation of glutamate receptors in unmyelinated axons of rat glabrous skin. Neurosci Lett 1995;197:25–28. deGroot J, Zhou S, Carlton SM: Peripheral glutamate release in the hindpaw following low and high intensity sciatic stimulation. Neuroreport 2000;11:497–502. Parsons CG: NMDA receptors as targets for drug action in neuropathic pain. Eur J Pharmacol 2001;429:71–78. Conn PJ, Pin JP: Pharmacology and functions of metabotropic glutamate receptors. Annu Rev Pharmacol Toxicol 1997;37:205–237. Dogrul A, Ossipov MH, Lai J, Malan TP Jr, Porreca F: Peripheral and spinal antihyperalgesic activity of SIB-1757, a metabotropic glutamate receptor (mGLUR(5)) antagonist, in experimental neuropathic pain in rats. Neurosci Lett 2000;292:115–118. Walker K, Reeve A, Bowes M, Winter J, Wotherspoon G, Davis A, Schmid P, Gasparini F, Kuhn R, Urban L: mGlu5 receptors and nociceptive function. II. mGlu5 receptors functionally expressed on peripheral sensory neurones mediate inflammatory hyperalgesia. Neuropharmacology 2001;40:9–10. Bhave G, Karim F, Carlton SM, Gereau RW4th: Peripheral group I metabotropic glutamate receptors modulate nociception in mice. Nat Neurosci 2001;4:417–423. Carlton SM, Neugebauer V: Peripheral metabotropic glutamate receptors as drug targets for pain relief. Expert Opin Ther Targets 2002;6:349–361.


315

70 71

72 73 74 75

76 77 78 79 80 81 82 83 84

85

86 87 88

89

90 91

92 93 94

Karim F, Bhave G, Gereau RW4th: Metabotropic glutamate receptors on peripheral sensory neuron terminals as targets for the development of novel analgesics. Mol Psychiatry 2001;6:615–617. Karim F, Wang CC, Gereau RW4th: Metabotropic glutamate receptor subtypes 1 and 5 are activators of extracellular signal-regulated kinase signaling required for inflammatory pain in mice. J Neurosci 2001;21:3771–3779. Malcangio M, Bowery NG: GABA and its receptors in the spinal cord. Trends Pharmacol Sci 1996;17:457–462. Carlton SM, Zhou S, Coggeshall RE: Peripheral GABA(A) receptors: Evidence for peripheral primary afferent depolarization. Neuroscience 1999;93:713–722. Sawynok J: GABAergic mechanisms of analgesia: An update. Pharmacol Biochem Behav 1987;26:463–474. Morello CM, Leckband SG, Stoner CP, Moorhouse DF, Sahagian GA: Randomized double-blind study comparing the efficacy of gabapentin with amitriptyline on diabetic peripheral neuropathy pain. Arch Intern Med 1999;159:1931–1937. Mao J, Chen LL: Gabapentin in pain management. Anesth Analg 2000;91:680–687. Taylor CP: Mechanisms of action of gabapentin. Rev Neurol (Paris) 1997;153(suppl 1):S39–S45. DeLeo JA, Yezierski RP: The role of neuroinflammation and neuroimmune activation in persistent pain. Pain 2001;90:1–6. Ferreira SH, Lorenzetti BB, Bristow AF, Poole S: Interleukin-1 beta as a potent hyperalgesic agent antagonized by a tripeptide analogue. Nature 1988;334:698–700. Watkins LR, Martin D, Ulrich P, Tracey KJ, Maier SF: Evidence for the involvement of spinal cord glia in subcutaneous formalin induced hyperalgesia in the rat. Pain 1997;71:225–235. Watkins LR, Wiertelak EP, Goehler LE, Smith KP, Martin D, Maier SF: Characterization of cytokine-induced hyperalgesia. Brain Res 1994;654:15–26. Wagner R, Myers RR: Endoneurial injection of TNF-alpha produces neuropathic pain behaviors. Neuroreport 1996;7:2897–2901. Sorkin LS, Doom CM: Epineurial application of TNF elicits an acute mechanical hyperalgesia in the awake rat. J Peripher Nerv Syst 2000;5:96–100. Copray JC, Mantingh I, Brouwer N, Biber K, Kust BM, Liem RS, Huitinga I, Tilders FJ, Van Dam AM, Boddeke HW: Expression of interleukin-1 beta in rat dorsal root ganglia. J Neuroimmunol 2001;118:203–211. Schafers M, Geis C, Svensson CI, Luo ZD, Sommer C: Selective increase of tumour necrosis factor-alpha in injured and spared myelinated primary afferents after chronic constrictive injury of rat sciatic nerve. Eur J Neurosci 2003;17:791–804. Shamash S, Reichert F, Rotshenker S: The cytokine network of Wallerian degeneration: Tumor necrosis factor-alpha, interleukin-1alpha, and interleukin-1beta. J Neurosci 2002;22:3052–3060. Lindholm D, Heumann R, Meyer M, Thoenen H: Interleukin-1 regulates synthesis of nerve growth factor in non-neuronal cells of rat sciatic nerve. Nature 1987;330:658–659. Winkelstein BA, Rutkowski MD, Sweitzer SM, Pahl JL, DeLeo JA: Nerve injury proximal or distal to the DRG induces similar spinal glial activation and selective cytokine expression but differential behavioral responses to pharmacologic treatment. J Comp Neurol 2001;439:127–139. Sweitzer SM, Colburn RW, Rutkowski M, DeLeo JA: Acute peripheral inflammation induces moderate glial activation and spinal IL-1beta expression that correlates with pain behavior in the rat. Brain Res 1999;829:209–221. Wang CX, Olschowka JA, Wrathall JR: Increase of interleukin-1beta mRNA and protein in the spinal cord following experimental traumatic injury in the rat. Brain Res 1997;759:190–196. Wagner R, Janjigian M, Myers RR: Anti-inflammatory interleukin-10 therapy in CCI neuropathy decreases thermal hyperalgesia, macrophage recruitment, and endoneurial TNF-alpha expression. Pain 1998;74:35–42. Johansson A, Bennett GJ: Effect of local methylprednisolone on pain in a nerve injury model. A pilot study. Reg Anesth 1997;22:59–65. Zadina JE, Hackler L, Ge LJ, Kastin AJ: A potent and selective endogenous agonist for the mu-opiate receptor. Nature 1997;386:499–502. Przewlocki R, Labuz D, Mika J, Przewlocka B, Tomboly C, Toth G: Pain inhibition by endomorphins. Ann NY Acad Sci 1999;897:154–164.

Chaudhary/Burchiel

316

95 Stein C, Machelska H, Binder W, Schafer M: Peripheral opioid analgesia. Curr Opin Pharmacol 2001;1:62–65. 96 Przewlocki R, Przewlocka B: Opioids in chronic pain. Eur J Pharmacol 2001;429:79–91. 97 Ueda H, Inoue M: Peripheral morphine analgesia resistant to tolerance in chronic morphinetreated mice. Neurosci Lett 1999;266:105–108. 98 Yashpal K, Dam TV, Quirion R: Quantitative autoradiographic distribution of multiple neurokinin binding sites in rat spinal cord. Brain Res 1990;506:259–266. 99 Yashpal K, Dam TV, Quirion R: Effects of dorsal rhizotomy on neurokinin receptor sub-types in the rat spinal cord: A quantitative autoradiographic study. Brain Res 1991;552:240–247. 100 Mantyh PW, DeMaster E, Malhotra A, Ghilardi JR, Rogers SD, Mantyh CR, Liu H, Basbaum AI, Vigna SR, Maggio JE: Receptor endocytosis and dendrite reshaping in spinal neurons after somatosensory stimulation. Science 1995;268:1629–1632. 101 Mantyh PW, Rogers SD, Honore P, Allen BJ, Ghilardi JR, Li J, Daughters RS, Lappi DA, Wiley RG, Simone DA: Inhibition of hyperalgesia by ablation of lamina I spinal neurons expressing the substance P receptor. Science 1997;278:275–279. 102 Levine JD, Fields HL, Basbaum AI: Peptides and the primary afferent nociceptor. J Neurosci 1993;13:2273–2286. 103 Kuraishi Y, Nanayama T, Ohno H, Minami M, Satoh M: Antinociception induced in rats by intrathecal administration of antiserum against calcitonin gene-related peptide. Neurosci Lett 1988;92:325–329. 104 Zhang X, Xu ZQ, Bao L, Dagerlind A, Hokfelt T: Complementary distribution of receptors for neurotensin and NPY in small neurons in rat lumbar DRGs and regulation of the receptors and peptides after peripheral axotomy. J Neurosci 1995;15:2733–2747. 105 Zhang X, Ji RR, Nilsson S, Villar M, Ubink R, Ju G, Wiesenfeld-Hallin Z, Hokfelt T: Neuropeptide Y and galanin binding sites in rat and monkey lumbar dorsal root ganglia and spinal cord and effect of peripheral axotomy. Eur J Neurosci 1995;7:367–380. 106 Yaksh TL: Spinal opiate analgesia: Characteristics and principles of action. Pain 1981;11:293–346. 107 Epelbaum J, Dournaud P, Fodor M, Viollet C: The neurobiology of somatostatin. Crit Rev Neurobiol 1994;8:25–44. 108 Patel YC: Molecular pharmacology of somatostatin receptor subtypes. J Endocrinol Invest 1997;20:348–367. 109 Selmer I, Schindler M, Allen JP, Humphrey PP, Emson PC: Advances in understanding neuronal somatostatin receptors. Regul Pept 2000;90:1–18. 110 Wettstein JG, Earley B, Junien JL: Central nervous system pharmacology of neuropeptide Y. Pharmacol Ther 1995;65:397–414. 111 Hua XY, Boublik JH, Spicer MA, Rivier JE, Brown MR, Yaksh TL: The antinociceptive effects of spinally administered neuropeptide Y in the rat: Systematic studies on structure-activity relationship. J Pharmacol Exp Ther 1991;258:243–248. 112 Cridland RA, Henry JL: Effects of intrathecal administration of neuropeptides on a spinal nociceptive reflex in the rat: VIP, galanin, CGRP, TRH, somatostatin and angiotensin II. Neuropeptides 1988;11:23–32. 113 Wiesenfeld-Hallin Z, Villar MJ, Hokfelt T: Intrathecal galanin at low doses increases spinal reflex excitability in rats more to thermal than mechanical stimuli. Exp Brain Res 1988;71:663–666. 114 Post C, Alari L, Hokfelt T: Intrathecal galanin increases the latency in the tail-flick and hot-plate test in mouse. Acta Physiol Scand 1988;132:583–584. 115 Wynick D, Thompson SW, McMahon SB: The role of galanin as a multi-functional neuropeptide in the nervous system. Curr Opin Pharmacol 2001;1:73–77. 116 Liu HX, Hokfelt T: The participation of galanin in pain processing at the spinal level. Trends Pharmacol Sci 2002;23:468–474. 117 Zhang X, Dagerlind A, Elde RP, Castel MN, Broberger C, Wiesenfeld-Hallin Z, Hokfelt T: Marked increase in cholecystokinin B receptor messenger RNA levels in rat dorsal root ganglia after peripheral axotomy. Neuroscience 1993;57:227–233. 118 Verge VM, Wiesenfeld-Hallin Z, Hokfelt T: Cholecystokinin in mammalian primary sensory neurons and spinal cord: In situ hybridization studies in rat and monkey. Eur J Neurosci 1993; 5:240–250.


317

119 Marceau F: Kinin B1 receptors: A review. Immunopharmacology 1995;30:1–26. 120 Calixto JB, Cabrini DA, Ferreira J, Campos MM: Kinins in pain and inflammation. Pain 2000; 87:1–5. 121 Dray A, Urban L: New pharmacological strategies for pain relief. Annu Rev Pharmacol Toxicol 1996;36:253–280. 122 Shin J, Cho H, Hwang SW, Jung J, Shin CY, Lee SY, Kim SH, Lee MG, Choi YH, Kim J, Haber NA, Reichling DB, Khasar S, Levine JD, Oh U: Bradykinin-12-lipoxygenase-VR1 signaling pathway for inflammatory hyperalgesia. Proc Natl Acad Sci USA 2002;99:10150–10155. 123 Narumiya S, FitzGerald GA: Genetic and pharmacological analysis of prostanoid receptor function. J Clin Invest 2001;108:25–30. 124 Narumiya S, Sugimoto Y, Ushikubi F: Prostanoid receptors: Structures, properties, and functions. Physiol Rev 1999;79:1193–1226. 125 Funk CD: Prostaglandins and leukotrienes: Advances in eicosanoid biology. Science 2001;294: 1871–1875. 126 FitzGerald GA, Patrono C: The coxibs, selective inhibitors of cyclooxygenase-2. N Engl J Med 2001;345:433–442. 127 Raymond JR, Mukhin YV, Gelasco A, Turner J, Collinsworth G, Gettys TW, Grewal JS, Garnovskaya MN: Multiplicity of mechanisms of serotonin receptor signal transduction. Pharmacol Ther 2001; 92:179–212. 128 Kroeze WK, Kristiansen K, Roth BL: Molecular biology of serotonin receptors structure and function at the molecular level. Curr Top Med Chem 2002;2:507–528. 129 Hill SJ, Ganellin CR, Timmerman H, Schwartz JC, Shankley NP, Young JM, Schunack W, Levi R, Haas HL: International Union of Pharmacology. XIII. Classification of histamine receptors. Pharmacol Rev 1997;49:253–278. 130 Li L, Deng YS, Zhou XF: Downregulation of TrkA expression in primary sensory neurons after unilateral lumbar spinal nerve transection and some rescuing effects of nerve growth factor infusion. Neurosci Res 2000;38:183–191. 131 Verge VM, Gratto KA, Karchewski LA, Richardson PM: Neurotrophins and nerve injury in the adult. Philos Trans R Soc Lond B Biol Sci 1996;351:423–430. 132 Lewin GR, Ritter AM, Mendell LM: Nerve growth factor-induced hyperalgesia in the neonatal and adult rat. J Neurosci 1993;13:2136–2148. 133 Apfel SC: Neurotrophic factors and pain. Clin J Pain 2000;16(suppl 2):S7–S11. 134 Boucher TJ, McMahon SB: Neurotrophic factors and neuropathic pain. Curr Opin Pharmacol 2001;1:66–72. 135 Bennett DL, Michael GJ, Ramachandran N, Munson JB, Averill S, Yan Q, McMahon SB, Priestley JV: A distinct subgroup of small DRG cells express GDNF receptor components and GDNF is protective for these neurons after nerve injury. J Neurosci 1998;18:3059–3072. 136 Ghatei MA, Takahashi K, Suzuki Y, Gardiner J, Jones PM, Bloom SR: Distribution, molecular characterization of pituitary adenylate cyclase-activating polypeptide and its precursor encoding messenger RNA in human and rat tissues. J Endocrinol 1993;136:159–166. 137 Kausz M, Arimura A, Koves K: Distribution and somatotopical localization of pituitary adenylate cyclase activating polypeptide (PACAP) in the trigeminal ganglion of cats and rats. Ann NY Acad Sci 1998;865:529–532. 138 Harmar AJ, Arimura A, Gozes I, Journot L, Laburthe M, Pisegna JR, Rawlings SR, Robberecht P, Said SI, Sreedharan SP, Wank SA, Waschek JA: International Union of Pharmacology. XVIII. Nomenclature of receptors for vasoactive intestinal peptide and pituitary adenylate cyclaseactivating polypeptide. Pharmacol Rev 1998;50:265–270. 139 Spengler D, Waeber C, Pantaloni C, Holsboer F, Bockaert J, Seeburg PH, Journot L: Differential signal transduction by five splice variants of the PACAP receptor. Nature 1993;365:170–175. 140 Vaudry D, Gonzalez BJ, Basille M, Yon L, Fournier A, Vaudry H: Pituitary adenylate cyclaseactivating polypeptide and its receptors: From structure to functions. Pharmacol Rev 2000; 52:269–324. 141 Dickinson T, Fleetwood-Walker SM: VIP and PACAP: Very important in pain? Trends Pharmacol Sci 1999;20:324–329.

