Gene Therapy of the Central Nervous System: From Bench to Bedside

Contributors

Thais Federici Lerner Research Institute, Cleveland Clinic Foundation: NB2-126A, 9500 Euclid Ave., Cleveland, OH 44195, USA.

Paul D. Acton Department of Radiology, Thomas Jefferson University, Philadelphia, PA, USA. Kaveh Asadi-Moghaddam Department of Neurological Surgery, The Ohio State University Medical Center N-1017 Doan HaU 410 W, 10th Avenue Columbus, OH 43210, USA.

David J. Fink University of Michigan, 1914 Taubman Center, 1500 East Medical Center Drive Ann Arbor, MI 48109-0316, USA.

Krystof Bankiewicz Department of Neurological Surgery, University of California, San Francisco, CA 94103, USA.

Helen L. Fitzsimons NY 10032, USA.

Gerard J. Boer Netherlands Institute for Brain Research, Meibergdreef 33,1105 AZ Amsterdam, The Netherlands.

Neurologix Research, Inc., New York,

John R. Forsayeth Department of Neurological Surgery, University of California, San Francisco, CA 94103, USA.

Nicholas M. Boulis Lemer Research Institute Cleveland Clinic Foimdation: NB2-126A, 9500 Euclid Ave., Cleveland, OH 44195, USA.

Cornel Fraefel Institute of Virology, University of Zurich, Zurich, Switzerland. Justin F. Eraser Department of Neurological Surgery, New York Presbyterian Hospital - Weill Cornell Medical Center, New York, USA.

Xandra O. Breakefield Department of Neurology and Radiology, Massachusetts General Hospital, and Program in Neuroscience, Harvard Medical School, Boston, MA 02114, USA.

Guangping Gao Gene Therapy Program, Department of Medicine, University of Pennsylvania School of Medicine, Philadephia, PA 19104, USA.

Peter Carmeliet Center for Transgene Technology and Gene Therapy Flanders Intenmiversitary Institute for Biotechnology KU Leuven, Campus Gasthuisberg, Herestraat 49, B-3000, Leuven, Belgium.

Joseph C. Glorioso Department of Molecular Genetics and Biochemistry, University of Pittsburgh, Pittsburgh, PA, USA.

E. Antonio Chiocca Department of Neurological Surgery, The Ohio State University Medical Center N-1017 Doan Hall 410 W, 10th Avenue Columbus, OH 43210, USA.

Steven A. Goldman Division of Cell and Gene Therapy, Department of Neurology, University of Rochester Medical Center, Rochester, NY 14580, USA.

Ronald G. Crystal Department of Genetic Medicine, Weill Medical College of Cornell University, New York, NY 10021, USA.

Thomas A. Green Department of Psychiatry and Center for Basic Neuroscience, The University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390, USA.

Jane Dunning Department of Neurological Surgery, Weill Medical College of Cornell University, New York, NY 10021, USA.

Neil R. Hackett Department of Genetic Medicine, Weill Medical College of Cornell University, New York, NY 10021, USA.

Matthew J. During Department of Neurological Surgery, New York Presbyterian Hospital, Weill Medical College of Cornell University New York, NY 10021, USA.

Piotr Hadaczek Department of Neurological Surgery, University of California, San Francisco, CA 94103, USA.

Marina E. Emborg Wisconsin National Primate Research Center and Department of Anatomy, University of Wisconsin - Madison, 1223 Capitol Court, Madison, WI53715, USA.

William T.J. Hendriks Netherlands Institute for Brain Research, Meibergdreef 33, 1105 AZ Amsterdam, The Netherlands.

IX

CONTRIBUTORS

Charles E. Inturrisi Department of Pharmacology Weill Medical College of Cornell University New York, NY 10021, USA.

Neurosurgery, Wiell Medical College of Cornell University, New York, NY 10021, USA.

Luc Jasmin Department of Anatomy and Neurological Surgery, University of California, San Francisco, CA 94143-0452, USA.

Eric J. Nestler Department of Psychiatry and Center for Basic Neuroscience, The University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390, USA.

Stephen M. Kaminsky Department of Genetic Medicine, Weill Medical College of Cornell University, New York, NY 10021, USA.

Francesco Noe Department of Neuroscience, Mario Negri Institute for Pharmacological Research, Via Eritrea 62, 20157 Milan, Italy

Michael G. Kaplitt Department of Neurological Surgery, New York Presbyterian Hospital, Weill Cornell Medical College of Cornell University New York, NY 10021, USA.

Peter T. O'Hara Department of Anatomy and W M . Keck Foimdation Center for Integrative Neuroscience, University of California, 513 Parnassus Ave., San Francisco, CA 94143-0452, USA.

Matthias Klugmann Department of Neurobiology, Interdisciplinary Center for Neurosciences (IZN), University of Heidelberg, Im Neuenheimer Feld 364, 69120 Heidelberg, Germany. Diether Lambrechts Center for Transgene Technology and Gene Therapy Flanders Interuniversitary Listitute for Biotechnology KU Leuven, Campus Gasthuisberg, Herestraat 49, B-3000, Leuven, Belgium. Patricia A. Lawlor Department of Molecular Medicine & Pathology, Faculty of Medical and Health Sciences, The University of Auckland, New Zealand. Claudia B. Leichtlein Department of Neurobiology, Interdisciplinary Center for Neurosciences (IZN), University of Heidelberg, Im Neuenheimer Feld 364, 69120 Heidelberg, Germany Neal Luther Department of Neurological Surgery, New York Presbyterian Hospital, Weill Cornell Medical College of Cornell University New York, NY, USA. Marina Mata Department of Neurology, University of Michigan and VA Ann Arbor Healthcare System, Ann Arbor, MI 48109, USA. Jerry R. Mendell Ohio State University, Center for Gene Therapy, Columbus Children Research Institute, 700 Colombus Drive, Columbus, OH 43205, USA. Anne Messer Wadsworth Center, New York State Department of Health, and Department of Biomedical Sciences, University of Albany Albany, NY 12201, USA. Andra Miller The Biologies Consulting Group 6113 Walhonding Road, Bethesda, MD 20816, USA.

Sonoko Ogawa Department of Kansei and Cognitive Brain Science, University of Tsukuba, Tsukuba, Japan. Donald W. Pfaff Laboratory of Neurobiology and Behavior, The Rockefeller University New York, NY 10021, USA. Harish Poptani Department of Radiology, School of Medicine, University of Pennsylvania, Philadelphia, PA, USA. Marc J. Ruitenberg Red's Spinal Cord Research Laboratory, School of Anatomy and Human Biology, & Western Australian Institute for Medical Research and UWA Centre for Medical Research, The University of Western Australia, Crawley, Perth, Western Australia. Claudia Senn Institute of Virology, University of Zurich, Zurich, Switzerland. Fraser Sim Division of Cell and Gene Therapy, Department of Neurology, University of Rochester Medical Center, Rochester, NY 14580, USA. Dolan Sondhi Department of Genetic Medicine, Weill Medical College of Cornell University, New York, NY 10021, USA. Mark M. Souweidane Department of Neurological Surgery, Weill Medical College of Cornell University, New York, NY 10021, USA. Qingshan Teng Lemer Research Institute, Cleveland Clinic Foundation: NB2-126A, 9500 Euclid Ave., Cleveland, OH 44195, USA. Luk H. Vandenberghe Department of Medicine, Gene Therapy Program, Katholieke Universiteit Leuven, Kapucijnenvoer 33, B-3000 Leuven, Belgium.

Todd W. Miller Wadsworth Center, New York State Department of Health and Department of Biomedical Sciences, University of Albany Albany, NY 12201, USA and Vanderbilt University 2200 Pierce Ave., PRB 618 Nashville, TN 37232, USA.

Joost Verhaagen Netherlands Institute for Brain Research, Meibergdreef 33,1105 AZ Amsterdam, The Netherlands.

Jeffrey Moirano Department of Medical Physics, University of Wisconsin - Madison, 1300 University Avenue, The Madison, WI 3706, USA.

Annamaria Vezzani Department of Neuroscience, Mario Negri Institute for Pharmacological Research, Via Eritrea 62, 20157 Milan, Italy

Sergei Musatov Laboratory of Neurobiology & Behavior, The Rockefeller University and Laboratory of Molecular

Charles H. Vite WF. Goodman Center for Comparative Medical Genetics and Department of Clinical Studies,

CONTRIBUTORS

School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA, USA. James M. Wilson Department of Medicine, Gene Therapy Program, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, USA.

XI

John Wolfe W.R Goodman Center for Comparative Medical Genetics and Department of Pediatrics, University of Pennsylvania, Children's Hospital of Philadelphia, Philadelphia, PA 19104, USA.

Preface

Genetic manipulation in behavioral neurobiology is among the more ambitious and complex research fields, yet the contributions here demonstrate the power of this technology to understand the molecular biology of behaviors such as drug addiction and sexual behavior. Finally, stem cells have become a major source of hope to many with debilitating diseases, yet many scientific hurdles limiting effective human therapeutic applications remain. As outlined here, the marriage of gene therapy and stem cells has the potential to facilitate translation of this important technology into clinical practice. We hope that this volume will prove valuable to anyone interested in gene therapy. To those new to the field, we have asked several authors to highlight important methodological or technical issues, which could facilitate successful application of gene therapy to various areas of basic and translational neurobiology research. But we also expect that even the most seasoned gene therapy veteran will find many new and interesting things here. Despite having worked in this field for more than 15 years, we nonetheless found that most of the chapters offered valuable insights or interesting perspectives on many issues, which we are facing in our own labs and clinical practice. This effort would not have been possible without the help and understanding of many people. First, we are extremely grateful to all of the authors who have contributed to this book. Although the number of excellent scientists in this area are far too large to be included in a single volume, we believe that the reader will find many of the pioneers of neurological gene therapy represented here. In addition, we have tried to identify those newer investigators whose creativity and energy are helping to increase the breadth and promise of this field for the future. We are also indebted to Johannes Menzel, Senior Publishing Editor

Ten years ago, an earlier version of this book, entitled "Viral Vectors: Gene Therapy and Neuroscience Applications" focused upon developing gene transfer technologies in the brain and potential applications largely in neurobiological research. The title of the current volume and the contents reflect the enormous strides made in this field over the past decade. Since the last book, several gene therapy societies have either begun or expanded around the world to become large, robust organizations. Those interested in research or clinical gene therapy applications in the nervous system are now among the largest constituencies at the annual meetings of these groups. At the time of the earlier version, the only clinical trials ongoing were in neurooncology, and those were in their infancy. While neurooncology remains an important area, which is represented here, at least three other chapters reflect ongoing or completed human clinical trials of gene therapy for Parkinson's disease. Batten disease and Canavan disease. Several other contributions outline areas which may be in clinical trial at or soon after the printing of this book, including pain and epilepsy. With the expansion of clinical gene therapy applications, new considerations have recently arisen which are comprehensively reviewed here. Among these are the role of the immune system in both the safety and efficacy of gene transfer, functional imaging to follow both gene expression and the consequences to various brain regions, and the method of gene delivery to the human brain. Although clinical trials and associated issues have become more prominent over the past 10 years, many important and fascinating basic science applications of gene transfer remain as well. Technical issues relating to the efficiency of gene packaging and delivery are addressed, but we have endeavored not simply to review well-documented issues but rather focus upon newer areas such as synthetic or chimeric viruses.

xui

XIV

at Elsevier, and his assistant, Maureen Twaig, for their outstanding work in overseeing the development and completion of this project. We and many of the chapter authors have both research and clinical responsibilities, and often we have not been the easiest group to work with, but this book would likely never have been completed without their balance of patience and diligence. Finally, this is one of the rare opportunities

PREFACE

that we have to thank our families for their support not only during the production of this book, but over the many years of training and subsequent long hours and frequent travel which have allowed us to participate in the evolution of this exciting field. Michael G. Kaplitt Matthew J. During

C H A P T E R

1 Design and Optimization of Expression Cassettes Including Promoter Choice and Regulatory Elements Helen L, Fitzsimons, Matthew J. During

Abstract: Pivotal to the success of studies involving recombinant adeno-associated virus (rAAV)-mediated gene transfer to the brain is the design of the rAAV expression cassette and the selection of the rAAV serotype. Many promoters have been isolated that differ in cell-type specificity, size and strength. In addition, novel AAV serotypes are continually being isolated and characterized in vivo. These will differ in their cell-type-specific tropism, the efficiency of cellular transduction and the level and spread of gene expression mediated by the recombinant vectors. To that end, this chapter provides an introduction and summary of the promoters, regulatory elements and serotypes that are available and a guide to assist in the design of rAAV cassettes and selection of the appropriate rAAV serotype for a particular application. Keywords: adeno-associated virus; promoter; gene expression; brain; regulatory element

I.

INTRODUCTION

In the 10 years since recombinant adeno-associated virus (rAAV) was first used successfully to transduce neurons (Kaplitt et al, 1994) it has proved to be a very efficient vector for gene transfer to the brain. The field is moving at a cracking pace with technical improvements in production and purification, cloning of new serotypes and also the selection and characterization of new promoters and regulatory elements. These advances have enabled transgenes to be targeted to specific cell types in focal or widespread areas of the brain and have dramatically increased the number of disease targets amenable to gene therapy. A myriad of neurological disorders including Parkinson's disease, Huntington's disease, epilepsy and Alzheimer's disease may now be treatable using rAAV-mediated gene therapy. Each disorder has different requirements in terms of the specific cell type to be transduced and the level and range of therapeutic protein necessary to fall within the therapeutic window for that particular disease.

Gene Therapy of the Central Nervous System: From Bench to Bedside

The level of transgene expression is dependent on a number of factors. The choice of rAAV serotype influences the cell-type specificity and the dose of vector combined with the transduction efficiency of that particular serotype controls the spread of rAAV transduction within the tissue. Also critical to the success of rAAV as a gene transfer vector is the design of the expression cassette, which once delivered by the rAAV vector, maintains control over the level and duration of transgene expression within that cell.

II.

DESIGN OF THE rAAV CASSETTE

The minimum requirements of an rAAV expression cassette are a promoter, a transgene and a polyadenylation site flanked on either end by AAV inverted terminal repeats. The 4.7 kb wild-type AAV genome is very tightly folded into the 20 nm AAV particle. Various analyses of the maximum size of the genome that can be accommodated with the particle Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

1. DESIGN AND OPTIMIZATION OF EXPRESSION CASSETTES INCLUDING PROMOTER CHOICE

have been carried out. Xu et al. (2001) demonstrated that a 5n kilobase (kb) rAAV expression cassette containing luciferase under control of the rat preproenkephalin promoter was packaged into functional AAV particles, which facilitated luciferase expression in primary rat neuronal cultures and in the rat brain. In addition, Hermonat et al. (1997) reported that 900 bp of stuffer sequence could be inserted into the 4.7 kb wildtype AAV genome (corresponding to a total genome size of 5.6 kb) without compromising the wild-type phenotype. In contrast, however, it has also been reported that rAAV packaging is optimal between 4.1 and 4.9 kb, with a sharp reduction in packaging efficiency up to 5.2 kb (Dong et al., 1996). Addition of further DNA sequence precluded AAV packaging. These discrepancies in the maximum size of a transgene cassette that can be packaged into a functional rAAV particle may be reconciled by the possibility that each expression cassette has different topological constraints based on the tertiary folding structure of specific DNA sequences. Of particular interest is the finding by Mastakov et al. (2002), who showed that the use of different promoters within the AAV expression cassette altered the antigenicity of the capsid. The efficacy of re-administration of rAAV vectors was tested by re-injecting an rAAV2-luciferase vector into the rat striatum at certain time points after the first administration (into the contralateral side). If the vector was re-administered at 2 or 4 weeks post-injection, neutralizing antibodies were detected and luciferase activity was reduced by 90%; however, if the second vector was injected after an interval of 3 months, luciferase expression was not altered. An unexpected caveat to the study was the finding that if the second dose of rAAV vector contained a different transgene or promoter to the first dose, there was no decrease in expression or production of neutralizing antibodies from the second vector. These data suggest that the outer structure of the virion is influenced by the vector genome sequence (Mastakov et al., 2002). A possibility that has yet to be examined is that alterations in the capsid structure may also influence the vector tropism. It is becoming more obvious that obtaining optimal expression in a specific cell type is not as simple as selecting a cell-type-specific promoter of the desired size and strength. In fact the major influence over celltype expression is in many cases not the promoter being used but the inherent tropism of rAAV It is therefore pertinent at this point to discuss the impact that the tropism of the rAAV capsid has on achieving rAAV-mediated cell- and tissue-specific expression in the brain.

III.

CELL^TYPE^SPECIFIC TROPISM OF rAAV

Eleven distinct AAV serotypes have been isolated to date (Atchison et al., 1965; Mayor and Melnick, 1966; Bantel-Schaal and Zur Hausen, 1984; Gao et al., 2002; Mori et al., 2004), as have hundreds of AAV cap gene sequences (each representing a unique serotype), which were amplified from human and non-human primate tissues (Gao et al., 2003, 2004). The transduction properties of the vast majority have not yet been characterized in the brain. Recombinant AAV serotype 2 (rAAV2) was the first rAAV vector to be used in the brain and its pattern of transduction has been the most widely characterized. The primary cell surface receptors of rAAV2 are membrane-bound heparan sulfate proteoglycans (Summerford and Samulski, 1998), which are present throughout the brain and on the surface of neurons and glial cells (Fuxe et al., 1994). Two co-receptors for rAAV2 have so far been identified, the aVj85 integrin receptor (Summerford et al., 1999) and the human fibroblast growth factor receptor 1 (Qing et al., 1999). Bartlett et al. (1998) demonstrated that AAV was preferentially taken up into neurons in the rat brain by fluorescently labeling the wild-type AAV particle and thereby proving that the lack of expression in glia was not due to absence of promoter activity but lack of uptake. Many analyses of cell-type-specific expression have been performed following rAAV2-mediated transduction of enhanced green fluorescent protein (EGFP) or other reporter genes into various brain regions. When gene expression was driven by neuron-specific promoters (see Section IV.B) including the neuron-specific enolase (NSE) promoter (Peel et al., 1997; Klein et al., 1998, 2002a), the platelet-derived growth factor j^-chain (PDGF) promoter (Peel et al., 1997; Paterna et al., 2000), all of the transduced cells co-localized with the neuronal marker NeuN and failed to co-localize with the astrocytic marker glial fibrillary acidic protein (GFAP). Similarly, no glial transduction was detected when the expression cassette was under control of the cellular hybrid cytomegalovirus (CMV) immediateearly enhancer/chicken j8-actin (CBA) promoter (see Section IV.C; Klein et al., 2002b; Burger et al., 2004). Under some conditions, rAAV2-mediated transduction of glia has been observed. When the viral CMV promoter was used (see Section IV.A), approximately 1-1.5% of transduced cells were astrocytes (Klein et al., 1998). This was also observed when the CMV promoter was used in combination with the human jS-globin second intron (Scammell et al., 2003), a small

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

CHOICE OF PROMOTER

portion of human j8-globin exon two and exon three (the MD promoter, Mandel et al., 1998) or the first hGH intron (Shen et a l , 2000). The cell-type-specific tropisms of rAAVl and rAAV5 have more recently been characterized in the brain. The VPl capsid proteins of AAVl are reasonably well conserved with AAV2, sharing 83% amino acid identity; however AAV5 is more divergent, sharing 58% and S7% amino acid identity with AAVl and AAV2, respectively. The three serotypes all differ with respect to their target cell surface receptors. Neither AAVl nor AAV5 bind heparin (Chiorini et al., 1999; Rabinowitz et al., 2002). AAV5 binds a-2,3-and a-2,6-N-linked sialic acid with high affinity (Walters et al., 2001; Kaludov et al., 2001) and its receptor was recently determined to be the platelet-derived growth factor-a receptor (DiPasquale et al., 2003). The membrane receptor(s) for AAVl have not been identified. Although they enter the cell by different receptors a n d / o r pathways, rAAVl and rAAV5 also preferentially target to neurons. Passini et al. (2003) designed an expression cassette containing the human jS-glucuronidase (GUSB) promoter and transgene, the SV40 splice donor/acceptor and polyA flanked by AAV2 inverted terminal repeats (ITRs) and crosspackaged it into either the rAAVl or rAAV5 capsid. Following intraventricular injection into neonates, an analysis of the cell types transduced by either of the rAAV vectors revealed that almost all the transduced cells contained the neuron-specific marker NSE and none co-localized with an astrocytic or oligodendrocytic marker. Burger et al. (2004) carried out a comprehensive analysis of rAAVl or rAAV5mediated cell-type specific expression in many areas of the rat brain. An EGFP expression cassette driven by the CBA promoter was cross-packaged into either the rAAVl or rAAVS capsid and injected into the hippocampus, substantia nigra, striatum, globus pallidus and spinal cord. Following immunohistochemistry with cell-type specific markers, cell counts revealed that all of the transduced cells from both rAAV serotypes were neuronal. The authors calculated that based on the number of cells that were analyzed, the level of astrocytic transduction could not be higher than 0.05%. At odds with this conclusion are the results from Wang et al. (2003), who found that injection of a crosspackaged rAAVl vector containing EGFP driven by the CMV promoter into the mouse brain resulted in considerable glial expression although the majority of expression was neuronal. Astrocytic expression in the striatum and corpus callosum as well as oligodendrocytic and

m^icroglial expression within the white matter was detected. Following immunohistochemistry with celltjrpe specific markers, quantitative cell counts of transduced cells lining the ventricular space showed that approximately 36% of transduced cells were astrocytes, 8% were oligodendrocytes and 5% were microglial cells (Wang et al., 2003). A possible explanation for the differences observed in the level of glial transduction between these studies is that rAAV has the ability to transduce glia at low frequencies and the level of glial transduction observed depends on the activity of the promoter in glia. It is clear from the rAAV2-based studies on cell-specific tropism that viral promoters (see Section IV.A) such as CMV promoter or the rous sarcoma virus (RSV) promoter appears to allow a higher level of expression in glial cells whereas the neuronal NSE promoter and the CBA promoter appear to be almost silent. The method of vector administration, site of injection and the use of dissimilar titers between research groups may also account for the differences observed. In addition, rAAV produced by different purification methods may produce different results as cellular contaminants in the vector stocks can alter transduction efficiency (Tenenbaum et a l , 1999). In summary, if rAAVl, rAAV2 or rAAV5 are injected into the brain with the transgene under control of a neuronal promoter the cell-specific pattern of expression will most likely be entirely neuronal. If a viral-derived promoter is used there may be a low level (probably less than 2%) of astrocytic a n d / o r oligodendrocytic/microglial expression, depending on which area of the brain is injected and at what titer. In many instances, although neuronal expression is desired, the level of glial expression will be low enough to be considered insignificant. A promoter that drives neuronspecific expression is thus not absolutely required and in selection of a promoter, equal if not more importance should be placed on its size and the level of expression it drives in the target brain area. IV>

CHOICE OF PROMOTER

The choice of promoter often depends upon a compromise between the levels of expression required, the target cell type and the size of promoter that can be accommodated by the cassette without overstretching the 4.7 kb rAAV packaging limit. As more efficient packaging and purification technologies are developed, leading to higher vector titers, the compromise is leaning less toward achieving an optimal level of

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

1. DESIGN AND OPTIMIZATION OF EXPRESSION CASSETTES INCLUDING PROMOTER CHOICE

expression but more toward the ability to target particular cell types. The following is a detailed description of the expression properties of viral, neuronal and hybrid promoters that have been most commonly used in concert with rAAV for transduction of the specific brain areas and cell types. A summary of the main characteristics of the promoters described in this chapter is provided in Table 1.

A,

Viral Promoters

1. The Human Cytomegalovirus Immediate/Early Qene Promoter and Enhancer (CMV promoter) The 0.7 kb CMV promoter has been the most widely used promoter for achievement of rAAV-mediated gene expression in the brain. Almost all the original studies examining the potential of rAAV as a vector for gene transfer to the brain employed the CMV promoter. In the first published example of rAAV-mediated transduction of the brain, Kaplitt et al. (1994) used an rAAV2 expression cassette consisting of the CMV promoter driving expression of tyrosine hydroxylase. Three weeks after injection of this vector into the rat striatum, approximately 1000 transduced neurons could be detected, however expression was significantly diminished 4 months post-injection (Kaplitt et al., 1994). Although rAAV-mediated gene expression driven by the CMV promoter can be detected 12 months following administration (Lo et a l , 1999; Tenenbaum et al., 2000), many researchers have reported a reduction of gene expression over time when using the CMV promoter (McCown et al., 1996; During et al., 1998; Klein et al., 1998; Lo et al., 1999; Tenenbaum et al., 2000). It is not entirely clear why this occurs, although silencing of the CMV promoter has been observed following methylation at CpG dinucleotides (Prosch et al., 1996). 2, The Rous Sarcoma Virus Long Terminal Repeat Promoter The small size of the 0.4 kb RSV promoter is an attractive property for use in rAAV expression cassettes. This promoter has been used to direct rAAV2, rAAV4 and rAAV5-mediated lacZ expression in the rat ependyma and striatum (Davidson et al., 2000). Recombinant AAV5-mediated expression was highest in the striatum, with approximately 5000 cells transduced in comparison to fewer than 100 for rAAV2 and rAAV4. In the ependyma, rAAV4 and rAAV5 both

facilitated transduction of around 200 ependymal cells 15 weeks post-injection. Nomoto et al. (2003) compared the performance of the RSV and CMV promoters in gerbil brain. Recombinant AAV2 and rAAV5 vectors containing lacZ under control of each of the RSV and CMV promoters were injected into the mongolian gerbil hippocampus (Nomoto et al., 2003). Transgene expression was examined 5 days post-injection by Xgal staining. Expression from the rAAV5 vectors was higher than that directed by the rAAV2 vectors. A comparison of CMV- versus RSV-driven expression for each serotype showed that while transduction rAAV2/CMV-lacZ resulted in a poor level of staining clustered around the stratum oriens, rAAV2/RSV-lacZ expression was more widespread in the pyramidal and granule cell layers. Recombinant AAV5/CMV-lacZ expression was present throughout the hippocampus whereas rAAV5/ RSV-lacZ expression was concentrated at high levels in the granule cell layer. A quantitative comparison of transgene expression was not carried out, although the overall level of gene expression directed by the RSV promoter appeared to be superior to that driven by the CMV promoter (Nomoto et al., 2003). The stability of the RSV promoter over a longer time frame has not been examined. B.

Neuron^Specific Promoters

1.

The Rat 'Neuron-Specific Enolctse Promoter

The 2.2 kb NSE promoter was originally shown to drive a high level of exclusively neuronal expression in the brains of NSE-lacZ transgenic mice (Forss-Petter et al., 1990). Use of the NSE promoter to drive EGFP expression from an rAAV2 vector resulted in a high level of EGFP expression in the rat spinal cord (Peel et al., 1997). Since then, this promoter has also been demonstrated to promote robust rAAV-mediated expression in many brain areas including the striatum (Mastakov et al., 2001; Xu et al., 2001) medial septum (Klein et a l , 1998), substantia nigra (Klein et al., 1998, 2002a; Peel and Klein, 2000; Xu et al., 2001) and the hippocampus (Klein et al., 1998, 2002a; Xu et al., 2001). Klein et al. (1998) compared rAAV2/NSE-driven expression in the rat striatum and substantia nigra to that controlled by the CMV promoter and found that expression driven by CMV declined to barely detected levels by 3 months post-injection (Klein et al., 1998) whereas NSE-driven expression was eightfold higher than that of CMV at its peak and furthermore expression remained stable over the 3-month

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

CHOICE OF PROMOTER

ON rt^ T-H 0 >

^"

ON

O

^1

^^ 0 0

§

d,

'^ ^ ^ o ^ 2 "S ^

o o +^ ' ^ ^ +^ •+e C|2 o) c o

^

CIH

T5

(« o)

^-^ 0

CO CN 0 ^^

'^l ' — C ^ :

X ;^

0 0

1^1

cd

^ 0 0 C o o

tj

'oJ CSCi

3 -13 + . ^^

0)

C

bp

13

1

0

0

rsi

'0

% bO

lu

a> ?^

2; Q ^

^ 0

0 rsj ir-^ 0 C 0

O

C» C3N ON

1^

_ J T3

ON ON ON

I

I

"a; "S

0) CU

^cu. s0). s0)

C

^

u

fii 3 3

4:;

a;

2

§

> t! O

g •? A

A

A

A

3

3

3

en a>

0

A

Q)

X OS

> ^

U

;-(

^>

TJ i^ CC!

73

^

CU ^T"

Si OH

CLI OH

X

w

<

C3 73 OJ

73 73 0)

O)

u ?J

CO

n

u cu

cu

w I .a

0)

CO 73 (U <J

13

^

^ s ^

^

B

CO

2(3

O

cu 73

en

o)

I? X

8-

o w

g

0 •43

73 73 CU CU 73

en a;

en

5^

13

:3

tt,

w

en 0

?^ X

en 0

0

OH

0

OH

z

w

OH

z

w

en

2

g

g

^

T1 CU

O '

>

u ti

^

"S "S

4^2 ^ 0 3^

ti

^o R ^

u

« > fd

73

C

0


0 O.

^ ^> cu •s

.5 C

0) OH

.j=; •*-" 5H

0)

^ u 0) X

rX^

u

x^ tb ^^ bJD

.^3 3

73 "^ 0 "^ ' ^ c:^

73 0

73 0) cu en

11

^

bO O

2 2-i ^ -a

•Jt ^

w

M-(

^H

VH CU OH C

cd ns

S

§

a3

iij

OH

VM

O

0) O C > b O'?3 -^ fd ;::i

5^^

W

J3

o

en

^ o ^ O ii CO t ^ l ^-> o >, T-H 4-; (d b>. ^ CU . ^

^

en §

u w ,«

P^

'G ;5 S Cfi

g

>

u

S ^ ^

CO

<

;3

I 2 0)

s ?? -«

2

H-J

a

cu

U

2 0

;3 cu

^

a»

0)

^

PH

^

CD ^

u U

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

"en

CU 0) .S '^ 3 -^3 g e 2 - | ;^O cuOH

o

Q P-H

I:

2 S

o

>

J

^ ^ X

2 bo2 >

0

g s en

pin

S g ^

(U

.s .s

gl ?>

^

;3

8

1. DESIGN AND OPTIMIZATION OF EXPRESSION CASSETTES INCLUDING PROMOTER CHOICE

duration of the study. Subsequently, the NSE promoter has been demonstrated to promote stable expression for over a year in the rat basal forebrain (Peel and Klein, 2000) and for over 25 months in the rat substantia nigra (Klein et al., 2002a). The strength of the NSE promoter was compared with eight other promoters in the rat cortex, hippocampus, substantia nigra and striatum by Xu et al. (2001). Recombinant AAV-mediated luciferase expression driven by the NSE promoter was highest in all brain areas, being up to 69-fold higher than CMV, and up to 8- and 20-fold higher than EFla and GFAP, respectively. 2. The Human Platelet-Derived Qrowth Factor p-Chain Promoter The PDGF promoter drives neuron-specific gene expression of a-synuclein in transgenic mice (Masliah et al., 2000). Recombinant AAV2/PDGF-mediated expression has been demonstrated in the rat spinal cord, with EGFP levels higher than that driven by the CMV promoter but three times lower than that driven by the NSE promoter. Interestingly, in motor neurons of the cervical enlargement, PDGF drove a higher level of EGFP than NSE (Peel and Klein, 2000). In addition, rAAV2/PDGF-luciferase expression was found to be approximately ninefold higher in the rat substantia nigra than that driven by rAAV2/CMV, however, expression levels were equivalent in the striatum (Wang et al., 2005). The data from these studies suggest that the activity of the PDGF promoter varies significantly across different brain areas. Paterna et al. (2000) also compared the strength of the PDGF and CMV promoters in the rat substantia nigra. By counting the percentage of dopaminergic neurons that were transduced, they estimated that the PDGF promoter was approximately threefold stronger than the CMV promoter. In a quantitative analysis of vector stability, rAAV2/PDGF was shown to promote stable transgene expression in the rat striatum for at least 12 weeks post-injection (Wang et al., 2005). The PDGF promoter has also been used in studies of lysosomal storage disorders. Delivery of rAAV2 vectors containing PDGF-EGFP or PDGF-CDCrell to the rat substantia nigra pars compacta resulted in expression of CDCrell or EGFP in most dopaminergic neurons 10 days following surgery (Dong et al., 2003). 3*

The Human oc-Synapsin-l (hSYN) Promoter

The 480 bp hSYN promoter was recently used to drive rAAV2-mediated EGFP expression in rat brain

parenchyma. EGFP was observed in up to 40% of rat striatal neurons. This promoter drove a higher level of expression than the 824 bp mouse CMV promoter in the rat cortex, although expression levels were similar in the thalamus. In contrast, a predominance of mouse CMV-driven expression was found in the striosomes (Kiigler et al., 2003). A quantitative in vivo comparison with other promoters commonly used in rAAV vectors has not yet been carried out although rAAV2/hSYN and rAAV2/CBA vectors drove similar levels of expression in primary hippocampal cultures (Shevtsova et al., 2004). 4* The Rat MCH Peptide Promoter This promoter drives expression of the melanin-concentrating hormone (MCH) peptide in a subpopulation of neurons within the dorsal lateral hypothalamus (Thompson and Watson, 1990). Injection of rAAV2 vectors containing EGFP under control of the 370 bp MCH promoter in the dorsal lateral hypothalamus resulted in transduction of almost exclusively MCH-positive neurons (Van den Pol et al., 2004). Ninety-eight percent of the EGFP-positive neurons were also MCH positive, whereas a different subset of adjacent neurons characterized by being immunopositive for hypocretin were not transduced. In addition, a control vector containing CMV-EGFP also transduced surrounding brains areas. These data suggest that the MCH promoter can restrict transgene expression almost exclusively to MCH-containing neurons although no transcriptional mechanism for exclusion from other neurons was given (Van den Pol et al., 2004). C.

Hybrid Promoters

L The CMV'Enhancer/ Human Platelet Derived Qrowth Factor P-Chain Promoter The hybrid 1.8 kb CMV/E-PDGF promoter was recently constructed and evaluated for its ability to enhance expression of rAAV2-luciferase over that driven by the PDGF or CMV promoters (Wang et al., 2005). Four weeks following injection of rAAV2 vectors containing each of the promoters into the rat striatum or substantia nigra, luciferase activity in striatal extracts was measured. The CMV/E-PDGF promoter was found to exhibit 10-fold higher activity than both promoters in the striatum and 3- and 17-fold higher activity than the PDGF and CMV promoters in the substantia nigra, respectively. Examination of the stability of the hybrid promoter over time revealed that there was no decrease in the luciferase activity of striatal extracts over a 6-month period (Wang et al., 2005).