Chaudhary/Burchiel

318

142 Fristad I, Heyeraas KJ, Kvinnsland IH: Neuropeptide Y expression in the trigeminal ganglion and mandibular division of the trigeminal nerve after inferior alveolar nerve axotomy in young rats. Exp Neurol 1996;142:276–286. 143 Noguchi K, Senba E, Morita Y, Sato M, Tohyama M: Prepro-VIP and preprotachykinin mRNAs in the rat dorsal root ganglion cells following peripheral axotomy. Brain Res Mol Brain Res 1989;6:327–330. 144 Cizkova D, Lukacova N, Marsala M, Marsala J: Neuropathic pain is associated with alterations of nitric oxide synthase immunoreactivity and catalytic activity in dorsal root ganglia and spinal dorsal horn. Brain Res Bull 2002;58:161–171. 145 Luo ZD, Cizkova D: The role of nitric oxide in nociception. Curr Rev Pain 2000;4:459–466. 146 Riedel W, Neeck G: Nociception, pain, and antinociception: Current concepts. Z Rheumatol 2001; 60:404–415. 147 Palecek J, Paleckova V, Dougherty PM, Willis WD: The effect of phorbol esters on the responses of primate spinothalamic neurons to mechanical and thermal stimuli. J Neurophysiol 1994;71: 529–537. 148 Mao J, Price DD, Phillips LL, Lu J, Mayer DJ: Increases in protein kinase C gamma immunoreactivity in the spinal cord dorsal horn of rats with painful mononeuropathy. Neurosci Lett 1995; 198:75–78. 149 Lin Q, Peng YB, Willis WD: Possible role of protein kinase C in the sensitization of primate spinothalamic tract neurons. J Neurosci 1996;16:3026–3034. 150 Mao J, Price DD, Mayer DJ, Hayes RL: Pain-related increases in spinal cord membrane-bound protein kinase C following peripheral nerve injury. Brain Res 1992;588:144–149. 151 Malmberg AB, Chen C, Tonegawa S, Basbaum AI: Preserved acute pain and reduced neuropathic pain in mice lacking PKCgamma. Science 1997;278:279–283. 152 Hundle B, McMahon T, Dadgar J, Messing RO: Overexpression of epsilon-protein kinase C enhances nerve growth factor-induced phosphorylation of mitogen-activated protein kinases and neurite outgrowth. J Biol Chem 1995;270:30134–30140. 153 Khasar SG, Lin YH, Martin A, Dadgar J, McMahon T, Wang D, Hundle B, Aley KO, Isenberg W, McCarter G, Green PG, Hodge CW, Levine JD, Messing RO: A novel nociceptor signaling pathway revealed in protein kinase C epsilon mutant mice. Neuron 1999;24:253–260. 154 Eaton MJ: Emerging cell and molecular strategies for the study and treatment of painful peripheral neuropathies. J Peripher Nerv Syst 2000;5:59–74. 155 Wu CL, Garry MG, Zollo RA, Yang J: Gene therapy for the management of pain. I. Methods and strategies. Anesthesiology 2001;94:1119–1132. 156 Wu CL, Garry MG, Zollo RA, Yang J: Gene therapy for the management of pain. II. Molecular targets. Anesthesiology 2001;95:216–240. 157 Mata M, Glorioso JC, Fink DJ: Targeted gene delivery to the nervous system using herpes simplex virus vectors. Physiol Behav 2002;77:483–488. 158 Antunes Bras JM, Epstein AL, Bourgoin S, Hamon M, Cesselin F, Pohl M: Herpes simplex virus 1-mediated transfer of preproenkephalin A in rat dorsal root ganglia. J Neurochem 1998;70: 1299–1303. 159 Shimeld C, Whiteland JL, Nicholls SM, Grinfeld E, Easty DL, Gao H, Hill TJ: Immune cell infiltration and persistence in the mouse trigeminal ganglion after infection of the cornea with herpes simplex virus type 1. J Neuroimmunol 1995;61:7–16. 160 Pohl M, Braz J: Gene therapy of pain: Emerging strategies and future directions. Eur J Pharmacol 2001;429:39–48. 161 Wilson SP, Yeomans DC, Bender MA, Lu Y, Goins WF, Glorioso JC: Antihyperalgesic effects of infection with a preproenkephalin-encoding herpes virus. Proc Natl Acad Sci USA 1999;96: 3211–3216. 162 Wilson SP, Yeomans DC: Virally mediated delivery of enkephalin and other neuropeptide transgenes in experimental pain models. Ann NY Acad Sci 2002;971:515–521. 163 Hao S, Mata M, Goins W, Glorioso JC, Fink DJ: Transgene-mediated enkephalin release enhances the effect of morphine and evades tolerance to produce a sustained antiallodynic effect in neuropathic pain. Pain 2003;102:135–142.


319

164 Beutler AS, Banck MS, Bach FW, Gage FH, Porreca F, Bilsky EJ, Yaksh TL: Retrovirus-mediated expression of an artificial beta-endorphin precursor in primary fibroblasts. J Neurochem 1995; 64:475–481. 165 Finegold AA, Mannes AJ, Iadarola MJ: A paracrine paradigm for in vivo gene therapy in the central nervous system: Treatment of chronic pain. Hum Gene Ther 1999;10:1251–1257. 166 Finegold AA, Perez FM, Iadarola MJ: In vivo control of NMDA receptor transcript level in motoneurons by viral transduction of a short antisense gene. Brain Res Mol Brain Res 2001; 90:17–25. 167 Collin E, Mantelet S, Frechilla D, Pohl M, Bourgoin S, Hamon M, Cesselin F: Increased in vivo release of calcitonin gene-related peptide-like material from the spinal cord in arthritic rats. Pain 1993;54:203–211. 168 Myers KJ, Dean NM: Sensible use of antisense: How to use oligonucleotides as research tools. Trends Pharmacol Sci 2000;21:19–23. 169 McManus MT, Sharp PA: Gene silencing in mammals by small interfering RNAs. Nat Rev Genet 2002;3:737–747. 170 Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, Mello CC: Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 1998;391:806–811. 171 Hammond SM, Caudy AA, Hannon GJ: Post-transcriptional gene silencing by double-stranded RNA. Nat Rev Genet 2001;2:110–119. 172 Krichevsky AM, Kosik KS: RNAi functions in cultured mammalian neurons. Proc Natl Acad Sci USA 2002;99:11926–11929. 173 Zlokovic BV, Apuzzo ML: Cellular and molecular neurosurgery: Pathways from concept to reality. II. Vector systems and delivery methodologies for gene therapy of the central nervous system. Neurosurgery 1997;40:805–812; discussion 812–813. 174 Ishii K, Isono M, Inoue R, Hori S: Attempted gene therapy for intractable pain: Dexamethasonemediated exogenous control of beta-endorphin secretion in genetically modified cells and intrathecal transplantation. Exp Neurol 2000;166:90–98. 175 Saitoh Y, Taki T, Arita N, Ohnishi T, Hayakawa T: Analgesia induced by transplantation of encapsulated tumor cells secreting beta-endorphin. J Neurosurg 1995;82:630–634. 176 Luo D, Saltzman WM: Synthetic DNA delivery systems. Nat Biotechnol 2000;18:33–37. 177 Luo Z: Molecular dissection of pain mediators. Pain Reviews 2000;7:37–64. 178 Sawynok J: Topical and peripherally acting analgesics. Pharmacol Rev 2003;55:1–20. 179 Nabel GJ: Development of optimized vectors for gene therapy. Proc Natl Acad Sci USA 1999; 96:324–326. 180 Parks RJ, Chen L, Anton M, Sankar U, Rudnicki MA, Graham FL: A helper-dependent adenovirus vector system: Removal of helper virus by Cre-mediated excision of the viral packaging signal. Proc Natl Acad Sci USA 1996;93:13565–13570. 181 Besson JM: The neurobiology of pain. Lancet 1999;353:1610–1615. 182 Henning SW, Beste G: Loss-of-function strategies in drug target validation. Curr Drug Discov 2002;May:17–21. 183 Verma IM, Somia N: Gene therapy – Promises, problems and prospects. Nature 1997;389:239–242. 184 Huang W, Simpson RK Jr: Antisense of c-fos gene attenuates Fos expression in the spinal cord induced by unilateral constriction of the sciatic nerve in the rat. Neurosci Lett 1999;263:61–64. 185 Lee CE, Kest B, Jenab S, Inturrisi CE: Effect of supraspinal antisense oligodeoxynucleotide treatment on delta-opioid receptor mRNA levels in mice. Brain Res Mol Brain Res 1997; 48:17–22. 186 Ji RR, Zhang X, Wiesenfeld-Hallin Z, Hokfelt T: Expression of neuropeptide Y and neuropeptide Y (Y1) receptor mRNA in rat spinal cord and dorsal root ganglia following peripheral tissue inflammation. J Neurosci 1994;14:6423–6434. 187 Hua XY, Chen P, Polgar E, Nagy I, Marsala M, Phillips E, Wollaston L, Urban L, Yaksh TL, Webb M: Spinal neurokinin NK1 receptor down-regulation and antinociception: Effects of spinal NK1 receptor antisense oligonucleotides and NK1 receptor occupancy. J Neurochem 1998;70:688–698. 188 Lai J, Gold MS, Kim CS, Bian D, Ossipov MH, Hunter JC, Porreca F: Inhibition of neuropathic pain by decreased expression of the tetrodotoxin-resistant sodium channel, NaV1.8. Pain 2002; 95:143–152.

Chaudhary/Burchiel

320

189

190 191

192

193 194

Porreca F, Lai J, Bian D, Wegert S, Ossipov MH, Eglen RM, Kassotakis L, Novakovic S, Rabert DK, Sangameswaran L, Hunter JC: A comparison of the potential role of the tetrodotoxin-insensitive sodium channels, PN3/SNS and NaN/SNS2, in rat models of chronic pain. Proc Natl Acad Sci USA 1999;96:7640–7644. Khasar SG, Gold MS, Levine JD: A tetrodotoxin-resistant sodium current mediates inflammatory pain in the rat. Neurosci Lett 1998;256:17–20. Levesque L, Dean NM, Sasmor H, Crooke ST: Antisense oligonucleotides targeting human protein kinase C-alpha inhibit phorbol ester-induced reduction of bradykinin-evoked calcium mobilization in A549 cells. Mol Pharmacol 1997;51:209–216. Rydh-Rinder M, Berge OG, Hokfelt T: Antinociceptive effects after intrathecal administration of phosphodiester-, 2⬘-O-allyl-, and C-5-propyne-modified antisense oligodeoxynucleotides targeting the NMDAR1 subunit in mouse. Brain Res Mol Brain Res 2001;86:23–33. Garry MG, Malik S, Yu J, Davis MA, Yang J: Knock down of spinal NMDA receptors reduces NMDA and formalin evoked behaviors in rat. Neuroreport 2000;11:49–55. Tao F, Tao YX, Gonzalez JA, Fang M, Mao P, Johns RA: Knockdown of PSD-95/SAP90 delays the development of neuropathic pain in rats. Neuroreport 2001;12:3251–3255.

Kim Burchiel, MD Chairman, Department of Neurological Surgery, L 472 Oregon Health & Science University, 3181 SW Sam Jackson Park Road Portland, OR 97239 (USA) Tel. ⫹1 503 494 4173, Fax ⫹1 503 494 7161, E-Mail [email protected]


321


Gene Transfer in the Treatment of Pain David Fink a, Marina Mataa, Joseph C. Gloriosob a

Department of Neurology, University of Michigan and VA Ann Arbor Healthcare System, Ann Arbor, Mich., and bDepartment of Molecular Genetics and Biochemistry, University of Pittsburgh, Pittsburgh, Pa., USA

Introduction

The Physiology of Pain Pain is an unpleasant sensory and affective experience that serves an essential biological role in alerting an organism to tissue damage. The perception of acute pain is essential for survival in a potentially hostile environment, and an elaborate set of specialized high threshold sensory transducers, nociceptors that respond to painful heat, cold, pressure, and alterations in the peripheral microenvironment are designed to detect these acute stimuli. In the setting of chronic tissue damage, changes in pH and ionic composition of the peripheral microenvironment and the release of bioactive peptides such as cytokines, growth factors, and kinins, all act to sensitize peripheral nociceptors. In addition, continued neural transmission through pain pathways leads to central changes at the level of the spinal cord and higher centers that together result in a heightened pain experience. While these changes may serve an adaptive role in preventing the use of an injured body part, thus promoting recovery and repair, the same processes lead to spontaneous or exaggerated pain that does not serve any functional biological purpose. Pain that persists beyond the course of the acute insult or pain that accompanies a chronic primary process that cannot be cured represents a major medical and social problem resulting in enormous cost to the individual and to society. Such chronic pain may result from continued peripheral injury (inflammatory or ‘nociceptive’ pain) or from damage to neural structures in the absence of peripheral tissue damage (neuropathic or central pain). Derivatives of the active agents extracted from the poppy seed (opiate drugs) and willow bark (nonsteroidal anti-inflammatory drugs) remain among our most effective and widely used analgesic drugs, but increased understanding of

the pathways involved in acute pain perception and the modifications that occur in chronic pain have now set the stage for the rational design of novel therapeutic agents to treat chronic pain. Peripheral nociceptors consisting of cells with unmyelinated (C-fibers) or thinly myelinated (A␦ fibers) axons represent a subclass of neurons whose cell bodies are located in the dorsal root ganglion (DRG). The central projection of the bipolar axons of the primary nociceptors synapse on ‘second order’ neurons in the dorsal horn of spinal cord in a regional and anatomically defined manner. Second order neurons located in the dorsal horn project rostrally to the thalamus and the parabrachial nucleus in the brainstem. Pain-related neurons in the thalamus project primarily to somatosensory cortex, conveying the discriminative aspects of the pain sensation; neurons in the parabrachial nucleus project to the hippocampus and amygdala (among other brain regions) to mediate the affective components of the pain experience. Descending pathways, integrated in the periaqueductal gray of the midbrain and relayed through the nucleus raphe magnus, project caudally to the dorsal horn of spinal cord to synapse with inhibitory interneurons. The principal neurotransmitter released from axons of the primary nociceptor at the dorsal horn is glutamate, although corelease of peptides including substance P, neuropeptide Y, dynorphin and galanin from the primary afferent serve to modulate the pain response. In the dorsal horn, intrinsic inhibitory interneurons may modulate the transmission of nociceptive information through the release of neurotransmitters such as ␥-aminobutyric acid (GABA) acting on GABAA and GABAB receptors, enkephalin acting through ␦ opioid receptors, and endomorphin 1 and 2 acting at the ␮ opioid receptor. Descending projections from brainstem nuclei control the activity of the inhibitory interneurons through the release of monoamine neurotransmitters including norepinephrine and serotonin. In states of chronic pain, there are post-translational or transcriptional changes in primary nociceptors that alter the threshold, excitability, or transmission properties of these neurons. A prolonged increase in the activity of peripheral nociceptors also results in the sensitization of second order neurons, through phosphorylation of ion channels and receptors as well as transcriptional changes in second order neurons. In addition, there may be remodeling in the dorsal horn, including sprouting of primary afferents after injury and loss of inhibitory interneurons. Imaging studies have also identified complex central changes in the activity of subcortical and cortical structures that occur in states characterized by chronic pain. The most direct approach to treating pain is to alleviate the primary inciting cause of the pain. But when the primary disease process cannot be cured, or the pain persists after the identified inciting cause has been treated, other approaches are required. Many of the neurotransmitters or neurotransmitter receptors that are the targets of pharmacological therapy are widely distributed

Gene Transfer in the Treatment of Pain

323

through the nervous system and may be present in other organs as well. As a result, the dose of many of the drugs that may be used to treat pain are limited by side effects resulting from the action of these agents on neural pathways unrelated to pain processing, or from the action of these agents directly on non-neural tissues. Opioid receptors, for example, are present on peripheral nociceptors, on the second order projection neurons in the dorsal horn of spinal cord, and in brainstem and brain centers involved in pain processing and other cognitive and affective functions as well as in the urinary bladder and the intestine. As a result, high-dose treatment with opiate drugs may be complicated by alterations in mood and/or cognition, urinary retention, and constipation that limit the dose that may be administered. Gene transfer offers the possibility to achieve local release of analgesic substances to act at the spinal or peripheral level to maximize the analgesic effectiveness while minimizing side effects.

Cell Transplantation for Pain Relief

One method of gene transfer involves the transplantation of cells that carry and express the gene of interest. Chromaffin cells of the adrenal medulla naturally express and release a number of neuroactive substances, many of which are involved in the pain-processing pathway at the spinal level. These include serotonin, GABA, galanin, and met-enkephalin [1, 2]. Accordingly, the injection of chromaffin cells into the lumbar subarachnoid space reduces pain-related behavior in models of neuropathic pain, and in rodent models, inflammatory pain induced by subcutaneous injection of a dilute solution of formalin [3, 4]. Such grafts reduce touch-induced elevation of c-fos expression in spinal cord [5] and prevent the loss of endogenous GABA synthesis in the dorsal horn [6]. Several different mechanisms have been implicated in the analgesic effect of chromaffin cell grafts. Chromaffin cell grafts release met-enkephalin, and the level of met-enkephalin in CSF is increased following grafting [7], but the grafts also elevate CSF catecholamine levels [8]. It is possible that indirect effects, such as catecholamine stimulation by the release of inhibitory neurotransmitters from dorsal horn interneurons might account for some of the effects of the graft. In a neurophysiological study Hentall et al. [9] demonstrated that the intrathecal transplantation of chromaffin cells prevented the normal development of ‘windup’, a phenomenon of electrophysiological potentiation that is characteristic of chronic pain, in second-order wide dynamic range neurons of the dorsal horn. It appears from those studies that interference with potentiation is due to the release of molecules that persistently block the NMDA receptor (or block cellular events mediated by these receptors), separate from the possible effect of inhibitory neurotransmitters from the graft [9].

Fink/Mata/Glorioso

324

Cells also can be modified to produce specific desired gene products. Wu et al. [10] showed that AtT-20 cells, a cell line which produces and releases ␤-endorphin, and AtT-20/hENK cells, an AtT-20 cell line transfected with the human proenkephalin gene (PE) and secreting enkephalin and ␤-endorphin, implanted into the mouse subarachnoid space produced an isoproterenolstimulated anti-nociceptive effect that was dose related and could be blocked by naloxone. Mice receiving AtT-20 cell implants developed tolerance to ␤-endorphin and the ␮-opioid agonist DAMGO, whereas mice receiving genetically modified AtT-20/hENK cell implants developed tolerance to the ␦-opioid agonist DPDPE. Genetically modified AtT-20/hENK cell implants, but not AtT-20 cell implants, reduced the development of acute morphine tolerance in the host mice [10]. Transplants of other cell lines modified to secrete substances that might act as inhibitory neurotransmitters at the spinal level to block nociceptive neurotransmission have included the demonstration that a neuronal cell line genetically modified to secrete galanin [11] or engineered to secrete GABA [12] are anti-nociceptive in models of chronic neuropathic pain. A cell line engineered to produce and release brain-derived neurotrophic factor has been shown to reduce allodynia and hyperalgesia in the chronic constriction injury model of neuropathic pain [13]. The mechanical alternative to cell transplantation is peptide delivery using an intrathecal pump. Both approaches target the pharmcological agent to the lumbar spinal cord in order to minimize the effect of these agents on the brain and on peripheral organs. A theoretical advantage of cell transplantation is the ability of the cells to deliver peptide neurotransmitter (such as enkephalin) in their natural conformation and not in a derivative state. Intrathecal pumps may be used to deliver opioid analgesic drugs such as morphine. Intrathecal delivery reduces the dose requirement and thus limits side effects, but tolerance may develop. Even intrathecal administration of the modified derivatives is limited by the very short half-life of these agents. While cell transplants produce a continuous release of the native peptide, release of additional substances (e.g., ␤-endorphin, serotonin) which may add to the effectiveness of these transplants in the animal models may complicate the practical implementation of the cells as agents for human treatment. In addition, the possibility of an immune reaction or nonspecific scar formation resulting from the injection of foreign material into the subarachnoid space is not fully known. Nonetheless, transplantation of encapsulated bovine chromaffin into the subarachnoid space has been tested in patients with severe chronic pain not satisfactorily managed with conventional therapies. The patients received no pharmacological immunosuppression. Histological analysis, immunostaining, and analysis of secretory function all confirmed survival and biochemical function of the encapsulated cells up to 6 months after


325

implantation. Reductions in morphine intake and improvement in pain ratings were observed in several patients [14]. Similar results were observed in an open phase II trial of patients with intractable pain from cancer who received adrenal medullary allografts [15]. While analgesic efficacy was suggested by a reduction or stabilization in opioid use, a controlled trial has not yet been reported.

Adenoviral Gene Transfer to Meningeal Cells in the Treatment of Pain

Some of the problems attendant on injection of foreign cells into the subarachnoid space can be avoided by the direct transfer of the gene of interest into meningeal cells lining the subarachnoid space. For example, Finegold et al. [16] used a first-generation replication deficient adenoviral (Ad5) vector containing the coding sequence for human ␤-endorphin to transduce cells of the pia mater. Ad5-mediated gene transfer resulted in the release of ␤-endorphin into the CSF and attenuated inflammatory hyperalgesia, measured as the thermal withdrawal latency in the carageenen model of inflammatory pain in the rodent, without affecting basal nociceptive responses. Although the inflammatory response elicited by the first-generation adenoviral vectors employed in this study limited the duration of transgene expression, a similar approach using later-generation vectors might be appropriate for patients with severe refractory pain in the terminal stages of a disease process.

Herpes-Mediated Gene Transfer in the Treatment of Pain

An alternative but related approach uses vector-mediated gene transfer to transduce neurons, rather than meningeal cells. For this purpose, herpes simplex virus (HSV) has proven to be a useful gene transfer vector. HSV is a neurotropic virus that naturally infects skin and mucous membranes. Following the initial epithelial infection, HSV is taken up by nerve terminals in the skin and carried by retrograde axonal transport to the cell bodies of sensory neurons in the DRG where the viral DNA is inserted through a nuclear pore into the nucleus. The uptake and transport of the virion from the skin is an efficient process mediated first by interactions between specific viral envelope glycoproteins and high-affinity receptors in the sensory nerve terminals in the skin [17, 18], followed by specific interactions of capsid and tegument proteins with dynein molecules in the axoplasm to mediate the retrograde axonal transport along microtubules to the cell body [19]. The highly efficient delivery of viral DNA to the DRG neuronal nucleus from an original infection in the skin coupled

Fink/Mata/Glorioso

326

with the natural ability of the viral genome to establish a life-long latent state as an intranuclear episomal element makes HSV a very effective gene transfer vector for the peripheral nervous system. Pohl et al. demonstrated that an HSV-based vector in which the HSV thymidine kinase sequence is replaced by the cDNA coding for human proenkephalin, injected under the skin in the paw of a rat, transduced DRG neurons to produce enkephalin [20]. Immunoreactive met-enkephalin is transported anterogradely (i.e., away from the cell body) in both directions in the bipolar axon from DRG neurons, towards the spinal cord and back towards the skin with a larger amount moving peripherally than centrally [21]. Electrically stimulated release of metenkephalin from nerve terminals can be demonstrated in an in vitro preparation [Pohl, pers. commun.]. Wilson et al. [22] then showed that subcutaneous inoculation of a similar tk-deleted HSV vector expressing PE reduces hyperalgesia measured by the sensitization of the foot withdrawal response after application of capsaicin (C fibers) or dimethyl sulfoxide (A␦ fibers). The effect of the vector persisted for at least 7 weeks after the inoculation of the vector subcutaneously into the dorsum of the foot. Baseline foot withdrawal responses to noxious radiant heat mediated by A␦ and C fibers were similar in animals infected with PE-encoding and ␤-galactosidase-encoding vectors, demonstrating that the PE-expressing vector selectively blocked hyperalgesia without disrupting the baseline sensory neurotransmission. This blockade of sensitization was reversed by the administration of the opioid antagonist naloxone, apparently acting in the spinal cord [22]. Deletion of HSV-TK impairs the ability of the virus to replicate in neurons while leaving replication characteristics in non-neuronal cells intact. HSV gene expression occurs in a rigid temporal sequence and only five of the more than eighty HSV genes that are expressed during the lytic replication cycle are characterized as ‘immediate early’ (IE) genes. The expression of IE genes begins immediately after the viral entry into the nucleus, activated by a viral protein (VP16) contained in the tegument, and does not require the de novo expression of other viral proteins. Deletion of even one essential IE gene from the HSV genome creates a recombinant that can be propagated in a complementing cell line that provides the essential IE gene product in trans. These IE gene-deleted HSV vectors are incapable of replication in noncomplementing cells [23]. Introduced into animals, such replication-incompetent vectors do not replicate, but instead traffic to DRG neurons and establish a persistent state in a fashion identical to that observed for the replication-competent recombinants. We have examined the pain-relieving properties of a replication-incompetent HSV vectorexpressing human PE in rodent models of inflammatory pain, neuropathic pain, and pain resulting from cancer in bone.


327

Cumulative pain score

10

8

*

6

** **

4

2 SHZ

7d

14 d

28 d

28 d ⫹7 d

IT naltrexone

SHPE

Fig. 1. Time course of the anti-nociceptive effect of SHPE inoculation in inflammatory pain (formalin test). The cumulative pain score during the delayed phase of the formalin test (10 min to one hour after the inoculation of formalin) was significantly reduced in animals inoculated with the PE-expressing vector SHPE. The vector-mediated effect was maximal one week after the inoculation of the vector, and was no longer statistically significant in animals tested 4 weeks after vector inoculation. However reinoculation of the vector at 4 weeks reestablished the analgesic effect (28 ⫹ 7 day group). Intrathecal administration of the ␦ opioid receptor antagonist naltrexone in animals tested one week after vector inoculation (IT naltrexone) blocked the vector-mediated analgesic effect. From [31].

Injected under the skin of the foot, the PE-expressing HSV vector was detected in the DRG by PCR using primers specific for the human PE sequence, and the expression of PE mRNA was detected by RT-PCR using the same sequences [24]. In the formalin test of inflammatory pain, injection of the PE-expresssing HSV vector reduced spontaneous pain behavior during the delayed phase (10–60 min after the injection of formalin) without affecting the acute pain score. This effect was reversed by the intrathecal administration of naltrexone [24], suggesting in agreement with Wilson et al. that the site of action of the released transgene product is in the dorsal horn of spinal cord. The analgesic effect was limited to the injected limb; formalin testing on the limb contralateral to the injection showed no analgesic effect [Glorioso, unpubl. observations], further suggesting that release of enkephalin from primary afferent terminals in the dorsal horn was limited to the region of their projections in dorsal horn. HSV vector-mediated analgesic effects persisted for several weeks and then waned. Animals tested 4 weeks after vector inoculation showed no significant reduction in pain-related behavior during the delayed phase of the formalin

Fink/Mata/Glorioso

328

SHPE 12

Reinoculation

Threshold (gms)

Inoculation

**

10

*

8

*

6

*

**

SHZ

Vehicle

** ** **

*

4 2

0

5

10

15

Weeks after spinal nerve ligation

Fig. 2. Time course of the anti-allodynic effect of SHPE inoculation in neuropathic pain (spinal nerve ligation model). Injection of SHPE (but not control vector SHZ) resulted in a sustained anti-allodynic effect in neuropathic pain that lasted for several weeks. Reinoculation of the vector 6 weeks after the initial inoculation re-established the antiallodynic effect. The vector-mediated pain-relieving effect was reversed by intraperitoneal administration of naloxone (data not shown). From [26].

test [24]. However, in animals reinoculated with the vector 4 weeks after the initial inoculation and then tested with formalin injection at 5 weeks, a substantial and significant anti-nociceptive effect was demonstrated, which was at least as great as the initial effect [24]. The time course of the vector-mediated effect is consistent with the known time course of the human cytomegalovirus immediate-early promoter that was used to drive transgene expression in these experiments. The fact that the reinjection could re-establish the initial analgesic effect suggests, but does not prove, that the animals had not developed tolerance to the vector-mediated release of enkephalin. The effectiveness of reinoculation also suggests that the exposure of animals to the HSV vector does not elicit a neutralizing immune response that would be capable of attenuating gene transfer from the vector inoculation. We have also examined the effect of transgene-mediated enkephalin release in the spinal nerve ligation model of neuropathic pain [25]. Isolated L5 spinal nerve ligation distal to the DRG results in a painful state that can be quantified by measures of mechanical and thermal hypersensitivity. Subcutaneous injection of the PE-expressing vector into the foot one week after spinal nerve ligation resulted in an anti-allodynic effect that lasted for several weeks [26]. The characteristics of this anti-allodynic effect were similar to those observed in the formalin model. Reinoculation of the vector at 6 weeks


329

Sham

Vehicle

SHZ

SHPE

a Ligation

60

fos-LI neurons

50

Sham Vehicle SHZ SHPE

40

*

30 20

**

*

**

10 0 I-II

III-IV

V-VI

VIII-X

Laminae

b

Fig. 3. Transduction with SHPE blocks touch-induced elevation in c-fos expression in the dorsal horn. Animals with neuropathic pain from spinal nerve ligation stimulated with gentle rubbing of the paw show a characteristic increase in c-fos expression in dorsal horn neurons seen in vehicle-treated animals (a, top right panel). Subcutaneous inoculation of SHPE one week after spinal nerve ligation substantially blocked this induction in c-fos expression (a, bottom right panel). The quantitative data are shown in b. From [26].

re-established the anti-allodynic effect; the magnitude of the effect produced by reinoculation was at least as great as that produced by the initial injection and the effect persisted for a longer time after the reinoculation than that produced by the initial inoculation. Intraperitoneal naloxone reversed the anti-allodynic effect.

Fink/Mata/Glorioso

330

Spontaneous ambulatory pain score

3 2.5

2 1.5

*

1

0.5 (11)

(10)

(9)

(4)

SHPE

SHZ

SHPE ⫹ NTX

0

tumor

Fig. 4. Transduction with SHPE reduces pain-related behavior in a model of pain resulting from cancer in bone. Rats with experimental osteogenic sarcoma of the distal femur (shown in right part of radiograph) demonstrate a spontaneous pain behavior that is reduced significantly in those animals that were inoculated subcutaneously in the foot with SHPE one week after tumor implantation. The analgesic effect of the vector was reversed by intrathecal naltrexone. From [28].

The effect of this vector-mediated anti-allodynic effect in neuropathic pain was continuous throughout the day. Animals tested repeatedly at different times through the day showed a similar elevation in threshold at all times tested [26]. Intraperitoneal morphine produced a greater anti-allodynic effect than the vector alone, but the inoculation of the maximum dose of morphine produced an anti-allodynic effect that persisted for only 1–2 h before waning. The effect of vector-mediated enkephalin (acting predominantly at ␦ opioid receptors) and morphine (acting predominantly at ␮ opioid receptors) was additive. The ED50 of morphine was shifted from 1.8 ␮g/kg in animals with neuropathic pain from spinal nerve ligation treated with PBS or inoculated with a control vector


331

expressing lacZ, to 0.15 ␮g/kg in animals that had been injected with the PE-expressing vector one week after spinal nerve ligation. Twice daily inoculation of morphine (10 mg/kg IP) in spinal nerve ligated animals resulted in the development of tolerance by one week; beyond that timepoint, the continued twice-daily administration of morphine had no anti-allodynic effect. Animals that had been inoculated with the PE-expressing vector one week after spinal nerve ligation continued to demonstrate the anti-allodynic effect of the vector despite the induction of tolerance to morphine [26]. We also examined the effect of vector-mediated expression of PE in a rodent model of pain resulting from cancer in bone [27]. Implantation of NTCT 2,472 cells into the distal femur resulted in a significant spontaneous pain-related behavior that increased between 2 and 3 weeks after tumor injection. Animals that received a subcutaneous inoculation of the PE-expressing vector into the plantar surface of the foot one week after tumor injection showed a significant reduction in the ambulatory pain score when compared to control vector-inoculated tumor-bearing animals at both 2 and 3 weeks after tumor injection, an effect that was reversed by intrathecal naltrexone [28]. Radiographical analysis of tumor-bearing mice inoculated with SHZ or SHPE demonstrated bone loss indicative of the presence of the osteolytic tumor, and there was no evidence that transgene expression had any effect on tumor growth [28]. Similar effects have been demonstrated in adjuvant-induced polyarthritis in the rat [29]. Using a replication-competent HSV vector expressing PE, Pohl et al. showed that subcutaneous inoculation of the vector not only markedly improved the locomotion and reduced hyperalgesia, but also that the release of enkephalin from the peripheral terminals of the DRG axons resulted in a slowing of the progression of bone destruction. In that model both the slowing of joint destruction as well as reversal of the analgesic effect by a peripherally acting substituted naloxone analog suggest that the site of action of transgenemediated enkephalin released from transduced neurons is at the peripheral projection rather than in the spinal cord. However, the effect on joint destruction appears to be unique to the model of arthritis employed [Pohl, pers. commun.]. Whether this approach will be effective in the treatment of human pain should be determined quite soon. A proposal for a phase I human trial to examine the safety and tolerability of subcutaneous inoculation of an HSV vector deleted for the IE genes ICP4, ICP22, ICP27 and ICP41, and expressing human PE under the control of the human cytomegalovirus immediate-early promoter was presented to the Recombinant Advisory Committee at the NIH (for details, see RAC Protocol #0201–529 at http://www.webconferences.com/nihoba/ 20–21_june_02.htm). The study protocol describes the enrollment of 18 patients with cancer metastatic to a vertebral body resulting in pain unresponsive to maximal conventional management, who will receive an inoculation of the

Fink/Mata/Glorioso

332

HSV vector subcutaneously in the dermatome corresponding to the radicular distribution of the pain. In this dose-escalation trial, 3 patients will be enrolled at each dose, increasing at half-log intervals. Pain treatment by transduction of DRG neurons using an HSV vector will be limited to pain syndromes that result in regional or focal pain. The noninvasive means of delivery of the vector, the focal effect of transgene expression, and the synergy with opiate drug treatment are chief advantages of this approach. However, once the transgene is introduced into the DRG neuron it cannot be removed, and using the current vectors, transgene expression is not regulated. This should not prove a problem for the use of enkephalin, and the transgene expression driven by the human cytomegalovirus promoter is transient. But in the use of other transgenes, or promoters designed to drive long term gene expression, these issues will resume further consideration and vector engineering. The fact that these patients are facing a fatal course of the disease with severe pain that is frequently not responsive even to high-dose systemic opioids strengthens the rationale for local, targeted gene transfer for pain relief.