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

CHOICE OF PROMOTER

2* The CMV'Enhancer/Chicken P-Actin Promoter The 1.7 kb CBA promoter has been shown to facilitate a high level of rAAV-mediated gene expression in many brain areas. It consists of the CMV enhancer and chicken jS-actin promoter fused to 90 nucleotides of exon one of the chicken jS-actin gene, 917 nucleotides of a hybrid chicken jS-actin/rabbit j6-globin intron and 55 nucleotides of rabbit jS-globin exon. Gene transfer of an rAAV/CBA expression cassette to the brain was first performed by Kaemmerer et al. (2000), which resulted in the transduction of numerous Purkinje cells in the mouse cerebellum. Since this time rAAV/CBA expression cassettes have been used in many studies to facilitate a high level of transgene expression. Extensive transduction of pyramidal cells in the rat hippocampus was observed 7 months after injection of rAAV2/CBA-EGFP (Peel and Klein, 2000), and therapeutic levels of jS-glucuronidase (GUSB) persisted for at least 1 year following intravenous administration of rAAV2/CBA-EGFP to MPSVII mice (Daly et al., 2001). Recombinant AAV/CBA mediated over-expression of EGFP or a-synuclein resulted in transduction of most of the neurons within the substantia nigra, with projection fibers visible throughout the striatum. The level of EGFP expression remained high for at least the yearlong duration of the study (Klein et al., 2002b). Comparison of rAAV2/CBA-EGFP with a corresponding rAAV2 vector containing EGFP under control of the NSE promoter resulted in a threefold higher level of EGFP-positive cells when the CBA promoter was used (Klein et al., 2002a), and combination of the CBA promoter with the woodchuck post-transcriptional regulatory element (WPRE, see Section I.J.I) resulted in a 7- to 50-fold higher transduction efficiency in the striatum and substantia nigra than expression driven by the MD promoter (which is composed of the CMV enhancer and a portion of the human jS-globin second and third exons, Bjorklund et al., 2000). In our laboratories we have used a shorter 933 bp version of the CBA promoter in many of our rAAV expression cassettes. This promoter is composed of 266 bp of the CMV enhancer and a 410 bp sequence containing exon one of chicken j?-actin, a hybrid chicken jS-actin/rabbit jS-globin intron and the 5' end of a rabbit j?-globin exon. Recombinant AAV vectors with transgenes under control of the 933 bp CBA promoter have been used in studies targeting robust expression to various brain regions including the subthalamic nucleus (During et al., 2003), hippocampus (During et al., 2003; Francis et al., 2004; Klugmann et al., 2005), striatum

(McBride et al., 2003), nucleus accumbens (Heusner et al., 2003), caudate putamen (Heusner et al., 2003; Cannon et al., 2004) and hypothalamus (Noordmans et al., 2004). D,

Targeting Glia

Due to the inherent neuronal tropism of AAV, glia have in most cases proved particularly difficult to transduce with any efficiency in the brain. Transduction of glia may be desirable for development of gene therapy-based treatments for demyelinating diseases such as multiple sclerosis and Canavan disease. Under certain conditions, however, rAAV2 has been shown to transduce glial cells very efficiently. Most of these examples are from in vitro experiments in which rAAV vectors were applied to glial cultures. Transduction of 70% of mouse BAS8.1 astrocytic cells and 10% of mouse N19 oligodendrocytic cells occurred after application of an rAAV vector expressing GAD65 under control of the CMV promoter (Mi et al., 1999). Similarly, after application of an rAAV/CBA-EGFP vector to primary rat brain astrocytes and microglia, transduction of 98% and 75% of the cells, respectively, was observed. However, when the same vector was applied to the astrocytes co-cultured with neurons, only the neurons were transduced (Gong et al., 2004). Significant transduction of glial cells has also been observed in white matter of the rodent brain. For example, delivery of rAAV2/CMV-EGFP to the posterior striatum resulted in neuronally restricted expression in the striatum, but glial expression was observed in the internal capsule, which is rich in glia (Tenenbaum et al., 2000) and furthermore, following injection of rAAV2/GFAP-EGFP into injured rat spinal cord, 15-30% of EGFP positive cells co-localized with astrocytic markers (Klein and Peel, 2000). These data indicate that there is no absolute preclusion of rAAV to transduce glial cells, rather it appears that significant transduction occurs only when neurons are absent, such as in glial cell cultures, sparse, such as in the corpus callosum, or severely depleted, such as after ischemia. Several glial promoters have been isolated and evaluated with respect to rAAV-mediated targeting of glia. £•

Glial Promoters

1 • The Human Qlial Fihrillary Acidic Protein Promoter In transgenic mice, the 2.2 kb GFAP promoter directs expression specifically to astrocytes (Brenner et al..

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

10

1. DESIGN AND OPTIMIZATION OF EXPRESSION CASSETTES INCLUDING PROMOTER CHOICE

1994), however use of a glial cell-specific promoter will not guarantee rAAV-mediated expression restricted exclusively to glia. In two separate studies, rAAV-mediated EGFP expression under control of the GFAP promoter resulted in primarily neuronal expression in rat spinal cord (Peel and Klein, 2000), hippocampus and striatum (Xu et al., 2001) with around 1-5% of transduced cells appearing glial in morphology (Peel and Klein, 2000; Xu et a l , 2001). This was highly surprising considering the astrocyte-specific expression in transgenic mice. The AAV2 ITRs contain an enhancer that can direct a low level of gene expression in the absence of a promoter (Flotte et al., 1993; Haberman et a l , 2000), therefore a probable explanation is that the enhancer activity of the ITRs is altering GFAP promoter activity to allow gene expression when it would otherwise be silent. 2.

The Mouse Myelin Basic Protein (MBP) Promoter

The MBP promoter drives oligodendrocyte-specific expression in transgenic mice (Gow et al., 1992). Injection of rAAV2/EGFP driven by the MBP promoter into the mouse cerebrum resulted in expression restricted to cells in the corpus callosum that were morphologically characteristic of oligodendrocytes (Chen et al., 1998,1999). Unlike that seen with the GFAP promoter, MBP promoter activity appeared to be silent in neurons. This is a very encouraging finding for researchers attempting to transduce oligodendrocytes, however the level of transduction was not particularly high. 3.

The Mouse F4/80 Microglial Promoter

F4/80 is a transmembrane spanning cell surface molecule that is a marker specific for macrophages (Cucchiarini et al., 2003). This protein is expressed in microglia, which are macrophages of the brain. The F4/80 promoter was isolated and inserted into an rAAV5 vector to control expression of red fluorescent protein (RFP). The vector was delivered to the rat striatum and RFP expression was analyzed 3 weeks post-injection. RFP was detected only in cells that also expressed the F4/80 antigen marker, confirming that gene expression was restricted to microglia in this cell population (Cucchiarini et al., 2003). The F4/80 promoter will therefore be a very useful tool for selective delivery of genes to microglia. F. Improving rAAV-Mediated Glial-Restricted Expression In order to boost the level of gene expression in glia and to achieve selective rAAV-mediated expression in

specific glial cell types, there are a few avenues that can be explored. The first is to further isolate and characterize glial promoters in order to find promoters that drive a high level of glial-restricted expression. The transduction profiles of newly isolated AAV serotypes will continue to be evaluated in the brain with one of the goals being to isolate a serotype that efficiently transduces glial populations, and preferably does not transduce neurons, which would allow a strong pan-cellular promoter such as CBA to be used. Another strategy is to alter the AAV capsid in order to produce a vector that is targeted specifically to glia. Much effort is being focused on mutation of the AAV capsid in order to target particular cell types, particularly for development of cancer treatments. For a review see Buning et al. (2003).

V.

REGULATORY ELEMENTS

A. The Woodchuck Post-Transcriptional Regulatory Element WPRE is a czs-acting RNA element originating from the woodchuck hepatitis B virus, which facilitates cytoplasmic accumulation and stability of RNA (Donello et al., 1998) therefore boosting the level of protein synthesis. Insertion of WPRE into an rAAV cassette between the transgene and the polyadenylation signal resulted in a six- to sevenfold increase in transgene expression following application of rAAV2 vectors to HEK293 cells and human primary fibroblasts (Loeb et al., 1999). The presence of WPRE also causes an increase in rAAVmediated transgene expression in vivo. Paterna et al. (2000) compared expression from rAAV/PDGF-EGFP vectors with or without WPRE. EGFP expression from the WPRE-containing vector was twofold higher in dopaminergic neurons of the substantia nigra and gene expression was stable over the 41-week duration of the study. The effect of WPRE on rAAV-mediated expression in other areas of the brain was evaluated by Xu et al. (2001). Addition of WPRE to an rAAV2/NSEluciferase vector resulted in a four- to ninefold increase in luciferase expression in the rat striatum, hippocampus, cortex and the nigra. This is in agreement with data from Klein et al. (2002a), who achieved an 11-fold increase in rAAV2/CBA-EGFP expression in the rat hippocampus on addition of WPRE. Many studies have now been published where robust rAAV-mediated transgene expression was achieved using expression cassettes containing WPRE in combination with the NSE promoter (Lin et al., 2003; Kells et al., 2004; Richichi et al., 2004) or the CBA

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

REGULATORY ELEMENTS

promoter (Arvidsson et al., 2003; Cao et al, 2003; During et al., 2003; Burger et al., 2004; Klein et al., 2004; Francis et al., 2004; Noordmans et al., 2004; Eslamboli et al., 2005; Klugmann et al., 2005). Furthermore, two human phase I clinical trials using these rAAV expression cassettes are also underway. They involve administration of rAAV2/ CBA-GAD65-WPRE to the subthalamic nucleus of Parkinson disease patients (During et al., 2001) and administration of rAAV2/NSE-ASPA-WPRE to the brains of patients with Canavan disease (Janson et al, 2002). Recently, attention was brought to the presence of a promoter and enhancer for the woodchuck hepatitis virus X protein and the sequence for the first 60 amino acids of this protein in the wild-type WPRE sequence. Hepadnavirus X proteins can alter expression of multiple cellular and viral genes and have been implicated in the generation of liver cancer (Kingsman et al., 2005). The authors state that although it is unlikely that the 60-amino-acid X protein peptide would be expressed as it requires the presence of a second enhancer, they advise modification of the X protein promoter and ATG in order to prevent any possible expression of X protein peptides. The WPRE element used routinely in our laboratories (and in the two clinical trials described above) was obtained from Thomas J. Hope (The Salk Institute for Biological Studies) who originally isolated and characterized WPRE (Donello et al., 1998). The X protein ATG and upstream regulatory sequences of this WPRE sequence have been mutated, therefore preventing X protein peptide expression. With the rAAV titers now far exceeding the titers that could be produced a few years ago (our laboratories routinely produce in excess of 10^^ rAAVl genomes from one Coming cell stack, which is equivalent to seven 15 cm^ dishes), and the resulting high levels of transgene expression that can now be achieved in vivo, in cases where the size of the expression cassette is an issue, the inclusion of WPRE may not be always desired and the inclusion of tissue-specific promoter or a bicistronic cassette may be favored. B.

Co-Expression of Two or More Genes

In some instances the delivery of two or more genes to the same cell is desired. This can be achieved either by injection of two separate rAAV vectors or by combining both transgenes into one rAAV vector. The latter scenario is more favorable as it does not depend on the assumption that both vectors wiU be co-delivered to all cells and at a 1:1 ratio. Co-expression of a potentially therapeutic gene with a reporter gene allows selective analysis of the effect of the therapeutic gene on the transduced cells.

11

Two genes can be co-expressed from the same vector to produce two separate proteins by one of three strategies: by inserting an internal ribosome entry site between the transgenes in order for both to be separately translated from the same mRNA; by inserting a cleavage site between the two genes which would allow co-translation and subsequent cleavage into the two protein products, or by co-expressing the two genes off a bidirectional promoter. C.

Internal Ribosome Entry Site (IRES)

An IRES initiates translation in a cap-independent manner, allowing synthesis of two proteins from a single bicistronic mRNA. Several different IRESs have been evaluated in rAAV vectors. 1 • The Encephalomyocarditis Virus Internal Rihosomal Entry Site The IRES in the EMCV 5'-untranslated region (UTR) was characterized by Jang and Wimmer (1990) and subsequently demonstrated by Ghattas et al. (1991) to direct co-expression of lacZ and choline acetyltransferase (CAT) from a recombinant provirus in chicken embryos. This 575 bp IRES was used in an rAAV2 expression cassette by During et al. (1998) to co-express tyrosine hydroxylase and aromatic acid decarboxylase in the monkey striatum. Expression of both genes was detected, however the EMCV IRES may not be the best choice of IRES as the translation of the second gene is less efficient than the first. This arrangement may be appropriate in some instances, such as when a reporter gene is placed in the downstream position and a level of expression comparable to the upstream gene is not required. 2» Poliovirus IRES The presence of an IRES in the poliovirus 5'-UTR was first described by Pelletier and Sonenberg (1988). Dirks et al. (1993) constructed a dicistronic cassette by placing a 628 bp sequence containing the poliovirus IRES between the luciferase and SEAP reporter genes under control of the SV40 promoter/enhancer. This lead to a 1:1 ratio of expression, regardless of the order of the reporter genes. The poliovirus IRES has been used in rAAV2 expression cassettes to co-express either of the potentially neurotrophic genes brain-derived neurotrophic factor (BDNF), GAP43, or nerve growth factor (NGF) with the EGFP reporter gene in the rat substantia nigra or basal forebrain. Expression of all the genes was detected and expression was stable over the duration of the studies (Klein et al., 1999a-c).

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

12

1. DESIGN AND OPTIMIZATION OF EXPRESSION CASSETTES INCLUDING PROMOTER CHOICE

3. Hepatitis C IRES

D.

A 230 bp IRES derived from hepatitis C, which was first described by Tsukiyama-Kohara et al. (1992), has also been used to co-express two genes via rAAVmediated gene transfer. Recombinant AAV2 vectors were engineered to express luciferase and blastocidin S deaminase (bsr), separated by the hepatitis C IRES (Urabe et a l , 1997). A corresponding vector with the EMCV IRES was also constructed. Following transduction of HEK293 cells with either of the rAAV vectors, there was no difference in luciferase activity, therefore expression of luciferase was not affected by the following IRESs. Selection of clones with blastocidin S was successful, indicating that both IRESs were capable of directing translation of the bsr gene (Urabe et al., 1997).

The 16-amino-acid-long 2A peptide of foot and mouth disease virus mediates cleavage of the foot-andmouth disease virus (FMDV) polyprotein to generate mature FMDV protein. This property was exploited by Furler et al. (2001) to allow co-translation and subsequent cleavage of proteins expressed from rAAV vectors. A fusion protein of EGFP and a-synuclein was expressed in the rat substantia nigra via rAAV-mediated transduction. The presence of the 48 nt 2A sequence between the two proteins allowed for efficient cleavage of the polyprotein as visualized by immunohistochemistry, although cleavage was not 100% complete. The level of EGFP that was produced was however higher than that directed by the EMCV IRES (Furler et al., 2001). The only potential disadvantage to this method is the residual 2A sequence that remains tagged to the 3' end of the upstream protein following cleavage. By performing site-directed mutagenesis of the FMDV 2A sequence, modified versions of 2A have been generated that direct greater than 99% cleavage in in vitro translation assays (Donnelly et al., 2001). The utility of the 2A sequence for co-expression of genes was highlighted by De Felipe and Ryan (2004), who created a tricistronic vector from which three discrete proteins from a single open reading frame could be co-ordinately produced and targeted to different subcellular locations. In an elegant experiment, a construct containing enhanced yellow fluorescent protein (EYFP), ECFP linked to a Golgi targeting signal and puromycin N-acetyltransferase (PAC), a puromycin resistance gene was transfected into HeLa ceUs. PAC conferred puromycin resistance, EYFP diffused throughout the cytoplasm and nucleus and ECFP was targeted to the Golgi apparatus, indicating that all three proteins were discretely produced (DeFelipe and Ryan, 2004).

4. Other IRESs Many other IRESs have been isolated that have yet to be evaluated in rAAV cassettes. A nine-nucleotide (nt) sequence from the 5'untranslated region (UTR) of the Gtx homeodomain protein was demonstrated to contain IRES activity by Chappell et al. (2000). Furthermore, multiple copies of the sequence were shown to increase IRES activity in a synergistic manner. Ten copies of the nine nt IRES, each separated by a nine nt spacer (approximately 200 bp in total) were placed between renilla luciferase and CAT under control of the SV40 promoter/ enhancer. This construct was transfected into neuro2a cells alongside a corresponding vector containing the EMCV IRES. Transfection of the EMCV-containing vector resulted in a 1:1 ratio of luciferase and CAT expression. Luciferase activity from the Gtx IREScontaining vector was 10-fold higher and CAT activity was 100-fold higher than that produced by the EMCV vector, leading to a luciferase/CAT ratio of 10-fold (Chappell et a l , 2000). Wong et al. (2002) compared the activity of several IRESs of human origin against the EMCV IRES in KB3-1 cells. A 338 nt IRES from Eukaryotic initiation factor 4 appeared to be a very potent IRES. When placed between CAT and lacZ, this IRES induced a 295-fold increase in ^-galactosidase expression over that of the EMCV IRES when normalized to CAT activity The superior performance of the EIF4C IRES was also observed in three other cell lines. The small sizes of the Gtx and EIF4C IRESs and the improved activities over the EMCV IRES suggest they may be valuable in the construction of dicistronic rAAV vectors.

E.

Cleavage of Polyproteins

Bidirectional Promoters

Constitutive bidirectional promoters are yet to be characterized in AAV vectors although analysis of the human genome revealed that up to 10% of promoters may control expression of two opposing genes (Trinklein et a l , 2004). VL

SUMMARY A N D CONCLUSION

The criteria for selection of a particular promoter and regulatory element(s) is based on the size of the transgene, the cell type to be transduced and the level of gene expression required for a particular application.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

SUMMARY AND CONCLUSION

The optimal packaging size of an rAAV vector is around 4.7 kb. Once the two 145 bp ITRs and a polyadenylation site such as the 250 bp bovine growth hormone or SV40 early polyA are taken into account, that leaves approximately 4.15 kb remaining to accommodate a promoter, transgene and possible regulatory elements. The most robust promoter that has currently been evaluated in the brain appears to be the 1 or 1.7 kb CBA promoter which drives high levels of neuronspecific gene expression in all brain areas it has been tested in. The 1.7 kb CMV/E-PDGF and 0.48 kb hSYN promoters also appear very promising based on their preliminary rAAV-mediated expression profiles and both warrant further characterization of their expression levels and stability in different brain areas. Addition of the WPRE element boosts gene expression from rAAV vectors by approximately 2- to 11-fold depending on the brain area and transgene. Combination of a strong promoter such as CBA with WPRE would leave room for a transgene of up to approximately 2.5 kb. If the transgene is larger, the WPRE element can be dropped, or alternatively the CBA promoter can be replaced with a shorter promoter such as the 0.48 kb hSYN promoter. For targeting cell types other than neurons, the F4/80 promoter drives microglia-specific expression, the MBP promoter drives oligodendrocyte-specific expression and the MCH promoter restricts expression to MCH neurons of the hypothalamus. As yet a promoter that restricts expression specifically to astrocytes has not been isolated. The GFAP promoter will facilitate some astrocytic expression however most of the cells expressing the transgene will be neurons. Three rAAV serotypes have currently been characterized in the brain. Recombinant AAVl and rAAV5 facilitate a similar level of expression in many brain areas, with both of them being 80-fold more efficient than AAV2 in the hippocampus and 8-10-fold higher in the striatimi (for a review see Burger et al., 2004). Therefore, rAAVl and rAAV5 are more ideal for applications in which widespread expression is desired, such as for transduction of the entire striatum. It should be noted that with the extremely high rAAV titers that can now be generated, care should be taken to assess the level of gene expression mediated by a particular viral stock before embarking on a large study. If the vector dose is too high it may spread outside the desired brain area or cause inadvertent cell death if the transgene is toxic. For applications in which small brain areas such as the substantia nigra pars compacta, subthalamic nucleus or hilar interneurons of the hippocampus are targeted, rAAV2 may be the best choice as it provides

13

efficient focal transduction without spreading to surrounding tissue. With these tools in hand, researchers will be well equipped to design rAAV cassettes appropriate for their particular applications and take a step closer toward successful gene therapy of neurological disorders. References Arvidsson, A., Kirik, D., Lundberg, C , Mandel, R.J., Andsberg, G., Kokaia, Z. and Lindvall, O. (2003) Elevated GDNF levels following viral vector-mediated gene transfer can increase neuronal death after stroke in rats. Neurobiol. Dis., 14: 542-556. Atchison, R.W., Casto, B.C. and Hammon, W.M. (1965) Adenovirusassociated defective virus particles. Science, 149: 754-756. Bantel-Schaal, U. and Zur Hausen, H. (1984) Characterization of the DNA of a defective human parvovirus isolated from a genital site. Virology, 134: 52-63. Bjorklund, A., Kirik, D., Rosenblad, C , Georgievska, B., Lundberg, C. and Mandel R.J. (2000) Towards a neuroprotective gene therapy for Parkinson's disease: use of adenovirus, AAV and lentivirus vectors for gene transfer of GDNF to the nigrostriatal system in the rat Parkinson model. Brain Res., 886: 82-98. Bartlett, J.S., Samulski, R.J. and McCown, T.J. (1998) Selective and rapid uptake of adeno-associated virus type 2 in brain. Hum. Gen. Ther., 9:1181-1186. Brenner, M., Kisseberth, W.C, Su, Y., Besnard, F. and Messing, A. (1994) GFAP promoter directs astrocyte-specific expression in transgenic mice. J. Neurosci., 14: 1030-1037. Buning, H., Ried, M.U., Perabo, L., Gemer, P.M., Huttner, N.A., Enssle, J. and Hallek, M. (2003) Receptor targeting of adeno-associated virus vectors. Gene Ther., 10:1142-1151. Burger, C , Gorbatyuk, O.S., Velardo, M.J., Peden, C.S., Williams, P , Zolotukhin, S., Reier, P.J., Mandel, R.J. and Muzyczka, N. (2004) Recombinant AAV viral vectors pseudotyped with viral capsids from serotypes 1, 2, and 5 display differential efficiency and cell tropism after delivery to different regions of the central nervous system. Mol. Ther., 10: 302-317. Cannon, CM., Abdallah, L., Tecott, L.H., During, M.J. and Palmiter, R.D. (2004) Dysregulation of striatal dopamine signaling by amphetamine inhibits feeding by hungry mice. Neuron, 44: 509-520. Chappell, S.A., Edelman, G.M. and Mauro, V.P (2000) A9-nt segment of a cellular mRNA can function as an internal ribosome entry site (IRES) and when present in linked multiple copies greatly enhances IRES activity. Proc. Natl. Acad. Sci. USA., 97:1536-1541. Chen, H., McCarty D.M., Bruce, A.T., Suzuki, K. and Suzuki, K. (1998) Gene transfer and expression in oligodendrocytes under the control of myelin basic protein transcriptional control region mediated by adeno-associated virus. Gene Ther., 5: 50-58. Chen, H., McCarty D.D., Bruce, A.T., Suzuki, K. and Sukuzi, K. (1999) Oligodendrocyte-specific gene expression in mouse brain: use of a myelin-forming cell type-specific promoter in an adenoassociated virus. J. Neurosci. Res., 55: 504-513. Chiorini, J.A., Kim, R, Yang, L. and Kotin, R.M. (1999) Cloning and characterization of adeno-associated virus type 5. J. Virol., 73: 1309-1319. Cucchiarini, M., Ren, X.L., Perides, G. and Terwilliger, E.E (2003) Selective gene expression in brain microglia mediated via adeno-associated virus type 2 and type 5 vectors. Gene Ther., 10: 657-667.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

14

1. DESIGN AND OPTIMIZATION OF EXPRESSION CASSETTES INCLUDING PROMOTER CHOICE

Daly, T.M., Ohlemiller, K.K., Roberts, M.S., Vogler, C.A. and Sands, M.S. (2001) Prevention of systemic clinical disease in MPS VII mice following AAV-mediated neonatal gene transfer. Gene Ther., 8: 1291-1298. Davidson, B.L., Stein, C.S., Heth, J.A., Martins, I., Kotin, R.M., Derksen, T.A., Zabner J., Ghodsi, A. and Chiorini, J.A. (2000) Recombinant adeno-associated virus type 2, 4, and 5 vectors: transduction of variant cell types and regions in the mammalian central nervous system. Proc. Natl. Acad. Sci., 97: 3428-3432. De Felipe, P. and Ryan, M.D. (2004) Targeting of proteins derived from self-processing polyproteins containing multiple signal sequences. Traffic, 5: 616-622. Di Pasquale, G., Davidson, B.L., Stein, C.S., Martins, I., Scudiero, D., Monks, A. and Chiorini, J.A. (2003) Identification of PDGFR as a receptor for AAV-5 transduction. Nat. Med., 10: 1306-1312. Dirks, W., Wirth, M. and Hauser, H. (1993) Dicistronic transcription units for gene expression in mammalian cells. Gene, 128: 247-249. Donello, J.E., Loeb, J.E. and Hope, T.J. (1998) Woodchuck hepatitis virus contains a tripartite posttranscriptional regulatory element. J. Virol., 72: 5085-5092. Donnelly, M.L., Hughes, L.E., Luke, G., Mendoza, H., ten Dam, E., Gani, D. and Ryan, M.D. (2001) The 'cleavage' activities of footand-mouth disease virus 2A site-directed mutants and naturally occurring '2A-like' sequences. J. Gen. Virol., 82(Pt 5): 1027-1041. Dong, J.Y., Fan, RD. and Frizzell, R.A. (1996) Quantitative analysis of the packaging capacity of recombinant adeno-associated virus. Hum. Gene Then, 7: 2101-2112. Dong, Z., Ferger, B., Patema, J.C., Vogel, D., Purler, S., Osinde, M., Feldon, J. and Bueler, H. (2003) Dopamine-dependent neurodegeneration in rats induced by viral vector-mediated overexpression of the parkin target protein, CDCrel-1. Proc. Natl. Acad. Sci. USA., 100: 12438-12443. During, M.J., Samulski, R.J., Elsworth, J.D., Kaplitt, M.G., Leone, P, Xiao, X., Li, J., Freese, A., Taylor, J.R., Roth, R.H., Sladek, J.R. Jnr., O'Malley, K.L. and Redmond, D.E. Jr. (1998) In vivo expression of therapeutic human genes for dopamine production in the caudates of MPTP-treated monkeys using an AAV vector. Gene Ther., 5: 820-827. During, M.J., Kaplitt, M.G., Stem, M.B. and Eidelberg, D. (2001) Subthalamic GAD gene transfer in Parkinson disease patients who are candidates for deep brain stimulation. Hum. Gene Ther., 12: 589-591. During, M.J., Cao, L., Zuzga, D.S., Francis, J.S., Fitzsimons, H.L., Jiao, X., Bland, R.J., Klugmann, M., Banks, W.A., Drucker, D.J. and Haile, C.N. (2003) Glucagon-like peptide-1 receptor is involved in learning and neuroprotection. Nat. Med., 9: 1173-1179. Eslamboli, A., Georgievska, B., Ridley, R.M., Baker, H.F., Muzyczka, N., Burger, C , Mandel, R.J., Annett, L. and Kirik, D. (2005) Continuous low-level glial cell line-derived neurotrophic factor delivery using recombinant adeno-associated viral vectors provides neuroprotection and induces behavioral recovery in a primate model of Parkinson's disease. J. Neurosci., 25: 769-777. Flotte, T.R., Afione, S.A., Conrad, C , McGrath, S.A., Solow, R., Oka, H., Zeitlin, PL., Guggino, W.B. and Carter, B.J. (1993) Stable in vivo expression of the cystic fibrosis transmembrane conductance regulator with an adeno-associated virus vector. Proc. Natl. Acad. Sci., 90: 10613-10617. Forss-Petter, S., Danielson, P.E., Catsicas, S., Battenberg, E., Price, J., Nerenberg, M. and Sutcliffe, J.G. (1990) Transgenic mice expressing beta-galactosidase in mature neurons under neuron-specific enolase promoter control. Neuron, 5: 187-197.

Francis, J.S., Dragunow, M. and During, M.J. (2004) Over expression of ATF-3 protects rat hippocampal neurons from in vivo injection of kainic acid. Brain Res. Mol. Brain Res., 124:199-203. Purler, S., Patema, J.-C, Weibel, M. and Bueler, H. (2001) Recombinant AAV vectors containing the foot and mouth disease virus 2A sequence confer efficient bicistronic gene expression in cultured cells and rat substantia nigra neurons. Gene Ther., 8: 864-873. Fuxe, K., Chadi, G., Tinner, B., Agnati, L.F., Pettersson, R. and David, G. (1994) On the regional distribution of heparan sulfate proteoglycan immunoreactivity in the rat brain. Brain Res., 636: 131-138. Gao, G.P, Alvira, M.R., Wang, L., Calcedo, R., Johnston, J. and Wilson, J.M. (2002) Novel adeno-associated viruses from rhesus monkeys as vectors for human gene therapy. Proc. Natl. Acad. Sci. USA, 99:11854-11859. Gao, G., Alvira, M.R., Somanathan, S., Lu, Y, Vandenberghe, L. H., Rux, J. J., Calcedo, R., Sanmiguel, J., Abbas, Z. and Wilson, J.M. (2003) Adeno-associated viruses undergo substantial evolution in primates during natural infections. Proc. Natl. Acad. Sci. USA, 100: 6081-6086. Gao, G.P, Vandenberghe, L.H., Alvira, M.R., Lu, Y, Calcedo, R., Xiangyang Zhou, X. and Wilson, J.M. (2004) Clades of adenoassociated viruses are widely disseminated in human tissues. J. Virol., 78: 6381-6388. Ghattas, I.R., Sanes, J.R. and Majors, J.E. (1991) The encephalomyocarditis virus internal ribosome entry site allows efficient coexpression of two genes from a recombinant provirus in cultured cells and in embryos. Mol. Cell. Biol., 12: 5848-5859. Gong, Y, Chen, S., Sonntag, C.R, Sumners, C , Klein, R.L., King, M.A., Hughes, J.A. and Meyer, E.M. (2004) Recombinant adeno-associated virus serotype 2 effectively transduces primary rat brain astrocytes and microglia. Brain Res. Brain Res. Protoc, 14:18-24. Gow, A., Friedrich, VL. Jr. and Lazzarini, R.A. (1992) Myelin basic protein gene contains separate enhancers for oligodendrocyte and Schwann cell expression. J. Cell. Biol. 119: 605-616. Haberman, R.P, McCown, T.J. and Samulski, R.J. (2000) Novel transcriptional regulatory signals in the adeno-associated virus terminal repeat A / D junction element. J. Virol, 74: 8732-8739. Hermonat, PL., Quirk, J.G., Bishop, B.M. and Han, L. (1997) The packaging capacity of adeno-associated virus (AAV) and the potential for wild-type-plus AAV gene therapy vectors. FEBS Lett., 407: 78-84. Heusner, C.L., Hnasko, T.S., Szczypka, M.S., Liu, Y, During, M.J. and Palmiter, R.D. (2003) Viral restoration of dopamine to the nucleus accumbens is sufficient to induce a locomotor response to amphetamine. Brain Res., 980: 266-274. Jang, S.K. and Wimmer, E. (1990) Cap-independent translation of encephalomyocarditis virus RNA: structural elements of the internal ribosomal entry site and involvement of a cellular 57-kD RNA-binding protein. Genes Dev, 9: 1560-1572. Janson, C , McPhee, S., Bilaniuk, L., Haselgrove, J., Testaiuti, M., Freese, A., Wang, D.J., Shera, D., Hurh, P., Rupin, J., Saslow, E., Goldfarb, O., Goldberg, M., Larijani, G., Sharrar, W , Liouterman, L., Camp, A., Kolodny, E., Samulski, J. and Leone, P (2002) Clinical protocol. Gene therapy of Canavan disease: AAV-2 vector for neurosurgical delivery of aspartoacylase gene (ASPA) to the human brain. Hum. Gene Ther., 13:1391-1412. Kaludov, N., Brown, K.E., Walters, R.W, Zabner, J. and Chiorini, J.A. (2001) Adeno-associated vims serotype 4 (AAV4) and AAV5 both require sialic acid binding for hemagglutination and efficient transduction but differ in sialic acid linkage specificity. J Virol., 75: 6884-6893.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

SUMMARY AND CONCLUSION Kaemmerer, W.R, Reddy, R.G., Warlick, C.A., Hartung, S.D., Mclvor, R.S. and Low, W.C. (2000) In vivo transduction of cerebellar Purkinje cells using adeno-associated virus vectors. Mol. Ther., 2: 446-457. Kaplitt, M.G., Leone, R, Samulski, R.J., Xiao, X., Pfaff, D.W., O'Malley, K.L. and During M.J. (1994) Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat. Genet., 8: 148-154. Kells, A.R, Pong, D.M., Dragunow, M., During, M.J., Young, D. and Connor, B. (2004) AAV-mediated gene delivery of BDNF or GDNF is neuroprotective in a model of Huntington disease. Mol. Ther., 9: 682-688. Kingsman, S.M., Mitrophanous, K. and Olsen, J.C. (2005) Potential oncogene activity of the woodchuck hepatitis post-transcriptional regulatory element (WPRE). Gene Ther., 12: 3-4. Klein, R.L., Meyer, P.M., Peel, A.L., Zolotukhin, S., Meyers, C., Muzyczka, N. and King, M.A. (1998) Neuron-specific transduction in the rat septohippocampal or nigrostriatal pathway by recombinant adeno-associated virus vectors. Exp. Neurol., 150: 183-194. Klein, R.L., Lewis, M.H., Muzyczka, N. and Meyer, P.M. (1999a) Prevention of 6-hydroxydopamine-induced rotational behavior by BDNP somatic gene transfer. Brain Res., 847: 314-320. Klein, R.L., McNamara, R.K., King, M.A., Lenox, R.H., Muzyczka, N. and Meyer, P.M. (1999b) Generation of aberrant sprouting in the adult rat brain by GAP-43 somatic gene transfer. Brain Res., 832:136-144. Klein, R.L., Muir, D., King, M.A., Peel, A.L., Zolotukhin, S., MoUer, J.C, Kruttgen, A., Heymach, J.V. Jr., Muzyczka, N. and Meyer, P.M. (1999c) Long-term actions of vector-derived nerve growth factor or brain-derived neurotrophic factor on choline acetyltransferase and Trk receptor levels in the adult rat basal forebrain. Neuroscience, 90: 815-821. Klein, R.L., Hamby M.P., Gong, Y, Hirko, A.C., Wang, S., Hughes, J.A., King, M.A. and Meyer, P.M. (2002a) Dose and promoter effects of adeno-associated viral vector for green fluorescent protein expression in the rat brain. Pxp. Neurol., 176, 66-74. Klein, R.L., King, M.A., Hamby M.E. and Meyer, P.M. (2002b) Dopaminergic cell loss induced by human A30P alpha-synuclein gene transfer to the rat substantia nigra. Hum. Gene Ther., 13: 605-612. Klein, R.L., Lin, W.L., Dickson, D.W., Lewis, J., Hutton, M., Duff, K., Meyer, P.M. and King, M.A. (2004) Rapid neurofibrillary tangle formation after localized gene transfer of mutated tau. Am. J. Pathol., 64: 347-353. Klugmann, M., Symes, C.W., Leichtlein, C.B., Klaussner, B.K., Dunning, J., Pong, D., Young, D. and During, M.J. (2005) AAV-mediated hippocampal expression of short and long Homer 1 proteins differentially affect cognition and seizure activity in adult rats. Mol. Cell Neurosci., 28: 347-360. Kugler, S., Lingor, R, SchoU, U., Zolotukhin, S. and Bahr, M. (2003) Differential transgene expression in brain cells in vivo and in vitro from AAV-2 vectors with small transcriptional control units. Virology, 311: 89-95. Lin, P.J., Richichi, C , Young, D., Baer, K., Vezzani, A. and During, M.J. (2003) Recombinant AAV-mediated expression of galanin in rat hippocampus suppresses seizure development. Pur. J. Neurosci., 18: 2087-2092. Lo, W.D., Qu, G., Sferra, TJ., Clark, R., Chen, R. and Johnson, PR. (1999) Adeno-associated virus-mediated gene transfer to the brain: duration and modulation of expression. Hum. Gene Ther., 10: 201-213.