Gene Transfer to the Brain in Models of Pain

There are other sites within the neuraxis where pain transmission may be interrupted by vector-mediated neurotransmitter expression. Injection of a replication-competent (tk-deleted) HSV vector expressing proenkepahlin into the amygdala bilaterally has been shown to reduce pain-related behavior in the delayed phase of the formalin test (animals tested 4 days after vector inoculation) [30], and injection of an HSV amplicon vector expressing glutamic acid decarboxylase to result in the release GABA in brain nuclei also reduces painrelated behaviors [Jasmin and Rabkin, pers. commun.]. While these experiments demonstrate that focal neurotransmitter effects can be achieved by gene transfer to the brain as well as the peripheral nervous system, applications to human clinical therapies are likely to take longer to develop, given the complicated neuropharmacology of CNS function.

Conclusion

Cell transplantation or vector-mediated gene transfer, by providing a means to target expression focally in the nervous system, may allow the use of short-lived macromolecules identical to the endogenous substances to enhance pain relief in specific situations. In the future, other macromolecules acting to interrupt the process responsible for the development of the pain


333

(such as anti-inflammatory cytokines or specific neurotrophic factors in neuropathic pain) may be delivered in a similar fashion.

Acknowledgments This work was supported by grants from the NIH (JCG and DJF), the Department of Veterans Affairs (MM and DJF), and the Juvenile Diabetes Foundation Research International (DJF).

References 1 2 3 4 5 6

7 8 9 10

11 12 13

14 15 16 17 18

Wilson SP, Chang KJ, Viveros OH: Opioid peptide synthesis in bovine and human adrenal chromaffin cells. Peptides 1981;2(suppl 1):83–88. Unsicker K: The trophic cocktail made by adrenal chromaffin cells. Exp Neurol 1993;123:167–173. Hama AT, Sagen J: Alleviation of neuropathic pain symptoms by xenogeneic chromaffin cell grafts in the spinal subarachnoid space. Brain Res 1994;651:183–193. Siegan JB, Sagen J: Attenuation of formalin pain responses in the rat by adrenal medullary transplants in the spinal subarachnoid space. Pain 1997;70:279–285. Sagen J, Wang H: Adrenal medullary grafts suppress c-fos induction in spinal neurons of arthritic rats. Neurosci Lett 1995;192:181–184. Ibuki T, et al: Loss of GABA-immunoreactivity in the spinal dorsal horn of rats with peripheral nerve injury and promotion of recovery by adrenal medullary grafts. Neuroscience 1997;76: 845–858. Sagen J, Kemmler JE: Increased levels of Met-enkephalin-like immunoreactivity in the spinal cord CSF of rats with adrenal medullary transplants. Brain Res 1989;502:1–10. Sagen J, Kemmler JE, Wang H: Adrenal medullary transplants increase spinal cord cerebrospinal fluid catecholamine levels and reduce pain sensitivity. J Neurochem 1991;56:623–627. Hentall ID, Noga BR, Sagen J: Spinal allografts of adrenal medulla block nociceptive facilitation in the dorsal horn. J Neurophysiol 2001;85:1788–1792. Wu HH, Wilcox GL, McLoon SC: Implantation of AtT-20 or genetically modified AtT-20/hENK cells in mouse spinal cord induced antinociception and opioid tolerance. J Neurosci 1994;14: 4806–4814. Eaton MJ, et al: Lumbar transplant of neurons genetically modified to secrete galanin reverse pain-like behaviors after partial sciatic nerve injury. J Peripher Nerv Syst 1999;4:245–257. Eaton MJ, et al: Transplants of neuronal cells bioengineered to synthesize GABA alleviate chronic neuropathic pain. Cell Transplant 1999;8:87–101. Cejas PJ, et al: Lumbar transplant of neurons genetically modified to secrete brain-derived neurotrophic factor attenuates allodynia and hyperalgesia after sciatic nerve constriction Pain. 2000; 86:195–210. Buchser E, et al: Immunoisolated xenogenic chromaffin cell therapy for chronic pain. Initial clinical experience. Anesthesiology 1996;85:1005–1012; discussion A29–A30. Lazorthes Y, et al: Human chromaffin cell graft into the CSF for cancer pain management: A prospective phase II clinical study. Pain 2000;87:19–32. Finegold AA, Mannes AJ, Iadarola MJ: A paracrine paradigm for in vivo gene therapy in the central nervous system: treatment of chronic pain. Hum Gene Ther 1999;10:1251–1257. Spear PG, Eisenberg RJ, Cohen GH: Three classes of cell surface receptors for alphaherpesvirus entry. Virology 2000;275:1–8. Shukla D, Spear PG: Herpesviruses and heparan sulfate: An intimate relationship in aid of viral entry. J Clin Invest 2001;108:503–510.

Fink/Mata/Glorioso

334

19 20 21

22 23

24 25 26 27 28 29 30 31

Smith GA, Enquist, LW: BREAK INS AND BREAK OUTS: Viral interactions with the cytoskeleton of mammalian cells. Annu Rev Cell Dev Biol 2002;18:135–161. Antunes Bras JM, et al: Herpes simplex virus 1-mediated transfer of preproenkephalin A in rat dorsal root ganglia. J Neurochem 1998;70:1299–1303. Antunes Bras J, et al: Met-enkephalin is preferentially transported into the peripheral processes of primary afferent fibres in both control and HSV1-driven proenkephalin A overexpressing rats. Neuroscience 2001;103:1073–1083. Wilson SP, et al: Antihyperalgesic effects of infection with a preproenkephalin-encoding herpes virus. Proc Natl Acad Sci USA 1999;96:3211–3216. DeLuca NA, McCarthy AM, Schaffer PA: Isolation and characterization of deletion mutants of herpes simplex virus type 1 in the gene encoding immediate-early regulatory protein ICP4. J Virol 1985;56:558–570. Goss JR, et al: Antinociceptive effect of a genomic herpes simplex virus-based vector expressing human proenkephalin in rat dorsal root ganglion. Gene Ther 2001;8:551–556. Kim SH, Chung JM: An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain 1992;50:355–363. Hao S, et al: Transgene-mediated enkephalin release enhances the effect of morphine and evades tolerance to produce a sustained antiallodynic effect. Pain 2003;102:135–142. Schwei MJ, et al: Neurochemical and cellular reorganization of the spinal cord in a murine model of bone cancer pain. J Neurosci 1999;19:10886–10897. Goss JR, et al: Herpes vector-mediated expression of proenkephalin reduces pain-related behavior in a model of bone cancer. pain. Ann Neurol 2002;52:662–665. Braz J, et al: Therapeutic efficacy in experimental polyarthritis of viral-driven enkephalin overproduction in sensory neurons. J Neurosci 2001;21:7881–7888. Kang W, et al: Herpes virus-mediated preproenkephalin gene transfer to the amygdala is antinociceptive. Brain Res 1998;792:133–135. Chen X, et al: Herpes simplex virus type 1 ICP0 protein does not accumulate in the nucleus of primary neurons in culture. J Virol 2000;74:10132–10141.

David Fink, MD 1914 Taubman Center/0316, 1500 E Medical Center Drive Ann Arbor, MI 48109-0316 USA Tel. ⫹1 734 936 9070, Fax ⫹1 734 763 5059, E-Mail [email protected]


335

Neurovascular Disorders Freese A, Simeone FA, Leone P, Janson C (eds): Principles of Molecular Neurosurgery. Prog Neurol Surg. Basel, Karger, 2005, vol 18, pp 336–376

Gene Discovery Underlying Stroke Frank C. Baronea, Simon J. Read b a

High Throughput Biology, GlaxoSmithKline, King of Prussia, Pa., USA, RIRA, Astra Zeneca, Mereside, Alderley Park, Macclesfield, Cheshire, UK

b

Introduction

As first-pass efforts to complete sequencing of the entire human genome are now concluded [1], there is an increased interest in the application of genomic approaches to aid in the discovery, development, and rationale use of drugs. As databases of differential gene expression have expanded, so has the expectation of identifying novel drug targets for disease intervention. Indeed, significant work has already been carried out to understand gene expression changes in many diseased organ systems, including the ischemic brain [2–6]. Early epidemiological studies of the 1970s provided initial evidence for a genetic influence in stroke. The Framingham study was one of the first studies to suggest that a positive parental history of stroke contributed significant risk to the offspring [7]. Thirty years later, stroke remains an area of substantial unmet medical need. The complexity of stroke undoubtedly reflects the heterogeneity of the human stroke population, the contribution of monogenic and polygenic disorders to this disease process, and the interactions of these with a multitude of environmental factors. This chapter focuses on genetics of risk and sensitivity to ischemic stroke. It will discuss how inheritance relates to the broader stroke population and provide a detailed discussion of the stroke genomics literature. It will describe how pre-clinical models of spontaneous stroke can be applied to humans to identify the chromosomal loci of risk, and how the changes in gene expression associated with stroke are associated with poststroke brain injury, resolution of brain injury, and brain recovery processes. In addition, it will provide a detailed discussion of several differential gene expression analyses techniques. This will include a detailed discussion of genes identified using different techniques and the importance of a stroke model that has been well characterized and representative of the

type of stroke most often observed in man. Issues of validation of potential stroke targets, the relevance of the expression of neuroprotective and neurodestructive genes and their specific timings, genes involved in endogenous brain protection and in brain recovery of function/plasticity, and the emerging problems with handling novel/unknown genes that may be discovered from these analyses of differential gene expression also will be addressed.

Ischemic Stroke

Stroke is the third largest cause of death in the USA, ranking only behind heart disease and cancer. It is the leading cause of disability in the USA and has the highest disease burden cost. Ischemic strokes comprise the majority of strokes, between 70–80% of all strokes. No medical treatment is approved for the treatment of acute ischemic stroke other than thrombolytic agents such as tPA, which for optimum results must be administered within 3 h after stroke onset. At most centers, only 1–2% of the stroke patients meet the criteria for treatment with this thrombolytic agent. Aspirin and anticoagulants (where embolic phenomena are documented) also are utilized as preventative therapy. Estimates indicate that there are about 775,000 new stroke cases per year in the USA, with a prevalence of about 4 million surviving, but at an increased risk of a secondary cardiovascular event. In the USA, stroke is costly, with an annual health care cost of $30–50 billion. Estimates indicate that stroke is responsible for half of all the patients hospitalized for acute neurological disease [8, 9]. Stroke risk factors include both genetic and environmental factors. Stroke risk factors that can be treated include high blood pressure, heart disease, cigarette smoking, transient ischemic attacks, and high red blood cell count. Risk factors for stroke that cannot be changed include age, gender (men have ⬃20% greater risk of stroke than women), race (African-Americans have a much higher risk of death and disability from stroke), diabetes mellitus, prior stroke, and family history of strokes. Other controllable risk factors, secondary risk factors, for stroke that also contribute to heart disease include high blood LDL-cholesterol and lipids, physical inactivity, and obesity [10].

Genetics of Increased Stroke Risk

The strongest evidence for a genetic risk to stroke comes from twin studies. Proband concordance rates have long been used to identify the heritability of a trait or disorder. The concept of concordance is that for a disorder of genetic predisposition, the rate will be higher for monozygotic twins than dizygotic twins.

Gene Expression in Stroke

337

Aside from genetic influence, it is assumed that other factors, such as environmental exposure will be approximately similar for both types of twins [11]. An elevated proband-wise concordance rate for stroke risk in monozygotic twins over dizygotic twins (17.7 vs. 3.6%) has confirmed a genetic predisposition to stroke with a role of environmental factors [12]. A more recent twin study has refined cohort analysis to stroke risk by assessing individual stroke phenotypes that may be influenced by genetic factors [13]. In this study, the phenotype of white matter hyperintensity volumes using magnetic resonance imaging (MRI) was applied and genetic factors accounted for 71% of the variation in this endpoint [13]. A large number of familial studies have verified that a history of paternal or maternal stroke is associated with an occurrence of stroke in offspring, and that a positive paternal history of stroke was an independent prognostic predictor of stroke [14–16]. In a cohort of men studied since 1913, maternal history of stroke increased relative stroke risk by 3-fold [17]. Similarly, it has been reported that a positive family history of stroke in any first-degree relative was an independent predictor of stroke mortality in women aged 50–79, but not in men [18]. Moreover, a family history of stroke in men was an independent predictor of coronary heart disease aged 50–64 years, indicating that genetic risk factors for stroke may be shared with other cardiovascular disorders that have a high genetic component [18]. Indeed, studies of the relative risk of other cerebrovascular diseases with less heterogeneous phenotypes are able to document strong patterns of inheritance. Subarachnoid hemorrhage occurs with a relative risk of 6.6 in first-degree relatives compared to second-degree relatives [19]. Defining specific stroke subtypes may be the key to elucidating the exact degree of genetic contribution to any particular phenotype. From twin studies, it appears that the extent to which genetic factors may contribute to stroke risk varies with age. These factors are caveats to the identification of therapeutic targets from candidate gene strategies, and one must remember that a candidate gene approach for inheritance of risk factors may only be relevant to a highly limited stroke subpopulation.

Simple Stroke-Like Diseases: Single Gene Mutations

Identification of possible genetic determinants of stroke risk has been hampered by the lack of similar patient populations. Mendelian disorders with specific stroke-like phenotypes have been explored as genetic models of the more general population. These disorders include CADASIL (cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy) [20], MELAS (mitochondrial encephalopathy, lactic acidosis-and stroke-like

Barone/Read

338

syndrome) [21], Sneddon’s syndrome [22], familial hemiplegic migraine [23] and hereditary coagulopathies [24]. Although these subgroups contribute little to the overall prevalence of stroke, genes identified from them are hoped to highlight potential commonalities in the wider patient population. Studies on CADASIL and MELAS are examples of such approaches. CADASIL was originally described as an inherited, autosomally dominant dementia with multiple infarcts [25]. Epidemiologically, CADASIL is limited to sporadic cases in Europe [26, 27] and North America [28, 29]. The principal symptoms of the CADASIL are migraine with aura, ischemic stroke, and psychiatric symptoms including dementia [30]. In these patients, T2-weighted MRI reveals small periventricular white matter hyperintensities often involving the internal capsule [31]. The CADASIL gene, identified as Notch 3, is located at the chromosomal loci 19p13.1–13.2 [27, 32]. The Notch genes regulate the lin12/sel-12 signaling pathway that is known to be important in development, although the normal adult function of Notch genes remains unknown [33]. An interesting association of the Notch 3 gene with Alzheimer’s disease has also been discovered. Notch gene products interact with the presenilin 1 pathway as substrates for ␥-secretase. This enzyme is known to have a key pathological role in the production of A␤ peptide, although the modulatory role that Notch 3 may have in this disease process is undefined [34, 35]. The Notch 3 gene encodes a transmembrane protein composed of 2,321 amino acids, presumed to have a receptor function and located primarily on smooth muscle cells [30]. In CADASIL, approximately 90% of patients have missense mutations in extracellular domains of the protein product, whilst in about 70% of patients, the mutation is located within exons 3 and 4 [35]. All known mutations associated with CADASIL result in the removal or addition of cysteine residues and it is proposed that the expression of these mutated Notch 3 proteins results in cerebral vascular smooth muscle dysfunction [36]. Whether abnormalities in Notch signaling impact on the broader stroke population is at present unknown, although the pathogenesis of CADASIL, characterized by the progressive disruption of vascular endothelium with secondary fibrosis and thrombosis is typical of some stroke subpopulations [37]. Anticoagulant therapy has been tried in CADASIL without positive results [30]. More broadly, CADASIL also has close relationships to Alzheimer’s disease; signaling components of the presenilin pathway are shared with the Notch pathway [38]. The presenilin-1-regulated ␥-secretase cleaves both the Notch intracellular domain and ␤-amyloid precursor protein for subsequent translocation to the nucleus and binding to DNA [38]. Therefore, whilst the pathology of CADASIL may bear similarity to stroke, the cell biology also has potential connections with Alzheimer’s disease. Since vascular risk factors and/or disease can impact on both vascular dementia and Alzheimer’s disease, these relationships are intriguing [39–41].