15

Loeb, J.P., Cordier, W.S., Harris, M.P., Weitzman, M.D. and Hope, T.J. (1999) Pnhanced expression of transgenes from adeno-associated virus vectors with the woodchuck hepatitis virus posttranscriptional regulatory element: implications for gene therapy. Hum. Gene Ther., 10: 2295-2305. Mandel, R.J., Rendahl, K.G., Spratt, S.K., Snyder, R.O., Cohen, L.K. and Leff, S.P. (1998) 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., 18: 4271-4284. Masliah, P., Rockenstein, P., Veinbergs, I., Mallory M., Hashimoto, M., Takeda, A., Sagara, Y, Sisk, A. and Mucke, L. (2000) Dopaminergic loss and inclusion body formation in alpha-S5muclein mice: implications for neurodegenerative disorders. Science, 287:1265-1269. Mastakov, M.Y, Baer, K., Xu, R., Pitzsimons, H. and During, M.J. (2001) Combined injection of rAAV with mannitol enhances gene expression in the rat brain. Mol. Ther., 3: 225-232. Mastakov, M.Y, Baer, K., Symes, C.W., Leichtlein, C.B., Kotin, R.M. and During, M.J. (2002) Immunological aspects of recombinant adeno-associated virus delivery to the mammalian brain. J. Virol., 76: 8446-8454. Mayor, H.D. and Melnick, J.L. (1966) Small deoxyribonucleic acidcontaining viruses (picornavirus group). Nature, 210: 331-332. McBride, J.L., During, M.J., Wuu, J., Chen, E.Y, Leurgans, S.P. and Kordower, J.H. (2003) Structural and functional neuroprotection in a rat model of Huntington's disease by viral gene transfer of GDNP Pxp. Neurol., 181: 213-223. McCown, T.J., Xiao, X., Li, J., Breese, G.R. and Samulski, R.J. (1996) Differential and persistent expression patterns of CNS gene transfer by an adeno-associated virus (AAV) vector. Brain Res., 713: 99-107. Mi, J., Chatterjee, S., Wong, K.K., Porbes, C., Lawless, G. and Tobin, A.J. (1999) Recombinant adeno-associated virus (AAV) drives constitutive production of glutamate decarboxylase in neural cell lines. J. Neurosci. Res., 57:137-148. Mori, S., Wang, L., Takeuchi, T. and Kanda, T (2004) Two novel adeno-associated viruses from cynomolgus monkey: pseudotyping characterization of capsid protein. Virology, 330: 375-383. Nomoto, T, Okada, T, Shimazaki, K., Mizukami, H., Matsushita, T, Hanazono, Y, Kume, A., Katsura, K., Katayama, Y and Ozawa, K. (2003) Distinct patterns of gene transfer to gerbil hippocampus with recombinant adeno-associated virus type 2 and 5. Neurosci. Lett., 340: 153-157. Noordmans, A.J., Song, D.K., Noordmans, C.J., Garrity-Moses, M., During, M.J., Pitzsimons, H.L., Imperiale, M.J. and Boulis, N.M. (2004) Adeno-associated viral glutamate decarboxylase expression in the lateral nucleus of the rat hypothalamus reduces feeding behavior. Gene Ther., 11: 797-804. Passini, M.A., Watson, D.J., Vite, C.H., Landsburg, D.J., Peigenbaum, A.L. and Wolfe, J.H. (2003) Intraventricular brain injection of adeno-associated virus type 1 (AAVI) in neonatal mice results in complementary patterns of neuronal transduction to AAV2 and total long-term correction of storage lesions in the brains of betaglucuronidase-deficient mice. J. Virol., 77\ 7034-7040. Paterna, J-C, Moccetti, T , Mura, A., Peldon, J. and Biieler, H. (2000) Influence of promoter and WHV post-transcriptional regulatory element on AAV-mediated transgene expression in the rat brain. Gene Ther., 7:1304-1311. Peel, A.L., Zolotukhin, S., Schrimsher, G.W., Muzyczka, N. and Reier, P.J. (1997) Pfficient transduction of green fluorescent protein in spinal cord neurons using adeno-associated virus vectors containing cell type-specific promoters. Gene Ther., 4: 16-24.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

16

1. DESIGN AND OPTIMIZATION OF EXPRESSION CASSETTES INCLUDING PROMOTER CHOICE

Peel, A.L. and Klein, R.L. (2000) Adeno-associated vims vectors: activity and applications in the CNS. J. Neurosci. Meth., 98: 95-104. Pelletier, J. and Sonenberg, N. (1988) Internal initiation of translation of eukaryotic mRNA directed by a sequence derived from poliovirus RNA. Nature, 334: 320-325. Prosch, S., Stein, J., Staak, K., Liebenthal, C , Volk, H. and Kriiger, D.H. (1996) Inactivation of the very strong HCMV immediate early promoter by DNA CpG methylation in vitro. Biol. Chem., 377: 195-201. Qing, K., Mah, C , Hansen, J., Zhou, S., Dwarki, V. and Srivastava, A. (1999) Human fibroblast growth factor receptor 1 is a co-receptor for infection by adeno-associated virus 2. Nat. Med., 5: 71-77. Rabinowitz, J.J., Rolling, R, Li, C , Conrath, H., Xiao, W., Xiao, X. and Samulski, R.J. (2002) Cross-packaging of a single adenoassociated virus (AAV) type 2 vector genome into multiple AAV serotypes enables transduction with broad specificity. J. Virol., 76: 791-801. Richichi, C , Lin, E.J., Stefanin, D., Colella, D., Ravizza, T., Grignaschi, G., Veglianese, P., Sperk, G., During, M.J. and Vezzani, A. (2004) Anticonvulsant and antiepileptogenic effects mediated by adeno-associated virus vector neuropeptide Y expression in the rat hippocampus. J. Neurosci., 24: 3051-3059. Scammell, T.E., Arrigoni, E., Thompson, M.A., Ronan, P.J., Saper, C.B. and Greene, R.W. (2003) Focal deletion of the adenosine Al receptor in adult mice using an adeno-associated viral vector. J. Neurosci., 23: 5762-5770. Shen, Y., Muramatsu, S., Ikeguchi, K., Fujimoto, K., Fan, D., Ogawa, M., Mizukami, H., Urabe, M., Kame, A., Nagatsu, I., Urano, R, Suzuki, T., Ichinose, H., Nagastu, T., Monahan, J., Nakano, L and Ozawa, K. (2000) 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., 11: 1509-1519. Shevtsova, Z., MaUk, J.M., Michel, U., Bahr, M. and Kugler, S. (2004) Promoters and serotypes: targeting of adeno-associated virus vectors for gene transfer in the rat central nervous system in vitro and in vivo. Exp. Physiol., 90: 53-59. Summerford, C. and Samulski, R.J. (1998) Membrane-associated heparan sulfate proteoglycan is a receptor for adeno-associated virus type 2 virions. J. Virol., 72: 1438-1445. Sununerford, C., Bartiett, J.S. and Samulski, R.J. (1999) AlphaVbeta5 integrin: a co-receptor for adeno-associated virus type 2 infection. Nat. Med., 5: 78-81. Tenenbaum, L., Hamdane, M., Pouzet, M., Avalosse, B., Stathopoulos. A., Jurysta, R, Rosenbaum, C., Hanemann, C O . , Levivier, M.

and Velu, T. (1999) Cellular contaminants of adeno-associated virus vector stocks can enhance transduction. Gene Ther., 6: 1045-1053. Tenenbaum, L., Jurysta, R, Stathopoulos, A., Puschban, Z., Melas, C., Hermens, W.T.J.M.C., Verhaagen, J., Pichon, B., Velu., T. and Levivier, M. (2000) Tropism of AAV-2 vectors for neurons of the globus pallidus. NeuroReport, 11: 2277-2283. Thompson, R.C. and Watson, S.J. (1990) Nucleotide sequence and tissue-specific expression of the rat melanin concentrating hormone gene. DNA Cell Biol., 9: 637-645. Trinklein, N.D., Aldred, S.R Hartman, S.J., Schroeder, D.I., Otillar, R.P and Myers, R.M. (2003) An abundance of bidirectional prometers in the human genome. Genome Res., 14: 62-66. Tsukiyama-Kohara, K., lizuka, N., Kohara, M. and Nomoto, A. (1992) Internal ribosome entry site within hepatitis C virus RNA. J. Virol., 66: 1476-1483. Urabe, M., Hasumi, Y, Ogasawara, Y, Matsushita, T., Kamoshita, N., Nomoto, A., Colosi, P., Kurtzman, G.J., Tobita, K. and Ozawa, K. (1997) A novel dicistronic AAV vector using a short IRES segment derived from hepatitis C virus genome. Gene, 200: 157-162. Van den Pol, A.N., Acuna-Goycolea, C , Clark, K.R. and Ghosh, PK. (2004) Physiological properties of hypothalamic MCH neurons identified with selective expression of reporter gene after recombinant virus infection. Neuron, 42: 635-652. Walters, R.W., Yi, S.M., Keshavjee, S., Brown, K.E., Welsh, M.J., Chiorini, J. A. and Zabner, J. (2001) Binding of adeno-associated virus type 5 to 2,3-lii\ked sialic acid is required for gene transfer. J. Biol. Chem., 276: 20610-20616. Wang, C , Wang, CM., Clark, K.R. and Sferra, T.J. (2003) Recombinant AAV serotype 1 transduction efficiency and tropism in the murine brain. Gene Ther., 10: 1528-1534. Wang, C.Y, Guo, H.Y, Lim, T.M., Ng, Y.K., Neo, H.P, Hwang, PY, Yee, W C . and Wang, S. (2005) Improved neuronal transgene expression from an AAV-2 vector with a hybrid CMV enhancer/ PDGF-beta promoter. J. Gene Med. Epub. Wong, E.T., Ngoi, S.M. and Lee, C.G. (2002) Improved co-expression of multiple genes in vectors containing internal ribosome entry sites (IRESes) from human genes. Gene Ther., 9: 337-344. Xu, R., Janson, C.G., Mastakov, M., Lawlor, PA., Young, D., Mouravlev, A.L, Fitzsimons, H.L., Choi, K., Ma, H., Dragunow, M., Leone, P , Chen, Q., Dicker, B. and During, M.J. (2001) Quantitative comparison of expression with adeno-associated virus (AAV-2) brain-specific gene cassettes. Gene Ther., 8: 1323-1332.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

CHAPTER

2 Identification of Novel Adeno-Associated Virus Serotypes for Use as Vectors Luk H. Vandenberghe, ]ames M. Wilson, Guangping Gao

Abstract: Adeno-associated virus (AAV)-based vectors can achieve stable gene transfer without noticeable vector-related toxicities. However, the AAV serotype 2-derived vector has a restricted tissue tropism and a low transduction efficiency, which hampers the further development of AAV as vector. The use of alternative serotypes demonstrated a significant improvement over AAV2. For serotypes 1, 4, 5 and 6, a pattern of very distinct tropisms was described leading to a serotype of choice depending on the target cell or tissue. In the search for more potent AAV vectors with enhanced performance profiles, polymerase chain reaction-based techniques were exploited to examine human and non-human primate tissues for the identification and isolation of endogenous AAVs. Over one hundred novel primate AAV proviral sequences were discovered. The molecular prevalence of endogenous AAVs was 18-19% in the tissues evaluated. Further analyses suggested that primate AAVs are clustered in clades according to their phylogenetic relatedness. The members of each clade share functional and serological similarities. Three novel clades, represented by serotypes AAV7, AAVS and AAV9, yielded vectors that outperformed previously known AAV vector systems for several different target tissues. Although, to date, no studies have been reported that employ these novel primate vectors for the central nervous systemdirected gene transfer, several other gene therapy applications for long-term phenotypic correction in mouse and canine disease models have already been accomplished. These newly identified AAV vectors represent a new generation of efficient vehicles for somatic gene transfer in different gene therapy applications. This review summarizes previously published work in this area. Keywords: AAV serotypes; AAV biology; AAV classification; AAV vectors; gene transfer; tissue tropism 1. A.

All AAVs carry ssDNA genome flanked by two palindromic hairpin sequences known as inverted terminal repeats (ITRs). The wild-type virus encodes 4 Rep and 3 Cap polypeptides from two open reading frames by elegantly combining mechanisms such as differential splicing and alternative start codon usage. The viral genome is surrounded by a near-spherical protein shell that consists of 60 capsid subunits arranged with T = 1 icosahedral symmetry. The three overlapping capsid components are referred to as viral protein (VP) 1-3 which only differ in their length of the N-terminus. They are represented in the viral architecture at a ratio of 1:1:10.

INTRODUCTION

AAV as a Virus

The adeno-associated virus (AAV) is a member of the Parvoviridae. As the single-stranded (ss) DNA viruses, members of this family share a similar genomic organization. These small animal viruses have a genome of about 5 kb and encode two types of gene products: non-structural (NS) and structural proteins. In the AAV world, these are referred to as replication (rep) and capsid (cap) gene products, respectively.

Gene Therapy of the Central Nervous System: From Bench to Bedside

17

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

18

2. IDENTIFICATION OF NOVEL ADENO-ASSOCIATED VIRUS SEROTYPES FOR USE AS VECTORS

AAV is a member of Dependovirus genus within the subfamily of the Parvovirinae. Their dependence on helper viruses for replication and their host tropism for vertebrae justify this classification. Natural AAV infections undergo a latent and a lytic phase in w^hich helper functions from adenovirus or herpes simplex virus are indispensable for AAV replication. Within the Dependoviridae, AAVs are formally classified per serotype. Most early isolates of AAVs, namely AAVl, AAV2, AAVS, AAV4 and AAV6, v^ere identified as contaminants in adenovirus preparations, hence their name. AAV5 was reported to be isolated from a human condylomatous wart (Atchison et al., 1965; Melnick et a l , 1965; Hoggan et al., 1966; Mayor and Melnick, 1966; Bantel-Schaal and Zur Hansen, 1984; Georg-Fries et al., 1984; Rutledge et al., 1998; Bantel-Schaal et al., 1999). Natural infections of AAV serotypes in primates were previously reported by serological analyses but no clinical sequelae have been associated with AAV infection (Blacklow et al., 1967, 1968a, b; Mayor and Ito, 1967). Serologically, AAV2 is the dominant AAV in humans (Chirmule et al., 1999). B.

AAV as a Vector

AAV was developed as a gene transfer vehicle through pioneering work by Samulski et al. after their isolation and characterization of an infectious molecular clone of AAV serotype 2 (Samulski et a l , 1982, 1983). AAV vectors currently used in gene therapy applications are designed to carry only minimal virus sequence devoid of viral genes. A minigene cassette is solely flanked in cis by the viral ITRs. The capsid components are provided in trans in the context of helper functions during vector packaging (Samulski et al., 1989; Xiao et a l , 1998). Earlier studies in several animal models of AAV2-based vectors demonstrated sustained transgene expression in different target tissues, but lack of vector-related toxicities and host immune responses to the transduced cells (Kotin, 1994; Rabinowitz and Samulski, 1998; Grimm et al., 2003). However, further development of AAV2-based vector for gene therapy applications was hindered by its restricted tissue tropism, low efficiency of gene transfer, delayed onset of gene expression and highly prevalent pre-existing immunity in human populations. The cloning of AAV genomes from alternative serotypes 1,3,4,5 and 6 and introduction of the transencapsidation — also referred to as pseudotyping — strategy in the mid-1990s made it possible to package the same AAV2 recombinant genomes with capsid proteins

from other AAV serotypes and evaluate capsid-specific vector biology in different target tissues. For instance, AAV2 vector genome transcapsidated with AAVl capsid led to 10- to 100-fold higher gene transfer in muscle, whereas AAV5 could significantly improve AAV vector-mediated gene transfer to limg and CNS (Xiao et a l , 1999; Davidson et al., 2000; Zabner et al., 2000; Auricchio et al., 2001; Chao et al., 2001).

C. AAV Biology Even though the different serotypes share substantial sequence homology and identity, their transduction pathway into the cell diverges at various points. This is reflected in the tropism and potency of the different serotype vectors in various settings (reviewed by Sanglioglu et al. (2001)). AAV2 is known to bind membrane-associated heparin sulfate proteoglycan as its primary attachment receptor (Summerford and Samulski, 1998). Fibroblast growth factor receptor and a^P^ integrin have also been shown to play a role in AAV2 viral entry (Qing et al., 1999; Summerford et al., 1999). AAV4 and AAV5 are known to bind sialic acid on the cell surface (Kaludov et al., 2001; Walters et al., 2001). Interactions of AAVs with different receptors then lead to endocytotic uptake of the virion through clathrin-coated pits. Endosomal acidification precedes release into the cytosol. Following these events, a perinuclear accumulation is described through an unknown mechanism. The pathway then ends with a slow nuclear entry (Bartlett et al., 2000). At the molecular level, the ssAAV genome then needs to be converted into its double stranded version to initiate transcription. The mechanism by which this occurs is still under debate. Second-strand synthesis has been studied intensively as one of the major hurdles for AAV to successfully express from its transgene cassette (Ferrari et al., 1996; Fisher et al., 1996; Qing et al., 2001). Recently, an alternative mechanism of single-strand annealing was proposed by Thomas et al. (2004). Why and where the road ends for one serotype but continues for another has yet to be fully understood. It is likely that the rate-limiting steps for transduction will be serotype-specific and tissuedependent. Elegant studies in liver- and lung-directed gene transfer have illustrated this (Duan et al., 2000; Thomas et al., 2004). Viral uncoating and escape from endosomal processing have been added to the list of potential hurdles on the post-entry pathway of AAVs. Another important aspect of AAV biology is related to its ability to establish stable infection or transduction as a wild-type virus or as a recon\binant vector.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

IDENTIFICATION AND ISOLATION OF NOVEL PRIMATE AAVS

The initial notion that AAV2 was capable of achieving specific integration in a locus on chromosome 19 was later found to be only true in the wild-type context and in the presence of rep gene products (Kotin et al., 1992). Thus far, there is no evidence that AAV vector genomes devoid of all viral genes integrate site specifically upon transduction. Extensive investigations on the mechanisms of molecular persistence of the AAV vector genome have revealed a convincing pattern of episomal colonization through such processes as intermolecular concatamerization and intramolecular circularization (Carter, 2004; McCarty et al., 2004). As the field voices concerns over possible insertional mutagenesis of viral vectors in gene therapy applications, episomal persistence makes AAV even more attractive for somatic gene transfer.

II.

A.

IDENTIFICATION A N D ISOLATION OF NOVEL PRIMATE AAVS The Hunt for AAVs

Several groups over the past years have probed a variety of species for novel AAVs. The common goal is to provide the gene therapy community with a potent and serologically less prevalent vehicle that is tropic for a tissue of interest. The hunt for novel AAVs was sparked by the discovery of two novel Non Human Primate (NHP) serotypes AAV7 and AAVS (Gao et al., 2002). In this study, a polymerase chain reaction (PCR) methodology was used to screen a large body of samples for the presence and subsequent cloning of novel AAV-like sequences. Where initially only Rhesus monkeys were included, the study was later expanded to describe the state of AAV in several other primate species including humans (Gao et al., 2003, 2004b). AAV was found to be widely prevalent in both NHPs and humans. In a PCR-based screening for AAV sequences, 18% of all human and 19% of all other primate tissue yielded AAV sequences. A seroepidemiological study was performed on a limited number of captive bred NHPs for the prevalence of AAV7 and AAVS. A very high frequency of sero-positive animals was recorded for most macaque species with individual species showing over S0% seropositivity. Slightly lower rates were documented in baboons with the lowest prevalence in chimpanzees (Gao et al., 2003). These findings have to be taken into consideration when using NHP models for AAV-mediated gene transfer. A comprehensive survey of the

19

sero-prevalence of the novel AAVs in human populations is needed. This is especially relevant for the clinical success of AAV gene therapy in populations where circulating antibodies are prevalent. The ongoing hunt for novel AAVs led to several serologically distinct isolates from different primate species: at least three novel primate AAV serotypes (AAV7-9) (Gao et al., 2002, 2004b) and a proposed two additional ones (AAVIO and 11) have been reported recently by Mori et al. (2004). In the process of identifying these serologically distinct clones, a much wider sequence diversity in AAV family was unveiled. Over 100 distinct viral cap sequences were described and isolated from human, chimpanzee, rhesus, pigtailed and cynomolgus macaques (respectively hu., ch., rh., pi. and cy AAVs). As a DNA virus, uncharacteristically, AAV seems to undergo substantial evolution in primates during natural infections. This rapid molecular evolution generates a diversity comparable to quasispecies common to RNA viruses (Gao et al., 2003). B. Molecular Identification and Rescue of Novel Primate AAV Sequences A critical stage in the life cycle of AAV during natural infections is to establish latency in the host until rescue by subsequent helper virus infections. Taking the advantage of this unique property of AAV, a molecular amplification technique was designed to retrieve pro virus sequences of endogenous AAVs from primate tissue (Gao et al., 2002). The key to successful recovery of yet unidentified AAV sequences by a PCRbased method was the design of the primers through sequence alignment and homology analysis of previously characterized genomes representing a wide range of AAVs and their close evolutionary relatives. The general principle in the primer selection was to target the conserved regions of AAV genomes and to flank regions of divergence. The primers derived from such analyses were not only used for quick identification of novel AAV sequences by a process called signature PCR but also for isolation of full length rep and cap sequences (Gao et al., 2002). Since the major goal of the novel AAV discovery effort was to develop novel vectors for gene transfer, the molecular rescue was primarily focused on fulllength cap sequences. Two strategies were employed to accomplish this goal. The first was to generate two overlapping PCR amplicons using primers to the conserved regions and to fuse them into a full-length gene sequence (Gao et al., 2002). The second strategy was to directly amplify a 3.1 kb region spanning a segment in

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

20

2. IDENTIFICATION OF NOVEL ADENO-ASSOCIATED VIRUS SEROTYPES FOR USE AS VECTORS

the 3'-end of the rep gene (about 800 bp) and the entire cap gene sequence. Primers used for the full-length cap gene amplification were located in a highly conserved sequence of rep and an untranslated region in the 3'-end of AAV genomes (Gao et al., 2003,2004b). In a different application of this molecular method for the study of the abundance and biodistribution of AAV provirus genomes in tissues and animals, sets of primers and probes were designed to anneal to short stretches of conserved regions in the rep or cap genes. In this case, the PCR amplicons were aimed at detecting AAV sequences that are highly conserved among known AAV genomes. Quantitative PCR amplification of AAV proviral sequences in total tissue DNAs was performed using the TaqMan technology (Gao et al., 2003). This methodology has yielded new AAV species from NHP and humans (Chen et al., 2004; Clark et al., 2004; Jensen et al., 2004) and some other lower animals (Arbetman et al., 2004a, b). Interestingly, a caprine AAV isolated in this study was highly reminiscent of AAV5, which was previously reported as a humanderived AAV serotype, whereas a bovine AAV was a recombination product between AAV4 and AAV5, which had been previously cloned and characterized by Chiorini's group (Bossis and Chiorini, 2003; Arbetman et al., 2004a, b; Colosi, 2004; Schmidt et al., 2004). Classical virus rescue and isolation was performed in selected cases where some novel proviral sequences were highly enriched in certain tissues. AAV-containing tissue DNA was first digested with a restriction enzyme not expected to cut within the viral genome. Subsequent to the restriction digest, the product was transfected into 293 cells for viral rescue in the presence of adenovirus helper functions. Verified on the basis of morphological and structural information provided by electron microscopy and immunoblotting, AAV virions could be isolated from the AAV-containing tissues (Gao et al., 2002, unpublished data). The molecular state of the AAV proviral genome has been an area of significant interest, since it was shown that AAV2 has the ability of site-specific integration into chromosome 19 at 19ql3.4 (Kotin et al., 1992). Southern-blot analyses of AAV-infected tissue suggested that AAV is mostly present in an episomal form (Gao et al., 2003). This allowed Johnson et al. to exploit linear rolling circle amplification (LRCA) (Lizardi et al., 1998) to directly rescue molecular clones of latent endogenous AAV genomes from primate tissue. This strategy was proven to be extremely efficient as shown by the isolation of five infectious AAV clones from six AAV positive tissues (Chen et al., 2004; Clark et al., 2004; Jensen et al., 2004).

III.

CLASSIFICATION OF NOVEL PRIMATE AAVS

Given their extensive biodiversity, a novel way of classifying the primate AAVs within the dependoviridae was proposed (Gao et al., 2004b). A previously suggested rationale (Lukashov and Goudsmit, 2001) based on phylogeny instead of the established classifiers like serology and host range was explored. A.

AAV Serotypes

Amongst the helper-dependent viruses, the scientific community has traditionally differentiated the available species into serotypes. If polyclonal antiserum generated against a particular virus is unable to productively neutralize another virus in a reciprocal manner, these viruses are considered serological distinct. A practical definition for discrimination of serotypes was borrowed from the adenoviral community. An independent serotype is established if the ratio of the heterologous neutralization titer versus the homologous titer is more than or equal to 16-fold for both the antisera. Neutralization is commonly measured at the level of 50% neutralization. Until recently, nine primate AAV serotypes have been proposed. Their serological cross-neutralization was re-evaluated by Gao et al. (2004b). For A A V l ^ and 7-9, their serological distinctiveness was confirmed. AAVl and AAV6, which only differ in six amino acids of their capsid sequence were foimd to belong to the same serotype. A more puzzling finding was made in the fact that AAVl and AAV5 showed significant cross reactivity. These two capsids are of the most divergent at the sequence level within the primate AAVs and were found to be structurally incompatible (Rabino-witz et al., 2004). Recently, other helper-dependent parvoviruses from non-primate species have been characterized and cloned for gene transfer purposes (Bossis and Chiorini, 2003; Farkas et al., 2004; Schmidt et al., 2004). B. Recombination as a Confounding Factor for Serotyping A mechanism like homologous recombination might allow for swapping of neutralizing antibody epitopes. This would be a confounding factor for a classification system based on serology. Two distinct scenarios provided evidence that molecular recombination was a mechanism for generating additional capsid diversity. One particular example of such recombination process was reported in a rhesus macaque that probably

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

CLASSIFICATION OF NOVEL PRIMATE AAVS

21

Therefore, the concept of clades was introduced. A clade is a cluster of viruses that share a common ancestry. A clade has to contain isolates from a minimum of three or more sources. Sufficient sequence diversity was found to establish six novel clades A-F (Fig. 1) (Gao et al., 2002; Gao et al., 2003; Gao et al., 2004b). All these clades cover existing serotypes. For example, Clade B is represented by AAV2. The range of heterogeneity within a clade varies with clades D and E clearly being the widest in diversity. Members within a certain clade are expected to share functional and serological properties. AAV seems to have evolved from avian parvoviruses (Brown et al., 1995; Zadori et al., 1995). Both AAV4 and AAV5 are most distinct from each other and the other primate dependoviruses with over 30% difference in capsid protein sequence, although they do share most homology with the autonomous avian parvoviruses. Characterized AAVs from non-primate mammalian sources have consistently clustered with AAV5 or AAV4. The molecular clone for bovine AAV is even described to be a hybrid of these two serotypes (Schmidt et al., 2004). All other primate AAVs are relatively closely related with over 80% identity in amino-acid sequence. Members within several indivi-dual clades have been isolated from several different species. For example, Clade E, represented by AAV serotj^e 8 contains closely

underwent a natural co-infection with two distinct isolates. These isolates then became the parental strains that gave rise to a dozen hybrid genomes (Gao et al., 2003). A second indication for the occurrence of recombination as a means of AAV evolution was provided when an entire monophylic group was identified in human tissue from eight different subjects. AAV from this clade showed a similar recombination profile. This group of viruses was found to be related to AAV2 towards the 5'-end of the capsid, while the 3'-end showed clear homology with AAVS with a breakpoint aroimd position at 1400 bp of the VPl sequence (Gao et al., 2004b). Recombination is an unlikely event requiring co-infection of the same animal, the same tissue and the same cell with two distinct viruses. Even more imlikely is the generation of a fit hybrid virus as an outcome of this event. The high prevalence of various AAVs in different species probably increases these odds dramatically, making recombination a viable strategy for the virus to evolve. C. Evolution, Clades and Cross^Species Transmission Even though serotypes are a functional way to describe the biodiversity of AAV, it falls short in describing the heterogeneity present in the AAV world.

Clade E (AAVS)

Clade D (AAV7) Clade F (AAV9) 4 Clade A (AAV1 & 6) AAV3 AAV3B ch.5 Clade B (AAV2)

I Clade C 1

AAV4

" Bovine AA V (Schmidt et al.) AAV11(Morietal.) rh.34 rh.32 rh.33

•t

— I

AAVS

- Avian A A V (Bossis et al.) • Serpentine AAV (Farkas etal.) • Goose Parvovirus 0.05

FIGURE 1 Clade Dendrogram of AAV Species. This Simplified Phylogenetic Tree Rooted With the Goose Parvovirus Lays Out the Different Clades Identified in Primates. Each Clade Contains Minimally Three Sequences Isolated from Distinct Sources.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

22

2. IDENTIFICATION OF NOVEL ADENO-ASSOCIATED VIRUS SEROTYPES FOR USE AS VECTORS

related members that were isolated from both several species of macaques as w^ell as humans. This particular finding indicates an event of cross-species transmission betv^een these species or from a common host of AAV. It also highlights the promiscuous nature of AAV from an epidemiological and biological point of viewr.

IV.

NOVEL AAVS AS GENE TRANSFER VECTORS

Discovery of an expanded AAV family has created unprecedented opportunities for the field of gene therapy to identify suitable AAV vectors for different applications. In the initial studies to evaluate these novel AAV-based gene transfer vectors in rodent models, AAVs 8,9, cy5, rh.2, rh.8, rh.20, rh.39 and rh.43 demonstrate liver transduction efficiencies that are several logs higher than AAV2. AAV rh.lO presents a strong lung tropism, whereas AAV7, AAVrh.8 and AAVrh.lO can achieve levels of muscle transduction that are similar to AAVl. Interestingly, AAV9 was the only rAAV that stands out in all tissues tested including the cardiac muscle (Gao et al., 2002, 2003, 2004b; Grimm et al., 2003). Follow-up studies of some representative novel AAVs in NHP models for muscle- and liver-directed gene transfer confirmed the findings in rodent models and revealed interesting vector biology in terms of host responses, transgene expression, molecular status and biodistribution of vector genomes (Chenuaud et al., 2004; Gao et al., 2004a). An interesting application with one of the vectors isolated from non-primates has been the use of the bovine AAV for targeting inner ear neuroepithelial cells (Di Pasquale et al., 2005). When some of those lead vectors were tested for gene therapy treatment of several mouse and canine disease models including Hemophilia A and B, FH and LGMD, long-term phenotypic correction of monogenic defects were successfully accomplished (Gregorevic et a l , 2004; Lebherz et al., 2004; Sarkar et al., 2004; Zhu, 2004; Wang et al., 2005). Newly identified AAV serotypes represent a new generation of efficient vehicles for somatic gene transfer in different gene therapy applications though their potential in the CNS-directed gene transfer remains to be assessed.

V.