339

MELAS are characterized by migraine-like headache, nausea, seizures, and stroke-like episodes. Lesions are most commonly found in the occipital and parietal regions, with high lactate levels found within lesions using proton nuclear magnetic resonance [42]. Patients typically have mutations of mitochondrial DNA for the tRNA-leu gene at an A-G transition mutation at nucleotide position 3243 [42, 43] and at a T-C transition at position 3271 [44]. It has been speculated that as mutations accumulate, a gradual mitochondrial dysfunction develops [45]. It is unclear how widespread such mutations are in the broader stroke population. Indeed, cases of MELAS have been reported without a family history, suggesting that these point mutations may be spontaneous [46]. Pharmacological interventions have reflected a unique nature of MELAS within stroke/cardiovascular disease subpopulations. Antithrombotic therapy has been employed in MELAS patients for cardiac complications associated with left ventricular dysfunction [47]. CADASIL and MELAS demonstrate that several relatively rare ‘strokelike’ syndromes can be used to explore potential genetic determinants of stroke. Parallel strategies have been adopted with similar success in other more complex, multifactorial polygenic traits such as hypertension [48–50]. Genes such as 11␤-hydroxylase in glucocorticoid-remediable aldosteronism have been shown to mediate hereditary hypertension in these patients [51]. However, as in stroke genetics, narrowing heterogeneity and studying single gene or Mendelian disorders may have limited application to the broader patient population.

Ischemic Stroke: Complex Genetic Associations

In common with many diseases, there are individuals with complex genetic profiles which confer vulnerability to stroke, as well as poststroke gene expression, which can contribute to increased cerebral ischemic stroke effects. Candidate gene studies in heterogeneous stroke populations minimize issues of limited patient population by the choice of a functionally relevant gene and its relationship with a particular phenotype. This is termed ‘association’ and is a statistical measure of the dependence of a particular phenotype (e.g., ischemic stroke with the presence of a particular candidate gene/allele). Therefore, association can be positive (i.e., significant statistical relationship/association between the gene of choice and phenotype), or negative (i.e., absence of significant relationship/association between gene/allele and phenotype). Candidate gene polymorphisms with a positive association with stroke include: apolipoprotein E [52–54], ACE ([55–58], fibrinogen [59], Factor V [60] and Prothrombin [61]. Those gene polymorphisms with a negative association with stroke include: eNOS [62], methyl-tetrahydrofolate [61, 63], angiotensinogen [55], Factor V

Barone/Read

340

[64, 65], Factor VII [65], Factor VIII [66] and prothrombin [67]. Atrial natriuretic peptide (ANP) has a particularly strong positive association with stroke [68–70]. Candidate gene choice is frequently driven by accepted stroke risk factors (e.g., hypertension, hemostasis and abnormalities in lipid metabolism) and indeed significant positive associations of numerous markers with ischemic stroke have been identified. At present, however, it is difficult to identify candidate gene associations with ischemic stroke [4]. Reproducibility of these gene expression associations in different patient populations (e.g., different race or genetic backgrounds) also is not known. The bewildering combination of possible outcomes for candidate gene association studies is related to the genomic and phenotypic heterogeneity of the global stroke population. Studies are typically designed with case controls or by cohorts to enable close approximation of phenotype between affected and nonaffected individuals. Superimposed upon these levels of variation are issues in the timing of stroke onset, in the variability of environmental influences and penetrance (i.e., not all individuals of a given genotype will express the phenotype). Finally, although the human genome project has been completed [1], identifying functionality of gene products lags significantly behind. Currently, it is estimated that only approximately 10% of the human genome has been ascribed function [24]. Certainly more work needs to be done in this area, and issues related to stroke genomics that include risk and the expression of genes underlying brain vulnerability and ischemic sensitivity must be considered.

Preclinical Models of Spontaneous Stroke

It is with these caveats in mind that studies have focused on animal models of spontaneous stroke, where environmental and genetic variability can be controlled. Bioinformatic approaches using synteny can facilitate the matching of ‘stroke loci’ found in stroke-prone rats to candidate genes on the human chromosome. Heterogeneity of risk factors and life events in humans has made it advantageous to study rodent models. Highly homogeneous populations of stroke-prone rats have been isolated from the incompletely inbred, spontaneously hypertensive rat (SHR) and then inbred further for this phenotype. Initial studies using this stroke-prone rat indicated that the degree of functional collateral blood flow after occlusion of the middle cerebral artery (MCAo) was inherited in an autosomally recessive manner [71–73]. The authors studied luminal diameters in vascular anastomoses between middle and anterior cerebral arteries and hypothesized that a single gene not directly linked to hypertension determined the collateral flow phenotype. Further genetic comparisons between strains were hampered by heterogeneity.


341

Narrowing the genotype by further crossing SHR rats with stroke-prone animals allowed cosegregation of genes defining various stroke phenotypes and for homogeneity of alleles for hypertension [74]. Two separate groups have utilized these inbred populations for identification of genes associated with manifestation of specific stroke phenotypes. A genome-wide screen was performed in an F2 cross-obtained by mating stroke-prone and SHR rats, in which latency to stroke was used as a phenotype [75]. This study identified three major quantitative trait loci (QTLs) designated, STR-1, STR-2, and STR-3. Of these, STR-2 and STR-3 conferred a protective effect against stroke in the presence of stroke-prone alleles and STR-2 colocalized with the candidate gene encoding ANP and brain natriuretic peptide (BNP). Furthermore, interactions between alleles from within STR-1 and STR-2 suggested that this phenotype was a reasonable model of the polygenicity of stroke in man. Follow-up sequencing to characterize ANP and BNP as candidates for stroke revealed point mutations in ANP and no differences in BNP. In vitro functional studies indicated lower ANP promoter activation in endothelial cells from stroke-prone rats versus SHR, with significantly lower ANP expression in the brain and no difference in BNP expression [68]. To determine the in vivo significance of the STR-2 lowered ANP promoter activation in stroke-prone animals, in comparison to stroke-resistant animals, a cosegregation analysis of stroke occurrence in SHR stroke-prone rats/SHR stroke-resistant F2 descendants and ANP expression was performed [69]. It was found that reduced expression of ANP did cosegregate with the appearance of early strokes in F2 animals [70]. Therefore, although lowered ANP expression may be part of the phenotype of the protective STR-2 QTL, it is unlikely that this is the primary protective mechanism in these animals. Parallel human studies of the role of ANP in cerebrovascular disease have confirmed that variation in the ANP gene may represent an independent risk factor for stroke in humans [4, 75] and emphasizes the utility of this cohort of animals as a model of ANP dysfunction in multiple subtypes of stroke. Two other groups have utilized a modified model of the stroke-prone animal, employing F2 hybrids derived from crossing the stroke-prone SHR with Wistar-Kyoto rats [76, 77]. One group [76] used brain weight poststroke as the phenotype for linkage analysis, after the discovery that F2 animals had higher levels of brain edema formation poststroke. This group found clear evidence of the linkage of phenotype to a gene on chromosome 4, which contributed to the severity of brain edema and was independent of blood pressure and STR3 identified by others [75]. The other group [77] designed studies to identify the genetic component responsible for large infarct volumes in the stroke-prone rat in response to a focal ischemic insult. To do this, they performed a genome scan in an F2 cross-derived from the stroke-prone rat and the normotensive Wistar-Kyoto rat [77]. Unlike others [75], they were only able to identify one

Barone/Read

342

major QTL responsible for large infarct volumes. This QTL was located on rat chromosome 5, and like STR-2, it colocalized with ANP and BNP, and was blood pressure independent. Unlike STR-2, this locus showed a much higher significance (lod 16.6) and accounted for greater (i.e., 67%) of phenotypic variance [77]. Subsequent studies identified that infarct volumes in the F1 rats were approximately identical to those of the stroke-prone animals, suggesting a dominant mode of inheritance [78]. Authors have argued over the significance of the overlap of STR-2 identified by some [55] with the QTL identified by others [77] on chromosome 5. It is unclear how the two phenotypes studied, latency to stroke (i.e., relative risk) [75] and size of infarct after occlusion (i.e., sensitivity to focal ischemia) [77], should physiologically relate to each other. However, this may only become apparent when individual genes are cosegregated with each phenotype. At present, altered ANP expression appears to play a role in the phenotype described by one group [75], but has been excluded from a role in the colony used by the other group [77, 79]. What can be concluded from each of these stroke-prone rat models? Certainly, each represents a unique and valid model of stroke for the study of inheritance, and for a role of candidate genes, in particular, stroke phenotypes. Neither colony represents a definitive model of human stroke, although linking identified candidate genes in these stroke-prone colonies to the human population has made progress [70]. One such research strategy that we have used is the analysis of genomic synteny between the rat and human genome. This bioinformatic approach seeks to align regions of homology using evolutionary conserved markers and has been applied with some success in relating animal models to human genetics in other disease paradigms such as non-insulin-dependent diabetes [80]. Relating identified loci from stroke-prone animals to the human genome offers a strategy for potential identification of candidate genes. For example, the STR-2 region of rat chromosome 5 shows well-conserved gene order and synteny with the human chromosome region 1p35–36. The high level of synteny between these regions makes this region ideal for rat-human comparative analysis. Sequence tagged sites localized to this region have been identified and mapped to human transcript clusters. As many as 132 transcripts have been identified in this region. The main candidates with some rationale for involvement in stroke are shown in table 1. Interestingly, only a few candidate genes identified at 1p35–36 have been examined in association studies. ANP has recently been assessed for association with multiple subtypes of stroke [70]. The polymorphism G664A, responsible for a valine-methionine substitution in proANP peptide was found to be positively associated with the occurrence of stroke [70]. In contrast, methylenetetrahydrofolate, another marker located at 1p35–36, was negatively associated


343

Table 1. 1p36–p35 positional candidates with a biological rationale in stroke* Homology

Rationale in stroke

CD30L receptor precursor

Nerve growth factor receptor superfamily Mouse KO increases neuronal damage in response to insults Depletion of complement system improves outcome following cerebral ischemia Role in cell adhesion May be related to sustained contraction during cerebral vasospasm Critical upstream activator of the caspase cascade in vivo Role in neuronal development

Tumor necrosis factor receptor 2 Human MASP-2 serine protease protein – complement processing protease gp40 mucin – putative influenza virus receptor Human Tropomyosin-related proteinexclusively neuronal/brain expression Human protease proMch6 (Mch6)/CASPASE-9 EPHRIN RECEPTOR EphA2/Tyrosine-protein kinase receptor ECK Human PDGF-associated protein Endothelin-converting enzyme E1 WNT4 protein precursor

EPHRIN RECEPTOR EphB2/Tyrosine-protein kinase receptor ERK Stathmin – v high brain expression Corticosteriod-binding protein – yeast putative bicistronic heat shock proteins Platelet-activating factor receptor Dishevelled-1

Atrial natriuretic peptide A Brain natriuretic peptide B Complement component 1, q subcomponent, alpha polypeptide (C1QA) 5,10-Methylene-tetrahydrofolate reductase

Brain-specific angiogenesis inhibitor (BAI2) Platelet phospholipase A2, group IIA

Barone/Read

Unknown Enzyme that produces potent vasoconstriction Possible role in synaptic plasticity. Linked to JNK signaling and indirectly to Notch (CADASIL) Role in neuronal development Phosphorylated by CAM kinase II Stress response Unknown Possible role in synaptic plasticity. Linked to JNK signaling and indirectly to Notch Localized with LOD peak hypertension, see text Localized with LOD peak hypertension, see text Depletion of complement system improves outcome following cerebral ischemia Heterozygous mutations are significant cause of stroke in general population, see text Regulator of angiogenesis Antiplatelet agents modify stroke risk

344

Table 1 (continued) Homology

Rationale in stroke

Sodium hydrogen exchanger-1

pH regulator of acidity associated with postischemic damage PAF involved in arterial thrombosis; Antiplatelet agents modify stroke risk

Platelet-activating factor receptor (PTAFR)

*This table documents the identification of human candidate genes that are syntenic to the STR2 region of rat chromosome 5 identified in a cohort of SHR-stroke prone animals. STR2 shows well-conserved gene order and synteny with the human chromosome region 1p35–36 (between D1S503–D1S2667). The high level of synteny between these regions makes this region ideal for rat-human comparative analysis. Sequence tagged sites localized to this region have been identified and mapped to human transcript clusters. Many transcripts, specifically 132 of them, have been identified in this region, with the main candidates listed above.

with occurrence of stroke [79]. Further studies may elucidate the predictability of markers of 1p35–36 and association with stroke for those genes listed in table 1. In contrast, rat-human synteny in the regions of the rat STR-1 and STR-3 loci are not well conserved, as several disruptions of synteny appear to have been introduced during evolution. It may be difficult to determine the exact regions of synteny between these rat loci and human chromosomal loci, and thus to extrapolate the candidate genes from rat to human. Human chromosomal regions syntenic with STR-1 span regions of two human chromosomes, around 16p11 and 19q13. Human synteny with the STR-3 region also appears to be disrupted, with regions of synteny mapping telomerically to opposite arms of chromosome 7 (7p21 and 7q35). Of course, this is a problem of animal modeling of human diseases in general and is not restricted only to ischemic stroke.

Stroke-Associated Gene Expression in the Evolution of Brain Injury

Cerebral ischemia is a powerful stimulus for the de novo expression and up-regulation of numerous genes [2, 4, 5, 81, 82]. In terms of isolation of gene candidates for a neuroprotective strategy, interpretation of expression changes has proven difficult. The multitude of animal models of ischemia with varying


345

genetic heterogeneity and infarct pathophysiology, is also complicated by spatial and temporal variations that have largely confounded interpretation. Furthermore, assays of differential expression have varying sensitivity to the relative fold-increase or -decrease in mRNA expression. As a result, ‘fishing’ exercises (i.e., differential expression detection studies) will often result in ‘catches’ (i.e., hits) of differential gene expression that vary depending on the assay employed. Bearing in mind this bewildering array of complexity, the next section addresses animal model(s) that might be utilized with differential gene expression analysis, target confirmation methodology that is necessary following the identification and confirmation of a differentially expressed gene (i.e., a ‘hit’), and the functional assessment of these genes in the disease process. A hierarchical critical path that depicts the path from target identification to target confirmation/validation is depicted schematically in figure 1.

Clinical Relevance of Ischemic Stroke Models

The failure of several putative neuroprotective agents in recent large multicentered clinical trials [83, 84] has led to critical re-examination of preclinical models of ischemia [85]. Heterogeneity in the human stroke population and the multitude of well-defined animal models of ischemia have led to attempts to refine model choice as related to patient subgroups [86, 87]. In an effort to stratify patient groups that can be predicted using specific animal models, authors have focused on the use of MRI signatures, particularly perfusionweighted or diffusion-weighted imaging (PWI, DWI) mismatches. Two main groups of acute stroke patients are identifiable: those with evolving infarcts in which lesion PWI ⬎ DWI, or those with a stabilized infarct where PWI ⱕ DWI [88, 89]. Such PWI/DWI assessments have been proposed to correlate to the extent of salvageable tissue, with approximately 70% of patients exhibiting PWI lesions ⬎ DWI at 6 h poststroke, and about 50% of patients exhibiting this mismatch at 24 h poststroke [89]. Applying the same imaging paradigms to animal models of focal ischemia should enable translation of preclinical pathophysiology into predictive outcomes in the appropriate patient population. However, detailed comparisons of the development of PWI/DWI signatures between animal models of ischemia are difficult to establish due to the use of various rat strains, anesthetics, and modes of ischemia induction. However, broad comparisons are possible by exploring the development of DWI signal as a marker of lesion volume with respect to time. Data in certain animal models of focal stroke can show a delay in the development of DWI hyperintensity (i.e., brain lesion size) that lags

Barone/Read

346

Critical path from expression to target validation Animal model

mRNA differential expression

RDA, Microarrays, SAGE, SSH, DD

Reproducibility/commonality of ‘hits’?