CONCLUSIONS

The recent identification of novel AAVs from a variety of species has enabled the gene therapy field

to tackle several longstanding hurdles in its successful application. Novel methodologies in the discovery of the smallest of mammalian viruses have increased sensitivity in the viral isolation process. These innovations have uncovered the extremely heterogeneous world of AAV and have introduced recombination and species transmissions as recurring motifs in AAV evolution. To properly describe the unanticipated spectrum of genomic diversity, a novel classification of the primate dependoviruses has been proposed by introducing clades. Even though very promising results have been obtained with novel AAVs in small and large animal models, no published reports have described their tropism and efficiency in the brain. The exploration of the wide biodiversity of the primate dependoviruses available allows the gene therapist to screen a panel of gene transfer vectors for desirable properties such as: preferential tropism, minimal toxi-city, adequate efficiency, minimal seroprevalence, etc. After this evaluation, a selection of lead candidates and further evaluation of these vectors in preclinical models with a focus on understanding their biology as vectors, their direct and indirect interactions with the host, as well as an understanding of the genomic persistence would be necessary prior to their clinical applications. References Arbetman, A., Lochrie, M., Randlev, B., Surosky, R., Zhou, S., Wellman, J., Pater, C , Lehmkuhl, H., Hobbs, L., Pierce, G. and Colosi P. (2004a) Isolation of a close AAVS relative from goat tissues: evidence of host promiscuity. ASGT Vllth Annual Meeting, Minneapolis, MN. Arbetman, A., Lochrie, M., Surosky, R., Randlev, B., Zhou, S., Wellman, J., Lemkuhl, H., Hobbs, L.A., Peierce, G. and Colosi P. (2004b) Characterization of novel caprine and bovine AAV capsids with unique transduction and neutralization properties. Xth Parvovirus Workshop, St. Petersburg, FL. Atchison, R.W., Casto, B.C. and Hammon, W.M. (1965) Adenovirusassociated defective virus particles. Science, 149: 754-756. Auricchio, A., Kobinger, G., Anand, V, Hildinger, M., O'Connor, E., Maguire, A.M., Wilson, J.M. and Bennett, J. (2001) Exchange of surface proteins impacts on viral vector cellular specificity and transduction characteristics: the retina as a model. Hum. Mol. Genet., 10(26): 3075-3081. Bantel-Schaal, U., Delius, H., Schmidt, R. and zur Hausen, H. (1999) Human adeno-associated virus type 5 is only distantly related to other known primate helper-dependent parvoviruses. J. Virol., 73(2): 939-947. Bantel-Schaal, U. and Zur Hausen, H. (1984) Characterization of the DNA of a defective human parvovirus isolated from a genital site. Virology, 134: 52-63. Bartlett, J.S., Wilcher, R. and Samulski, R.J. (2000) Infectious entry pathway of adeno-associated virus and adeno-associated virus vectors. J. Virol., 74(6): 2777-2785.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

CONCLUSIONS Blacklow, N.R., Hoggan, M.D., Kapikian, A.Z., Austin, J.B. and Rowe, W.R (1968a) Epidemiology of adenovirus-associated virus infection in a nursery population. Am. J. Epidemiol., 88: 368-378. Blacklow, N.R., Hoggan, M.D. and Rowe, W.R (1967) Isolation of adenovirus-associated viruses from man. Proc. Natl. Acad. Sci., 58: 1410-1415. Blacklow, N.R., Hoggan, M.D. and Rowe, W.R (1968b) Serologic evidence for human infection with adenovirus-associated viruses. J. Natl. Cancer Inst., 40(2): 319-327. Bossis, I. and Chiorini, J.A. (2003) Cloning of an avian adeno-associated virus (AAAV) and generation of recombinant AAAV particles. J. Virol., 77(12): 6799-6810. Brown, K.E., Green, S.W and Young, N.S. (1995) Goose parvovirus — an autonomous member of the dependovirus genus? Virology, 210(2): 283-291. Carter, B.J. (2004) Adeno-associated virus and the development of adeno-associated virus vectors: a historical perspective. Mol. Ther. 10(6): 981-989. Chao, H., Monahan, RE., Liu, Y., Samulski, R.J. and Walsh, C.E. (2001) Sustained and complete phenotype correction of hemophilia B mice following intramuscular injection of AAVl serotype vectors. Mol. Ther., 4(3): 217-222. Chen, C , Jensen, R., Schnepp, B., Clark, K.R. and Johnson, R (2004) Characterization of adeno-associated virus sequences in human tissues. Mol. Ther., 9(Suppl 1): 132. Chenuaud, P., Larcher, T., Rabinowitz, J.E., Provost, N., Cherel, Y, Casadevall, N., Samulski, R.J. and MouUier, R (2004) Autoimmune anemia in macaques following erythropoietin gene therapy Blood, 103(9): 3303-3304. Chirmule, N., Propert, K., Magosin, S., Qian, Y, Qian, R. and Wilson, J. (1999) Immune responses to adenovirus and adenoassociated virus in humans. Gene Ther., 6(9): 1574-1583. Clark, K.R., Chen, C , Jensen, R., Schnepp, B. and Johnson, P. (2004) Characterization of wild-type adeno-associated viruses isolated from human tissues. Xth Parvovirus Workshop, St. Petersburg, PL. Colosi, E.A. (2004) Xth Parvovirus Workshop, St. Petersburg, PL. Davidson, B.L., Stein, C.S., Heth, J.A., Martins, I., Kotin, R.M., Derksen, T.A., Zabner, J., Ghodsi, A. and Chiorini, J.A. (2000) Recombinant adeno-associated virus type 2, 4, and 5 vectors: transduction of variant cell types and regions in the mammalian central nervous system. Proc. Natl. Acad. Sci. USA, 97(7): 3428-3432. Di Pasquale, G., Rzadzinska, A., Schneider, M.E., Bossis, I., Chiorini, J.A. and Kachar, B. (2005) A novel bovine virus efficiently transduces inner ear neuroepithelial cells. Mol. Ther., 11(6): 849-855. Duan, D., Yue, Y, Yan, Z., Yang, J. and Engelhardt, J.F. (2000) Endosomal processing limits gene transfer to polarized airway epithelia by adeno-associated virus. J. Clin. Invest., 105(11): 1573-1587. Parkas, S.L., Zadori, Z., Benko, M., Essbauer, S., Harrach, B. and Tijssen, P. (2004) A parvovirus isolated from royal python (Python regius) is a member of the genus Dependovirus. J. Gen. Virol., 85(Pt 3): 555-561. Ferrari, RK., Samulski, T, Shenk, T. and Samulski, R.J. (1996) Second-strand synthesis is a rate-limiting step for efficient transduction by recombinant adeno-associated virus vectors. J. Virol., 70(5): 3227-3234. Fisher, K.J., Gao, G.R, Weitzman, M.D., DeMatteo, R., Burda, J.R and Wilson J.M. (1996) Transduction with recombinant adenoassociated virus for gene therapy is limited by leading-strand synthesis. J. Virol, 70(1): 520532.

23

Gao, G., Alvira, M.R., Somanathan, S., Lu, Y, Vandenberghe, L.H., Rux, J.J., Calcedo, R., Sanmiguel, J., Abbas Z., and Wilson, J.M. (2003) Adeno-associated viruses undergo substantial evolution in primates during natural infections. Proc. Natl. Acad. Sci. USA, 100(10): 6081-6086. Gao, G.R, Alvira, M.R., Wang, L., Calcedo, R., Johnston, J. and Wilson, J.M. (2002) Novel adeno-associated viruses from rhesus monkeys as vectors for human gene therapy. Proc. Natl. Acad. Sci. USA, 99(18): 11854-11859. Gao, G., Lebherz, C , Weiner, D.J., Grant, R., Calcedo, R., McCuUough, B., Bagg, A., Zhang Y, and Wilson, J.M. (2004a) Erythropoietin gene therapy leads to autoimmune anemia in macaques. Blood, 103(9): 3300-3302. Gao, G., Vandenberghe, L.H., Alvira, M.R., Lu, Y, Calcedo, R., Zhou, X. and Wilson, J.M. (2004b) Glades of adeno-associated viruses are widely disseminated in human tissues. J. Virol., 78(12): 6381-6388. Georg-Fries, B., Biederlack, S., Wolf, J. and zur Hansen, H. (1984) Analysis of proteins, helper dependence, and seroepidemiology of a new human parvovirus. Virology, 134(1): 64-71. Gregorevic, P., Blankinship, M.J., Allen, J.M., Crawford, R.W, Meuse, L., Miller, D.G., Russell, D.W. and Chamberlain, J.S. (2004) Systemic delivery of genes to striated muscles using adeno-associated viral vectors. Nat. Med., 10(8): 828-834. Grimm, D. and Kay, M.A. (2003) From virus evolution to vector revolution: use of naturally occurring serotypes of adeno-associated virus (AAV) as novel vectors for human gene therapy. Curr. Gene Ther., 3(4): 281-304. Grimm, D., Zhou, S., Nakai, H., Thomas, C.E., Storm, T.A., Fuess, S., Matsushita, T, Allen, J., Surosky, R., Lochrie, M., Meuse, L., McClelland, A., Colosi, R and Kay, M.A. (2003) Preclinical in vivo evaluation of pseudotyped adeno-associated virus vectors for liver gene therapy Blood, 102(7): 2412-2419. Hoggan, M.D., Blacklow, N.R. and Rowe, W.R (1966) Studies of small DNA viruses found in various adenovirus preparations: physical, biological, and immunological characteristics. Proc. Natl. Acad. Sci. USA, 55(6): 1467-1474. Jensen, R., Schnepp, B., Johnson, P. and Clark, K. (2004) Adeno-associated virus infection in non-human primates. Xth Parvovirus Workshop, St.-Pete's Beach, FL. Kaludov, N., Brown, K.E., Walters, R.W., Zabner J. and Chiorini, J.A. (2001) Adeno-associated virus serotype 4 (AAV4) and AAV5 both require sialic acid binding for hemagglutination and efficient transduction but differ in sialic acid linkage specificity. J. Virol., 75(15): 6884-6893. Kotin, R.M. (1994) Prospects for the use of adeno-associated virus as a vector for human gene therapy Hum. Gene Ther., 5(7): 793-801. Kotin, R.M., Linden, R.M. and Berns, K.I. (1992) Characterization of a preferred site on human chromosome 19q for integration of adeno-associated virus DNA by non-homologous recombination. EMBO J., 11(13): 5071-5078. Lebherz, C , Gao, G., Louboutin, J.R, Millar, J., Rader, D. and Wilson, J.M. (2004) Gene therapy with novel adeno-associated virus vectors substantially diminishes atherosclerosis in a murine model of familial hypercholesterolemia. J. Gene Med., 6(6): 663-672. Lizardi, RM., Huang, X., Zhu, Z., Bray-Ward, R, Thomas, D.C. and Ward, D.C. (1998) Mutation detection and single-molecule counting using isothermal rolling-circle amplification. Nat. Genet., 19(3): 225-232. Lukashov, VV. and Goudsmit, J. (2001). Evolutionary relationships among parvoviruses: virus-host coevolution among autonomous

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

24

2. IDENTIFICATION OF NOVEL ADENO-ASSOCIATED VIRUS SEROTYPES FOR USE AS VECTORS

primate parvoviruses and links between adeno-associated and avian parvoviruses. J. Virol., 75(6): 2729-2740. Mayor, H.D. and Ito, M. (1967) Distribution of antibodies to Type 4 adeno-associated satellite virus in simian and human sera. P. S. E. B. M., 26: 723-725. Mayor, H.D. and Melnick, J.L. (1966) Small deoxyribonucleic acidcontaining viruses (Picodnavirus group). Nature, 210: 331-332. McCarty, D.M., Young Jr., S.M. and Samulski, R.J. (2004) Integration of adeno-associated virus (AAV) and recombinant AAV vectors. Annu. Rev Genet., 38: 819-845. Melnick, J.L., Mayor, H.D., Smith, K.O. and Rapp, F. (1965) Association of 20 millimicron particles with adenoviruses. J. Bacteriol. 90: 271-274. Mori, S.,Wang, L., Takeuchi, T. and Kanda, T. (2004) Two novel adeno-associated viruses from cynomolgus monkey: pseudotyping characterization of capsid protein. Virology, 330(2): 375-383. Qing, K., Hansen, J., Weigel-Kelley K.A., Tan, M., Zhou, S. and Srivastava, A. (2001) Adeno-associated virus type 2-mediated gene transfer: role of cellular FKBP52 protein in transgene expression. J. Virol, 75(19): 8968-8976. Qing, K., Mah, C , Hansen, J., Zhou, S., Dwarki, V. and Srivastava, A. (1999) Human fibroblast growth factor receptor 1 is a co-receptor for infection by adeno-associated virus 2. Nat. Med. 5(1): 71-77. Rabinowitz, J.E., Bowles, D.E., Faust, S.M., Ledford, J.G., Cunningham, S.E. and Samulski, R.J. (2004) Cross-dressing the virion: the transcapsidation of adeno-associated virus serotypes functionally defines subgroups. J. Virol., 78(9): 4421-4432. Rabinowitz, J.E. and Samulski, J. (1998) Adeno-associated virus expression systems for gene transfer. Curr. Opin. BiotechnoL, 9(5): 470-475. Rutledge, E.A., Halbert, C.L. and RusseU, D.W. (1998) Infectious clones and vectors derived from adeno-associated virus (AAV) serotypes other than AAV type 2. J. Virol., 72(1): 309-319. Samulski, R.J., Bems, K.I., Tan, M. and Muzyczka, N. (1982) Cloning of adeno-associated virus into pBR322: rescue of intact virus from the recombinant plasmid in human cells. Proc. Natl. Acad. Sci. USA, 79(6): 2077-2081. Samulski, R.J., Chang, L.S. and Shenk, T. (1989) Helper-free stocks of recombinant adeno-associated viruses: normal integration does not require viral gene expression. J. Virol., 63(9): 3822-3828. Samulski, R.J., Srivastava, A., Bems, K.I. and Muzyczka, N. (1983) Rescue of adeno-associated virus from recombinant plasmids: gene correction within the terminal repeats of AAV Cell, 33(1): 135-143. SanUoglu, S., Monick, M.M., Luleci, G., Hunninghake, G.W. and Engelhardt, J.R (2001) Rate limiting steps of AAV transduction

and implications for human gene therapy. Curr. Gene Ther. 1(2): 137-147. Sarkar, R., Tetreault, R., Gao, G., Wang, L., Bell, R, Chandler, R., Wilson, J.M. and Kazazian Jr., H.H. (2004) Total correction of hemophilia A mice with canine FVIII using an AAV 8 serotype. Blood, 103(4): 1253-1260. Schmidt, M., Katano, H., Bossis, I. and Chiorini, J.A. (2004) Cloning and characterization of a bovine adeno-associated virus. J. Virol., 78(12): 6509-6516. Summerford, C , Bartlett, J.S. and Samulski, R.J. (1999) AlphaVbeta5 integrin: a co-receptor for adeno-associated virus type 2 infection. Nat. Med., 5(1): 78-82. Summerford, C. and Samulski, R.J. (1998) Membrane-associated heparan sulfate proteoglycan is a receptor for adeno-associated virus type 2 virions. J. Virol. 72(2): 1438-1345. Thomas, C.E., Storm, T.A., Huang, Z. and Kay, M.A. (2004) Rapid uncoating of vector genomes is the key to efficient liver transduction with pseudotyped adeno-associated virus vectors. J. Virol., 78(6): 3110-3122. Walters, R.W., Yi, S.M., Keshavjee, S., Brown, K.E., Welsh, M.J., Chiorini, J.A. and Zabner, J. (2001). Binding of adeno-associated virus type 5 to 2,3-linked sialic acid is required for gene transfer. J. Biol. Chem. 276(23): 20610-20616. Wang, L., Calcedo, R., Nichols, T., Bellinger, D., Dillow, A., Verma, I. and Wilson, J. (2005) Sustained correction of disease in naive and AAV2-pretreated Hemophilia B dogs: AAV2/8 mediated, liverdirected gene therapy Blood, 105(8): 3079-3086. Xiao, W , Chirmule, N., Berta, S.C, McCuUough, B., Gao, G. and Wilson, J.M. (1999) Gene therapy vectors based on adeno-associated virus type 1. J. Virol., 73(5): 3994-4003. Xiao, X., Li, J. and Samulski, R.J. (1998) Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus. J. Virol., 72(3): 2224-2232. Zabner, J., Seller, M., Walters, R., Kotin, R.M., Fulgeras, W , Davidson, B.L. and Chiorini, J.A. (2000) Adeno-associated virus type 5 (AAV5) but not AAV2 binds to the apical surfaces of airway epitheUa and facilitates gene transfer. J. Virol., 74(8): 3852-3858. Zadori, Z., Stefancsik, R., Ranch, T. and T. Kisary, T. (1995) Analysis of the complete nucleotide sequences of goose and muscovy duck parvoviruses indicates common ancestral origin with adeno-associated virus 2. Virology, 212(2): 562-573. Zhu (2004) Profound cardiac and whole-body functional recovery in heart failure/muscular dystrophy hamsters by systemic delivery of AAV8 vectors via I.V or I.P. route. Xth Parvovirus Workshop, St. Petersburg, FL.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

CHAPTER

3 HSV Amplicon Vectors for Gene Delivery to the Nervous System Claudia Senn, Cornel Fraefel, Xandra O. Breakefield

Abstract: The HSV amplicon vector incorporates features of HSV-1, including a 150 kb transgene capacity, a viral origin of DNA replication and packaging signal, and virion proteins. The large transgene capacity is one of the most distinguishing features of this vector system, which allows incorporation of multiple transgenes and large genomic fragments, as well as informational elements from other virus vectors, including AAV, EBV, and retrovirus. Vectors have been modified to include elements, which increase infection of specific cell types and allow retention of transgene sequences, either as replicating episomal elements or through site-specific integration into the cell genome, and provide the ability to control transgene expression. The virion itself includes proteins that can be used to track infection and deliver fusion proteins to cells. Within the nervous system, these vectors are especially useful due to the natural neurotropism of the virus, with a strong retrograde component, and minimal perturbation of neuronal physiology. Vectors have been used to deliver proteins to facilitate fluorescence, bioluminescent, and magnetic resonance imaging, as well as to monitor neuronal functions in animal models involving learning/memory and addiction paradigms. Vectors have been designed to ameliorate symptoms in models of neurologic disease, including protection of neurons from toxic insults and replacement of genetically deficient proteins, as well as in treatment of brain tumors. Amplicon vectors are considered highly compatible with clinical trials due to their intrinsic lack of toxicity, but methods of production need to be improved to generate high titers and clinically compatible vector stocks. Keywords: amplicon; HSV; nervous system; hybrid vectors; brain L A,

Based on their biological and genetic properties, herpesviruses are divided into three subfamilies (Alpha-, Beta-, and Gammaherpesvirinae). Herpes simplex virus type 1 (HSV-1), the prototypical member of the Alphaherpesvirinae subfamily, is a widespread pathogen in humans and has been intensively investigated (Roizman and Pellett, 2001). HSV-1 has the ability to infect dividing, as well as nondividing cells, of many types. In vivo, they are characterized by their neurotropism and neuroinvasion, which, in rare cases, may result in severe retinal and central nervous system infections (Izumi and Stevens, 1990). Entry of HSV-1 is a multistep process involving several viral glycoproteins and a number of different

BASICS OF HSV-1 AMPLICON

Characteristics of Herpes Simplex Virus type 1

Herpesviruses represent a very large family of enveloped double-stranded DNA viruses. The tj^ical herpes virion consists of a dense core, which contains the viral genome. The core is covered by an icosahedral capsid composed of 162 tubular capsomers. The capsid, which is 100-110 nm in diameter (Homa and Brown, 1997), is surrounded by an amorphous proteinaceous layer called the tegioment. The envelope of the virus is a lipid bilayer that contains glycoproteins spikes and has a diameter of approximately 200 nm (Grunewald et al., 2003; Fig. 1).

Gene Therapy of the Central Nervous System: From Bench to Bedside

25

Copyright© 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

26

3. HSV AMPLICON VECTORS FOR GENE DELIVERY TO THE NERVOUS SYSTEM

FIGURE 1 Electron micrograph of HSV-1 virion. This virion (approximately 200 nm in diameter) includes a lipid bilayer envelope (solid arrow head), tegument space including about 12 HSVderived proteins (open arrow head), and an internal icosahedral capsid containing the viral DNA. Magnification 265,000. Kindly provided by Drs. Elisabeth Schraner and Peter Wild, Institute of Veterinary Anatomy, University of Zurich, Switzerland.

cellular receptors. The initial contact of HSV-1 with a host cell is binding of glycoproteins B (gB) and C (gC) to cell surface glucosaminoglycans, mainly heparan sulfate but also dermatan sulfate (Shieh et al., 1992; Banfield et al., 1995). Attachment of virions to the cell surface is followed by gD-mediated binding to secondary cell receptors. To date, three classes of HSV-1 gD receptors have been identified: herpesvirus entry mediator A (HveA or HVEM), a member of the tumor necrosis factor receptor family; nectin-1 and nectin-2, two members of the immunoglobulin superfamily; and specific sites in heparan sulfate generated by certain isoforms of 3-O-sulfotransferase (Spear et al., 2000). Any one of these cell surface molecules can bind to gD triggering viral penetration into cells. Upon viral entry into the cell, tegument proteins and capsids are transported along microtubules to the nuclear pores, where the viral genome is released into the nucleoplasm and remains as an episome (Sodeik et al., 1997). The HSV-1 genome is approximately 152 kbp in size and consists of two covalently linked components, UL and Ug, each flanked by inverted repeats (Stevens, 1975). During replication, the viral DNA is synthesized by a rolling circle mechanism, yielding concatemers, which are subsequently cleaved at pac sequences into

unit-length monomers upon packaging when the viral capsids are filled with DNA (Severini et al., 1994). HSV-1 can either enter a lytic life cycle, leading to viral replication and cell death, or establish latency characterized by the persistence of episomal viral DNA within sensory neurons (Preston, 2000). During lytic infection, approximately 85 open reading frames of HSV-1 are transcribed and translated in a tightly regulated and orderly manner. The genes are characterized in accordance with their time of expression as immediate early (IE), early (E), and late (L) genes (Rajcani et al., 2004). The tegument protein VP16 is transported to the nucleus after cell entry and acts as a transactivator (Dalrymple et al., 1985; Wysocka and Herr, 2003; Jonker et al., 2005). In turn, these IE genes initiate expression of both E and L genes, which regulate the replication of the virus genome and contribute structural proteins for virus assembly, respectively. Alternatively, HSV-1 can establish a lifelong latent infection within neurons. During latency, no viral proteins are detectable and transcription of the episomal viral genome is limited to the latency-associated transcripts (LATs). Periodic reactivation of the latent virus leads to a new spread of viral infection (Mitchell et al., 2003). B.

HSV-1 Amplicon Vector

Viruses represent successful intracellular parasites, which have evolved efficient mechanisms to introduce and express their genetic information in recipient cells. Therefore, viruses represent natural vectors for the transfer of foreign genetic material into cells. Much effort has gone into the development of safe and effective viral vector systems, which have become promising tools for gene therapy (Smith, 1995; Advani et al., 2002). Spaete and Frenkel (1982) were the first to document the presence of defective HSV-1 particles upon high-multiplicity passaging of wild-type HSV-1. DNA molecules isolated from these defective particles were found to be composed of a viral genome with multiple reiterations of specific HSV-1 sequences, including a DNA cleavage/packaging signal {pac) and an origin of DNA replication (pri) (Spaete and Frenkel, 1985). The analysis of these defective HSV-1 particles established the concept of herpes virus amplicons. The essential components for amplicon vectors include: both bacterial and viral origins of replication necessary for cloning in gene bacteria and viral replication/packaging in mammalian cells, respectively; an antibiotic resistance gene for selection in bacteria; the pac sequence responsible for directing cleavage and packaging into

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

BASICS OF HSV'l AMPLICON

virions; and finally, the transcription unit(s) for transgene expression (Fig. 2). In the presence of HSV-1 helper functions, amplicons are replicated and packaged into HSV-1 virions resulting in lysis of the host cell. In 1985, the first HSV-1 amplicon vector-mediated expression of a transgene in mammalian cells was reported (Kwong and Frenkel, 1985). Over the past 20 years, several research groups have focused on the further development of the amplicon vector system to improve efficiency and safety of gene delivery (Maguire-Zeiss et a l , 2001; Epstein and Manservigi, 2004; Oehmig et al., 2004).

The production of amplicon vector stocks requires helper virus functions for delivering regulatory as well as structural proteins. In the beginning of the amplicon era, wild-type virus was used as helper virus (Spaete and Frenkel, 1982; Kwong and Frenkel, 1985). However, the fact that amplicon stocks were contaminated with cytotoxic helper virus limited the use of amplicon vectors for gene delivery. The next step to improve the safety of amplicon vectors was the replacement of wild-type HSV-1 by replication-conditional mutant HSV-1 helper viruses. Such replication-conditional viruses typically carry mutations in essential

Transgene cassette

HSV-l

27

amp'

colEl Co-transfection into X permissive cells

Progeny

FIGURE 2 Helper virus-free packaging of HSV-1 Amplicon Vectors into HSV-1 particles. The amplicon vector contains the HSV-1 origin of DNA replication {ori), the HSV-1 packaging signal {pac), the Escherichia coli origin of DNA replication (colEl), an antibiotic resistance gene and a transgene cassette. The /:7flc-deleted HSV-1 genome is cloned into a bacterial artificial chromosome (BAC). The HSV-1 BAC Apac DNA provides all the functions necessary for replication and packaging of the amplicon vector. Co-transfection of permissive eukaryotic cells w^ith amplicon vector and packaging-defective HSV-1 BAC Apac DNA results in helper virus-free stocks of packaged amplicon vectors.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

28

3. HSV AMPLICON VECTORS FOR GENE DELIVERY TO THE NERVOUS SYSTEM

immediate early genes and can support amplicon packaging in complementing cell lines (Geller et al., 1990; Lim et a l , 1996). Unfortunately, methods for purification of amplicon vectors could not eliminate the presence of contaminating helper virus, which can induce cytotoxicity and inflammatory responses. Therefore, Logvinoff and Epstein (2001) developed a Cre-lox system to excise the packaging sequences from the helper virus genome for the production of large amounts of amplicon vector stocks with low helper virus contamination. They created a recombinant HSV-1 helper virus that contains the pac signal flanked by loxP sites. By performing the final passage of amplicon packaging in a Cre-expressing cell line, the loxP-flanked pac signal in the helper virus genome is excised to inhibit packaging. This method can generate high-titer amplicon stocks with low helper virus contamination. To circumvent helper virus-associated problems such as toxicity, conversion of the helper virus into a wild-type HSV-1 phenotype, and potential interactions with endogenous viruses, a helper virusfree packaging system was developed. A set of five cosmids, containing the entire HSV-1 genome with deleted pac signals, was the basis of the first helper virus-free packaging system (Fraefel et al., 1996). These cosmids do not produce replication-competent virus progeny, but can efficiently replicate and package co-transfected amplicon DNA. With this method, helper virus-free amplicon stocks could be produced, but with the drawback of relatively low virus titers. This cosmid-based method was further simplified by cloning the entire p^c-deleted HSV-1 genome as a bacterial artificial chromosome (BAG) (Fig. 2) (Stavropoulos and Strathdee, 1988; Saeki et al., 1998; Horsburgh et al., 1999). However, the 1 kb sequence homology between the ori on the HSV-1 BAG DNA and the ori on the amplicon allowed the generation of packaging-competent HSV-1 helper genomes via homologous recombination events. This resulted in amplicon vector stocks contaminated with replicationcompetent HSV-1, albeit at low levels. To eliminate this possibility, an essential HSV-1 gene was deleted and additional coding sequences were added to HSV-1 BAG to produce a viral genome too large to be packaged into a viral capsid (Saeki et al., 2001). Amplicon stocks produced by this method have no detectable HSV helper virus contamination (Saeki et al., 2001) and induce no detectable immune responses after in vivo administration, even at high doses (Olschowka et al., 2003).

IL

ADVANTAGES OF HSV-1 AMPLICON

The application of amplicon vectors for gene delivery into cells both in culture and in vivo are well established (Sandler et al., 2002; Wang et al., 2002c; Seijffers and Woolf, 2004; Muller et al., 2005). HSV-1 amplicons have many advantages including simple manipulation, a large transgene capacity, easy production, minimal or no cytopathic effect or immune response, and broad cellular tropism. However, transgene expression from amplicon vectors in dividing cells is not stable (Fraefel et al., 1997; Johnston et al., 1997). Improvements to the amplicon vector include the use of the virion itself to deliver proteins and target infection, maximizing the transgene payloads, directing the fate of amplicon DNA in host cells, and the use of cell type-specific and regulatable promoters to control transgene expression. A»

Modification of Virions

The HSV-1 amplicon vector is composed of the same compartments as the wild-type HSV-1 virion. All three compartments — envelope, tegument, and capsid — can be modified and adapted for specific purposes. The broad cellular tropism of HSV-1 amplicons is useful, but for some applications it would also be advantageous to target gene delivery to specific cell populations. Entry of HSV-1 into host cells requires different viral glycoproteins. Attachment of the viral envelope to the cell membrane is initiated by the binding of gB and gG to cell surface proteoglycans. Attachment enables gD to interact with one of several cellular surface receptors, resulting in recruitment of gB, gH, and gL to induce virus entry (Turner et al., 1998). Foreign ligands engineered into gB a n d / o r gG have been demonstrated to target tropism of HSV-1 to specific cell types (Laquerre et al., 1998; Grandi et al., 2004a, b). Grandi et al. (2004a, b) replaced the gG heparan sulfate-binding domain with a hexameric histidine tag. The resulting amplicons had enhanced transduction efficiency in cells expressing specific his receptors as compared to standard amplicons. Restriction of the viral host range has also been achieved by gD mutations. gD binds to several cell surface receptors, including HVEM, nectin-1, nectin-2, and 3-O-sulfated heparan sulfate (Spear, 2004). These cellular receptors interact with gD at different sites (GonnoUy et al., 2002, 2005; Yoon et al., 2003). Mutagenesis of gD can abrogate binding to one specific receptor with no effect on interactions with

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

ADVANTAGES OF HSV4 AMPLICON

other cellular receptors (Manoj et al., 2004; Yoon and Spear, 2004). An interesting approach has been taken by Nakano et al. (2005), who developed a method for targeting wild-type HSV-1 to specific cell surface receptors without modification of the envelope glycoproteins. An adapter protein consisting of the nectin-1 gD-binding domain fused to an epidermal growth factor receptor (EGFR)-specific singlechain antibody was used to target HSV-1 infection to cells expressing this receptor. In the presence of the specific adapter molecule, HSV-1 could even enter into HSV-resistant cells expressing EGFR. Although the adapter protein binds to the nectin-1 gD-binding region, cell entry via nectin-1 receptor was not inhibited in the presence of the adapter molecule (Nakano et al., 2005). Amplicons are not only efficient gene delivery vehicles, but they can also deliver peptides and full-length functional proteins. Fusion of structural virus proteins to foreign proteins (creating chimeric proteins) provides a means to deliver proteins to cells. For example, the tegument protein VP22 can be used as an intercellular protein shuttle (Elliott and O'Hare, 1997) to deliver therapeutic proteins such as p53 or HSV-1-thymidine kinase into neighboring cells (Phelan et al., 1998; Dilber et al., 1999). This intercellular movement can be visualized with a fluorescent protein-tagged tegument protein (Elliott and O'Hare, 1999a, b). Intracellular movement of these chimeric proteins can also be visualized, such as the retrograde axonal transport of capsids and associated tegument proteins, by monitoring the movement of a VP16-GFP fusion protein incorporated into the tegument (Bearer et al., 2000). VP16-GFP-labeled virions have also been used to monitor their diffusion and movement within tumors over time in a tumor window model (McKee et al., submitted). Capsid proteins can also be used to label virions and target proteins to the nucleus. The capsid protein VP26 was fused to GFP and incorporated into virions (Desai and Person, 1998). GFP did not affect the cellular distribution of VP26 and was detected within the nucleus within minutes after infection. These experiments indicate that therapeutic and imaging proteins fused to capsid or tegument proteins have the potential to be transported efficiently to the nucleus and to carry out functional activities. Such protein cargo is ideal for tracking virions in vivo, and especially suitable to deliver proteins, which are ideally present in the infected cell for only a limited time after vector delivery.

29

B. Transgene Capacity of HSV-1 Amplicon Vectors The most outstanding property of amplicons is their large transgene capacity. Up to 150 kb of foreign DNA can be delivered to the nucleus of transduced cells. There is no comparable mammalian vector system displaying such a high efficiency and capacity of gene delivery as the HSV-1 amplicon vector. Initial experiments were carried out with a monocistronic transgene HSV-1 amplicon. Subsequently, bicistronic vectors were designed, harboring an internal ribosomal entry site (IRES) from the encephalomyocarditis virus, allowing the expression of two gene products (Jacobs et al., 2003). More recently, amplicon vectors supporting the expression of multicistronic DNA (Bujold et al., 2002) or multiple transgenes under different controlling elements (Strathdee and McLeod, 2000; Wang et a l , 2003) have been developed. Amplicons have also proven to be an efficient mode of delivery of genomic DNA cloned in BAG libraries that contain fragments in the 150 kbp range, with a loxP site in the amplicon and BAG used to facilitate incorporation in the vector in the presence of Gre recombinase (Wade-Martins et al., 2001). Wade-Martins et al. (2003) demonstrated packaging of human genomic DNA of approximately 135 kbp into amplicon vectors and efficient delivery to human cells in culture. The advantage of this delivery system is that entire genes (up to 135 kb) can be delivered with their cognate 5' and intronic regulatory elements sequences, thus allowing physiologic regulation, for example, of the low-density lipoprotein receptor (Wade-Martins et al., 2003). This can be critical in recapitulating complex loci function, where differential splicing yields distinct functional proteins, as in the GDKN2A/ GDKN2B region (Inoue et al., 2004).

C.

Hybrid Amplicons

A drawback of amplicons is the lack of stable retention of their DNA in the host cell. As the amplicon DNA does not integrate into host chromosomes, it is lost in dividing cells over a few generations (Johnston et al., 1997). Even in nondividing cells, decay in levels of transgene expression can be observed over time, probably due, at least in part, to degradation a n d / o r silencing of the amplicon genome. This drawback has been overcome by designing hybrid amplicons. Hybrid amplicons carry HSV-1 pac and ori sequences and elements from other viral vectors, which can support

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

30

3. HSV AMPLICON VECTORS FOR GENE DELIVERY TO THE NERVOUS SYSTEM

either genomic integration or episomal persistence of transgenes, all packaged within the HSV-1 virion. For example, Epstein-Barr virus (EBV) is classified as gammaherpesvirus and infects epithelial cells and B cells. In B-lymphocytes, the EBV genome is stably maintained as an extrachromosomal element replicating in parallel with the host cell genome (Hammerschmidt and Sugden, 2004). The genetic elements required for long-term episomal retention are the EBV latent origin of replication (oriP) and the EBV nuclear antigen-1 (EBNA-1). EBNA-1 interacts with oriP and with host cell chromosomes, promoting nuclear retention, cellcycle-synchronous replication and segregation with chromosomes at mitosis (Reisman et al., 1985). The incorporation of these EBV elements into the HSV-1 amplicon (Fig. 3A) resulted in a hybrid vector with a broad host range that could replicate and segregate

A

in dividing cells and, therefore, persist as an episomal element for prolonged times (Wang and Vos, 1996). HSV/EBV hybrid vectors have been used to sustain longer transgene expression in culture and in vivo, as compared to standard amplicon vectors (Wade-Martins et al., 2001; MuUer et al., 2005). Reporter gene expression was observed for at least 6 weeks in transduced human hepatocytes using HSV/EBV hybrid vector, while conventional amplicon vectors expressed the reporter gene for only 2 weeks in these cells. Although EBNA-1 and oriP are not fully functional in rodent cells, as compared to human cells, transgene expression in mouse liver was detected over 3 weeks after HSV/EBV hybrids vector injection into the liver (MuUer et al., 2005), and Wade-Martins et al. (2003) have reported prolonged maintenance of transgene cassettes of up to 135 kbp by HSV/EBV hybrid vectors in dividing cells.