Expression studies

Confirm differential expression Expression analysis

Northern analysis Quantitative RT-PCR In situ hybridization (ISH)

mRNA localization

Confirmation and localization of protein expression

ELISA, Western analysis, Immunohistochemistry (IHC)

Functional studies

Genetic/transcriptional modification

In-Vivo pharmacological studies

Targeted gene KO Antisense Adenoviral transfection Appropriate PK tool compound treatment

Fig. 1. Critical pathway for target identification and confirmation. Following selection of an animal model appropriate to clinical subpopulation, broad mRNA differential expression strategies are adopted employing differential expression assays such as RDA, microarrays, subtractive hybridization (SSH), serial analysis of gene expression (SAGE) and/or differential display (DD). Across these assays, reproducibility in identified hits are explored as a technique of prioritizing subsequent studies to confirm differential expression. Comprehensive expression analysis using RT-PCR or Northern analysis allows confirmation of identified hits and fully quantified temporal profiling. This can also include RNA localization using in-situ hybridization (ISH). Protein confirmation by ELISA, Western analysis and immunohistochemistry (IHC) follows mRNA profiling, to confirm translation. In an ischemic brain, pooling of mRNA and uncoupling of translation can occur as indicated in the text. Finally, functional studies encompassing target gene knockout, adenoviral transfection and in vivo pharmacology complete validation of a potential target gene. Appropriate chemistry directed against the verified biological target could result in the eventual discovery of a drug for stroke.

behind a perfusion deficit (i.e., PWI changes associated with stroke or focal ischemia). This delay is attributable, in part, to the relative contribution of insufficient collateral flow and the peri-infarct depolarization that injures the poorly perfused, ischemic brain during infarct evolution [87, 90].


347

Direct MCAo by proximal electrocoagulation of the middle cerebral artery produces an expanding DWI lesion, with an initial marked expansion at 4 h followed by a small increase from 4 to 24 h [91]. Electrocoagulation occlusion of the distal MCA produces a much more rapidly evolving infarct with a near maximal DWI lesion observed by 1–2 h [92, 93]. Available embolic models of focal stroke using intra-arterial injection of thrombin [94] or aged [95] or fibrin-rich [96] clots have reported similar expansion of DWI hyperintensity. For example, following thrombin injection, DWI hyperintensity is apparent at 80 min postadministration, with the volume gradually expanding up to 24 h [94]. Intraluminal suture occlusion produces a range of DWI lesion progression dependent on whether the filament is introduced via the common carotid artery [97] or the external carotid artery [98]. Permanent MCAo via common carotid artery suture produces a rapid evolution of DWI hyperintensity within minutes, followed by maximal expansion by 2 h [99, 100]. In comparison, permanent MCAo without occluding the common carotid artery (i.e., via the external common carotid artery) [98], evokes an initial rapid expansion of DWI hyperintensity over the first 30 min followed by final infarct volume reached at 7 h [101–103]. A close inspection of this model identifies it as exhibiting a flow mismatch similar to that in man, i.e., it represents the type of evolving infarct that should provide information relevant to human stroke [87]. For this reason a number of groups [6, 104, 105] have decided to use this model for differential gene expression studies.

Differential Gene Expression Methodologies

The detection of genes that are differentially expressed due to stroke can be identified using simpler techniques such as Northern blotting, RT-PCR, or in situ hybridization. These techniques involve the selection and study of a specific gene of interest based on previous data that provides a biological rationale for study in stroke or another specific disease. However, more sophisticated screening techniques are now available that can identify groups of differentially expressed genes, both known and unknown. These screening techniques include subtractive hybridization, differential hybridization, representational differential analysis (RDA), serial analysis of gene expression (SAGE), and differential display [106]. The identification of differentially expressed genes in stroke has employed the simpler techniques as well as these newer techniques. Variations in assay and threshold of detection can result in the isolation of gene sets that differ according to assay selection. Therefore, to ensure maximum confidence in the detection of adaptive up-regulation of gene expression,

Barone/Read

348

reproducibility of genes and pathways should be verified across several differential expression techniques and independent cross-validation of a gene’s up-regulation should be emphasized. A summary of stroke-associated gene expression discovered by all the techniques is listed in table 2. Only those gene/message expression changes in the rat following MCAo were confirmed using more than one expression detection technique. A brief summary of the more complex techniques is provided below. Differential Display Differential display is a means of comparing all poly-A mRNA between experimental and control populations. In this technique, mRNA is converted into first strand cDNA with reverse transcription followed by polymerase chain reaction (PCR) with multiple sets of primers. The PCR products are then displayed with control and experimental samples side-by-side on high-resolution denaturing gel. In this way, differential gene expression is apparent. This technique has been applied with success to isolate differentially expressed products following experimental MCAo. For example, differential display was used after rat MCAo to discover a gene that encodes adrenomedullin, a member of the calcitonin gene-related peptide family [113]. This analysis was followed by temporal studies using Northern analysis, which confirmed that expression of mRNA levels increased in the ischemic cortex at 3 and 6 h after MCAo and levels remained elevated for up to 15 days. Immunohistochemical studies to confirm protein expression then localized adrenomedullin to ischemic neuronal processes. In functional studies, synthetic adrenomedullin microinjected into the preconstricted rat pial arteries produced dose-dependent relaxation of the vessels. In addition, intracerebroventricular administration of adrenomedullin, prior to and after MCAo, increased the degree of focal ischemic injury. Other groups have also used this technique in ischemia to identify differentially expressed mRNAs such as a zinc transporter gene [144] and an ADP-ribosylation factor like gene [145]. Other examples include the transcription factor SEF-2 [146], proteosome p112 [146], and ST-38 chemokine [147] following rat MCAo. Differential display is useful but very labor intensive. It is most useful for examining several RNA samples simultaneously and has been used extensively for temporal, doseresponse, and multiple treatment studies. Although differential display is ‘semiquantitative,’ only relatively a small amount of total RNA (approximately 15 ␮g) are required. Some problems include high false positive rates that cannot be confirmed by RT-PCR or Northern blotting. Modifications such as subtracted differential display, which removes unregulated cDNA by mRNA subtraction prior to differential display [148], represent improvements. Confidence in an isolated candidate gene can be improved by using independent follow-up assays of gene expression in parallel.


349

Table 2. Summary of differential gene expression changes determined by techniques that measure transcription differences following focal ischemia stroke in the rat* Barone/Read

Gene

350

Stroke pMCAo ⫽ p tMCAo ⫽ t

Early RNA expression ⱕ24 h

Later RNA expression ⱖ24 h

RNA/Protein cross-validation techniques

Protein demonstrated/ verified

Reference(s)

Immediate early genes NGFI-A NGFI-B NGFI-C NGFI-C Nurr-1 erg-2 erg-3 Zif 268, c-fos NF-␬B

p p p p p p p p t

Yes Yes Yes Yes Yes Yes Yes Yes Yes

No No No No No No No No No

No No No No No No No No Yes

107 107 107 3 107 107 107 108 109

NF-␬B

t

Yes (subunit specific)

Yes (subunit specific)

Yes

110

Activating transcription factor c-fos, c-jun, zif 268 ATF-3

t

No

Yes

111

p p, t

Yes (decrease) Yes Yes

No ND

ISH ISH ISH ISH ISH ISH ISH Northern analysis IHC Gel shift analysis Western blotting IHC Gel shift analysis IHC Western analysis Northern analysis RT-PCR, ISH, IHC

No No

108 105

Cytokines IL-1 receptor

p

No

RT-PCR

No

113

IL-1RA

p

Yes

RT-PCR

No

113

Yes (subunit specific) Yes


IL-1RA IL-1B IL-1B IL-1␤ IL-1␤ IL-2 IL6, Zif 268, c-fos CINC/IL-2 IL-10 TNF-␣ TNF-␣ TNF-␣ TNF-␣

p p p p t p p p p p t p p

Yes Yes Yes Yes Yes No change Yes Yes Yes (6 h) Yes Yes Yes Yes

ND Yes No No Yes No change No ND No Yes Yes No Yes

LIF

t

Yes

Yes

LIF SOCS-3 IL-1␤ TNF␣

p p p, t p, t

Yes Yes Yes Yes

Yes Yes No No

Inflammation COX-1 COX-2 COX-2 MCP-1 MCP-1

t t p t p

No change Yes

No change Yes Yes Yes Yes

Yes Yes

RT-PCR Northern analysis ISH RT-PCR Northern analysis RT-PCR Northern analysis ISH RT-PCR ISH, IHC, RT-PCR Northern analysis RT-PCR Northern analysis IHC RT-PCR Western blot IHC RT-PCR RT-PCR RT-PCR, ISH RT-PCR, ISH RT-PCR RT-PCR, IHC RT-PCR ELISA Nothern analysis RT-PCR, IHC

No No No No No No No No No Yes No No Yes

114 115 116 117 118 117 108 119 117 120 108 117 121

Yes

122

No No No No

104 104 105 105

No Yes No Yes Yes

123 123 104 124 125

351

Table 2 (continued) Barone/Read 352

Gene






Reference(s)

MCP-1 MCP-1 iNOS iNOS MCP-3 IP10 CXCR3 Heme oxygenase LPS-binding protein

p p t t p, t p p p, t p

No Yes Yes Yes Yes Yes Yes Yes No

Yes Yes No Yes Yes Yes Yes ND Yes

Northern analysis RT-PCR RT-PCR RT-PCR, IHC Northern analysis Northern analysis Northern analysis RT-PCR, ISH RT-PCR

No No No Yes No No No ND ND

126 104 123 127 128 129 130 105 6

Apoptosis Bax Caspase 1,6,7,8,11 Caspase 2,9 Caspase 3 Caspase-3 Fas, Fas-L TR3-death receptor p75 NGF-R Arc BIS Arc

t p p t p p p p p p, t p

Yes Yes No change No Yes Yes Yes Yes Yes Yes

No ND ND Yes No ND ND ND No ND Yes

ISH RT-PCR RT-PCR ISH RT-PCR RT-PCR RT-PCR RT-PCR ISH RT-PCR, ISH RT-PCR

No No No No No No No No Yes No No

131 132 133 131 133 133 133 133 3 105 104

Growth factors VEGF

p

Yes

No

ISH, IHC

Yes

134

Gene Expression in Stroke 353

VEGF

t

Yes

Yes

VEGF receptor VGF BDNF TGF␤-1 TGF-␤1 Narp Cathepsin C Cystatin B Agrin JAK-2 cpg-21 Narp

p p t p p p, t p p p p p p

No Yes Yes No ND Yes No No Yes Yes Yes Yes

Yes Yes No Yes Yes ND Yes Yes No No No No

Other genes Adrenomedullin

p

Yes

Yes

HIF-1 Hsp-27 Hsp-70 Hsp-70

p p p t

No ND ND Yes

Yes Yes Yes No

GADD 45

t

Yes

No

MIP-1␣

p

Yes

Yes

MIP-1␣ MIP-3␣ CRH

p, t p p

Yes

No Yes No

Yes

Northern analysis Western blotting ISH, IHC RT-PCR ISH Northern analysis RT-PCR RT-PCR, ISH RT-PCR RT-PCR RT-PCR RT-PCR RT-PCR RT-PCR Northern analysis IHC ISH, IHC RT-PCR RT-PCR In situ autoradiography In situ autoradiography Northern analysis RT-PCR, IHC In situ RT-PCR ISH

Yes

135

Yes No No No No No No No No No No No

134 104 136 143 104 105 6 6 6 6 6 6

Yes

112

Yes No No No

134 104 104 137

No

137

Yes

125

No No No

138 104 139

Table 2 (continued) Barone/Read 354

Gene






Reference(s)

NT-3 Trk-B Osteopontin Osteopontin

t t p p

Yes (decrease) Yes Yes No

No No Yes Yes

No No Yes Yes

136 136 140 141

Osteoactivin TIMP-1 TIMP-1

p p p

ND Yes Yes

Yes No Yes

No No No

104 142 142

TIMP-1 CD14 CD44 GADD45␥ Xin Hsp-70 Cyr61 Lox-1 Rad G33A HYCP2 Mim-3 CELF

p p p p p p, t p, t p, t p p p p p

ND ND ND ND

Yes Yes Yes Yes Yes ND ND ND No No No No No

ISH ISH ISH, IHC Northern analysis IHC RT-PCR Northern analysis Subtractive cDNA libraries Southern analysis RT-PCR RT-PCR RT-PCR RT-PCR RT-PCR RT-PCR, ISH RT-PCR, ISH RT-PCR, ISH RT-PCR RT-PCR RT-PCR RT-PCR RT-PCR

No No No No No No No No No No No No No

104 104 104 104 104 105 105 105 6 6 6 6 6

Yes Yes Yes Yes Yes Yes Yes Yes


Tenascin DAF Cip-26 PHAS-I TBFII Spr Glycerol-3 phosphate dehydrogenase PRG1

p p p p p p p

Yes Yes No No No No No

p

No

Yes Yes

RT-PCR RT-PCR RT-PCR RT-PCR RT-PCR RT-PCR RT-PCR

No No No No No No No

6 6 6 6 6 6 6

Yes

RT-PCR

No

6

No No Yes Yes

355

*This table only lists increased message expression that has been validated within or between labs independently. The importance of this validation in addition to the verification of translated protein for candidate genes is emphasized in the text. Permanent MCAo ⫽ pMCAo; transient MCAO ⫽ tMCAo; ND ⫽ not determined; ISH ⫽ in situ hybridization; IHC ⫽ immunohistochemistry.

Subtractive Hybridization Subtractive hybridization compares qualitative differences in gene expression between two experimental paradigms. This is usually achieved by hybridization of biotinylated ‘driver’ cDNAs to the mRNA pool from the target tissue. Duplexes of driver cDNAs and target mRNAs are then removed, resulting in a pool of target mRNAs expressed only by the target tissue [149]. Down-regulated mRNAs are determined by carrying out the reaction in reverse. Modifications to the assay include suppression subtractive hybridization and RDA, where the polymerase chain reaction replaces physical subtraction methods to enrich for differentially expressed transcripts. Such modifications emphasize differential mRNA of both low and high abundance, rather than biasing selection of only highly expressed genes as is the case with the more basic subtractive hybridization methodology. Suppression subtractive hybridization has been used to identify candidate genes with putative roles in experimental cerebral ischemia. For example, the induced expression of a rat homolog to human monocyte chemotactic protein-3 (MCP-3) was identified in the ischemic brain [128]. Independent Northern analysis identified increases in MCP-3 mRNA observed at 12 h postischemia, with 49-fold and 17-fold increases over control in permanent and temporary MCAo, respectively. Significant induction of MCP-3 in the ischemic cortex was sustained up to 5 days after ischemic injury. In other models, subtractive hybridization has been less widely used to identify candidate genes, perhaps due to the technique demands of the subtraction approach, although false positives are less frequent. The subtractive hybridization approach also has been used to successfully identify a novel cDNA clone (pGSH3), expressed only after ischemia in the gerbil cortex [149], which turned out to be a homolog of a 72-kilodalton human heat-shock protein (hsp70) gene. Basal cortical levels were found to be low, but 8 h after a 10-min transient forebrain ischemia, gene expression became prominent in the cerebral cortex. RDA RDA is a relatively novel PCR-coupled, genome subtractive process [150] that until recently [104, 105] had not been used to assay differential expression in models of cerebral ischemia. RDA is conceptually similar to subtractive hybridization, but the unavailability of a commercially produced kit for RDA has meant that it has been less broadly exploited. RDA was originally established to monitor differences in genomic DNA content between individuals, it was later modified to identify differences in gene expression [150]. The robust gene expression changes that characterize the MCAo model are detectable with RDA, as we have recently been able to show [104]. Subtracting ischemic cortex from rats 24 h following permanent MCAo from similarly treated tissue

Barone/Read

356

from sham-operated animals, we identified candidate ischemia-regulated transcripts. Primary confirmation of the accumulation of these gene products in the ischemic cortex was confirmed using SYBR Green RT-PCR, followed by the more comprehensive time course analysis using TaqMan RT-PCR in selected cases. Several genes identified through this approach had previously been reported to increase following MCAo, such as heat shock proteins (hsp 27 and hsp 70) and others (MCP-1, MIP3␣, COX-2, TGF-␤1, tissue inhibitor of matrix metalloproteinase (TIMP-1) and Arc), but some were newly identified as MCAo-induced genes in this study (LIF, SOCS-3, VGF, CD44, CD14, CD81, osteoactivin, GADD45␥ and Xin) [104]. In another study [105] using this technique, 128 unique gene fragments were isolated and 13 were selected for RT-PCR analysis. Many transcripts were verified to be differentially expressed by RT-PCR, including four genes not previously implicated in stroke: neuronal activity-regulated pentraxin (Narp), cysteine rich protein 61 (cyr61), Bcl2binding protein Bis (Bcl-2-interacting death suppressor), and lectin-like ox-LDL receptor (Lox-1). Microarray Analysis All the above strategies identify relatively small numbers of differentially expressed genes. Large numbers of DNA fragments (110–450 bp) are produced in the process, which need to be confirmed and frequently extended to full lengths to obtain gene identity. Although all of these technologies are useful for isolating candidate genes, they are of limited utility in providing a broad characterization of the expression of large number of genes within a particular model. Array-based technology, on the other hand, allows large-scale and prospective analysis of gene expression as well as time-response profiling and drug treatment analysis (pharmacogenomics). Whether using arrays of oligonucleotides [151, 152] or gene fragments [153], the array technology allows parallel expression monitoring of numerous genes at the same time [154]. The limitations and biases of the technique are obviously in the selection of genes to study on the array. The power and quality of microarrays has continued to improve significantly, and now thousands of genes can be evaluated at a time. The pioneer study [3] using this technique in the context of stroke was applied to studying gene expression in a proximal MCAo electrocautery model [155]. Oligonucleotide probe arrays were employed with 750 predetermined genes optimized for gene expression in rat bone and cartilage. The gene chip (Roche, ROEZ002) was used to monitor gene expression after 3 h of permanent focal ischemia. To determine genes differentially expressed as a consequence of ischemia, the authors took tissue from the ipsilateral frontal and parietal cortices and compared their expression to corresponding regions on the contralateral side. A significant change in transcription was defined as a 2-fold or greater


357

increase or decrease in expression compared to the contralateral hemisphere. The authors described a significant up-regulation of 24 genes in the parietal and frontal cortices and striatum, with particularly robust changes in c-fos, NGFI-A, NGFI-B, NGFI-C, Krox20, Nor-1, cyclooxygenase-2, and Arc. This study clearly demonstrated the utility of array technology in the analysis of gene regulation following experimental cerebral ischemia. The use of arrays optimized for bone and cartilage genes unfortunately limited the usefulness to 15% of the total gene representation on the array. Nevertheless, key gene families such as the phosphatases (MKP-1 and MKP-3) and the chemokines (MCP-1 and MIP-1␣) were represented and expression profiles agreed with previous findings [156, 157]. No change in ‘housekeeping’ genes GAPDH and ␤-actin were found (ipsilateral vs. contralateral). Another outstanding study [6] used the rat Affymetrix U34A oligonucleotide array to assess 8740 transcripts in the peri-infarction rat cortex 24 h after thread MCAo, a model representing slow stroke evolution, as in man [98]. Using strict analysis criteria, less than 4% of transcripts were regulated (e.g., 264 were up-regulated and 64 were down-regulated), of which 163 had not been reported to be modified in stroke previously. In terms of functional groups, G-protein-related genes were least variable, while cytokines, chemokines, stress proteins and cell adhesion and immune molecules were most modulated. Quantitative RT-PCR of selected genes identified early up-regulated genes including Narp, Rad, G33A, HYCP2, Pim-3, Cpg21, Jak2, CELF, Tenascin, and DAF. Late up-regulated genes (⬎24 h) included cathepsin C, Cip-26, cystatin B, PHAS-I, TBFII, Spr, PRG1, and LPS-binding protein [6]. Implications of up-regulated glycerol 3-phosphate dehydrogenase, plasticity-related transcripts, gene regulation related to cell survival, death and tissue repair and functional recovery, and biochemical pathways related to gene changes were evaluated [6]. SAGE SAGE yields information about absolute transcript numbers of many, if not all, genes expressed in a given tissue and therefore allows for the identification of differentially expressed genes when applied to tissues in different conditions [158–160]. The technique is based on the reduction of each expressed transcript sequence to short (14–15 bp), yet representative, sequences (tags) at a defined position, which are concatenated into long molecules. Sequencing these molecules reveals the identity of multiple transcripts simultaneously. The number of times a particular tag is detected in a SAGE library, therefore, provides a quantitative and digital measure of gene expression [158–160]. In a recent and an elegant study, differentially expressed genes in mouse brain 14 h after the induction of focal cerebral ischemia were determined using SAGE [160]. From the estimated 30,000 genes of the mouse genome, at least 24,590 genes were detected

Barone/Read

358

by SAGE. Analysis of ⬎60,000 transcripts revealed 83 up-regulated and 94 down-regulated transcripts, defined as greater than or equal to 8-fold difference from baseline. Up-regulated genes were classified as transport and secretion (4%), ribosomal proteins (4%), DNA/RNA metabolism (4%), cytoskeletal (6%), protein folding and degradation (12%), intermediary metabolism (16%), signal transduction (22%), or unclassified (32%). Metallothionein-II (MT-II) was found to be the most significantly up-regulated transcript in the ischemic hemisphere. MT-I and MT-II both appear to be induced by metals, glucocorticoids, and inflammatory signals in a coordinated manner, yet their function remains unknown. Up-regulation of both MT-I and MT-II was confirmed by Northern blotting. MT-I and MT-II mRNA expression increased immediately after 2 h of transient ischemia, with a maximum after 16 h. Western blotting and immunohistochemistry revealed MT-I/-II up-regulation in the ischemic hemisphere, whereas double-labeling demonstrated colocalization of MT with markers for astrocytes as well as for monocytes/macrophages. The completeness of this study (fig. 1) was demonstrated by the use of MT-I- and MT-II-deficient mice, which developed an approximately 3-fold larger infarcts than wild-type mice and a significantly worse neurological outcome [160].

Assay Variation and Confidence in Identified Gene Expression

In the preceding sections we discussed many of the techniques available for the detection of differential gene expression and some of the ‘within assay’ issues associated with each technique. Next, we will discuss in more detail model-to-model differences, the importance of the poststroke timings of RNA sampling, and different experimental paradigms in stroke that might help us discover genes that have roles in brain protection or tolerance, and studies that can be used to look for genes that might contribute to the recovery/plasticity of the brain postinjury. There are several issues which warrant discussion including the significant variability between techniques, and the identification of false positive and false negative results. Assays of differential expression have an inherent variability dependent on assay methodology, sensitivity, and reaction efficiency. When exploring disease paradigms which are powerful stimulators of gene expression such as cerebral ischemia, the usual tendency is to highlight gene sets or functional groups that are up-regulated and differentially expressed. Given the large numbers of genes identified, it is difficult to confirm all the differentially expressed genes and false positive and false negative differential gene expression becomes an issue. False positives can broadly be defined as genes whose differential expression is not subsequently confirmed by an independent study


359

(i.e., RT-PCR, Northern analysis, in situ hybridization, etc.). False negatives, in contrast, are genes that are, in fact, differentially expressed, but not detected as such by the complex assay employed (e.g., subtractive hybridization, RDA, DNA microarray). To manage these issues, we have employed the strategy of using multiple assays of differential expression on the same RNA pool and then cross-validating all of the numerous differentially expressed products between assays. Identifying commonalities in expression across assays increases confidence in particular results, and genes identified across two or more assays receive a higher priority for confirmatory studies. A table of descending confidence in ‘hits’ can then be constructed. This technique for handling large numbers of ‘hits’ avoids issues of biasing the identification of differential expression to a single assay and also interassay variability. Subsequent analysis by Taqman RT-PCR has confirmed that robust differential gene expression identified across all assays had particularly high-fold increases in differential expression. This strategy is also useful for identifying false negatives (i.e., products differentially expressed but not detected as such in assays) [4]. The ‘complimentary-techniques’ approach at target validation maximizes the coverage of differential gene expression by minimizing the losses due to the technical vagaries of any single technique.

Confirmation: An Integral Part of Differential Gene Expression

The techniques cited above for the identification of differentially expressed mRNAs represent starting points for the study of gene expression following stroke. All data derived by these methods require confirmation from an independent study to remove false positives, and this usually forms part of a broader analysis of expression of the gene that has been identified [3–6, 104–106, 160]. Traditional methods for analyzing gene expression include techniques such as Northern blotting, RNAse protection, in situ hybridization, and semi-quantitative RT-PCR. All of these methodologies have been used to study the expression of individual genes or small groups of genes in stroke models. Differential screening methodologies ideally generate large numbers of ‘hits’ which require rapid confirmation in a high throughput system. Recently, real-time quantitative RT-PCR techniques, such as ‘Taqman’ probes or SYBR green [161] to monitor an accumulating PCR product in real time, allow an accurate comparison of initial PCR template numbers. These assays can be carried out in 96 or 384 well formats and can utilize robots, reducing operator time and error. With these techniques, it is possible to carry out rapid confirmation of many differentially expressed genes simultaneously

Barone/Read

360

or to undertake a more detailed expression analysis of a single ‘hit’ [142]. For example, Taqman RT-PCR has been extensively applied to temporal profiling of caspase expression following MCAo in rats [132, 133] and SYBR green has been used to confirm differentially expressed genes identified by RDA [104]. Taqman is a PCR-based technique that is more sensitive than other confirmatory technologies such as Northern blotting. Typically, Taqman RT-PCR uses approximately 50 ng of total RNA per gene, while Northern blotting uses 10–20 ␮g. Additionally, Taqman RT-PCR is as sensitive as in situ hybridization, with the added advantage of higher throughput. Perhaps most importantly for paradigms such as MCAo, where gene expression can exceed 600-fold over that observed in naïve animals, Taqman PCR can quantitate gene expression over five to six orders of magnitude without multiple dilution series, as necessitated by other assays [115]. Clearly, PCR-based technologies such as Taqman RT-PCR and SYBR Green RT-PCR, whilst in their infancy in application to the study of cerebral ischemia [104, 132, 133], offer advantages for confirmation and expansion of data on differential gene expression. These techniques are also of value in testing hypotheses about genes already known to be regulated in stroke models, where differential expression is suggested by other biological evidence/data. The sensitivity of PCR-based methodologies suggests that sufficient RNA can be isolated from a single animal to allow the simultaneous assessment of several hundred genes. A large body of data can be amassed and the expression of many different genes compared in a single study, without drawbacks such as variation between studies, operators and cohorts of animals. The major drawback of high-throughput quantitative RT-PCR is that while it allows for the rapid assessment of changes in gene expression at the level of mRNA, it is not able to provide information on the precise cellular localization of such changes. A detailed understanding of stroke models utilized for differential gene expression is essential (i.e., the models have to be adequately characterized over time for cellular changes). The cell-type and intracellular locations of changes in gene expression are important. For example, neurons and oligodendrocytes die within the ischemic infarct [162, 163] particularly after 12 h of ischemia. Astrocytes and microglia are decreased in number in the core region of the lesion and proliferation of both of these cell types occurs in the marginal areas [164]. Polymorphonuclear leukocytes, and later macrophages, invade the lesion after around 12 h and for days after [163–165]. Changes in gene expression have to be understood in the context of these evolving cell types present at any given time after stroke. Ultimately, expression profiling must involve techniques such as in situ hybridization and immunohistochemistry, which allow the localization of expression to be viewed in relation to the structure of the evolving lesion, and the identification of the types of cells in which expression is occurring.


361

Confirmation of expression and time course of protein translation is essential for verification that up-regulated gene transcription has proceeded to the formation of protein. This point is especially pertinent given the severe energy perturbations in the ischemic brain. During this state, transcription and translation can become uncoupled due to energetic demands of assembling protein. This effect is temporally and spatially dependent and has been extensively reviewed [5, 81], including the uncoupling of mRNA transcription and protein translation following MCAo [4]. Issues associated with the sensitivity of protein-detection assays must be considered for some proteins. Ultimately, it is proteins which are pivotal in cellular function, and thus proteomic analysis in addition to mRNA analysis will point the way ahead.

Functional Studies,Transgenic Studies, and in vivo Pharmacology

There are already many examples from the literature where transgenic animal studies and/or pharmacological studies have coincided with gene expression studies to demonstrate the involvement of specific gene expression in focal stroke injury or protection. For example, cyclooxygenase-2-deficient mice are known to exhibit reduced susceptibility to brain injury [166]. IL-1ra was shown to be neuroprotective in brain injury [167] well before the altered expression of the IL-1 system in stroke was demonstrated [114]. IL-6 also has been shown to be neuroprotective in stroke [168, 169]. In addition, it has been shown that blocking thyrotropin-releasing hormone provides a significant protection against ischemic brain damage and associated neurological deficits [170, 171]. Following stroke, treatment with BDNF reduces brain injury in the MCAo model [172]. Genes for all these proteins have been shown to be up-regulated in stroke models. One recent example is Metallothionene-II as a major neuroprotective gene in mouse focal cerebral ischemia [160]. In this study, changes in the metallothionene-II gene were verified using multiple methods including immunohistochemistry, in situ hybridization and Western blot. In addition, stroke in the metallothionene knockout mouse resulted in stroke three times larger than in wild-type mice [160].

Additional Models for Discovering Gene Targets in Brain Injury

Neuroprotective or Neurodestructive Gene Expression As pointed out previously [2], focal ischemia stimulates multiple gene expression changes. Focal ischemia is a very powerful reformatting and reprogramming stimulus for the brain. There are broad and robust gene expression

Barone/Read

362

responses that occur following the focal stroke that are exhibited as temporal episodes or ‘waves’ of expression of different groups of genes [2]. These waves are largely comprised of increased expression of inflammatory cytokines including IL-6 and IL-1ra. In addition, growth factors (e.g., BDNF) that might be expected to play a neuroprotective role following stroke also increase in this time frame. The increased cytokine gene expression appears to drive leukocyte infiltration, a poststroke brain response to injury, and is associated with secondary brain injury and repair processes following stroke. Later waves of new gene expression include mediators, which appears to be important in tissue remodeling (i.e., resolution of ischemic tissue injury) and perhaps recovery of function. These issues are important in relation to the models suggested below that may provide new directions in future differential gene expression analysis. Preconditioning Stress in Brain Tolerance Strategies Certain stimuli that can cause injury will protect the brain against subsequent, severely injuring stimuli if applied at a low intensity (i.e., subthreshold for injury) prior to that severe injury. This phenomenon involves complex processes involved in endogenous organ protection. For ischemic stimuli, this phenomenon has been termed ischemic preconditioning (PC) or ischemic tolerance (IT). PC is a reaction to a potentially noxious stimulus such as hypoxia, ischemia, or inflammation. A short ischemic preconditioning event can result in a resistance to severe ischemic tissue injury. This phenomenon has been described in brain and heart, and may represent a fundamental cell response to certain types or levels of injury [173–176]. PC or IT in the brain is associated with a protected state that develops over hours, persists for days or longer and involves de novo protein synthesis [177, 178]. MRI can demonstrate the evolution of infarction and its reduction by tolerance induction [179–181]. Functional or motor effects are protected by PC [177]. It is interesting that the stroke-prone rat (discussed earlier in the chapter as a model of spontaneous stroke and used as an experimental model for gene associations) are significantly more sensitive to cerebral ischemia [92] and exhibit greater brain injury to cerebral ischemia and also exhibit a significantly reduced degree of IT to PC [179–181]. The brain changes associated with brain ischemia involve a progression of both injurious and protective processes as brain injury evolves and is then repaired following an insult such as a stroke. Focal cerebral ischemia induces a complex series of mechanisms [2, 173, 176] that result in infarcted tissue, a situation in which neurodestruction has overwhelmed neuroprotection. The major pathophysiological mechanisms of tissue destruction in stroke involve acute mechanisms of excitotoxicity and delayed mechanisms of inflammation and


363

apoptosis. The corresponding protective tissue responses include up-regulation of neurotrophic factors and protective or ‘stress’ genes/proteins. Later regenerative or restorative processes may allow recovery of some brain function poststroke [2], but generally are not sufficient to correct the initial damage. Because IT paradigms also provide insight into the mechanisms of endogenous brain neuroprotection in the absence of any destructive processes, it is believed that understanding the signaling and mechanisms involved in precondition-induced IT can discover new targets and approaches to protect the brain and other end organs from the injury/disease. IT has been demonstrated in the human brain [182, 183], and so differential gene expression strategies in IT experimental models have physiological relevance for studying apoptosis and other poststroke complications [184]. This endogenous brain protection phenomenon appears to represent a fundamental protective response to injury following prior stress [173, 176, 185, 186]. Models of brain protection provide an opportunity to identify novel protective gene expression associated with the development of brain tolerance. Data from differential gene expression suggests that neurotrophic factors, stress proteins, and cytokines contribute to the tolerance response to ischemia and other forms of stress in the brain [173, 174, 176–178, 187–189]. For example, IT is associated with an increased expression of the neuroprotective protein IL-1ra [177] and heat shock proteins [178] and a reduced postischemic expression of the early response genes, c-fos and zif268 [177]. A number of techniques including suppressive subtractive hybridization methodology have been applied to discover genes responsible for IT following PC [190]. With suppressive subtractive hybridization, tissue inhibitor of matrix metalloproteinase (TIMP-1) was identified as one candidate molecule in the stroke response and IT. Northern analysis confirmed that TIMP-1 mRNA was significantly elevated at 24 h and 2 days after PC, which corresponded well to the onset of IT [190]. Strategies in Brain Recovery, Plasticity and Recovery of Function Gene changes occurring one or more days after stroke might provide insight into reparative and recovery processes. While neurological functional deficits occur following stroke, there may be a recovery of brain function that occurs spontaneously or improves with training following stroke [2, 191–193]. Sampling tissue during functional brain recovery in animal models (i.e., at later time periods poststroke) might be expected to provide an opportunity to identify novel genes important for long-term brain regeneration or plasticity. These studies would be amenable to differential gene expression analyses, but would be profiled at later poststroke timepoints or under treatment conditions shown to facilitate such brain regeneration/recovery, for example after the introduction