AAV ITR

EBV EBNA-1

Transgene cassette

Transgene cassette

AAV ITR EBV oriP

colEJ

colEl

gag-pol-env EBV EBNA-1

AAV ITR LTR

D

LTR

Transgene cassette

1ransgene ^ cassette

_

LTR ^HSV/EBV/Mo-MLV HSV-1 H ^ ^ EBY oriP ori ^ HSV-1 amp'

1=^=1 LTR AAV ITR

HSV-1 ori

pac colEl

colEl

FIGURE 3 HSV-1 hybrid amplicons. (A) Structure of the HSV/EBV hybrid amplicon. In addition to the standard amplicon elements, HSV/EBV hybrid amplicons contain sequences from EBV, in particular the EBNA-1 gene and the EBV origin of DNA replication oriP, which support episomal maintenance and segregation in transduced cells. (B) HSV/AAV amplicons contain a gene of interest that is flanked by the AAV ITRs and the AAV rqp, which mediate the integration of the ITR-flanked transgene into a specific site on chromosome 19. (C) HSV/EBV/Mo-MLV tribrid amplicons contain genetic elements from three different viruses, the HSV-1 elements (ori and -pac), the EBV elements (EBNA-1 and oriP), and gag-pol-env genes and the LTR from Mo-MLV retrovirus, gag-pol-env genes are responsible for integration and production of the LTR-flanked transgene cassette. (D) HSV/AAV/Mo-MLV tribrid amplicons contain in addition to the HSV/AAV hybrid elements, the Mo-MLV-specific elements (LTR a n d gag-env-pol).

i

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

ADVANTAGES OF HSV-l AMPLICON

Another way to insure stable retention of transgenes is by insertion into the host cell genome. Since random insertions can yield variable transgene expression and potentially oncogenic events, efforts have focused on introducing elements into HSV amplicons to achieve site-specific integration. Site-specific integration of amplicon DNA into the host genome has been achieved by inserting genetic elements from adeno-associated virus (AAV) into the HSV amplicon vector (Fig. 3B). AAV belongs to the Parvoviridae family and causes no diseases in human. The AAV genome is a linear, single-stranded DNA of 4.7 kb, which has the unique ability to integrate into a specific site (AAVSl) on human chromosome 19ql3.3 without causing any apparent cytopathic effects (Berns and Linden, 1995). Two inverted terminal repeats (ITRs) a n d / o r the p5 promoter and the AAV Rep proteins, either Rep68 or Rep78, are sufficient to mediate site-specific integration (Linden et a l , 1996; Philpott et al., 2002a, b). The small transgene capacity of AAV-based vectors has placed some limitations on their use as gene delivery vectors. However, inserting the rep gene and the ITRs of AAV into an HSV-1 amplicon vector combines advantages of both vector systems, including the potential for large transgene delivery and integration of ITR-flanked transgene cassettes into a specific site on human genome 19 (Heister et al., 2002; Wang et al., 2002c). HSV/AAV hybrid vectors, in comparison with standard HSV-1 amplicon vectors, extend transgene expression in dividing and nondividing cells in culture and in vivo (Fraefel et al., 1997; Johnston et al., 1997). Site-specific integration of an AAV ITR-flanked reporter gene into chromosome 19 following transduction of human cells with HSV/AAV hybrid has been demonstrated to occur in 5-15% of infected cells (Heister et al., 2002; Wang et al., 2002c). Stable reporter gene expression was observed for at least 12 months in dividing 293 cells without chemical selection (Heister et al., 2002). Random integration of the ITR cassette along with sequences from the vector backbone also occurred, but most integration events were at the AAVSl site. One remaining hurdle of HSV/AAV hybrid vectors is the low packaging efficiency and titers when rep sequences are included in the amplicon vector (Heister et a l , 2002; Wang et al., 2002c). It has previously been reported that Rep can interfere with HSV-1 replication (Heilbronn et al., 1990). Another type of hybrid vector was designed to launch the production of retrovirus vectors by cells infected with HSV-1 amplicon vectors (Sena-Esteves et al., 1999,2002). Retroviruses are small single-stranded RNA viruses, which convert into double-stranded

31

DNA in the cytoplasm and integrate randomly into the host genome. Production of retroviral progeny occurs without killing the host cell. Some retroviruses, such as Moloney murine leukemia virus (Mo-MLV), require dividing cells for the viral DNA to be transported into the nucleus and integrated. A tribrid vector was developed, which contains genetic elements from three different viruses: HSV-1 ori and pac for replication and packaging, the long terminal repeats (LTRs) and gag-pol-env genes from Mo-MLV for production of retrovirus vectors and transgene integration, and either EBV oriP and EBNA-1 or the AAV ITR and rep to stabilize the transgene sequences in host cells (Figs. 3C and D). Tribrid vectors packaged in HSV-1 virions can infect dividing and nondividing cells resulting in the on-site production of retrovirus vectors, which can spread to neighboring cells. Following LTR-mediated integration, the LTR-flanked transgene cassette replicated along with the host genome. However, since the integration of the LTR cassette occurs randomly, there remains the potential risk of mutagenesis at the site of integration. Injection of HSV/EBV/Mo-MLV tribrid vector into pre-established tumors in mice showed a 4-fold increase of transgene expression in tumor cells 10 days postinjection compared to control amplicons, which lacked the retrovirus elements due to secondary retrovirus infection (Hampl et al., 2003). D.

Gene Regulation

Conventional amplicon vectors commonly use viral promoters, such as the HSV-1 immediately early 4 / 5 promoter, the human cytomegalovirus immediate early promoter or the SV40 early promoter (Ho et al., 1993; Smith et al., 1995; Hoshi et al., 2000) to control transgene expression. These viral promoters drive high levels of transgene expression and their activities are not restricted to specific cells and tissues. However, these promoters are subject to silencing after a relatively short period post-infection by as yet undefined mechanisms (Fraefel et al., 1997, 2005). Long-term gene expression has, however, been observed with promoter sequences responsible for cell or tissue typespecific expression (Kaplitt et al., 1994; Jin et al., 1996; Song et al., 1997). The potential of HSV-1 amplicon vectors to carry large regulatory regions makes the system valuable for cell type-specific expression of the transgene in different organs. Neuronal-specific preproenkephalin promoter sequences yielded expression of transgenes in neuronal populations for at least 2 months, although expression levels dropped over time (Kaplitt et al., 1994). Up to 10 weeks of stable

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

32

3. HSV AMPLICON VECTORS FOR GENE DELIVERY TO THE NERVOUS SYSTEM

reporter gene expression was observed in rat neuronal cells, when the reporter gene was under the control of the 9 kbp tyrosine hydroxylase promoter region, with expression being anatomically restricted to catecholaminergic neurons (Jin et al., 1996; Song et al., 1997). Specific transgene expression in hepatocytes was achieved by using liver-specific promoter and enhancer sequences (Fraefel et a l , 1997). The use of cell- or tissue-specific promoters in amplicon vectors can prolong transgene expression, but does not provide a means of controlling levels of expression. Since regulation of gene expression can be critical in gene therapy, several mammalian gene regulation systems using readily bioavailable drugs have been explored. Glucocorticoid-inducible expression systems were the first prototype regulation systems implemented into the amplicon vector. The transcription unit contained a promoter with glucocorticoid-responsive elements and a transgene. In the presence of glucocorticoids, up to 50-fold induction of transgene expression was achieved in transduced primary rat hepatocytes (Lu and Federoff, 1995). The tetracycline-responsive expression system has also been examined in the amplicon vector allowing both drug-inducible, as well as drug-suppressible, gene expression (Ho et al., 1996; Fotaki et al., 1997). A new version of the tetracycline-inducible amplicon incorporates silencing proteins for the tetracyclineresponsive promoter, which bind in the absence of the drug (Freundlieb et al., 1999), as well as insulator elements on either side of the transgene cassette (Sena-Esteves et al., unpublished results). This tetracycline-inducible amplicon system was used to regulate expression of human torsinA in cultured cells, a protein, which underlies early-onset torsion dystonia in humans (Ozelius et al., 1997). Levels of torsinA increased in a dose-dependent manner with the amount of doxycycline administered (Bragg et al., 2004). In vivo experiments with an amplicon incorporating the tetracycline-suppressible system confirmed that tetracycline could pass through the blood-brain barrier (Ho et al., 1996; Fotaki et al., 1997). Early versions of the tetracycline-responsive system, as well as the glucocorticoid-regulated gene expression system, displayed a high level of basal activity even in the noninduced state. This leakiness could be due to the presence of tegument proteins introduced during amplicon vector infection. The tegument components include many proteins that affect gene expression. For example, the tegument protein VP16 is a strong transactivator of a viral enhancer sequence present

in the amplicon (Jonker et al., 2005). Furthermore, the pac and the ori sequence of HSV-1 within the amplicon contain transcriptional regulatory motifs, which can affect promoter specificity (McKnight and Tjian, 1986; Lu and Federoff, 1995). Also the fact that, depending on the size of the seed amplicon plasmid, HSV-1 amplicon vectors deliver multiple copies of transgene cassettes as a concatenate to cells can perturb regulated gene expression. Another approach for dose-dependent transgene regulation in an amplicon vector is based on the rapamycin-induced dimerization system to control transcription (Wang et al., 2003). A DNA-binding domain and an activation domain are fused to two different rapamycin-binding proteins. Upon addition of rapamycin, heterodimerization is induced forming a functional transcription factor, which activates transgene expression. Cell cultures transduced with an HSV-1 amplicon vector containing all components of the rapamycin regulatable dimerizer system showed a dose-dependent expression of the marker gene, with a maximum 25-fold increase. This reporter gene induction was also observed in the rodent brain after injection of HSV-1 amplicon vector and systemic delivery of rapamycin. This system shows very low baseline activity, but the level of induction is modest limiting its potential in its current form (Wang et al., 2003). Amplicon vectors are also amenable to high-throughput screening assays. Promoters, enhancer elements, insulator or even genes responsible for specific functions can be identified in culture or in vivo. Huang and Brandt (2003) developed an amplicon vector system capable of screening for promoter elements (Huang and Brandt, 2003). They cloned a library of random genomic DNA fragments upstream of a promoterless GFP open reading frame into an amplicon vector. The recombinant amplicon library was transduced into cells or tissue, and GFP-positive cells were isolated by fluorescence activated cell sorter (FACS). Cells could only express GFP when the inserted elements supported gene expression.

IIL A,

USES IN NERVOUS SYSTEM

Gene Delivery

HSV amplicon vectors have a broad tropism for many cell types through a basic binding mechanism to heparan sulfate on the cell surface and entry through at least four different receptors (Frampton et al., 2005). For most cells in culture infection of 70-90% of cells can usually be achieved with a multiplicity of infection

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

USES IN NERVOUS SYSTEM

(MOI) 1-5 transducing units (tu) per cell with transgene expression detected within 24 h and continuing on for weeks in culture (months in vivo). Although the vector itself appears to have no toxicity, cell lysate debris in stocks can have toxic effects. In vivo, the range of delivery is limited to some extent by the large size of the virions, which restricts diffusion, and by binding to heparin moieties in the extracellular matrix. To avoid damage to nervous tissue, small volumes of inoculum are injected in the range of 1-2 |Lil, which for a typical titer of 1 X 10^ t u / m l is equivalent to about 100,000 tu. Usually, thousands of cells are infected at the injection site, but likely each of those cells receives multiple infections (Fig. 4). In the brain, all cell types in an area are infected, with neurons being the predominant type, with fewer astrocytes and oligodendrocytes (e.g. Costantini et al., 1999; Agudo et al., 2002; Sandler et al., 2002). Cell type-specific, and frequently longer term, expression can be achieved using mammalian promoters specific to the cell of interest. For example, the preproenkephalin (ENK) promoter supports expression in neurons normally expressing this endorphin (Kaplitt et al., 1994), the tyrosine hydroxylase promoter in catecholaminergic neurons (Wang et al., 1999), and a hybrid promoter combining elements of the ENK promoter with that of the neurofilament heavy chain in striatal neurons (Wang et a l , 2004).

33

Morphology and electrophysiology of viable cells in slices from infected regions of the brain were found to be completely normal (Sandler et al., 2002; Rumpel et al., 2005) with minimal-low inflammatory response and no immune cell infiltration (Olschowka et al., 2003). Within the brain, the vector is efficiently moved within neuronal processes by active dynein-mediated retrograde transport to cell bodies at some distance from the injection site, for example, from the striatum to dopaminergic cell bodies in the substantia nigra (Costantini et al., 1999) and from the inferior olive to the cerebellar nuclei, with high infectivity of Furkinje cells (Agudo et al., 2002). Given that sensory neurons in the periphery are typically "home" to HSV, it is not surprising that these neurons are highly infectable (Marsh et al., 2000). Other neuronal populations that have been explored include cells in the inner ear (Derby et al., 1999), with injection through the utriculus providing delivery to both inner and outer hair cells responsible for sensory transduction (Praetorius et al., 2002). Within the eye, HSV amplicon vectors have been found to efficiently infect retinal ganglion, retinal pigment epithelial and photoreceptor cells (Wang et al., 2002b; Fraefel et al., 2005). The development of informative imaging reporters now allows refined assessment of intracellular dynamics, as well as parameters of gene delivery in vivo.

FIGURE 4 HSV amplicon-mediated gene delivery to mouse brain. An amplicon vector encoding GFP under the CMV-jS actin promoter (CBA; Daly et al., 1999), 2 X 10^ tu in 2 |il, was injected into the cortex of a nude mouse brain and visualized by fluorescence microscopy in frozen sections 21 days after injection. (A) Injection tract, (B) brain adjacent to tract. Magnification 10 X. Kindly Provided by Dr. Sam Wang, Mass. General Hospital, Boston, MA.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

34

3. HSV AMPLICON VECTORS FOR GENE DELIVERY TO THE NERVOUS SYSTEM

Intracellular reporters include a wide variety of fusion constructs in which cellular proteins are fused to fluorescent proteins so as to monitor dynamic changes in their location in cells in response to different signals (van Roessel and Brand, 2002; Tsien, 2005), as well as to proteins containing chemically modifiable sequences, like the tetracysteine arsenate-binding motif and biotinylation sequences (Sosinsky et al., 2003; Howarth et al., 2005; Tannous et al., in preparation). Other reporters, such as caged luciferases (Laxman et al., 2002) and near-infrared-fluorescence biopolymers, incorporate peptide sequences that are cleaved by specific proteases thereby activating the reporter. Such reporters can be delivered by a variety of vector types with the choice depending on the size of transgene cassettes and tropism of virions. Examples in which amplicon vectors have been used for cell biology imaging include GFP fusions to the cyclic AMP-response-element-binding protein (CREB; Olson et al., 2005) and the AMPA-type glutamate receptor (Rumpel et al., 2005) to monitor their intracellular movement in response to drugs and learning paradigms, respectively. Near-infrared-imaging probes have been designed to monitor activities of the apoptotic protein, caspase I (Messerli et al., 2004), and HIV-1 protease (Shah et al., 2004) delivered by amplicon vectors to tumor cells in culture and in vivo. Amplicon vectors encoding luciferase have also been used to monitor the extent of gene delivery into tumors by bioluminescence imaging (Shah et al., 2003), and an engineered transferrin receptor serves to monitor endocytosis of paramagnetic particles by magnetic resonance imaging (Ichikawa et a l , 2002). Amplicon vectors have also been used to monitor expression of a therapeutic prodrug-activating enzyme, HSVthymidine kinase (TK) in tumors using a radioactive substrate, which is retained by cells after activation by phosphorylation using positron emission tomography (Jacobs et al., 2003).

B*

Elucidating Neuronal Functions

HSV amplicon vectors provide an ideal vehicle to explore neuronal functions based on their efficient gene delivery to neurons, minimal impact on cellular physiology, and ease of generation. They have been employed by neuroscientists as a probe to deliver wild-type and rautant proteins, as well as probes to explore cell biology of neurons and functions involved in learning and memory, drug addition and psychiatric behavioral models. In such studies, the vectors become primarily a tool and hence are not always referred to per se. Some examples of use of these vectors in this

context include: delivery of an axonal marker — human placental alkaline phosphatase to monitor neurite outgrowth of sensory neurons (Seijffers and Woolf, 2004) and delivery of fibroblast growth factor-2 to steer the fate of embryonic precursor cells toward differentiated neurons (Vicario and Schimmang, 2003). These vectors also provide an efficient means to express mutant proteins in neuronal cells, as has been done to show formation of membrane inclusions in the cytoplasm of cells by the mutant form of torsinA, which is responsible for most cases of early onset torsion dystonia (Bragg et al., 2004), and aggregation of full-length mutant huntingtin protein (350 kDa) in which the extended polyglutamine tract associated with Huntington's disease is exposed (Persichetti et a l , 1999). HSV amplicon vectors have proven their worth in a number of studies in which synaptic dynamics within specific regions of the brain are probed, with the advantages being that they can target specific neurons by the mode of injection, achieving focal delivery to a specific region, and they do not perturb basic synaptic activity Interesting examples of this approach include expression of a functional GFP-CREB fusion protein or a dominant-negative form of CREB in the ventral tegmental area of the brain, which serves as a major reward region, to identify subzones affecting drug reward behavior and to monitor effects on levels of tyrosine hydroxylase and the AMPA glutamate receptor subunit, GluRl (Olson et al., 2005). This dominant negative form of CREB was also used to demonstrate CREB involvement in the nucleus accumbens in anxiety-like behavior (Barrot et al., 2005). Delivery of a functional GFP-GluRl receptor documented its increased incorporation at synapses in the amygdala engaged in long-term potentiation associated with learning and memory (Rumpel et al., 2005). Further, increased levels of GluRl-enhanced rectification associated with associative learning, as measured electrophysiologically in viable slices prepared from the injected brain region, and expression of a dominant negative GluRlrepressed rectification. Other learning paradigms explored with HSV amplicon vectors include demonstration of improved reacquisition of an auditory discrimination task by delivery of a constitutively active form of protein kinase C into the hippocampus (Neill et al., 2001); loss of performance in paradigms of habituation to open field and avoidance of foot shock by expression of anti-sense RNA to the NRl subunit of the NMDA receptor (Adrover et al., 2003); and increased spatial learning performance after hippocampal injection of a vector expressing nerve growth factor into mice.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

USES IN NERVOUS SYSTEM

which were haploinsufficient for this growth factor (Brooks et a l , 2000). C. Gene Therapy for Neurologic Disease and Brain Tumors The high efficiency of gene delivery in the nervous system, large transgene capacity, and sustained expression of HSV amplicon vectors have stimulated their use in a number of therapeutic paradigms for neurologic diseases and brain tumors. For neurologic disease, approaches have included delivery of protective or compensatory proteins in models of environmental insult, replacement of missing proteins in recessive genetic diseases, and removal of toxic proteins in dominant genetic diseases. In the protection model, delivery of growth factors has been shown to help neurons survive stress conditions. For example, the p75 neurotrophin receptor, which responds to NGF, served to protect neurons from the toxic effects of extracellular A-jS protein, which accumulated in Alzheimer's disease (Zhang et al., 2003), and glialderived neurotrophic factor (GDNF) protected cortical neurons from neuronal loss following cerebral ischemia, thereby preventing motor deficits in mice (Harvey et al., 2003). Interestingly neurons were also found to be protected from oxidative stress not only by expression of glutamic acid decarboxylase (GAD67) but also, to a lesser extent, by the empty vector itself (Lamigeon et al., 2003), suggesting the possibility that virion proteins may have some beneficial influence on cellular physiology. The impact of neuronal loss is nowhere more critical or widespread than that of death of hair cells underlying deafness. Delivery of the growth factor, neurotrophin-3 has been shown to protect supporting neurons in the ear from the toxic effects of the chemotherapeutic drug, cisplatin (Bowers et a l , 2002) and to stimulate survival of auditory neurons during developmental time windows (Carnicero et al., 2002). "Replacement" therapy can include delivery of compensatory proteins or the missing proteins themselves. For example, correction of motor abnormalities and increased dopamine synthesis was achieved by amplicon-mediated delivery of cDNAs for tyrosine hydroxylase and aromatic acid decarboxylase in a lesion model of Parkinson's disease (Sun et al., 2003). Amplicon vectors are the vector of choice for large cDNAs, such as that for ATM kinase (9 kbp), which is defective in ataxia telangiectasia. This kinase is critical in signaling events underlying protective responses to DNA-damaging agents, as well as in cell cycle

35

control and genomic stability (Rotman and Shiloh, 1999). Delivery of the ATM cDNA to null cells from patients has been found to yield functional recovery of kinase activity, appropriate cell cycle arrest in response to y-irradiation and protection from oxidative stress (Cortes et a l , 2003; Qi et al., 2004). Recent studies show that the amplicon vector can be used to restore dystrophin function encoded in a 17.2 kb cDNA to muscle following delivery to myoblasts and subsequent grafting into dystrophin-deficient mice (Bujold et al., 2002). HSV/AAV amplicon vectors hold great promise for stable delivery of these large cDNAs or genomic fragments into target regions of the genome for sustained restoration of function. In a novel approach to restorative therapy, HSV amplicon vectors, which have a high infectivity for antigen-presenting dendritic cells, have been used to vaccinate animals against toxic proteins (Bowers et al., 2005). In initial studies, peripheral administration of A-j8 expressing vectors was found to decrease A-P deposition in the brains of transgenic mice overexpressing this protein, albeit with some associated inflammatory response. HSV amplicon vectors have also been explored as a means of treatment for experimental brain tumors. Both recombinant and amplicon HSV-1 vectors have proven highly infective for gliomas with transduction efficiencies ranging from 3 to 42% in primary human cultures at an MOI 1 tu/cell (Rueger et al., 2005), and with commonly used human glioma cell lines having infectivities in the range of 11-80% (Lam et al., 2002). HSV amplicon vectors armed with antitumor agents have been shown to inhibit growth and cause regression of gliomas and other tumor types with single and multiple injections. Anti-tumor agents have included: a secretable version of the apoptotic protein, TRAIL (Shah et al., 2004); prodrug-activating enzymes, cytochrome P450B1 (Rainov et al., 1998) and HSV-TK (Wang et a l , 2002a); anti-angiogenic agents, such as soluble liver kinase-1 (sFlk-1; Pin et al., 2004); the tissue inhibitor of metalloproteinase-1 (TIMP-1), which blocks invasive growth of glioma cells (Hoshi et al., 2000); and immune-enhancing cytokines, such as granulocyte-macrophage-colony-stimulating factor (GM-CSF; Herrlinger et a l , 2000; Toda et al., 2000) and IL12 (Jarnagin et al., 2003). Three lines of investigation demonstrate the unique potential of amplicon vectors for tumor therapy. First, within the amplicon genome, which carries only a few regulatory elements, in contrast to a viral genome, which is loaded with them, it has been possible to achieve fidelity of gene expression in tumor cells using

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

36

3. HSV AMPLICON VECTORS FOR GENE DELIVERY TO THE NERVOUS SYSTEM

1999) (Fig. 5). Anti-tumor efficacy was increased, for example, by combined delivery of the oncolytic vector G207 and an amplicon vector encoding the cytokine, IL2 (Zager et al., 2001). As discussed above, elements of AAV, EBV or retrovirus vectors can be incorporated into amplicon vectors to increase retention of transgene sequences in dividing cells and potentially to amplify transgene sequences after infection.

both a hypoxia-inducible promoter (Pin et al., 2004) and a cell-cycle-regulated promoter further modified for specificity to glial-derived cells by incorporation of a glial acidic fibrillary (GFAP) enhancer element (Ho et al., 2004). Second, the large transgene capacity can be used to encode multiple therapeutic agents, such as in the case of delivery of a combination of immuneenhancing proteins — RANTES, B7.1, and GM-CSF — to maximize the immune response to tumor antigens (Delman et al., 2002). This strategy becomes even more potent when combined w^ith the high-transduction efficiency of dendritic cells and delivery of tumor-specific antigens (Willis et al., 2001). Third, HSV amplicon vectors can be combined with other vectors. For example, they can be propagated together with oncolytic FISV recombinant virus vectors, which are currently being evaluated clinically for glioblastoma therapy (Markert et al., 2000; Chiocca, 2002), and replication of these two HSV vector types can be coupled (Pechan et al..

Recombinant virus vector

IV.

Recent major achievements have promoted a "Renaissance" of the HSV-1 amplicon as promising vector for gene delivery. These include: (i) the development of packaging systems that allow the preparation of vector stocks with little or no helper virus contamination; (ii) the inclusion of elements from other viruses that confer physical stability to

HSV amplicon vector o

virus replication (cell death and vector production) (stable episome)

PERSPECTIVES

transduction (death due to therapeutic gene)

(not toxic to neurons)

FIGURE 5 HSV vectors for brain tumor therapy. For tumor therapy vectors are designed to be selectively toxic to dividing cells, as most normal cells in the brain are not dividing. Recombinant HSV vectors have mutations, which can be complemented in dividing cells, but not in neurons. Infection of neurons is thus benign (column 1), while infection of tumor cells yields virus replication and cell lysis (column 2). HSV amplicon vectors can also be designed such that they carry transgene cassettes, which are selectively toxic to tumor cells, e.g. encoding TRAIL. Such vectors can kill tumor cells in their own right (column 4) or can be combined with HSV recombinant vectors to allow propagation of both within tumor cells (column 3). When this same amplicon vector infects a normal ceU, it is not toxic (column 5).

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

PERSPECTIVES

the transgene in the infected cell nucleus; and (iii) the demonstration that amplicons can deliver functional transgenes of up to 150 kbp, thus making them a tool for genomics. However, important further technological improvements to the amplicon vector system will be critical to allow the transition from experimental to clinical gene transfer protocols. These improvements comprise: (i) the ability to produce large amounts of pure and high-titer vector stocks; (ii) the targeting of amplicon vectors to specific cells or, at least, the cell type-specific expression of the delivered transgene; and (iii) strategies to evade antiviral responses of the host cell/organism. Current helper virus-free packaging systems use replication-competent, packaging-defective HSV-1 helper genomes (cloned as either sets of cosmids or bacterial artificial chromosomes) to provide the HSV-1 proteins required for amplicon replication and packaging. The titers of vector stocks are, therefore, directly influenced by the transfection efficiency. The identification of highly transfectable cell lines and the development of more efficient transfection protocols, which for instance use viral fusogenic proteins (Kaneda et al., 1989; Saeki et al., 1997; Abe et al., 1998) or DNA-binding nuclear proteins (Kaneda et al., 1989), should serve to increase amplicon titers. However, as the ultimate goal, a helper virus-free packaging cell line would eliminate many difficulties with large scale production. The development of such a cell line depends on efficient means to shut-off helper virus gene expression in the un-induced "latent" state and to control viral gene expression following induction. Moreover, either in an integrated state or episomally, the "latent" helper virus genome should replicate and segregate at a relatively low but stable copy number in concert with the cellular genome. Upon "reactivation", for instance, by infection with a seed amplicon vector, the packaging-defective helper virus genomes should express the viral genes in a temporally ordered fashion and replicate autonomously from the cellular genome to provide sufficient helper functions to support efficient replication and packaging of the amplicon DNA. In addition to delivery of foreign genetic information, the HSV-1 particle can deliver peptides and proteins of interest fused with components of the envelope, tegument, or capsid. The delivery of proteins, in addition to transgenes, could be used to (i) catalytically prime reactions, (ii) track virions by in vivo imaging, (iii) activate promoters, (iv) mediate amplification or integration of transgenes in the host cell nucleus, and (v) target infection to specific cell types by modifying envelope glycoproteins.

37

Antiviral responses by the host cell and toxic effects caused by the amplicon have been largely reduced by the introduction of helper virus-free packaging systems. However, certain virion proteins, such as the transcriptional transactivator VP16 and the virion host shut-off protein VHS, may cause some cytopathic effects. While both proteins contain domains that are important to maintain virion structure, the host shutoff domain of VHS and the transactivating domain of VP16 may be modified or deleted. Amplicons may also induce some immune responses, both innate and adaptive, which can result in the silencing of transgene expression, and the elimination of the vector or the transduced cell. Therefore, it may be beneficial to include certain viral genes known to downregulate or to inhibit antiviral and immune responses, such as ICP47 (Fruh et al., 1995) in the amplicon. Improvements on HSV/AAV hybrid vectors can also be envisaged, for example, to address the problem of interaction of the competing replication machineries of two different viruses. The titers of the first generation of HSV/AAV hybrid vector stocks were markedly lower than those obtained with standard HSV-1 amplicons, presumably because the AAV Rep protein interferes with the HSV-1 replication machinery (Heister et al., 2002; Wang et al., 2002c). Novel HSV/AAV hybrid vectors may use the AAV p5 promoter, not to control expression of the AAV rep gene, but as an independent element to mediate (i) site-specific integration (in place of the ITRs), and (ii) HSV/AAV hybrid vector replication during packaging in place of the HSV-1 origin of DNA replication {ori). The p5 promoter, which controls expression of the rep68/78 genes in wild-type AAV, has been shown to mediate efficient. Rep-dependent and ITR-independent site-specific integration into AAVSl on human chromosome 19 (Philpott et al., 2002a, b). As in the first generation HSV/AAV hybrid vectors, the p5 promoter was used to control expression of the rep gene (Johnston et a l , 1997; Heister et al., 2002; Wang et al., 2002c), it may have interfered with site-specific integration of the ITR-flanked transgene cassette and promoted undesired vector backbone integration. Future hybrid vectors may therefore use either the ITRs or the p5 promoter to mediate high frequency site-specific integration and a heterologous promoter, such as T7, to drive low level rep expression (Recchia et al., 1999; Philpott et al., 2002a). The p5 promoter has also been shown to act as an AAV Rep- and helper virus-dependent replication origin (Nony et al., 2001). With HSV-1 as the helper virus, this ori activity is particularly high (Glauser and Fraefel, unpublished results). Moreover, p5 promoter-bearing plasmids are efficiently amplified

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

38

3. HSV AMPLICON VECTORS FOR GENE DELIVERY TO THE NERVOUS SYSTEM

to form large head-to-tail concatamers, which are readily packaged in HSV-1 virions, if an HSV-1 DNA packaging/cleavage signal is provided in cis (Glauser and Fraefel, unpublished results). This opens the possibility to design novel HSV/AAV hybrid vectors, w^hich use the p5 promoter to mediate site-specific integration and serve as the HSV-1-AAV Rep-dependent replication origin in place of the HSV-1 origin. Because replication from the p5 promoter requires only a small fraction of the amount of Rep protein required for replication from the ITRs (Glauser and Fraefel, unpublished results), rep expression during vector production could be kept at a low^ rate so as not to interfere with replication of the helper virus genome. Alternatively, replication of the helper virus genome itself could be switched from Repinhibited to Rep-dependent by replacing the HSV-1 replication origin with the p5 promoter. Also, Rep protein may be delivered by the vector particle as a fusion with a viral tegument protein, such as VHS or VP16 (Oehnnig and Breakefield, unpublished results), which are both abundant tegument components that localize to the cell nucleus upon infection and, therefore, could support Rep-mediated genomic integration via ITRs or p5 promoter. This strategy would eliminate the risk of undesired integration of the rep gene into the host cell genome, which could cause toxicity or subsequent excision of transgene sequences. The most unique feature of HSV-1 amplicons, the large transgene capacity, can also be further exploited. Methods to convert bacterial and phage artificial chromosomes (BACs and PACs) that carry large fragments of genomic human DNA into HSV-1 amplicons have been described (Wade-Martins et al., 2001) and the functional delivery of transgenes with sizes of up to 135 kb has been demonstrated (Wade-Martins et al., 2003; Inoue et al., 2004). This opens the possibility to construct infectious libraries of genomic DNA for large scale functional analyses in high-throughput screens in different cell types. In addition to revealing variations in gene regulation and splicing, in response to different conditions and cell types, it would provide a means to assess genes critical to dynamic processes, such as differentiation (Xing et al., 2004) and to reveal the functional consequences of variations in genomic sequences, which are linked to disease states. In summary, the versatility of the HSV-1 amplicon vector has motivated many research groups to make important contributions toward its use as a safe and efficient gene therapy vector. Further improvements, most importantly on manufacturing, are difficult and time consuming, but will be crucial for the application of HSV-1 amplicon vectors to clinical uses.

ACKNOWLEDGMENTS We thank Suzanne McDavitt for skilled editorial assistance in preparation of this manuscript; Deborah Schuback for help with preparation of figures; and Drs. Elisabeth Schraner, Peter Wild, and Sam Wang for providing us with unpublished images of their work. References Abe, A., Miyanohara, A. and Friedmann, T. (1998) Enhanced gene transfer with fusogenic liposomes containing vesicular stomatitis virus G glycoprotein. J. Virol., 72: 6159-6163. Adrover, M.F., Guyot-Revol, V., Cheli, V.T., Blanco, C , Vidal, R., Alche, L., Komisiuk, E., Epstein, A.L. and Jerusalinsky, D. (2003) Hippocampal infection with HSV-1-derived vectors expressing an NMDARl antisense modifies behavior. Genes Brain Behav., 2: 103-113. Advani, S.J., Weichselbaum, R.R., Whitley, R.J. and Roizman, B. (2002) Friendly fire: redirecting herpes simplex virus-1 for therapeutic applications. Clin. Microbiol. Infect., 8: 551-563. Agudo, M., Trejo, J.L., Lim, F , Avila, J., Torres-Aleman, I., DiazNido, J. and Wandosell, F (2002) Highly efficient and specific gene transfer to Purkinje cells in vivo using a herpes simplex virus I amplicon. Hum. Gene Ther., 13: 665-674. Banfield, B.W., Leduc, Y, Esford, L., Schubert, K. and Tufaro, F (1995) Sequential isolation of proteoglycan synthesis mutants by using herpes simplex virus as a selective agent: evidence for a proteoglycan-independent virus entry path. J. Virol., 69: 3290-3298. Barrot, M., Wallace, D.L., Bolanos, C.A., Graham, D.L., Perrotti, L.I., Neve, R.L., Chambliss, H., Yin, J.C. and Nestier, E.J. (2005) Regulation of anxiety and initiation of sexual behavior by CREB in the nucleus accumbens. Proc. Natl. Acad. Sci. USA, 102: 8357-8362. Bearer, E.L., Breakefield, X.O., Schuback, D., Reese, T.S. and LaVail, J.H. (2000) Retrograde axonal transport of herpes simplex virus: evidence for a single mechanism and a role for tegument. Proc. Natl. Acad. Sci. USA, 97: 8146-8150. Bems, K.I. and Linden, R.M. (1995) The cryptic life style of adenoassociated virus. Bioessays, 17: 237-245. Bowers, W.J., Chen, X., Guo, H., Frisina, D.R., Federoff, H.J. and Frisina, R.D. (2002) Neurotrophin-3 transduction attenuates cisplatin spiral ganglion neuron ototoxicity in the cochlea. Mol. Ther., 6: 12-18. Bowers, W.J., Mastrangelo, M.A., Stanley, H.A., Casey, A.E., Milo, L.J.J, and Federoff, H.J. (2005) HSV amplicon-mediated Abeta vaccination in Tg2576 mice: differential antigen-specific immune responses. Neurobiol. Aging, 26: 393-407. Bragg, D.C., Camp, S.M., Kaufman, C.A., Wilbur, J.D., Boston, H., Schuback, D.E., Hanson, PL, Sena-Esteves, M. and 6Breakefield, X.O. (2004) Perinuclear biogenesis of mutant torsinA inclusions in cultured ceUs infected with tetracycline-regulated herpes simplex virus type 1 amplicon vectors. Neuroscience, 125: 651-661. Brooks, A.L, Cory-Slechta, D.A., Bowers, W.J., Murg, S.L. and Federoff, H.J. (2000) Enhanced learning in mice parallels vectormediated nerve growth factor expression in hippocampus. Hum. Gene Ther., 11: 2341-2352. Bujold, M., Caron, N., Camiran, G., Mukherjee, S., Allen, P.D., Tremblay, J.P and Wang, Y (2002) Autotransplantation in mdx mice of mdx myoblasts genetically corrected by an HSV-1 amplicon vector. Cell Transplant., 11: 759-767.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

ACKNOWLEDGMENTS Camicero, E., Knipper, M., Tan, J., Alonso, M.T. and Schimmang, T. (2002) Herpes simplex virus iype 1-mediated transfer of neurotrophin-3 stimulates survival of chicken auditory sensory neurons. Neurosci. Lett., 321:149-152. Chiocca, E.A. (2002) Oncolytic viruses. Nature, 2: 938-950. Connolly S.A., Landsburg, D.J., Carfi, A., Wiley D.C., Eisenberg, R.J. and Cohen, G.H. (2002) Structure-based analysis of the herpes simplex virus glycoprotein D binding site present on herpesvirus entry mediator HveA (HVEM). J. Virol., 76: 10894-10904. Connolly S.A., Landsburg, D.J., Carfi, A., Whitbeck, J.C, Zuo, Y, Wiley D.C., Cohen, G.H. and Eisenberg, R.J. (2005) Potential nectin-1 binding site on herpes simplex virus glycoprotein d. J. Virol, 79: 1282-1295. Cortes, M.L., Bakhenist-CJ., DiMaria, M.W., Kastan, M.B. and Breakefield, X.O. (2003) HSV-1 amplicon vector-mediated expression of ATM cDNA and correction of the ataxia-telangiectasia cellular phenotype. Gene Ther., 10:1321-1327. Costantini, L.C., Jacoby D.R., Wang, S., Fraefel, C , Breakefield, X.O. and Isacson, O. (1999) Gene transfer to the nigrostriatal system by hybrid herpes simplex virus/adeno-associated virus amplicon vectors. Hum. Gene Ther., 10: 2481-2494. Dalrymple, M.A., McGeoch, D.J., Davison, A.J. and Preston, C M . (1985) DNA sequence of the herpes simplex virus type 1 gene whose product is responsible for transcriptional activation of immediate early promoters. Nucl. Acids Res., 13: 78657879. Delman, K.A., Zager, J.S., Bennett, J.J., Malhotra, S., Ebright, M.I., McAuliffe, PR, Halterman, M.W., Federoff, H.J. and Fong, Y. (2002) Efficacy of multiagent herpes simplex virus ampliconmediated immimotherapy as adjuvant treatment for experimental hepatic cancer. Ann. Surg., 236: 337-342. Derby, M.L., Sena-Esteves, M., Breakefield, X.O. and Corey, D.P (1999) Gene transfer of lacZ into the mammalian and amphibian inner ears: a comparison study of three viral vectors. Hearing Res., 134:1-8. Desai, P. and Person, S. (1998) Incorporation of the green fluorescent protein into the herpes simplex virus type 1 capsid. J. Virol, 72: 7563-7568. Dilber, M.S., Phelan, A., Aints, A., Mohamed, A.J., Elliott, G., Edvard Smith, C.I. and O'Hare, P (1999) Intercellular delivery of thymidine kinase prodrug activating enzyme by the herpes simplex virus protein, VP22. Gene Ther., 6:12-21. Elliott, G. and O'Hare, P. (1997) Intercellular trafficking and protein delivery by a herpesvirus structural protein. Cell, 88: 223-233. Elliott, G. and O'Hare, P (1999a) Intercellular trafficking of VP22GFP fusion proteins. Gene Ther., 6:149-151. Elliott, G. and O'Hare, P. (1999b) Live-cell analysis of a green fluorescent protein-tagged herpes simplex virus infection. J. Virol, 73: 4110-4119. Epstein, A.L. and Manservigi R. (2004) Herpesvirus/retrovirus chimeric vectors. Curr. Gene Ther., 4: 409-416. Fotaki, M.E., Pink, J.R. and Mous, J. (1997) Tetracycline-responsive gene expression in mouse brain after amplicon-mediated gene transfer. Gene Ther., 4: 901-908. Fraefel, C , Jacoby D.R., Lage, C , Hilderbrand, H., Chou, J.Y, Alt, RW., Breakefield, X.O. and Majzoub, J.A. (1997) Gene transfer into hepatocytes mediated by helper virus-free HSV/AAV hybrid vectors. Mol Med., 3: 813-825. Fraefel, C , Mendes-Madeira, A., Mabon, O., Lefebvre, A., Le Meur, G., Ackermann, M., MouUier, P. and Rolling, R (2005) In vivo gene transfer to the rat retina using herpes simplex virus type 1 (HSV-l)-based amplicon vectors. Gene Ther., 12: 1283-1288.

39

Fraefel C , Song, S., Lim, R, Lang, P, Yu, L., Wang, Y, Wild, P and Geller, A. (1996) Helper virus-free transfer of herpes simplex virus type 1 plasmid vectors into neural cells. J. Virol, 70: 7190-7197. Frampton, A.R.J., Goins, WR, Nakano, K., Burton, E.A. and Glorioso, J.C. (2005) HSV trafficking and development of gene therapy vectors with applications in the nervous system. Gene Ther., 12: 891-901. Freundlieb, S., Schirra-Muller, C. and Bujard, H. (1999) A tetracycline controlled activation/repression system with increased potential for gene transfer into mammalian cells. J. Gene Med., 1: 4-12. Fruh, K., Ahn, K., Djaballah, H., Sempe, P, can Endert, PM., Tampe, R., Peterson, PA. and Yang, Y (1995) A viral inhibitor of peptide transporters for antigen presentation. Nature, 375: 415-418. Geller, A.I., Keyomarski, K., Bryan, J. and Pardee, A.B. (1990) An efficient deletion mutant packaging system for defective HSV-1 vectors; potential applications to neuronal physiology and human gene therapy Proc. Nati. Acad. Sci. USA, 87: 8950-8954. Grandi, P , Spear, M., Breakefield, X.O. and Wang, S. (2004b) Targeting HSV amplicon vectors. Methods, 33:179-186. Grandi, P., Wang, S., Schuback, D., Krasnykh, V, Spear, M., Curiel D., Manservigi, R. and Breakefield, X.O. (2004a) HSV-1 virions engineered for specific binding to cell surface receptor. M o l Ther., 9: 419-427. Grunewald, K., Desai, P, Winkler, D.C., Heymann, J.B., Belnap, D.M., Baumeister, W and Steven, A. C. (2003) Three-dimensional structure of herpes simplex virus from cryo-electron tomography. Science, 302: 1396-1398. Hammerschmidt, W. and Sugden, B. (2004) Epstein-Barr virus sustains Burkitt's lymphomas and Hodgkin's disease. Trends Mol. Med., 10: 331-336. H a m p l J.A., Camp, S.M., Mydlarz, W.K., H a m p l M., Ichikawa, T., Chiocca, E.A., Louis, D.N., Sena-Esteves, M. and Breakefield, X.O. (2003) Potentiated gene delivery to tumor using herpes simplex virus/Epstein Barr/RV tribrid amplicon vectors. Hum. Gene Ther., 14: 611-626. Harvey B.K., Chang, C.R, Chiang, Y.H., Bowers, W.J., Morales, M., Hoffer, B.J., Wang, Y and Federoff, H.J. (2003) HSV amplicon delivery of glial cell line-derived neurotrophic factor is neuroprotective against ischemic injury. Exp. Neurol, 183: 47-55. Heilbronn, R., Burkel, A., Stephan, S. and zur Hansen, H. (1990) The adeno-associated virus rep gene suppresses herpes simplex virus-induced DNA amplification. J. Virol, 64: 3012-3018. Heister, T., Held, l , Ackermann, M. and Fraefel C. (2002) Herpes simplex virus type 1/adeno-associated virus hybrid vectors mediate site-specific integration at the adeno-associated virus preintegration site, AAVSl, on human chromosome 19. J. Virol, 76: 7163-7173. Herrlinger, U., Jacobs, A., Quinones, A., Woiciechowsky, C , SenaEsteves, M., Rainov, N.G., Fraefel C. and Breakefield, X.O. (2000) Helper virus-free herpes simplex virus type 1 amplicon vectors for granulocyte-macrophage colony-stimulating factor-enhanced vaccination therapy for experimental glioma. Hum. Gene Ther., 11:1429-1438. Ho, D.Y, McLaughlin, J.R. and Sapolsky R.M. (1996) Inducible gene expression from defective herpes simplex virus vectors using the tetracycline responsive promoter system. Brain Res. Mol. Brain Res., 41: 200-209. Ho, D.Y, Mocarski, E.S. and Sapolsky, R.M. (1993) Altering central nervous system physiology with a defective herpes siniplex virus vector expressing the glucose transporter gene. Proc. Natl. Acad. Sci. USA, 90: 3655-3659.

I GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

40

3. HSV AMPLICON VECTORS FOR GENE DELIVERY TO THE NERVOUS SYSTEM

Ho, I.A., Hui, K.M. and Lam, RY. (2004) Glioma-specific and cell cycle-regulated herpes simplex virus type 1 amplicon viral vector. Hum. Gene Then, 15: 495-508. Homa, RL. and Brown, J.C. (1997) Capsid assembly and DNA packaging in herpes simplex virus. Rev. Med. Virol., 7: 107-122. Horsburgh, B.C., Hubinette, M.M. and Tufaro, R (1999) Genetic manipulation of herpes simplex virus using bacterial artificial chromosomes. Methods RnzymoL, 306: 337-352. Hoshi, M., Harada, A., Kawase, T., Uyemura, K. and Yazaki, T. (2000) Antitumoral effects of defective herpes simplex virusmediated transfer of tissue inhibitor of metaUoroteinases-2 gene in malignant glioma U87 in vitro: consequences for anti-cancer gene therapy. Cancer Gene Ther., 7: 799-805. Howarth, M., Takao, K., Hayashi, Y. and Ting, A.Y. (2005) Targeting quantum dots to surface proteins in living cells with biotin ligase. Proc. Natl. Acad. Sci. USA, 102: 7583-7588. Huang, C.Y. and Brandt, C.R. (2003) Rapid in vivo isolation of gene expression elements using and HSV amplicon system. J. Virol. Meth. 113: 1-12. Ichikawa, T., Hogemann, D., Saeki, Y, Tyminski, E., Terada, K., Weissleder, R., Chiocca, E.A. and Basilion, J. R (2002) MRI of transgene expression: correlation of therapeutic gene expression. Neoplasia, 4: 523-530. Inoue, R., Moghaddam, K.A., Ranasinghe, M., Saeki, Y, Chiocca, E.A. and Wade-Martins, R. (2004) Infectious delivery of the 132 kb CDKN2A/CDKN2B genomic DNA region results in correctly spliced gene expression and growth suppression in glioma cells. Gene Ther., 11: 1195-1204. Izumi, K.M. and Stevens, J.G. (1990) Molecular and biological characterization of a herpes simplex virus type 1 (HSV-1) neuroinvasiveness gene. J. Exp. Med., 172: 487-496. Jacobs, A.H., Winkeler, A., Hartung, M., Slack, M., Dittmar, C , Kummer, C , Knoess, C , Galldiks, N., Vollmar, S., Wienhard, K. and Heiss, W.D. (2003) Improved herpes simplex virus type 1 amplicon vectors for proportional coexpression of positron emission tomography marker and therapeutic genes. Hum. Gene Ther., 14: 277-297. Jamagin, V^.R., Zager, J.S., Klimstra, D., Dehnan, K.A., Malhotra, S., Ebright, M., Little, S., DeRubertis, B., Stanziale, S.R, Hezel, M., Federoff, H. and Fong, Y (2003) Neoadjuvant treatment of hepatic malignancy: an oncolytic herpes simplex virus expressing IL-12 effectively treats the parent tumor and protects against recurrence-after resection. Cancer Gene Ther., 10: 215-223. Jin, B.K., Belloni, M., Conti, B., Federoff, H.J., Starr, R., Son, J.H., Baker, H. and Joh, T.H. (1996) Prolonged in vivo gene expression driven by a tyrosine hydroxylase promoter in a defective herpes simplex virus ampHcon vector. Hum. Gene Ther., 7: 2015-2024, Johnston, K.M., Jacoby, D., Pechan, P., Fraefel, C , Borghesani, P., Schuback, D., Dunn, R., Smith, R and Breakefield, X.O. (1997) HSV/AAV hybrid amplicon vectors extend transgene expression in human glioma cells. Hum. Gene Ther., 8: 359-370. Jonker, H.R., Wechselberger, R.W., Boelens, R., Folkers, G.E. and Kaptein, R. (2005) Structural properties of the promiscuous VP16 activation domain. Biochemistry, 44: 827-839. Kaneda, Y, Twai, K. and Uchida, T. (1989) Increased expression of DNA cointroduced with nuclear protein in adult rat liver. Science, 243: 375-378. KapUtt, M.G., Kwong, A.D., Kleopoulos, S.R, Mobbs, C.V, Rabkin, S.D. and Pfaff, W.D. (1994) Preproenkephalin promoter yield regionspecific and long-term expression in adult brain after in vivo gene transfer via a defective herpes simplex viral vector. Proc. Natl. Acad. Sci. USA, 91: 8979-8983.

Kwong, A.D. and Frenkel, N. (1985) The herpes simplex virus amplicon. IV. Efficient expression of a chimeric chicken ovalbumin gene amplified within defective virus genomes. Virology, 142: 421-425. Lam, P, Hui, K.M., Wang, Y, Allen, RA., Louis, D.N., Yuan, C.J. and Breakefield, X.O. (2002) Dynamics of transgene expression in human glioblastoma cells mediated by herpes simplex virus/ adeno-associated virus amplicon vectors. Hum. Gene Ther., 13: 2147-2159. Lamigeon, C , Prod'Hon, C , De Frias, V., Michoudet, C. and Jacquemont, B. (2003) Enhancement of neuronal protection from oxidative stress by glutamic acid decarboxylase delivery with a defective herpes simplex virus vector. Exp. Neurol., 184: 381-392. Laquerre, S., Anderson, D.B., Stolz, D.B. and Glorioso, J.C. (1998) Recombinant herpes simplex virus type 1 engineered for targeted binding to erythropoietin receptor-bearing cells. J. Virol., 72: 9683-9697. Laxman, B., Hall, D.E., Bhojaru, M.S., Hamstra, D.A., Chenevert, T.L., Ross, B.D. and RehemtuUa, A. (2002) Noninvasive realtime imaging of apoptosis. Proc. Natl. Acad. Sci. USA, 99: 16551-16555. Lim, R, Hartley, D., Starr, P, Lang, P, Song, S., Yu, L., Wang, Y and Geller, A.I. (1996) Generation of high-titer defective HSV-1 vectors using an IE2 deletion mutant and quantitative study of expression in cultured cortical cells. BioTechniques, 20: 458-469. Linden, R.M., Ward, P., Giraud, C , Winocour, E. and Berns, K.I. (1996) Site-specific integration by adeno-associated virus. Proc. Natl. Acad. Sci. USA, 93:11288-11294. Logvinoff, C. and Epstein, A.L. (2001) A novel approach for herpes simplex virus type 1 amplicon vector production, using the CreloxP recombination system to remove helper virus. Hum. Gene Ther., 12: 161-167. Lu, B. and Federoff, H.J. (1995) Herpes simplex virus type 1 amplicon vectors with glucocorticoid-inducible gene expression. Hum. Gene Ther., 6: 419-428. Maguire-Zeiss, K.A., Bowers, W.J. and Federoff, H.J. (2001) HSV vector-mediated gene delivery to the central nervous system. Curr. Opin. Mol. Ther., 3: 482-490. Manoj, S., Jogger, C.R., Myscofski, D., Yoon, M. and Spear, PG. (2004) Mutations in herpes simplex virus glycoprotein D that prevent cell entry via nectins and alter cell tropism. Proc. Natl. Acad. Sci. USA, 101:12414-12421. Markert, J.M., Medlock, M.D., Rabkin, S.D., Gillespie, G.Y, Todo, T., Hunter, W.D., Plamer, C.A., Feigenbaum, R, Tomatore, C , Tufaro, F. and Martuza, R.L. (2000) Conditionally replicating herpes simplex virus mutant, G207 for the treatment of malignant glioma: results of a phase I trial. Gene Ther., 7: 867-874. Marsh, D.R., Dekaban, G.A., Tan, W , Strathdee, C.A. and Weaver, L.C. (2000) Herpes simplex viral and amplicon vector-mediated gene transfer into glia and neurons in organotypic spinal cord and dorsal root ganglion cultures. Mol. Then, 1: 464-478. McKee, T., Grandi, P., Mok, W , Alexandrakis, G., Boucher, Y, Breakefield, X.O. and Jain, R.K. Collagen degradation augments oncolytic herpesvirus therapy of solid tumors by improving viral vector distribution, submitted. McKnight, S. and Tjian, R. (1986) Transcriptional selectivity of viral genes in mammalian cells. Cell, 46: 795-805. Messerli, S.M., Prabhakar, S., Tang, Y, Shah, K., Cortes, M.L., Murthy v., Weissleder, R., Breakefield, X.O. and Tung, C.-H. (2004) A novel method for imaging apoptosis using a caspase-1 nearinfrared fluorescent probe. Neoplasia, 6: 95-105.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

ACKNOWLEDGMENTS Mitchell, B.M., Bloom, D.C., Cohrs, RJ., Gilden, D.H. and Kennedy, RG. (2003) Herpes simplex virus-1 and varicella-zoster virus latency in ganglia. J. NeuroviroL, 9:194-204. Muller, L., Saydam, O., Saeki, Y, Held, I. and Fraefel, C. (2005) Gene transfer into hepatocytes mediated by herpes simplex virus-Epstein-Barr virus hybrid amplicons. J. Virol. Methods, 123: 65-72. Nakano, K., Asano, R., Tsumoto, K., Kwon, H., Goins, W.R, Kumagai, I., Cohen, J.B. and Glorioso, J.C. (2005) Herpes simplex virus targeting to the EGF receptor by a gD-specific soluble bridging molecule. Mol. Ther., 11: 617-626. Neill, J.C, Sarkisian, M.R., Wang, Y, Liu, Z., Yu, L., Tandon, R, Zhang, G., Holmes, G.L. and Geller, A.I. (2001) Enhanced auditory reversal learning by genetic activation of protein kinase C in small groups of rat hippocampal neurons. Brain Res. Mol. Brain Res., 93:127-136. Nony R, Tessier, J., Chadeuf, G., Ward, R, Giraud, A., Dugast, M., Linden, R.M., Moullier, P. and Salvetti, A. (2001) Novel cis-acting replication element in the adeno-associated virus type 2 genome is involved in amplification of integrated rep-cap sequences. J. Virol., 75: 9991-9994. Oehmig, A., Fraefel, C , Breakefield, X.O. and Ackermann, M. (2004) Herpes simplex virus type 1 amplicons and their hybrid virus partners, EBV, AAV, and retrovirus. Curr. Gene Ther., 4: 385-408. Olschowka, J.A., Bishop, G.A., Hurley, S.D., Mastrangelo, M.A. and Federoff, H.J. (2003) Helper-free HSV-1 amplicons elicit a markedly less robust innate immune response in the CNS. Mol. Ther., 7: 218-227. Olson, V.G., Zabetian, C.R, Bolanos, C.A., Edwards, S., Barrot, M., Eisch, A.J., Hughes, T., Self, D.W., Neve, R.L. and Nestler, E.J. (2005) Regulation of drug reward by cAMP response elementbinding protein: evidence for two functionally distinct subregions of the ventral tegmental area. J. Neurosci., 25: 5553-5562. Ozelius, L.J., Hewett, J., Page, C , Bressman, S., BCramer, R, Shalish, C , de Leon, D., Brin, M., Raymond, D., Corey, D.R, Fahn, S., Risch, N., Buckler, A., GuseUa, J.R and Breakefield, X.O. (1997) The early-onset torsion dystonia gene (DYTl) encodes an ATP-binding protein. Nat. Genet., 17: 40-48. Pechan, P., Herrlinger, U., Aghi, M., Jacobs, A. and Breakefield, X.O. (1999) Combined HSV-1 recombinant and amplicon piggyback vectors: replication-competent and defective forms, and therapeutic efficacy for experimental gliomas. J. Gene Med., 1: 176-185. Rersichetti, R, Trettel, R, Huang, C.C, Fraefel, C , Timmers, H.T., Gusella, J.R and MacDonald, M.E. (1999) Mutant huntingtin forms in vivo complexes with distinct context-dependent conformations of the polyglutamine segment. Neurobiol. Dis., 6: 364-375. Phelan, A., Elliott, G. and O'Hare, R (1998) Intercellular delivery of functional p53 by the herpesvirus protein VP22. Nat. Biotech., 16: 440-443. Rhilpott, N.J., Giraud-Wali, C , Dupuis, C , Gomos, J., Hamilton, H., Bems, K.I. and Falck-Pedersen, E. (2002b) Efficient integration of recombinant adeno-associated virus DNA vectors requires a p5-rep sequence in cis. J. Virol., 76: 5411-5421. Rhilpott, N.J., Gomos, H., Bems, K.I. and Flack-Pedersen, E. (2002a) A p5 integration efficiency element mediates Rep-dependent integration into AAVSl at chromosome 19. Proc. Natl. Acad. Sci. USA, 99:12381-12385. Pin, R.H., Reinblatt, M., Bowers, W.J., Federoff, H.J. and Fong, Y (2004) Herpes simplex virus amplicon delivery of a hypoxia-

41

inducible angiogenic inhibitor blocks capillary formation in hepatocellular carcinoma. J. Gastrointest. Surg., 8: 812-822. Praetorius, M., Knipper, M., Schick, B., Tan, J., Limberger, A., Camicero, E., Alonso, M.T. and Schimmang, T. (2002) A novel vestibular approach for gene transfer into the inner ear. Audiol. NeurootoL, 7: 324-334. Preston, C M . (2000) Repression of viral transcription during herpes simplex virus latency. J. Gen. Virol., 81:1-19. Qi, J., Shackelford, R., Manuszak, R., Cheng, D., Smith, M . C , Link, CJ. and Wang, S. (2004) Functional expression of ATM gene carried by HSV amplicon vector in vitro and in vivo. Gene Ther., 11: 25-33. Rainov, N.G., Sena-Esteves, M., Fraefel, C , Dobberstein, K.U., Chiocca, E.A. and Breakefield, X.O. (1998) A chimeric fusion protein of cytochrome CYR4B1 and green fluorescent protein for detection of pro-drug activating gene delivery and for gene therapy in malignant glioma. In: Walden, P. (Ed.), Gene Therapy of Cancer, Vol. 61. Plenum Press, New York, pp. 393-403. Rajcani, J., Andrea, V. and Ingeborg, R. (2004) Peculiarities of herpes simplex virus (HSV) transcription: an overview. Virus Genes, 28: 293-310. Recchia, A., Parks, R.J., Lamartina, S., Toniatti, C , Pieroni, L., Palombo, R, Ciliberto, G., Graham, F.L., Cortese, R., La Monica, N. and Colloca, S. (1999) Site-specific integration mediated by a hybrid adenovirus-adeno/associated virus. Proc. Natl. Acad. Sci. USA, 96: 2615-2620. Reisman, D., Yates, J. and Sugden, B. (1985) A putative origin of replication of plasmids derived from Epstein-Barr virus is composed of two cis-acting components. Mol. Biol. Cell, 5:1822-1832. Roizman, B. and Pellett, RE. (2001) The family herpesvirdai: a brief introduction (3rd ed.). In: Knipe D.M. and Howley P.M. (Eds.), Fields Virology, Vol. 2. Lippencott-Raven, Philadelphia, pp. 2231-2295. Rotman, G. and Shiloh, Y (1999) ATM: a mediator of multiple responses to genotoxic stress. Oncogene, 18: 6135-6144. Rueger, M.A., Winkeler, A., Miletic, H., Kaestle, C , Richter, R., Schneider, G., Hilker, R., Heneka, M.T., Ernestus, R.L, Hampl, J.A., Fraefel, C and Jacobs, A.H. (2005) Variability in infectivity of primary cell cultures of human brain tumors with HSV-1 amplicon vectors. Gene Ther., 12: 588-596. Rumpel, S., LeDoux, J., Zador, A. and Malinow, R. (2005) Postsynaptic receptor trafficking underlying a form of associative learning. Science, 308: 83-88. Saeki, Y, Fraefel, C , Ichikawa, T., Breakefield, X.O. and Chiocca, E.A. (2001) Improved helper virus-free packaging system for HSV amplicon vectors using an ICP27-deleted, oversized HSV-1 DNA in a bacterial artificial chromosome. Mol. Ther., 3: 591-601. Saeki, Y, Ichikawa, T., Saeki, A., Chiocca, E.A., Tobler, K., Ackermann, M., Breakefield, X.O. and Fraefel, C (1998) Herpes simplex virus type 1 DNA amplified as bacterial artificial chromosome in Escherichia coli: rescue of replication-competent virus progeny and packaging of amplicon vectors. Hum. Gene Ther., 9: 2787-2794. Saeki, Y, Matsumoto, N., Nakano, Y, Mori, M., Awai, K. and Kaneda, Y (1997) Development and characterization of cationic liposomes conjugated with HSJ (Sendai virus): reciprocal effect of cationic lipid for in vitro and in vivo gene transfer. Hum. Gene Ther., 8: 2133-2141. Sandler, V, Wang, S., Angelo, K., Lo, H., Breakefield, X.O. and Clapham, D.E. (2002) Modified herpes simplex virus delivery of enhanced GFP into the central nervous system. J. Neurosci. Methods, 121: 211-219.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

42

3. HSV AMPLICON VECTORS FOR GENE DELIVERY TO THE NERVOUS SYSTEM

Seijffers, R. and Woolf, C.J. (2004) Utilization of an HSV-based amplicon vector encoding the axonal marker hPLAP to follow neurite outgrowth in cultured DRG neurons. J. Neurosci. Methods, 132: 169-176. Sena-Esteves, M., Hampl, J.A., Camp, S.M. and Breakefield, X.O. (2002) Generation of stable retrovirus packaging cell lines after transduction with herpes simplex virus hybrid amplicon vectors. J. Gene Med., 4: 229-239. Sena-Esteves, M., Saeki, Y, Camp, S., Chiocca, E.A. and Breakefield, X.O. (1999) Single-step conversion of cells to retrovirus vector producers with herpes simplex virus-Epstein-Barr virus hybrid ampUcons. J. Virol., 73: 10426-10439. Severini, A., Morgan, A.R., Tovell, D.R. and Tyrrell, D.L. (1994) Study of the structure of replicative intermediates of HSV-1 DNA by pulsed-field gel electrophoresis. Virology, 200: 428-435. Shah, K., Bureau, E., Kim, D.E., Yang, K., Tang, Y, Weissleder, R. and Breakefield, X.O. (2004) Glioma therapy and real-time imaging of neural precursor cell migration and tumor regression. Ann. Neurol., 57: 34-41. Shah, K., Tang, Y, Breakefield, X.O. and Weissleder, R. (2003) Realtime imaging of TRAIL-induced apoptosis of glioma tumors in vivo. Oncogene, 22: 6865-6872. Shieh, M.T., Wudunn, D., Montgomery, R.I., Esko, J.D. and Spear, RG. (1992) CeU surface receptors for herpes simplex virus are heparan sulfate proteoglycans. J. Cell Biol., 116: 1273-1281. Smith, A.E. (1995) Viral vectors in gene therapy. Annu. Rev. Microbiol., 49: 807-838. Smith, R.L., Geller, A.I., Escudero, K.W. and Wilcox, C.L. (1995) Long-term expression in sensory neurons in tissue culture from herpes simplex virus type 1 (HSV-1) promoters in an HSV-1derived vector. J. Virol., 69: 4593-4599. Sodeik, B., Ebersold, M.W. and Helenius, A. (1997) Microtubulemediated transport of incoming herpes simplex virus 1 capsids to the nucleus. J. Cell Biol, 136:1007-1021. Song, S., Wang, Y, Bak, S.Y, Lang, R, UUrey, D., Neve, R.L., O'Malley K.L. and Geller, A.I. (1997) An HSV-1 vector containing the rat tyrosine hydroxylase promoter enhances both long-term and cell type-specific expression in the midbrain. J. Neurochem., 68: 1792-1803. Sosinsky G.E., Gaietta, G.M., Hand, G., Deerinck, T.J., Han, A., Mackey, M., Adams, S.R., Bouwer, J., Tsien, R. Y and EUisman, M.H. (2003) Tetracysteine genetic tags complexed with biarsenical ligands as a tool for investigating gap junction structure and dynamics. Cell Commun. Adhes., 10:181-186. Spaete, R. and Frenkel, N. (1982) The herpes virus amplicon: a new eucaryotic defective-virus cloning-amplifying vector. Cell, 30: 295-304. Spaete, R.R. and Frenkel, N. (1985) The herpes simplex virus amplicon: analyses of cis-acting repUcation functions. Proc. Natl. Acad. Sci. USA, 82: 694-698. Spear, PG. (2004) Herpes simplex virus: receptors and ligands for cell entry. Cell Microbiol., 6: 401-410. Spear, PG., Eisenberg, R.J. and Cohen, G.H. (2000) Three classes of cell surface receptors for alpha herpes virus entry. Virology, 275: 1-8. Stavropoulos, T.A. and Strathdee, C.A. (1998) An enhanced packaging system for helper-dependent herpes simplex virus vectors. J. Virol., 72: 7137-7143. Stevens, J.G. (1975) Latent herpes simplex virus and the nervous system. Curr. Top. Microbiol. Immunol., 70: 31-50. Strathdee, C.A. and McLeod, M.R. (2000) A modular set of helperdependent herpes simplex virus expression vectors. Mol. Ther., 1: 479-485.