Barone/Read

364

of putative neuroprotective or neuroregenerative agents. It is again interesting that stroke-prone rats that are more sensitive to and exhibit greater brain injury to cerebral ischemia also exhibit a greater degree of neurological deficits with significantly less spontaneous recovery of neurological deficits poststroke [92, 180, 181]. Recent data indicating complete recovery of neurological function in the SHR, and by comparison the lack of improvement/recovery of neurological function in the stroke-prone rat (i.e., with similar degrees of absolute brain injury) suggests that the stroke-prone rat might be a valuable model for the evaluation of neurodegenerative drugs poststroke [181]. From the same point of view, differential gene expression studies could be used in the future to compare SHR-SP and SHR in order to elucidate mechanisms and to discover new targets that facilitate neurobehavioral recovery of the injured brain. Gene Therapy and Gene Target Validation In recent years, clinical progress in gene therapy has proceeded in parallel with in vivo gene transfer for physiological studies and gene validation. Therapeutic neovascularization for ischemic diseases has been one particularly encouraging area of study. For example, animal models involving intramuscular injection of naked plasmid or adenoviral carried DNA encoding vascular endothelial growth factor (VEGF) have shown promotion of angiogenesis in ischemic limbs [194, 195]. Similarly, in clinical trials, VEGF gene transfer augments the population of circulating endothelial progenitor cells and transiently increases plasma levels of VEGF [196]. Furthermore, myocardial transfer of naked plasmid DNA phVEGF(165) has been found to augment perfusion of ischemic myocardium and reduces the size of defects documented at rest by single-photon emission CT-imaging [197]. Ex vivo gene transfer, employing the modification of cultured cells and subsequent implantation into a host organism, is a proven strategy for recovery from CNS injury and could incorporate cells expressing genes that confer protection from stroke. An analogous area of research has been in recovery from long-term rodent hemiparkinsonism by implantation of cells following 6-OHDA lesioning, which has been found to improve behavioral deficit for up to 13 months [198]. In vivo gene transfer, the delivery of a gene directly to recipient somatic cells, has also been explored for neuroprotection and recovery from CNS injury. The delivery of the proto-oncogene bcl-2 has been examined in gerbil models using adeno-associated virus vectors. Transduction of both pre- and postforebrain ischemia was found to prevent DNA fragmentation in hippocampal CA1 neurons, commonly associated with cell death induced by ischemia [199]. Adenoviral transfection of the endogenous cytokine antagonist IL-1ra also has demonstrated neuroprotection in transient focal cerebral ischemia and


365

reperfusion models in the mouse [200, 201]. In addition, the neuroprotective potential of heat shock protein-70 on brain injury, including that produced by brain ischemia, has been tested with viral vectors. Use of transgenic animals and gene transfer technology to overexpress heat shock protein-70 clearly protects the brain [202, 203] in the context of stroke.

The Elusiveness of Novel Gene Targets and the Future

Initially, the movement to discover differential gene expression in brain destruction or protection was driven by the hope of discovering novel genes that would provide pathways to therapeutics. The complexity of understanding and applying resources to gene fragments in the hopes of ultimately reaching this ‘therapeutic nirvana’ has not yet occurred. We have identified several genes that have been differentially expressed in ischemic or tolerant brain tissue, and which appear to be important, but there still exists a therapeutic void. In pharmaceutical development, the odds of developing a successful drug are generally better for the pursuit of known genes as therapeutic targets, although the new science of pharmacogenetics as a lead optimization tool may change this manner of operation. Many other factors can make investing resources into work on unknown genes costly, risky, and difficult to pursue. Some of these include the absence of any understanding of an identifiable function for the unknown protein, and if it is in fact associated with a novel gene, and lack of any concrete, supporting information that the novel gene/protein has any involvement in the pathophysiology of stroke rather than being an epiphenomenon. In spite of all this, it is clear that the therapeutic targets of the future exist in the complex patterns of gene expression underlying stroke. As such, unknown or novel genes represent a potential source of collaboration between academic and industrial laboratories. Potential novel gene products could be evaluated for tissue distribution, function and relevance in tissue injury and protection, in a collaborative and productive setting. Clearly, the methodology is available to identify differential gene expression in stroke in addition to other conditions of brain disease. However, the methodology needs to be developed with the caveats discussed above. If one operates as we have suggested, using cross-validating technologies in animal models and human tissue to validate ‘hits,’ there is potential for many significant opportunities for the biological discovery, not only related to stroke, but in many other conditions such as end organ failure in various cardiovascular diseases, in areas such as oncology, and perhaps extending further to other very complex problems such as substance abuse, tolerance, addiction and drug dependency.

Barone/Read

366

Acknowledgements The authors would like to thank their collaborators in this research, especially Dr. Andy Parsons, Dr. David Harrison, Dr. Karen Philopia, Dr. Karen Kabnick, Dr. Shawn O’Bien, Dr. Steve Clark, Dr. Mary Brawner, Dr. Giora Feuerstein, Dr. Xinkang Wang, Dr. Ray White, Dr. Stewart Bates, Dr. Jeff Legos and Dr. Israel Gloger. It has been a pleasure working with all of them in this exciting area.

References 1

2 3

4

5 6 7 8 9 10

11 12 13

14

15 16 17

Stanton LW, Garrard LJ, Damm D, Garrick BL, Lam A, Kapoun AM, Zheng Q, Protter AA, Schreiner GF, White RT: Altered patterns of gene expression in response to myocardial infarction. Circ Res 2000;86:939–945. Barone FC, Feuerstein GZ: Inflammatory mediators and stroke: New opportunities for novel therapeutics (Review). J Cereb Blood Flow Metab 1999;15:819–834. Soriano MA, Tessier M, Certa U, Gill R: Parallel gene expression monitoring using oligonucleotide probe arrays of multiple transcripts with an animal model of focal ischemia. J Cereb Blood Flow Metab 2000;20:1045–1055. Read SJ, Parsons AA, Harrison DC, Philpott K, Kabnick K, Brawner M, O’Brien S, Clark S, Bates S, Gloger I, Legos JJ, Barone FC: Stroke genomics: Approaches to identify, validate and understand adaptive gene expression changes in ischemic stroke (Review). J Cereb Blood Flow Metab 2001;21:755–778. Sharp FR, Lu A, Tang Y, Millhorn DE: Multiple molecular penumbras after focal cerebral ischemia. J Cereb Blood Flow Metab 2000;20:1011–1032. Lu A, Tang Y, Ran R, Clark JF, Aronow BJ, Sharp FR: Genomics of the periinfarction cortex after focal cerebral ischemia. J Cereb Blood Flow Metab 2003;23:786–810. Kannel WB, Wolf PA, Verter J, McNamara PM: Epidemiologic assessment of the role of blood pressure in stroke. The Framingham study. JAMA 1970;214:301–310. Stephenson J: Rising stroke rates spur efforts to identify risks, prevent disease. JAMA 1998; 279:1239–1240. Fisher M, Bogousslavsky J: Further evolution toward effective therapy for acute ischemic stroke. JAMA 1998;279:1298–1303. Pancioli AM, Broderick J, Kothari R, Brott T, Tuchfarber A, Miller R, Khoury J, Jauch E: Public perception of stroke warning signs and knowledge of potential risk factors. JAMA 1998;279: 1288–1292. Hrubec Z, Robinette CD: The study of human twins in medical research. N Engl J Med 1984; 310:435–441. Brass LM, Isaacsohn JL, Merikangas KR, Robinette CD: A study of twins and stroke. Stroke 1992;23:221–223. Carmelli D, DeCarli C, Swan GE, Jack LM, Reed T, Wolf PA, Miller BL: Evidence for genetic variance in white matter hyperintensity volume in normal elderly male twins. Stroke 1998;29: 1177–1181. Jousilahti P, Rastenyte D, Tuomilehto J, Sarti C, Vartiainen E: Parental history of cardiovascular disease and risk of stroke. A prospective follow-up of 14371 middle-aged men and women in Finland. Stroke 1997;28:1361–1366. Kiely DK, Wolf PA, Cupples LA, Beiser AS, Myers RH: Familial aggregation of stroke. The Framingham Study. Stroke 1993;24:1366–1371. Liao D, Myers R, Hunt S, Shahar E, Paton C, Burke G, Province M, Heiss G: Familial history of stroke and stroke risk. The Family Heart Study. Stroke 1997;28:1908–1912. Welin L, Svardsudd K, Wilhelmsen L, Larsson B, Tibblin G: Analysis of risk factors for stroke in a cohort of men born in 1913. N Engl J Med 1987;317:521–526.


367

18 19

20

21 22 23

24 25 26

27

28

29

30 31 32

33 34 35

36 37 38

39

Khaw KT, Barrett-Connor E: Family history of stroke as an independent predictor of ischemic heart disease in men and stroke in women. Am J Epidemiol 1986;123:59–66. Bromberg JE, Rinkel GJ, Algra A, Greebe P, van Duyn CM, Hasan D, Limburg M, ter Berg HW, Wijdicks EF, van Gijn J: Subarachnoid haemorrhage in first and second degree relatives of patients with subarachnoid. Brit Med J 1995;311:288–289. Tournier-Lasserve E, Joutel A, Melki J, Weissenbach J, Lathrop GM, Chabriat H, Mas JL, Cabanis EA, Baudrimont M, Maciazek J: Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy maps to chromosome 19q12. Nat Genet 1993;3: 256–259. Hirano M, Pavlakis SG: Mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes (MELAS): Current concepts. J Child Neurol 1994;9:4–13. Lossos A, Ben Hur T, Ben Nariah Z, Enk C, Gomori M, Soffer D: Familial Sneddon’s syndrome. J Neurol 1995;242:164–168. Joutel A, Bousser MG, Biousse V, Labauge P, Chabriat H, Nibbio A, Maciazek J, Meyer B, Bach MA, Weissenbach J: A gene for familial hemiplegic migraine maps to chromosome 19. Nat Genet 1993; 5:40–45. Hassan A, Markus HS: Genetics and ischemic stroke. Brain 2000;123:1784–1812. Sourander P, Walinder J: Hereditary multi-infarct dementia. Lancet 1977;1:1015. Chabriat H, Tournier-Lasserve E, Vahedi K, Leys D, Joutel A, Nibbio A, Escaillas JP, Iba-Zizen MT, Bracard S, Tehindrazanarivelo A: Autosomal dominant migraine with MRI white-matter abnormalities mapping to the CADASIL locus. Neurology 1995;45:1086–1091. Dichgans M, Mayer M, Muller-Myhsok B, Straube A, Gasser T: Identification of a key recombinant narrows the CADASIL gene region to 8 cM and argues against allelism of CADASIL and familial hemiplegic migraine. Genomics 1996;32:151–154. Hedera P, Friedland RP: Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy: Study of two American families with predominant dementia. J Neurol Sci 1997;146:27–33. Desmond DW, Moroney JT, Lynch T, Chan S, Chin SS, Shungu DC, Naini AB, Mohr JP: CADASIL in a North American family: Clinical, pathologic, and radiologic findings. Neurology 1998;51:844–849. Viitanen M, Kalimo H: CADASIL: Hereditary arteriopathy leading to multiple brain infarcts and dementia. Ann NY Acad Sci 2000;903:273–284. Chabriat H, Levy C, Taillia H, Iba-Zizen MT, Vahedi K, Joutel A, Tournier-Lasserve E, Bousser MG: Patterns of MRI lesions in CADASIL. Neurology 1998;51:452–457. Joutel A, Corpechot C, Ducros A, Vahedi K, Chabriat H, Mouton P, Alamowitch S, Domenga V, Cecillion M, Marechal E, Maciazek J, Vayssiere C, Cruaud C, Cabanis EA, Ruchoux MM, Weissenbach J, Bach JF, Bousser MG, Tournier-Lasserve E: Notch3 mutations in CADASIL, a hereditary adult-onset condition causing stroke and dementia. Nature 1996;383:707–710. Artavanis-Tsakonas S, Rand MD, Lake RJ: Notch signaling: Cell fate control and signal integration in development. Science 1999;284:770–776. Levitan D, Greenwald I: Facilitation of lin-12-mediated signalling by sel-12, a Caenorhabditis elegans S182 Alzheimer’s disease gene. Nature 1995;377:351–354. Joutel A, Vahedi K, Corpechot C, Troesch A, Chabriat H, Vayssiere C, Cruaud C, Maciazek J, Weissenbach J, Bousser MG, Bach JF, Tournier-Lasserve E: Strong clustering and stereotyped nature of Notch3 mutations in CADASIL patients. Lancet 1997;350:1511–1515. Joutel A, Tournier-Lasserve E: Notch signalling pathway and human diseases. Semin Cell Dev Biol 1998;9:619–625. Ruchoux MM, Maurage CA: Endothelial changes in muscle and skin biopsies in patients with CADASIL. Neuropathol Appl Neurobiol 1998;24:60–65. De Strooper B, Annaert W, Cupers P, Saftig P, Craessaerts K, Mumm JS, Schroeter EH, Schrijvers V, Wolfe MS, Ray WJ, Goate A, Kopan R: A presenilin-1-dependent gamma-secretaselike protease mediates release of Notch intracellular domain. Nature 1999;398:518–522. Kudo T, Imaizumi K, Tanimukai H, Katayama T, Sato N, Nakamura Y, Tanaka T, Kashiwagi Y, Jinno Y, Tohyama M, Takeda M: Are cerebrovascular factors involved in Alzheimer’s disease? Neurobiol Aging 2000;21:215–224.

Barone/Read

368

40 41 42

43

44

45

46 47 48

49

50

51

52 53

54

55

56

57 58

59

60

Schmidt R, Schmidt H, Fazekas F: Vascular risk factors in dementia. J Neurol 2000;247:81–87. Skoog I: Vascular aspects in Alzheimer’s disease. J Neural Transm Suppl 2000;59:37–43. Ciafaloni E, Ricci E, Shanske S, Moraes CT, Silvestri G, Hirano M, Simonetti S, Angelini C, Donati MA, Garcia C: MELAS: Clinical features, biochemistry, and molecular genetics. Ann Neurol 1992;31:391–398. Macmillan C, Lach B, Shoubridge EA: Variable distribution of mutant mitochondrial DNAs (tRNA(Leu[3243])) in tissues of symptomatic relatives with MELAS: The role of mitotic segregation. Neurology 1993;43:1586–1590. Sakuta R, Goto Y, Horai S, Nonaka I: Mitochondrial DNA mutations at nucleotide positions 3243 and 3271 in mitochondrial myopathy, encephalopathy, lactic acidosis, and stroke-like episodes: A comparative study. J Neurol Sci 1993;115:158–160. Kovalenko SA, Tanaka M, Yoneda M, Iakovlev AF, Ozawa T: Accumulation of somatic nucleotide substitutions in mitochondrial DNA associated with the 3243 A-to-G tRNA(leu)(UUR) mutation in encephalomyopathy and cardiomyopathy. Biochem Biophys Res Commun 1996;222:201–220. Rastenyte D, Tuomilehto J, Sarti C: Genetics of stroke – A review. J Neurol Sci 1998;153:132–145. Kosinski C, Mull M, Lethen H, Topper R: Evidence for cardioembolic stroke in a case of KearnsSayre syndrome. Stroke 1995;26:1950–1952. Dominiczak AF, Negrin DC, Clark JS, Brosnan MJ, McBride MW, Alexander MY: Genes and hypertension: From gene mapping in experimental models to vascular gene transfer strategies. Hypertension 2000;35:164–172. Jeffs B, Clark JS, Anderson NH, Gratton J, Brosnan MJ, Gauguier D, Reid JL, Macrae IM, Dominiczak AF: Sensitivity to cerebral ischaemic insult in a rat model of stroke is determined by a single genetic locus. Nat Genet 1997;16:364–367. Brosnan MJ, Clark JS, Jeffs B, Negrin CD, Van Vooren P, Arribas SM, Carswell H, Aitman TJ, Szpirer C, Macrae IM, Dominiczak AF: Genes encoding atrial and brain natriuretic peptides as candidates