Sun, M., Zhang, G.R., Kong, L., Holmes, C , Wang, X., Zhang, W, Goldstein, D.S. and Geller, A.I. (2003) Correction of a rat model of Parkinson's disease by coexpression of tyrosine hydroxylase and aromatic amino acid decarboxylase from a helper virus-free herpes simplex virus type 1 vector. Hum. Gene Ther., 14: 415-424. Tannous, B.A., Grirm, T., Weissleder, R. and Breakefield, X.O. Metabolic biotinylation of receptors in mammalian cells for tumor imaging in vivo, in preparation. Toda, M., Martuza, R.L. and Rabkin, S.D. (2000) Tumor growth inhibition by intratumoral inoculation of defective herpes simplex virus vectors expression granulocyte-macrophage colony-stimulating factor. Mol. Ther., 2: 324-329. Tsien, R.Y (2005) Building and breeding molecules to spy on cells and tumors. FEBS Lett., 579: 927-932. Turner, A., Bruun, B., Minson, T. and Browne, H. (1998) Glycoproteins gB, gD, and gHgL of herpes simplex virus type 1 are necessary and sufficient to mediate membrane fusion in a Cos cell transfection system. J. Virol., 72: 873-875. van Roessel, P. and Brand, A.H. (2002) Imaging into the future: visualizing gene expression and protein interactions with fluorescent proteins. Nat. Cell Biol., 4: E15-E20. Vicario, I. and Schimmang, T. (2003) Transfer of FGF-2 via HSV-1based amplicon vectors promotes efficient formation of neurons from embryonic stem cells. J. Neurosci. Methods, 123: 55-60. Wade-Martins, R., Saeki, Y and Chiocca, E.A. (2003) Infectious delivery of a 135-kb LDLR genomic locus leads to regulated complementation of low-density lipoprotein receptor deficiency in human cells. Mol. Ther., 7: 604-612. Wade-Martins, R., Smith, E.R., Tyminski, E., Chiocca, E.A. and Saeki, Y (2001) An infectious transfer and expression system for genomic DNA loci in human and mouse cells. Nat. Biotechnol., 19: 1067-1070. Wang, S., Petravicz, J. and Breakefield, X.O. (2003) Single HSVamplicon vector mediates drug-induced gene expression via dimerizer system. Mol. Ther., 7: 790-800. Wang, S., Qi, J., Smith, M.C. and Link, C.J. (2002a) Antitumor effects on human melanoma xenografts of an amplicon vector transducing the herpes thymidine kinase gene followed by ganciclovir. Cancer Gene Ther., 9: 1-8. Wang, S. and Vos, J.-M. (1996) A hybrid herpes virus infectious vector based on Epstein-Barr virus and herpes simplex virus type 1 for gene transfer into human cells in vitro and in vivo. J. Virol., 70: 8422-8430. Wang, S.W., Mu, X., Bowers, WJ. and Klein, W H . (2002b) Retinal ganglion cell differentiation in cultured mouse retinal explants. Methods, 28: 448^56. Wang, X., Kong, L., Zhang, G.R., Sun, M. and Geller, A.I. (2004) A preproenkephalin-neurofilament chimeric promoter in a helper virus-free herpes simplex virus vector enhances long-term expression in the rat striatum. Neurobiol. Dis., 16: 596-603. Wang, Y, Mukherjee, S., Fraefel, C , Breakefield, X.O. and Allen, PD. (2002c) Herpes simplex virus type 1 amplicon vector-mediated gene transfer to muscle. Hum. Gene Ther., 13: 261-273. Wang, Y, Yu, L. and Geller, A.I. (1999) Diverse stabilities of expression in the rat brain from different cellular promoters in a helper virus-free herpes simplex virus type 1 vector system. Hum. Gene Ther., 10:1763-1771. Willis, R.A., Bowers, W.J., Turner, M.J., Fisher, T.L., Abdul-Alim, C.S., Howard, D.R, Federoff, H.J., Lord, E.M. and Frelinger, J.G. (2001) Dendritic cells transduced with HSV-1 amplicons expressing prostate-specific antigen generate antitumor immunity in mice. Hum. Gene Ther., 12: 1867-1879.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

ACKNOWLEDGMENTS

Wysocka, J. and Herr, W. (2003) The herpes simplex virus VP16induced complex: the makings of a regulatory switch. Trends Biochem. Sci., 28: 294-304. Xing, W., Baylink, D., Kesavan, C. and Mohan, S. (2004) HSV-1 amplicon-mediated transfer of 128-kb BMP-2 genomic locus stimulates osteoblast differentiation in vitro. Biochem. Biophys. Res. Commun., 319: 781-786. Yoon, M. and Spear, P.G. (2004) Random mutagenesis of the gene encoding a viral ligand for multiple cell entry receptors to obtain viral mutants altered for receptor usage. Proc. Natl. Acad. Sci. USA, 101: 17252-17257. Yoon, M., Zago, A., Shukla, D. and Spear, PG. (2003) Mutations in the N termini of herpes simplex virus type 1 and 2 gDs alter

43

functional interactions with the entry/fusion receptors HVEM, nectin-2, and 3-O-sulfated heparan sulfate but not with nectin-1. J. Virol., 77\ 9221-9231. Zager, J.S., Delman, K.A., Malhotra, S., Ebright, M.I., Bennett, J.J., Kates, T., Halterman, M., Federoff, H. and Fong, Y. (2001) Combination vascular delivery of herpes simplex oncolytic viruses and amplicon mediated cytokine gene transfer is effective therapy for experimental liver cancer. Mol. Med., 7: 561-568. Zhang, Y, Hong, Y, Bounhar, Y, Blacker, M., Roucou, X., Tounekti, O., Vereker, E., Bowers, W.J., Federoff, H.J., Goodyer, C.G. and LeBlanc, A. (2003) p75 neurotrophin receptor protects primary cultures of human neurons against extracellular amyloid beta peptide cytotoxicity. J. Neurosci., 23: 7385-7394.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

C H A P T E R

4 Influence of the Immune System on Central Nervous System Gene Transfer John R. Forsayeth

Abstract: The brain relies on the innate immune system to defend it against a wide variety of pathogens. The principle mechanism by which this system operates is activation of complement in response to the presence of pathogen-associated molecular patterns or apoptotic-cell-associated molecular patterns . This ancient mechanism monitors the presence of abnormal antigens associated with pathogenic organisms, but can also recognize damaged neurons. The principle reactive cells are glia (astrocytes and microglia). When activated, they release inflammatory cytokines and other reactive substances such as nitric oxide. Gene therapies are being developed for use in various pathological conditions, e.g. Parkinson's disease, in which this system is activated to some degree. Gene therapists need to take innate immunity into account when considering treatment of neurodegenerative conditions. However, the substantial absence of the adaptive immune system provides many advantages for treatment of brain disease, in contrast to significant immunological difficulties encountered in other tissues. Keywords: innate immune Parkinson's viral vector antibodies; microglia; astrocytes

that inhibits infiltration of armed (immunocompetent) lymphocytes, and promotes apoptosis of infiltrating lymphocytes and neutrophils through abundant expression of Tumor Necrosis Factor (TNF)-related death ligands (Gasque et al., 2000) and interleukin-6 (IL-6) (Van Wagoner et al., 1999). It is important at the outset to appreciate the rapid advance in our understanding of the dynamic nature of microglial and astrocytic function in the brain. The very name "glia" from the Greek word for glue suggests a static, structural role. We now know that astrocytes play critical roles in neurotransmission (Newman, 2003), synaptic potentiation and inhibition (Slezak and Pfrieger, 2003) and in the response of the brain to trauma and infection (Minagar et al., 2002). Microglia are often thought of as the macrophages of the brain. Although this is undoubtedly true, it is equally important to note that these cells can exist in at least three morphologically and functionally distinct states: (i) resting, (ii) activated, non-phagocytic found in areas

The delivery of genes into the Central Nervous System (CNS) takes place within the context of an organ that has learnt to survive somewhat independently of the world of adaptive immunity that protects the rest of the body. The brain relies on an ancient form of the immune system called innate immunity that does not rely upon priming of lymphocytes or the generation of antibodies in order to detect and eliminate infectious agents. The distinctly different world of brain immunity provides both advantages and challenges for gene therapy. The blood-brain barrier (BBB) keeps most cells of the adaptive immune system out of direct contact with neurons, astrocytes and microglia. Under certain conditions, this barrier can allow cells to cross, and in fact also possesses a selective transport mechanism for various biomolecules (Pardridge, 2002, 2003). There is a close physical and functional association of astrocytes with the BBB (Nedergaard et al., 2003). In addition to this physical barrier, there is an immunological barrier mediated by astrocytes and microglia

Gene Therapy of the Central Nervous System: From Bench to Bedside

45

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

46

4. INFLUENCE OF THE IMMUNE SYSTEM ON CNS GENE TRANSFER

of inflammation and (iii) reactive, phagocytic found in areas of trauma and infection. The second state (ii) is partly characterized by very significant production of nitric oxide in response to pathogens, lipopolysaccharides (LPS) and cytokines (Chao et a l , 1992,1993; Possel et al., 2000). It is this activity (Langston et al., 1999) that may be in part the source of nitrotyrosine found in protein aggregates in Parkinson's disease (PD) (Duda et al., 2000; Giasson et a l , 2000). These shifts between different states can be temporary in the case of some mild trauma or infection that resolves; however microglia can also remain activated for prolonged periods of time, particularly in the aged brain, and this should be borne in mind when gene therapy for neurodegenerative diseases is considered. The innate immune system relies on the ability of glial cells to recognize pathogen-associated molecular patterns (PAMPs) by means of toll-like receptors (TLRs) on the surface of these cells (Akira and Sato, 2003). PAMPs comprise invariant molecules associated with microbial pathogens such as LPS or peptidoglycan. In addition to PAMPs, it has been shown that brain cells (neurons, glia, endothelia) recognize apoptotic-cellassociated molecular patterns (ACAMPs) as a means by which target cells can be eliminated (Francis et a l , 2003). This recognition system feeds into activation of complement (C), a network of some 30 proteins organized into at least three major pathways (classical, alternative, lectin) that ultimately leads to C5b-9-mediated lysis of non-self cells or target organisms. The classical pathway, involving the CI complex, C4, C2 and C3, is activated primarily by antigen-antibody complexes, but can also be activated by non-immune molecules such as LPS, DNA, viral membranes and certain small polysaccharides, and this antibody-independent mechanism is usually dominant in the brain. The alternative C pathway involves C3 deposition on target cells where opsinization converts it first into C3b, and then by cleavage into the highly stable iC3b. Recognition of iC3b by the C receptors, CR3 (CDllb), ClqRp and CR4, on microglia is the key event in the elimination of pathogens, toxic debris and apoptotic cells. Anti-CDllb staining is often used as a histochemical marker of microglial activation and, therefore, of neuroinflammation. Frequently, CDllb can be detected along needle tracks for some weeks after adeno-associated virus (AAV) vector infusion (Sanftner et al., 2004). This system can recognize changes in cell morphology like blebbing that indicates cellular damage. The lectin pathway is the primary means by which the innate immune system responds to microbial carbohydrates. The key (but probably not the sole) lectin is mannan-binding lectin (MBL) that

is activated by oligosaccharides found commonly in pathogens, although its presence has not been reported in brain. It is possible that other, still uncharacterized lectins take the place of MBL in the brain. It should be noted that the innate immune system also relies on the presence of C inhibitors that serve to limit inflammatory responses, protecting bystander cells while still responding briskly to pathogens. These inhibitors act either to inhibit the C5b-9 lytic complex or the C3-cleaving enzymes, and are either secreted (CI inhibitor, S protein) or are membrane-bound {Decay Accelerating Factor (DAF), Membrane Co-factor Protein (MCP)) (Morgan and Meri, 1994). The C system is, however, robust and widespread in the brain, and nearly all the cells of the CNS including neurons, astrocytes and microglia can synthesize C proteins. Activation of the system has also been documented in a number of pathological states such as experimentally induced stroke (Schafer et al., 2000), and in Alzheimer's and Huntington's disease (Casque et al., 2000). Indeed, it has been shown that the toxic, fibrillar form of AjS (but not the non-toxic, diffuse form) binds to Clq resulting in C activation. In contrast, little upregulation of C inhibitors is observed in neurodegenerative disorders. In the review cited above. Casque et al. advance the concept that, at least in the early stages of neurodegenerative diseases, mild activation of the C pathway helps to remove cellular debris. But they also point out that C activation leads to anaphylotoxin-mediated activation of receptors (C3aR, C5aR) on glia and neurons that results in increased levels of pro-inflammatory cytokines, chemokines, adhesion molecules and C components that can greatly exacerbate pathology. One might argue that neurodegeneration in many cases advances more because of the absence of compensatory inhibition of C activation rather than of the upregulation of C. Obviously, in a neuropathological situation in which such a system is at least partially activated, the ability of viral vectors to influence C is potentially important. The essential problem, however, is that animal models in which vector safety and efficacy studies are undertaken may not adequately reflect innate immune status as it relates to a specific disease. The ability of some viruses to activate C in peripheral tissues has been documented. For example, hepatic administration of both recombinant adenovirus (rAd) (Cichon et al., 2001; Zaiss et al., 2002) and AAV (Zaiss et al., 2002) activated the innate immune system measured by induction of chemokines and infiltration of CDllb^ cells, the effect of adenovirus being both more prolonged and potent than that of AAV. These data suggest that this area should receive more attention insofar as brain gene therapy is concerned.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

47

ADAPTIVE IMMUNITY AND VIRAL VECTORS

L

ADAPTIVE IMMUNITY A N D VIRAL VECTORS •

A.

\

Adeno-Associated Virus

The simplest and most easily studied interaction between viral vectors and the immune system is that of pre-existing immunity against AAV. Circulating antibodies against AAV capsid are common in human populations, and are present at significant levels in about 80% of the population. An important consideration is the fact that antibodies are largely, though not completely, excluded from the brain parenchyma. High levels of pre-existing circulating antibodies directed at AAV2 inhibit transduction of striatal neurons in the rat (Sanftner et al., 2004; Fig. 1). This type of inhibition is not associated with any inflammation as determined by markers of glial activation (GFAP, CDllb), or T-lymphocyte infiltration (CD4, CDS). It is possible to induce such inflammation by infusion of certain research-grade formulations of AAV. Remarkably, however, even this does not appear to result in significant neuronal damage. Presumably, the contaminants that triggered chemotactic signals do not result in targeting of neurons, etc. Thus, the invading T-cells encounter no further instructions because there is no upregulation of MHC (Major Histocompatibility Complex) on neurons and no glial transduction by the AAV vector. There is also considerable evidence that AAV does not trigger a cell-mediated response partly because it does not transduce immature dendritic cells (DC), as does adenovirus (Jooss et al., 1998), and does not synthesize viral proteins, as do early-generation adenoviral vectors. The resistance of DCs to AAV infection limits the immune response to a B-cell-mediated humoral response. In humans, such a response is likely to be realized in terms of activation of a memory response dependent on activation of T-cells. In primate studies, Chirmule et al. (2000) showed that the activation of a memory response against intramuscular delivery of AAV2 in primates could be blocked by passive immunization with an anti-CD4 antibody. In the clinical situation, however, if low doses of vector are delivered under conditions that carefully restrict leakage of AAV out of the parenchyma, even such memory response should be quite mild. Generally, rats routinely show a much greater primary humoral response to intraparenchymal deli-very of AAV than do primates (Forsayeth and Bankiewicz, unpublished results). This difference appears to be driven by some absolute differences in tissue volume and surface area, together with the larger proportional surgical invasiveness in the rat compared with the primate brain. Hence,

m

^

FIGURE 1 Immunohistochemical detection of AADC after vector infusion. A representative, coronal brain section from each group through the AAV-hAADC (2.5 X 10^° vg/hemisphere) infusion site on day 42 shows that immunostaining is localized to the medium spiny neurons in the rat striatum. Animals immunized with AAVnull, 1 X 10^ vg (B) or 5 X 10^° vg (C) showed a decreased level of AADC expression in comparison to the control (A) that received only excipient. Group B had a mean neutralizing antibody titer of 350 ± 192 (n = 6), and group C a mean titer of 1208 ± 332 (n = 6). Note that the term Vg' is defined as the number of copies of viral genomes present in a suspension of AAV, and is usually determined by quantitative PCR. In our laboratory, we have started to use a more facile unit-based system where 1 Unit AAV = 1 X 10^ vg.

immune responses in the rat should be considered as worst case rather than typical of what one would expect in humans. Nevertheless, in all cases, the inhibitory effect of anti-AAV antibodies on transduction in the brain is very muted by comparison with their effect on

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

48

4. INFLUENCE OF THE IMMUNE SYSTEM ON CNS GENE TRANSFER

transduction in the liver. Remarkably, Scallan et al. (2004) found that dilutions of pooled, human immunoglobulin were highly effective at blocking liver transduction in SCID (Severe Combined Immune Deficiency) mice at unexpectedly low titers (<1:10). This result suggested that peripheral AAV gene therapy may be much more problematic than previously thought, but underscores the considerable insensitivity of AAV transduction in the brain to inhibition by anti-AAV antibodies. This problem is being addressed by exploring the use of novel or hybrid serotypes that may have reduced recognition by the immune system (Grimm and Kay, 2003). B.

Adenovirus Vectors

Adenovirus vectors have been used to transduce brain cells. Unlike AAV2, rAd vectors are not neuron-specific (Le Gal La Salle et al., 1993), and expression from first generation vectors lasts for some weeks at least (Davidson et al., 1993). All adenovirus vectors cause some degree of neuroinflammation that includes upregulation of MHC-1 on neurons (Thomas et al., 2000). The advent of "gutless" or high-capacity (HC-Ad) vectors that contain no viral open-reading frames has helped (Lowenstein et al., 2002; Sakhuja et al., 2003). However, the presence of contaminating helper virus, albeit low, and toxicity of capsid protein (Yang et al, 1996), are still of concern for CNS applications (Reddy et al., 2002). The interactions of these later generations of adenovirus with immune system components are still being evaluated (Tuettenberg et al., 2004), and much more work remains to be done in order to determine the ultimate utility of this vector system in the CNS. Whatever the clinical value of adenovirus in the long term, the use of rAd vectors has been very instructive in research into the immunology of gene therapy Unlike AAV, rAd vectors used to transduce mouse-lung tissue triggered an inflammatory response mediated by the induction of cytoxic CD8+ T-lymphocytes (CTL) (Yang et al., 1996). This inflammatory effect as well as B-lymphocyte response can be suppressed by treatment with cyclophosphamide (Jooss et al., 1996). This pronounced cell-mediated immunity is triggered by transduction of dendritic cells resident in the target tissue. Adoptive transfer of immature D C s transduced with rAd-LacZ into animals, with stable AAV-mediated j5-galactosidase expression in skeletal muscle, caused a robust cytotoxic elimination of LacZpositive muscle fibers (Jooss et al., 1998). When a similar experiment was done with AAV-LacZ, apart from the much poorer ability of AAV to transduce immature D C s , cell-mediated response was transferable to

wild-type C57BL6 mice but not to CD40-/- mice (Zhang et al., 2000). However, the response was directed against the LacZ transgene, not the AAV capsid. Thus, even in a situation biased toward dendritic cell priming with AAV capsid, no cell-mediated response against capsid could be detected. C.

Herpes Simplex Virus (HSV) Vectors

The natural lifecycle of HSV involves infection of sensory neurons, particularly trigeminal neurons. The virus maintains a latent state for the life of the host. Although wild-type virus may be reactivated from latency under the influence of a variety of stresses, completely replication-defective viruses retain the ability to establish persistent-quiescent genomes in neurons that are unable to reactivate in the nervous system subsequently. These persistent genomes are devoid of lytic gene expression, but retain the ability to express latency-associated transcripts (LATs). The LAT promoter can be used to drive prolonged transcription of transgenes in neurons (Fink and Glorioso, 1997). Although 40-90% of individuals have antibodies against HSV, depending on demography (Kamiyama et al., 2001), the avidity of the virus for neurons permits the use of extremely low multiplicities of infection (Wolfe et al., 2004). In addition, HSV is efficiently retrogradedly transported from nerve endings to soma, which makes HSV a remarkable tool for targeting peripheral neurons. When rats were immunized with a live attenuated virus sufficient to induce a robust neutralizing antibody response against a non-replicating reporter vector, intramuscular or subcutaneous injection of vector resulted in similar levels of anterior motor horn transduction indicated by expression of j8-galactosidase activity (Kamiyama et al., 2001). This apparent imperviousness to antibodies has encouraged the development of therapies for the treatment of chronic pain by repeated administration of an HSV vector that drives pro-enkephalin expression (Glorioso et al., 2003). Since peripheral administration represents the greatest possible exposure to antibodies, it bodes well for applications in which delivery to the brain itself is envisaged such as in PD (Fink et al., 1997; Burton et al., 2003). D,

Lentivirus Vectors

Lentiviruses are a type of retrovirus that can target post-mitotic cells such as neurons. The best known of the lentivirus vectors is based on HIV (Kordower et al., 1999). They are suited to the transduction of neurons, can carry quite large inserts of up to 10 kb, and do not

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

49

ADAPTIVE IMMUNITY AND VIRAL VECTORS

express any viral proteins (Zufferey et al., 1998). There appear to be iew if any known issues v\rith respect to this class of vector insofar as immunity is concerned. Perhaps the most significant issue w^ith these vectors

Group I

Group 2

has been their obligatory integration into the host genome. However, with the development of self-inactivating viruses, this is less likely to be a significant issue in the future (Blomer et a l , 1997; Miyoshi et al., 1998).

Group 3

Positive Control

FIGURE 2 Immunostaining for CD4, CDS, GFAP and CDllb. Representative sections from each group were stained with antibodies against CD4, CDS, C D l l b (microglial activation) and GFAP (astrocytic activation). High magnification fluorescent microscopy of sections from the right hemisphere from each group showed no detectable cytotoxicity, and no CD4- or CD8positive lymphocytes. The positive control animals showed infiltration of both CD4- and CDS-positive T-lymphocytes within the striatum. Similar images of sections from the right hemisphere showed no elevation in astrocytic or microglial activation, and no difference among groups 1, 2 and 3. Data from positive control animals showed robust microglial and astrocyte activation responses. Note the mild reactive gliosis along the needle track. Group 1 = control animals. Group 2 = low-antibody titer. Group 3 = high-antibody titer.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

50

4. INFLUENCE OF THE IMMUNE SYSTEM ON CNS GENE TRANSFER

IL

IMMUNOLOGY OF TRANSGENES

This aspect of gene therapy is generally considered to be an obvious issue that is dealt with early in the life of a project. There are at least two ways in which immunity might come into play with respect to even native human transgenes. The first relates to the breaking of tolerance. Our immune systems generally recognize human proteins as ''self." We do not mount a response against human proteins except under a few circumstances. The most common immune response against a transgene might be expected in genetic disorders in which there is little or no expression of the native protein. For gene therapists, the most easily recognizable is in hemophilia where the development of inhibitory antibodies occurs in approximately 30% of factor VIII hemophiliacs receiving recombinant or plasma-derived factor (Kaplan et a l , 2000). The development of such inhibitors is far more common in individuals who have no endogenous factor. Clearly, this is a potential issue in individuals with genetic neurological diseases such as Canavan disease (Surendran et al., 2003). Of course, the development of antibodies against a recombinant protein, may have limited consequences. Injection of recombinant human acid sphingomyelinase (ASM) into a mouse model of Niemann-Pick disease corrected the visceral disease, but also generated anti-ASM antibodies. Interestingly, none of the antibodies were inhibitors of ASM activity (Miranda et al., 2000). In diseases where point mutations cause biological inactivity but there is still significant protein produced, one might expect that gene therapy would be successful. In the absence of any endogenous protein, however, this could be a problem. But in the brain, the advantages of being sequestered from the adaptive immune system come to the fore. The innate immune system is not likely to respond to a transgene product that is not per se a modulator of that system. However, because gene therapy for such diseases typically involve attempts to obtain the widest possible expression in the brain (Janson et al., 2002), some contact with the adaptive immune system may be unavoidable. The consequences of such escape, particularly of viral vectors, will depend on a number of factors such as immunogenicity of the transgene product, whether brain-specific promoters are used. Such studies should be part of any pre-clinical program in which there is a reasonable possibility of lack of tolerance in the target population. A second issue concerns the breaking of tolerance due to prolonged over-expression of a therapeutic

protein. In order to break tolerance, at least two conditions must be satisfied: the presence of mature antigen-presenting cells (APC) and their interaction with T-cells. In the brain, the APC would be microglia, and they can be matured by the presence of PAMPs (see above) that activate Complement. The absence of T-cells in the parenchyma makes this a moot point for the most part. Under situations in which microglia are stimulated to become professional APC and come into contact with T-cells, the possibility of breaking tolerance exists. Microglia can be induced to undergo such maturation by activation of toll-like receptors, and this could occur through the presence of contaminants in vector preparations (Fig. 2). Such effects clearly result in the infiltration of CD4 and CDS lymphocytes into the parenchyma. In clinical grade (CMP) vector material, the presence of contaminating material is highly unlikely. But once again it should be noted that neurological diseases frequently present an environment in which the innate immune system is partially activated. IIL

CONCLUSION

Immunologically, the brain presents an inviting target for gene therapy. The innate immune system has a limited capacity to respond to the presence of viral vectors, and pre-existing anti-viral antibodies are largely screened out of the brain. Early generation adenoviral vectors presented a significant problem in terms of cell-mediated immunity, but these have been greatly ameliorated by the development of gutless vectors. However, there are still technical issues related to viral production that need addressing, unacceptably high levels of contaminating DNA in AAV preparations, for example, or low-level contamination of adenoviral vectors with helper virus. Other vector systems such as HSV and Lentivirus appear to be even less visible to the immune system, and are not likely to be a problem. Immunologically, HSV seems particularly suited to peripheral nervous system expression in both dorsal and ventral spinal ganglia. It should be borne in mind, however, that gene therapies are not developed for healthy people but for those with serious, chronic diseases. In this setting, activation of the innate immune system in the brain is of concern. The ability of viral transduction, either through viral proteins or through transgene expression, to influence complement activation, cytokine synthesis and glial activity, over quite long periods of time should not be ignored. So far, there has not been sufficient attention

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

ACKNOWLEDGMENTS

paid to the influence of brain gene therapy on innate immune function in the diseased brain. ACKNOWLEDGMENTS I thank Janet Cunningham, Laura Sanftner and Brian Suzuki for their work on some of these issues at Avigen Inc. I also thank Ciaran Scallan for helpful discussions concerning the passive immunity experiments. The advice and encouragement of members of the Bankiewicz laboratory is also gratefully acknowledged. References Akira, S. and Sato, S. (2003) Toll-like receptors and their signaling mechanisms. Scand. J. Infect. Dis., 35: 555-562. Blomer, U., Naldini, L. et al. (1997) Highly efficient and sustained gene transfer in adult neurons with a lentivirus vector. J. Virol., 71: 6641-6649. Burton, E.A., Glorioso, J.C. et al. (2003) Gene therapy progress and prospects: Parkinson's disease. Gene Ther., 10: 1721-1727. Chao, C.C., Hu, S. et al. (1992) Activated microglia mediate neuronal cell injury via a nitric oxide mechanism. J. Immunol., 149: 2736-2741. Chao, C.C., Anderson, W.R. et al. (1993) Activated microglia inhibit multiplication of Toxoplasma gondii via a nitric oxide mechanism. Clin. Immunol. Immunopathol., 67:178-183. Chirmule, N., Xiao, W. et al. (2000) Humoral immunity to adenoassociated virus t5q?e 2 vectors following administration to murine and nonhuman primate muscle. J. Virol., 74: 2420-2425. Cichon, G., Boeckh-Herwig, S. et al. (2001) Complement activation by recombinant adenoviruses. Gene Ther., 8:1794-1800. Davidson, B.L., Allen, E.D. et al. (1993) A model system for in vivo gene transfer into the central nervous system using an adenoviral vector. Nat. Genet., 3: 219-223. Duda, J.E., Giasson, B.I. et al. (2000) Widespread nitration of pathological inclusions in neurodegenerative synucleinopathies. Am. J. Pathol., 157:1439-1445. Fink, D.J. and Glorioso, J.C. (1997) Engineering herpes simplex virus vectors for gene transfer to neurons. Nat. Med., 3: 357-359. Fink, D.J., Poliani, PL. et al. (1997) Development of an HSV-based vector for the treatment of Parkinson's disease. Exp. Neurol. 144: 103-121. Francis, K., Van Beek, J. et al. (2003) Innate immunity and brain inflammation: the key role of complement. Expert Rev. Mol. Med., 2003: 1-19. Gasque, P., Dean, YD. et al. (2000) Complement components of the irmate immune system in health and disease in the CNS. Immunopharmacology, 49:171-186. Giasson, B.I., Duda, J.E. et al. (2000) Oxidative damage linked to neurodegeneration by selective alpha-synuclein nitration in synucleinopathy lesions. Science, 290: 985-989. Glorioso, J.C, Mata, M. et al. (2003) Gene therapy for chronic pain. Curr. Opin. Mol. Ther., 5: 483-488. Grimm, D. and Kay, M.A. (2003) From virus evolution to vector revolution: use of naturally occurring serotypes of adeno-associated virus (AAV) as novel vectors for human gene therapy. Curr. Gene Ther., 3: 281-304.

51

Janson, C , McPhee, S. et al. (2002) Clinical protocol. Gene therapy of Canavan disease: AAV-2 vector for neurosurgical delivery of aspartoacylase gene (ASPA) to the human brain. Hum. Gene Ther., 13:1391-1412. Jooss, K., Yang, Y et al. (1996) Cyclophosphamide diminishes inflammation and prolongs transgene expression following delivery of adenoviral vectors to mouse liver and lung. Hum. Gene Ther., 7: 1555-1566. Jooss, K., Yang, Y et al. (1998) Transduction of dendritic cells by DNA viral vectors directs the immune response to transgene products in muscle fibers. J. Virol, 72: 4212-4223. Kamiyama, H., Kurimoto, M. et al. (2001) Effect of immunity on gene delivery into anterior horn motor neurons by live attenuated herpes simplex virus vector. Gene Ther., 8: 11801187. Kaplan, J., Genyea, C. et al. (2000) Factor VIII inhibitors. Potential for prevention of inhibitor formation by immune tolerance. Semin. Thromb. Hemost., 26:173-178. Kordower, J.H., Bloch, J. et al. (1999) Lentiviral gene transfer to the nonhuman primate brain. Exp. Neurol, 160:1-16. Langston, J.W., Fomo, L.S. et al. (1999) Evidence of active nerve cell degeneration in the substantia nigra of humans years after l-methyl-4-phenyl-l,2,3,6-tetrahydropyridine exposure. Ann. Neurol, 46: 598-605. Le Gal La Salle, G., Robert, J.J. et a l (1993) An adenovirus vector for gene transfer into neurons and glia in the brain. Science, 259: 988-990. Lowenstein, PR., Thomas, C.E. et al. (2002) High-capacity helperdependent, ''gutless'' adenoviral vectors for gene transfer into brain. Methods Enzymol, 346: 292-311. Minagar, A., Shapshak, P. et al. (2002) The role of macrophage/ microglia and astrocytes in the pathogenesis of three neurologic disorders: HIV-associated dementia, Alzheimer disease, and multiple sclerosis. J. Neurol. Scl, 202:13-23. Miranda, S.R., He, X. et al. (2000) Infusion of recombinant human acid sphingomyelinase into niemann-pick disease mice leads to visceral, but not neurological, correction of the pathophysiology. Faseb J., 14: 1988-1995. Miyoshi, H., Blomer, U. et al. (1998) Development of a self-inactivating lentivirus vector. J. Virol, 72: 8150-8157. Morgan, B.P. and Meri, S. (1994) Membrane proteins that protect against complement lysis. Springer Semin. Immunopathol, 15: 369-396. Nedergaard, M., Ransom, B. et al. (2003) New roles for astrocytes: redefining the functional architecture of the brain. Trends Neuroscl, 26: 523-530. Newman, E.A. (2003) New roles for astrocytes: regulation of S5maptic transmission. Trends Neurosci., 26: 536-542. Pardridge, W.M. (2002) Targeting neurotherapeutic agents through the blood-brain barrier. Arch. Neurol, 59: 35-40. Pardridge, W.M. (2003) Molecular biology of the blood-brain barrier. Methods Mol. Med., 89: 385-399. Possel, H., Noack, H. et al. (2000) Selective upregulation of inducible nitric oxide s)aithase (iNOS) by lipopolysaccharide (LPS) and cytokines in microglia: in vitro and in vivo studies. Glia, 32: 51-59. Reddy, PS., Sakhuja, K. et a l (2002) Sustained human factor VIII expression in hemophilia A mice following systemic delivery of a gutless adenoviral vector. Mol. Ther., 5: 63-73. Sakhuja, K., Reddy, PS. et al. (2003) Optimization of the generation and propagation of gutless adenoviral vectors. Hum. Gene. Ther., 14: 243-254.

I GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

52

4. INFLUENCE OF THE IMMUNE SYSTEM ON CNS GENE TRANSFER

Sanftner, L.M., Suzuki, B.M. et al. (2004) Striatal delivery of rAAVhAADC to rats with preexistii\g immunity to AAV. Mol. Ther., 9: 403-409. Scallan, CD., Liu, T., Patarroyo-White, S., Jiang, Li., Zhou, S., Pierce, G. and Couto, L.B. (2004) Evaluation of AAV neutralizing antibodies using a passive immunioty model in SCID mice. American Society for Gene Therapy Annual Meeting, Minneapolis, MN. Schafer, M.K., Schwaeble, W.J. et al. (2000) Complement Clq is dramatically up-regulated in brain microglia in response to transient global cerebral ischemia. J. Immunol., 164: 5446-5452. Slezak, M. and Pfrieger, F.W. (2003) New roles for astrocytes: regulation of CNS synaptogenesis. Trends Neurosci., 26: 531-535. Surendran, S., Matalon, K.M. et al. (2003) Molecular basis of Canavan's disease: from human to mouse. J. Child Neurol., 18: 604-610. Thomas, C.E., Schiedner, G. et al. (2000) Peripheral infection with adenovirus causes unexpected long-term brain inflammation in animals injected intracranially with first-generation, but not with high-capacity, adenovirus vectors: toward realistic long-term neurological gene therapy for chronic diseases. Proc. Natl. Acad. Sci. USA, 97: 7482-7487.

Tuettenberg, A., Jonuleit, LL. et al. (2004) Early adenoviral gene expression mediates immunosuppression by transduced dendritic cell (DC): implications for immunotherapy using genetically modified DC. J. Immunol., 172:1524-1530. Van Wagoner, N.J., Oh, J.W et al. (1999) Interleukin-6 (IL-6) production by astrocytes: autocrine regulation by IL-6 and the soluble IL-6 receptor. J. Neurosci., 19: 5236-5244. Wolfe, D., Niranjan, A. et al. (2004) Safety and biodistribution studies of an HSV multigene vector following intracranial delivery to non-human primates. Gene Ther., 11: 1675-1684. Yang, Y., Su, Q. et al. (1996) Role of viral antigens in destructive cellular immune responses to adenovirus vector-transduced cells in mouse lungs. J. Virol., 70: 7209-7212. Zaiss, A.K., Liu, Q. et al. (2002) Differential activation of irmate immune responses by adenovirus and adeno-associated virus vectors. J. Virol., 76: 4580-4590. Zhang, Y, Chirmule, N. et al. (2000) CD40 ligand-dependent activation of cytotoxic T lymphocytes by adeno-associated virus vectors in vivo: role of immature dendritic ceUs. J. Virol., 74: 8003-8010. Zufferey, R., Dull, T. et al. (1998) Self-inactivating lentivirus vector for safe and efficient in vivo gene delivery. J. Virol., 72: 9873-9880.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

CHAPTER

5 Targeted Induction of Endogenous Neural Stem and Progenitor Cells: A New Strategy for Gene Therapy of Neurological Disease Steven A. Goldman, Fraser]. Sim Abstract: Neural stem and progenitor cells are distributed in several discrete niches in the adult vertebrate brain. These cells may respond to injury or disease with a limited degree of compensatory neurogenesis, but the clinical importance of this response seems at best limited. Nonetheless, resident progenitor cells can be induced to generate substantial numbers of new neurons by stimulation with both delivered and virally expressed growth factors. Such directed induction may be particularly efficacious in the subcortical and striatal neurodegenerations such as Huntington's disease, in which dysfunctional medium spiny neurons may be replenished from progenitors lining the striatal ventricular wall. Indeed, our increasing understanding of the molecular control of neural progenitor cell mobilization provides a new operational basis for using induced neuronal replacement as a means of treating neurodegenerative disease. Similarly, we may hope to tap the large reservoir of glial progenitor cells in the adult brain parenchyma to replete lost oligodendrocytes in diseases of acquired demyelination. More broadly, using gene therapeutic strategies to target endogenous progenitor cells, both for directed phenotypic differentiation and for discrete metabolic correction, as in the leukodystrophies and lysosomal storage disorders, offers great promise for harnessing resident progenitor cells for therapeutic benefit. Keywords: neurogenesis; ventricular zone; stem cells; gene therapy; neurological disease L

genesis that include both vascular and ependymoglial components (Palmer et al., 2000; Louissaint et al., 2002; Song et al., 2002). In mammals, active neurogenic niches are located within the rostrolateral ventricular wall anteriorly (Lim et al., 2000; Alvarez-Buylla and Garcia-Verdugo, 2002), and the subgranular zone of the dentate gyrus posteriorly (Louissaint et al., 2002; Palmer, 2002), leading to ongoing neurogenesis in both the olfactory bulb (Lois and Alvarez-Buylla, 1993; Luskin, 1993) and hippocampus (Roy et al., 2000a; Seaberg and van der Kooy, 2002) respectively. In addition, a larger pool of glial progenitors also pervades both the ventricular zone and tissue parenchyma (Levine et al., 2001). Although initially construed as astroglial and oligodendrocytic precursors, progenitor cells of

NEURAL STEM AND PROGENITOR CELLS OF THE HUMAN BRAIN

Neural stem cells and more restricted neuronal and glial progenitor cells are dispersed widely throughout the adult vertebrate brain (for reviews, see Goldman, 1998; Gage, 2000; Alvarez-Buylla et al., 2001; Goldman, 2003). Multipotential neural stem cells continue to line the forebrain ventricles long after development (Morshead et al., 1994; Weiss et al., 1996), while committed neuronal progenitor cells also remain within the ventricular wall, overlying the neostriata of both birds and mammals. These neuronally restricted progenitors appear to reside primarily in discrete niches for neuro-

Gene Therapy of the Central Nervous System: From Bench to Bedside

53

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

54

5. TARGETED INDUCTION OF ENDOGENOUS NEURAL STEM AND PROGENITOR CELLS: A NEW STRATEGY FOR GENE THERAPY

the adult brain parenchyma also retain multilineage competence, and can become robustly neurogenic when cultured in isolation (Palmer et al., 1999; Roy et al., 1999). As such, these may best be viewed as multipotential progenitor cells, restricted to generate glia by virtue of the adult tissue environment (Kondo and Raff, 2000; Belachew et al., 2003; Nunes et al., 2003). Each of these progenitor cell phenotypes persists in adult humans as well (Kirschenbaum et al..

1994; Pincus et al., 1997,1998; Kukekov et al., 1999; Roy et a l , 1999, 2000a,b; Arsenijevic et al., 2001; Nunes et al., 2003; Sanai et al., 2004a; Goldman and Sim, 2005). Together, these different classes of progenitors, along with the ventricular zone neural stem cells from which they co-derive, constitute the major known categories of neural precursor cells in the adult human CNS (Goldman, 2003; Goldman and Sim, 2005) (Fig. 1).

astrocyte

oligodfifidlrocyte

FIGURE 1 Stem and progenitor cells of the adult human brain. This cartoon illustrates the basic categories of progenitor cells in the adult brain and their lineal relationships, as well as markers and combinations thereof that permit their enrichment. The human temporal lobe is schematized here; it includes periventricular neural stem cells (red), that generate at least three populations of potentially neurogenic transit-amplifying progenitors of both neuronal and glial lineages (yellow). These include the neuronal progenitor cells of the ventricular subependyma, those of the subgranular zone of the dentate gyrus, and the glial progenitor cells of the subcortical white matter. Each transit-amplifying pool may then give rise to differentiated progeny appropriate to their locations, including neurons (purple), oligodendrocytes (green) and astrocytes (blue). Antigenically defined subependymal astrocytes are also noted, of which the neural stem cell population may comprise a subpopulation; the relationship of stem cells to the larger pool of subependymal astrocytes remains unclear (Doetsch et al., 1999; Alvarez-Buylla et al., 2001; Garcia et al., 2004). A-D cell stage terminology derived from (Alvarez-Buylla et al., 2001; Alvarez-Buylla and Garcia-Verdugo, 2002). Markers defining each stage have been reviewed previously in Svendsen et al. (2001); figure adapted from Svendsen et al. (2001) and Goldman (2003).

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

NEURAL PROGENITOR CELLS ARE WIDELY DISTRIBUTED IN THE ADULT HUMAN BRAIN

II-

CELL GENESIS IN THE ADULT CNS SHARES COMMON THEMES WITH THAT IN OTHER SOLID ORGANS

The neuronal progenitor cells of the adult human brain may considered akin to "transit-amplifying cells/' which have now been described as such in a variety of solid tissues. As initially defined in the skin and GI mucosae, transit-amplifying cells comprise the phenotypically biased, still-mitotic progeny of uncommitted stem cells (Potten and Loeffler, 1990; Loeffler and Potten, 1997; Watt, 2001; Niemann and Watt, 2002). As stem cell progeny depart these localized regions of stem cell expansion, their daughters may commit to more restricted lineages, still mitotic but phenotypically delimited and ultimately subject to senescence, which comprise the transit-amplifying pools. By this definition, the neuronal progenitor cell of the forebrain ventricular subependyma was first proposed as a transit-amplifying cell type, on the basis of its neuronal bias during mitotic expansion, and its ability to replenish the stem cell pool under appropriate mitotic stimulation (Menezes et al., 1995; Garcia-Verdugo et al., 1998; Doetsch et al., 2002). Similarly, the neuronally committed progenitor cells of the subgranular zone of the dentate gyrus, which also continue to divide while migrating (Palmer et al., 1997), may also best be considered transit-amplifying CNS progenitors (Goldman, 2003). As noted, glial progenitor cells also reside within both the ventricular zone and tissue parenchyma. In humans, these cells are dispersed throughout the subcortical white matter (Scolding, 1998; Scolding et al., 1998; Roy et al., 1999), just as in other infraprimate mammals (Levine et al., 2001). Although initially construed as oligodendrocyte progenitors (Scolding et al., 1998, 1999; Roy et al., 1999), adult glial progenitor cells retain multilineage competence (Richards et al., 1992; Palmer et al., 1999), and can become robustly neurogenic under appropriate conditions. In the adult human, potentially neurogenic but nominally glial progenitor pools have now been identified in both the cortex (Arsenijevic et al., 2001) and subcortical white matter (Nunes et al., 2003). Upon removal to low-density, serum-free culture, in which the cells are effectively removed from both autocrine and paracrine influences, these parenchymal progenitors generate neurons as well as astrocytes and oligodendrocytes, and remain propagable for several months in vitro (Nunes et al., 2003). These data suggest that the glial progenitor of the adult human white matter is a multipotential neural progenitor cell, restricted to generate

55

glia by the adult tissue environment, and not because of any autonomous lineage commitment. As such, the glial progenitor cell, and in particular that of the adult white matter, may also be considered a transit-amplifying cell; it is able to divide and regenerate multipotential progenitors, and yet more typically generates variably restricted daughter cells, that though still mitotic have diminished self-renewal capacity (Nunes et al., 2003). III.

NEURAL PROGENITOR CELLS ARE WIDELY DISTRIBUTED IN THE ADULT HUMAN BRAIN

In adult humans, both neural stem cells and competent neuronal progenitors persist within the olfactory subependyma and the striatal wall (Kirschenbaum et al., 1994; Pincus et al., 1997, 1998), as well as within the hippocampus (Eriksson et al., 1998; Roy et al., 2000a). Importantly, whereas human subependymal progenitors are heavily biased to gliogenesis in vivo (Sanai et al., 2004b), their hippocampal coimterparts continue to generate neurons in vivo (Eriksson et al., 1998). Yet notwithstanding the persistence of these subependymal and hippocampal progenitors, the glial progenitor cell appears to comprise by far the most abundant progenitor phenotype of the human brain (Windrem et al., 2002, 2003; Nunes et al., 2003). These cells comprise an abundant reservoir of widely dispersed, cycling progenitors, that though restricted to glial phenotype in vivo, are multipotential and neurogenic. Adult glial progenitors now appear to comprise as many as 4% of all cells in the adult white matter, yielding remarkably high estimates for their absolute incidence (Roy et al., 1999; Scolding et al., 1999; Nimes et al, 2003). Though the incidence of analogous parenchymal progenitor cells in the adult human gray matter has not yet been evaluated, these cells are abundant in the gray matter of adult rodents (Palmer et al., 1999), and there is no reason to think that they are any less so in himians. Together, these different classes of progenitors constitute the major known categories of neural precursor cells in the adult human CNS (reviewed in (Goldman, 2003,2005a; Goldman and Sim, 2005). Their therapeutic significance lies in the inherent potential of e^ch to be targeted, whether genetically or pharmacologically, for mobilization and directed induction to therapeutically required phenotypes. As such, adult progenitor cells lie at the intersection of cell- and gene-based therapeutic strategies. Gene therapeutics may be used to target specific progenitor phenotypes, to mobilize them to expand

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

56

5. TARGETED INDUCTION OF ENDOGENOUS NEURAL STEM AND PROGENITOR CELLS: A NEW STRATEGY FOR GENE THERAPY

and differentiate in response to expressed growth and differentiation factors, and to transduce both resident progenitors and their daughters to express potentially therapeutic transgenes.

IV. PROGENITOR CELLS MAY BE USEFUL AS B O T H ENGRAFTABLE A N D INDUCIBLE SUBSTRATES FOR REPAIR Neural stem cells and lineage-restricted progenitor cells can each be viewed as vectors for cell replacement via transplantation, and as targets for endogenous induction in situ. The former represents the broad category of cell-based therapy of the nervous system, which has been extensively reviewed elsewhere (Chmielnicki and Goldman, 2002; Goldman et al., 2002; Windrem et al., 2003; Lindvall et a l , 2004; Goldman, 2005b). In contrast, the latter view encompasses the broad category of gene therapeutic approaches to progenitor induction and repair, which have not been extensively discussed to date. That lack of emphasis notwithstanding, the persistence of both neural stem cells and their neuronal progeny in the wall of the lateral ventricle suggests the potential utility of these cells in restoring lost neuronal populations (Goldman, 2004). In general terms, although cell delivery strategies have been evaluated for over two decades, both experimentally and clinically, little effective clinical translation has yet been reported. Yet only in the past few years have neuronal progenitors of relatively homogeneous phenotype become available, as derived from both tissue-derived and embryonic stem cell-derived neural progenitor cells. As the promoter-based isolation (Li et al., 1998; Wang et al., 1998; Roy et al., 1999, 2000a,b; Keyoung et al., 2001) of therapeutically desired phenot3q:>es become more refined, we may envisage the acquisition of progenitors for all major phenotypes of the human CNS. As a result, neurodegenerative diseases as diverse as Parkinson's, Huntington's, and Alzheimers's may prove feasible targets for cell-based therapy, as progenitors restricted to midbrain dopaminergic, striatal GABAergic, and forebrain cholinergic fate, respectively, become available. Indeed, as a therapeutic modality, the transplantation of neuronal progenitors would seem of greatest potential therapeutic efficacy for neurodegenerative diseases such as Parkinson's, which is largely attributable to the loss of a single neural phenotype concentrated in a single location, and the metabolic oligodendrocytic disorders, in which a relatively homogeneous phenotype is affected within a single brain compartment. On the other hand, diseases of

multiple phenotypes or multifocal pathology may be less appropriate for cell therapy, and rather more amenable to strategies intended to induce resident progenitors to replace injured or lost cell populations in situ, in response to both delivered and endogenous cues.

V. E N D O G E N O U S PROGENITOR CELLS ARE MOBILIZED BY INJURY A N D DISEASE The replacement of neostriatal neurons from resident progenitors was first identified in songbirds, in which neurons of the adult vocal control nucleus are seasonally replaced by a newly generated cohort (Goldman and Nottebohm, 1983; Nottebohm, 1985, 2002). However, neostriatal neurogenesis does not appear to occur in the normal adult mammalian brain, despite the production of new neurons along the striatal wall that migrate rostrally to join the olfactory stream (Lois and Alvarez-BuyUa, 1994; Lois et al., 1996). Yet neostriatal neuronal production has been reported in adult marmnals in response to injury (Lindvall et al., 2004), as have been limited instances of compensatory cortical neurogenesis (Magavi et al., 2000; Chen et al., 2004) (Fig. 2). Compensatory neuronal addition to the neostriatum in particular has been reported by several groups following experimental focal stroke (Arvidsson et al., 2002; Parent et al., 2002; Jin et al., 2003). Similarly, Nakafuku and co-workers described compensatory replacement of hippocampal pyramidal neurons - which, despite their apparent dissimilarity from striatal medium spiny cells, comprise another periventricular subcortical neuronal pool (Nakatomi et al., 2002). Other groups have recently reported apparent compensatory neurogenesis in the striatum of Huntington's disease patients (Curtis et al., 2003), and increased dentate neurogenesis in the hippocampus of Alzheimer's patients (Jin et al., 2004). Together, these instances of compensatory neurogenesis in the striatum and hippocampus, though quantitatively modest, suggest the potential for recruiting new neurons from endogenous progenitor cells as a therapeutic strategy in each of these subcortical structures.

VL RESIDENT PROGENITOR CELLS C A N BE MOBILIZED PHARMACOLOGICALLY A N D GENETICALLY Many studies in restorative neurology of the past decade have focused on the use of transplanted neurons

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

RESIDENT PROGENITOR CELLS CAN BE MOBILIZED PHARMACOLOGICALLY AND GENETICALLY

Lod ©f coffipsinsatcir^ neyfog^mesli^ 1. Striatal ney ronal iscrwitinent pa:^-$^#t ^^ i f i d In Hyfitinitofff patltnti ^ 2Xoftlcil ntwunal adcfffion after tarp^tii Intmc-srtlcal aWatton * ^ B^Nippocampal p^antWil fieurcmat idditiQn 4.0€Maie In A t e l m ^ f s patfents ^

57

~ll' 7h pyramidal,

^

fitwnl "" psicynof '# 1

A T

nigriiipi nwi^wt

tW^MUMPf^ -WtUPWI

'k V

mmm pmmii mmm^ i

Luci of iprfyced r»y ri>geriesis: S.Str»^I - ovtrwprmslop «f i P N F ^ f c ^ i n ptWifHf Individ r i t y m f t n t s i s * mnd sypprtsslcm of ilic3f«n#sis apd a^nal pipjictbn to f l ^ b u i psiHIdys ^ i , ' ^ » i c ^ l i a l k : i f ^ ^ p t i l n«un2^p«iisfol0wmg higN««l«^t SWF irrfuitofi in mmtmA rats ^' 7. ^ n f e t t r««ro§tf»sis incltic»d by IQF1 ir» I^pci|5iysectetfnl3^ rife ^^ SSite Isftrc^ontrik r»ptil«t inhlbitosl ^^ MIS InlilWon ¥la bNAME ^ W ^ . S C ? ^ ^ ,

FIGURE 2 Compensatory and induced neuronal recruitment to the adult brain. This schematic illustrates both the described loci of compensatory and experimenterinduced neurogenesis in the adult rat brain, with relevant references listed for each. Loci of experimental compensatory neurogenesis include the neostriatum and hippocampal pyramidal layer following stroke. In patients, compensatory neurogenesis has similarly been reported in response to neurodegeneration in both Huntington's disease and Alzheimer's disease, in the striatum and dentate gyrus, respectively. Loci of induced neurogenesis include the neostriatum and diencephalon in response to BDNF, with potentiation of the striatal response with concurrent BMP-suppression via noggin, and the dentate gyrus of the hippocampus, in response to IGFl and VEGF, as well as to NOS inhibition and serotinergic agonists.

and glia, and more recently of stem and progenitor cells, as therapeutic agents. Relatively less effort has been devoted to utilizing or mobilizing endogenous progenitor populations. Yet several major populations of accessible progenitor populations persist in the adult brain. These pools are individually accessible, and may be mobilized through a variety of both pharmacological and gene therapeutic strategies, that result in the mitotic expansion of resident stem and progenitor cells. A number of humoral growth factors have been identified as modulating the mitotic expan-

sion and differentiated fate of neural stem cells. EGF and FGF2, which each have mitogenic effects on neuronal progenitor cells of the adult subependyma, can both potentiate neuronal replacement in the presence of permissive signals for neuronal differentiation (Kuhn et al., 1997). TGFa, a membrane-bound EGF-like ligand, has also been shown to achieve a similar effect in the adult striatum, in which TGFa exposure in the setting of catecholaminergic cell injury was found sufficient to induce the heterotopic recruitment of new dopaminergic neurons to the striatum (Fallon et a l , 2000). Yet EGF

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

58

5. TARGETED INDUCTION OF ENDOGENOUS NEURAL STEM AND PROGENITOR CELLS: A NEW STRATEGY FOR GENE THERAPY

and FGF2 appear to act solely as mitogens in the adult ventricular subependyma. EGF stimulation led largely to gliogenesis (Craig et al., 1996) and FGF2 infusion increased neuronal recruitment to the olfactory bulb, but nowhere else; neurons generated under the sole influence of FGF2 did not depart their tj^ical migratory paths to enter any other subcortical structures along their migratory route (Kuhn et al., 1997). A number of other ligands for receptor tyrosine kinases have been found to drive mitotic expansion by neural stem and progenitor cells, including vascular endothelial growth factor and stem cell factor, through the VEGFR2 and c-kit receptors, respectively (Jin et al., 2002a,b). Similarly, the inhibition of nitric oxide, an agent that appears to tonically suppress progenitor turnover in the adult brain, increased neuronal production in the olfactory bulb and dentate gyrus (Packer et al., 2003), doing so in a BDNF-dependent manner (Cheng et al., 2003). Yet despite these many approaches to influencing progenitor cell mobilization and neurogenesis within the forebrain subependyma, none of these strategies has been shown to be associated with heterotopic neuronal addition in vivo, i.e., the recruitment of new neurons into otherwise nonneurogenic regions of the adult brain. VIL

I N D U C E D NEUROGENESIS IN THE ADULT NEOSTRIATUM

To achieve the addition of new neurons to the mature brain, we and others have focused on delivering the trkB ligand BDNF to adult progenitor cells. BDNF had previously been shown to stimulate the production and survival of new neurons from adult precursor cells in vitro (Ahmed et a l , 1995; Kirschenbaum and Goldman, 1995; Goldman, 1997; Goldman et al., 1997). On the basis of these studies in culture, Luskin and colleagues next demonstrated that BDNF given intraventricularly could potentiate the addition of new neurons to the adult olfactory bulb (Zigova et al., 1998). On this basis, we then used adenoviral gene therapy to deliver the gene encoding BDNF to the adult rat ventricular wall, in an effort to assess the ability of BDNF to promote neurogenesis in otherwise non-neurogenic regions of the forebrain. We found that a single injection into the forebrain ventricles of replication-incompetent adenoviral BDNF induced the production of new neurons from neural progenitor cells in the ventricular subependyma (Benraiss et al., 2001). Most of the new neurons migrated to the olfactory bulb, but a large number also invaded the neostriatum.

a region of the brain that does not typically recruit new neurons in the normal, uninjured brain. Importantly, the new neurons integrated largely as medium spiny neurons, precisely the phenotype typically lost in Huntington's disease. Moreover, once integrated into the existing striatal network, the newly generated cells survived independently of periventricular BDNF overexpression (Benraiss et al., 2001). Together, these observations suggested that AdBDNF-induced neurons might directly replace the very phenotype lost in the course of Huntington's. Of note, Luskin and colleagues similarly demonstrated BDNF-associated neuronal addition to the adult neostriatum (Pencea et al., 2001), using BDNF protein infusion rather than virus. Interestingly, the high dose and sustained protein infusion used in this case led to neuronal addition to other subcortical limbic and diencephalic structures as well, the significance of which remains to be explored. VIIL THE DEVELOPMENT OF GLIAL SUPPRESSIVE STRATEGIES E N H A N C I N G HETEROTOPIC NEUROGENESIS Most neural stem cells differentiate as glia unless otherwise challenged. Chmielnicki et al. asked whether BDNF-stimulated striatal neuronal addition might be enhanced by the concurrent suppression of glial differentiation. Since gliogenesis by neural stem cells appears to be mediated by the pro-gliogenic bone morphogenetic proteins (BMPs) (Gross et al., 1996; Zimmerman et al., 1996), Chmielnicki et al. attempted to suppress adult glial production by periventricular overexpression of a potent BMP inhibitor, noggin. Noggin is a developmental inhibitor of the BMPs, and it continues to be expressed in regions of ongoing neurogenesis in vivo (Lim et al., 2000). In this study, a noggin mutein was used in which the heparin-binding domain was deleted, so as to ensure that ependymally expressed noggin would permeate the ventricular wall to achieve dissemination throughout regions of subependymal cell genesis (Economides et al., 2000). While adenovirally delivered noggin (AdNoggin) alone did not trigger the production of new striatal neurons, AdNoggin co-injected with AdBDNF strongly potentiated the latter's induction of striatal neurogenesis (Chmielnicki and Goldman, 2002). Within a month after viral injection, animals so treated added >350 confocal-confirmed new neurons per mm^, three-fold the number of new neurons observed in rats given AdBDNF alone (Chmielnicki and Goldman, 2002; Chmielnicki et al., 2004). These data indicated that the viral overexpression of noggin indeed suppressed

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

INDUCED NEUROGENESIS AS A RESTORATIVE STRATEGY FOR THE HIPPOCAMPAL ATROPHIES

gliogenesis in the adult subependyma, thereby expanding the BDNF-responsive pool of potentially neurogenic SVZ progenitor cells. The concurrent inhibition of glial differentiation and promotion of neuronal differentiation, via the intraventricular delivery of adenovectors overexpressing noggin and BDNF, thus appears to be an effective means of inducing progenitors to add new^ neurons to the adult forebrain. IX. INDUCED NEUROGENESIS AS A THERAPEUTIC STRATEGY IN HUNTINGTON^S DISEASE To assess the feasibility of induced neurogenesis for treating neurodegenerative diseases, we asked if AdBDNF and AdNoggin could stimulate the addition of medium spiny neurons into the neostriata of R6/2 mice. These mice were generated to include a 150 CAG-repeat polyglutamine expansion in the first exon of the Huntington gene; they display a progressive and severe behavioral phenotype, associated with striatal atrophy (Mangiarini et al., 1996), and as such provide a robust model of Huntington's disease. In preliminary studies, Cho et al. noted that R6/2 mice treated with AdBDNF and AdNoggin indeed exhibit substantial striatal neuronal addition, and recruit new medium spiny neurons throughout the medial neostriatum (Cho et al., 2004). These newly induced medium spiny neurons extend fibers to their normal efferent targets in the globus pallidus, and include both enkephalinergic and Substance P-defined striopallidal projection neurons. On this basis, we may postulate that the induction of striatal neuronal addition in the Huntington mutant brain may permit the functional replacement of those neurons lost to disease. Indeed, since BDNF may be used to stimulate human precursor cells as well as those of rodents (Pincus et al., 1998; Roy et al, 2000b), it is a real possibility that the BDNF-mediated, nogginenhanced induction of striatal neuronal recruitment from endogenous progenitor cells might prove a viable treatment strategy for Huntington's disease, as well as for such other causes of striatal neuronal loss as striatonigral degeneration and lenticulostriate stroke. X.

PROGENITOR CELL TARGETING IN PARKINSON^S DISEASE

As in Huntington's disease, nigral neurons lost to Parkinson's disease might optimally be regenerated from a patient's own store of endogenous progenitors, rather

59

than delivered as an allograft. But in regions lacking contiguity to a source of ventricular zone progenitors, it remains unclear whether an inductive approach to nigral regeneration is feasible. The mesencephalic ventricular zone continues to harbor neural stem cells, and these are especially biased toward dopaminergic neurogenesis (Sawamoto et al., 2001). Indeed, when isolated and expanded in vitro, mesencephalic neural stem cells give rise to dopaminergic neurons in sufficient numbers and proportions that they may be used to restore dopaminergic innervation to the 6-OHDA-lesioned striatum (Sawamoto et al., 2001). Nonetheless, no means of inducing the endogenous mesencephalic stem cell pool to in situ neurogenesis has yet been identified, so that no credible strategy of induced neuronal recruitm^t comparable to that established in the striatimi has yet been defined. As an alternative but still highly speculative approach, several groups have begim to focus on inducing dopaminergic neurogenesis from resident progenitors within the parenchyma of the substantia nigra. Neural progenitor cells indeed persist within the nigra (Lie et al., 2002; Zhao et al., 2003), as they do throughput much of the adult brain, and they are able to generate neurons in vitro (Palmer et al., 1997; Kondo and Raff, 2000). However, whether these parenchymal progenitor cells may be induced to generate neurqns in vivo remains unknown, and whether they may be stimulated specifically to generate dopaminergic neurons, and dopaminergic neurons competent to extend axons to the striatum no less, remains problematic. XL INDUCED NEUROGENESIS AS A RESTORATIVE STRATEGY FOR THE HIPPOCAMPAL ATROPHIES The adult dentate gyrus exhibits persistent constitutive neurogenesis throughout life in animals, and appears to do so as well in humans (Altman and Das, 1965; Eriksson et al., 1998; Roy et al., 2000a). New neurons are added to the adult dentate from progenitors in the subgranular zone (SGZ) of the hippocampus, a layer which is developmentally contiguous with the most posterior reaches of the subependymal zone of the lateral ventricle (Gage et al, 1998). The SGZ progenitors appear committed to neuronal phenotype, although they may derive from less committed multipotential progenitors (Seaberg and van der Kooy, 2002). SGZ progenitors respond to FGF2, IGFl, and VEGF with mitotic expansion (Palmer et al., 1995; Aberg et al., 2000; Jin et al., 2002a), the efficacy of which may increase in the setting of antecedent factor depletion

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

60

5. TARGETED INDUCTION OF ENDOGENOUS NEURAL STEM AND PROGENITOR CELLS: A NEW STRATEGY FOR GENE THERAPY

or injury. VEGF in particular has been an agent of much recent interest, in that it may act upon neural progenitors and vascular endothelial cells in concert, cementing the interaction of these phenotypes, which are frequently found dividing in co-association (Louissaint et al., 2002; Palmer, 2002; Palmer et al., 2002). In this regard, adeno-associated viral overexpression of VEGF in the adult hippocampus has been associated with both enhanced dentate neurogenesis and improved cognitive performance (Cao et al., 2004). Sonic hedgehog, a morphogen more often implicated in phenotypic induction and regionalization of the developing nervous system, also appears to regulate both the proliferation and differentiation of adult neural progenitor cells (Rowitch et al., 1999), including those of the SGZ in vivo (Lai et al., 2003). Besides the peptide growth factors, other positive regulators of hippocampal neurogenesis include environmental enrichment, exercise, and serotonin agonists (Gould, 1999; Nilsson et al., 1999; van Praag et al., 1999; Brezun and Daszuta, 2000; Malberg et al., 2000), all of which have been associated with improved performance in a variety of mood and memory-dependent tasks. The very malleability of hippocampal neurogenesis argues that SGZ progenitor cells should be especially amenable to genetic and pharmacologic modulation. The modulation of SGZ neurogenesis may thus prove beneficial not only in the affective disorders, but also in the degenerative dementias associated with hippocampal atrophy. XII. PARENCHYMAL GLIAL PROGENITORS ARE ATTRACTIVE TARGETS FOR EXOGENOUS MOBILIZATION Glial progenitor cells (GPCs) are widespread throughout the adult brain, and are competent to differentiate as both oligodendrocytes and astrocytes after transplantation. Glial pathologies may therefore lend themselves to cell-based therapy even more readily than neuronal disorders, given the relative homogeneity and accessibility of the major glial phenotypes, oligodendrocytes, and astrocytes, and the abundance of their progenitors. Predictably then, they have been used for cell-based therapy of diseases of myelin (Archer et al., 1997; Duncan et al., 1997; Windrem et al., 2002, 2003, 2004). Glial progenitors have proven competent to engraft both adult targets of acquired demyelination (Windrem et al., 2002), and perinataUy, in disorders of myelin formation or maintenance, such as the congenital leukodystrophies (Duncan et al., 1997).

A, Endogenous Glial Progenitors as Targets for Induction Given their prevalence and distribution, glial progenitor cells present one of the more exciting targets for cell-directed gene therapy in the adult CNS. As a result, a number of investigators have attempted to induce oligodendrocyte production and myelinogenesis from resident GPCs, as a means of restoring structural and functional integrity to demyelinated foci in diseases of acquired demyelination, such as multiple sclerosis. Indeed, from the standpoint of structural repair, disease targets as diverse as the vascular leukoencephalopathies in adults, and cerebral palsy in children, may prove amenable to therapies based on glial progenitor cell induction. For instance, cerebral palsy with perventricular leukomalacia appears largely due to a perinatal loss of oligodendrocytes and their precursors (Back et al., 2001; Back and Rivkees, 2004; FoUett et al., 2004; Robinson et al., 2005), which may prove amenable to replenishment from local progenitor stores, appropriately mobilized. Yet though straightforward in concept, the targeted induction of myelinogenesis by resident glial progenitors has proven difficult. To be sure, a variety of agents, delivered both as protein growth factors and as competent expression vectors thereof, have been used to stimulate endogenous glial progenitors, but to variable and generally modest effect (reviewed in Levine et al., 2001). Indeed, even those studies that have reported induced remyelination or improved functional competence have not been able to causally attribute the effect of experimental treatment to progenitor cell mobilization, so much as to broader paracrine effects on the disease environment and immune response (Franklin et al., 2001). For example, NTS and BDNF-expressing fibroblasts were reported to potentiate oligodendrocytic production and myelination in the contused rat spinal cord (McTigue et al., 1998). However, whether such effects are due to progenitor cell mobilization and attendant myelinogenesis, or rather to BDNF and NT-3-associated support of host axons, increasing their own availability and paracrine support for myelination, has proven difficult to define. Similarly, IGFl was reported to reduce lesion incidence and improve compensatory remyelination in models of experimental allergic encephalomyelitis (EAE), attendant with an IGFl-dependent increase in the number of oligodendrocytes (Yao et al., 1995). Yet IGFl's effects in EAE include attenuating vasculitic damage to the blood-brain barrier, thus acting as an immune modulator (Liu et al., 1997, 1998). Systemic

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

TRANSDUCED NEURAL PROGENITORS AS VECTORS FOR ENZYMATIC REPLETION IN THE STORAGE DISEASES

delivery of the neuregulin glial grov^th factor-2 (GGF2) w^as likewise associated with improved remyelination and clinical benefit in EAE, but whether this was causally dependent upon progenitor mobilization and myelinogenesis was similarly unclear (Cannella et al., 1998). Indeed, long-term GGF2 infusion improved neither the incidence nor rate of remyelination following ethidium bromide-induced demyelination in aged rats (Penderis et al., 2003). As a result, the search continues for effective means of activating resident GPCs, particularly in regards to defining those ligand-receptor interactions that may signal their mobilization and myelinogenesis. B.

Gene Delivery to Parenchymal Progenitor Cells

Besides identifying effective ligands competent to specifically activate parenchymal glial progenitor cells, we also need to identify vectors able to deliver transgenes to these cells, whether stably or transiently. Although both oligodendrocytes and their progenitors can be readily transduced with adenoviral vectors, they are relatively sensitive to viralinduced cytotoxic effects; as such adenoviruses have to be carefully titrated (Franklin et al., 1999). In contrast, lentiviral vectors have been used to successfully infect adult human GPCs in vitro, and these cells proved non-immunogenic and stably transduced after transplantation to both perinatal and adult demyelinated hosts (Windrem et al., 2002; Nunes et al., 2003). In adddition, NG2-defined GPCs were transduced with lentiviral lacZ in vivo, with stable transgene expression and little associated demyelination (Zhao et al., 2003). These promising results are offset though by the often modest efficiency of transgene transcription and restricted dispersal of lentiviral vectors in the adult white matter. To address these latter concerns, a number of investigators have used adeno-associated viruses (AAVs) as non-immunogenic vectors for stably targeting oligodendrocytes and their progenitor cells (Chen et al., 1998,1999). In particular, by taking advantage of the exquisite phenotypic specificity of different AAV serotypes, we might hope to establish vectors selectively competent to infect parenchymal progenitor cells. For instance, the identification of PDGFaR and PDGFjSR as potential receptors for AAV5 led to the observation that AAVS can stably and efficiently infect glial progenitors (Di Pasquale et al., 2003), although the promiscuity of PDGFAR expression does not yet permit the level of cell-type specificity required for selective gene delivery to GPCs. More likely, some combination of progenitor-accessible AAV serotypes

61

and GPC-selective promoters, such as CNP2 (Gravel et al., 1998; Roy et al., 1999) and PLP (Mallon et al., 2002), will permit the selective delivery of therapeutic transgenes to parenchymal glial progenitor cells. XIII. T R A N S D U C E D N E U R A L PROGENITORS AS VECTORS FOR ENZYMATIC REPLETION I N THE STORAGE DISEASES Glial progenitor cells may have additional value besides structural repair and myelination, in that their widespread dispersal and efficient integration into recipient brain suggests their use as cellular delivery vehicles of wild-type or overexpressed gene products. This function may be of particular utility in the congenital metabolic diseases of the CNS, especially those due to enzyme dysfunction or depletion, such as the mucopolysaccaridoses, the gangliosidoses, and other lysosomal lipid storage disorders (Kaye, 2001; Powers, 2004). Indeed, the congenital leukodystrophies due to lysosomal storage disorders present especially attractive targets for using genetically modified progenitor cells as therapeutic vectors, since wild-type lysosomal enzymes may be released by integrated donor cells, and picked up by enzyme deficient host cells through the mannose-6-phosphate receptor pathway (Urayama et al., 2004). As a result, only a relatively small number of donor cells may be needed within a much larger volume of diseased host cells, to provide sufficient enzymatic activity to correct the underlying host catalytic deficit and storage disorder. That being said, the enzymatic activity of implanted wild-type cells may be insufficient to achieve regional correction, and the resultant shortfall in enzymatic activity may be only incompletely addressed by increased donor cell dosage. To address this problem, donor GPCs might be transduced to overexpress therapeutic transgenes, specifically those encoding enzymes deficient in the diseased host. Genomically integrating retroviruses, AAVs and lentiviruses have been developed that express genes implicated in the metabolic and hereditary leukodystrophies, and several have been assessed with regards to their ability to restore normal phenotype after intracerebral injection. Mucopolysaccharidosis VII (MPSVII) has been an especially fruitful experimental model in this regard, and feline immunodeficiency virus expressing jS-glucoronidase, the enzyme deficient in MPSVII, has been shown to

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

62

5. TARGETED INDUCTION OF ENDOGENOUS NEURAL STEM AND PROGENITOR CELLS: A NEW STRATEGY FOR GENE THERAPY

improve survival after intrastriatal viral injection (Brooks et al., 2002). However, the widespread nature of these diseases argues that the intracerebral injection of replication-incompetent viral vectors, whose effects are necessarily limited to their effective infection radius, may be insufficient to achieve the widespread and uniform degrees of enzymatic correction required throughout the neuraxis. As an alternative, systemic administration of hematopoietic stem cells (HSCs) transduced to overexpress arylsulfatase A has recently been reported as an approach to treatment in mouse models of metachromatic leukodystrophy (Biffi et al., 2004). This strategy depends upon the infiltration of the CNS and PNS with donor-derived microglia and endoneural macrophages, respectively, carrying the lentivirally delivered transgene. Yet despite the inherent promise of this approach, the penetration of peripheral macrophages into the adult CNS remains limited to perivascular structures, which may sharply limit the range of potential enzymatic deficiencies amenable to correction by transduced HSCs. In contrast, unlike HSCs, glial progenitors enjoy widespread dispersal in the CNS (Windrem et al., 2004). As such, implanted glial progenitors may be capable of achieving high donor: host cell ratios throughout the recipient brain parenchyma. Indeed, it is conceivable that for some enzymatic disorders, wild-type unmodified glial progenitors may be sufficient to restore enzymatic activity throughout the CNS of affected hosts. Alternatively, glial progenitors may be transduced to overexpress the deficient gene, for the purpose of engrafting the diseased host with a cell type able to both deliver its transduced gene products at high levels throughout the neuraxis, while meaningfully contributing to host cytoarchitecture.

XIV.

OVERVIEW

The use of viral expression vectors to mobilize resident neural stem and progenitor cells may prove an effective strategy for treating a wide variety of neurological disease, particularly the geographically and phenotypically restricted neurodegenerative diseases. In these disorders, the reconstruction of precise neural circuits may depend upon the development of new neurons in situ, within the local context in which they will ultimately reside, and from which they will need to both attract and extend site-specific afferents and efferents. As such, the mobilization of endogenous progenitor cells by gene therapeutic vectors, and the directed differentiation of their daughter cells into discrete

neuronal and glial phenotypes in situ, may prove an especially attractive strategy for eliciting CNS repair.

ACKNOWLEDGMENTS Work discussed in the Goldman lab is supported by NIH/NINDS, the National Multiple Sclerosis Society, the NY State Spinal Cord Research Program, the AtaxiaTelangiectasia Children's Project, The CNS Foundation, Merck Research Labs and Berlex Bioscience. References Aberg, M. et al. (2000) Peripheral infusion of IGF-1 selectively induces neurogenesis in the adult rat hippocampus. J. Neurosci., 20: 2896-2903. Ahmed, S., Reynolds, B.A. and Weiss, S. (1995) BDNF enhances the differentiation but not the survival of CNS stem cell- derived neuronal precursors. J. Neurosci., 15(8): 5765-5778. Altman, J. and Das, G.D. (1965) Autoradiographic and histological evidence of postnatal hippocampal neurogenesis in rats. J. Comp. Neurol., 124(3): 319-335. Alvarez-Buylla, A. and Garcia-Verdugo, J.M. (2002) Neurogenesis in adult subventricular zone. J. Neurosci., 22(3): 629-634. Alvarez-Buylla, A., Garcia-Verdugo, J.M. and Tramontin, A. (2001) A unified hypothesis on the lineage of neural stem cells. Nat. Rev. Neurosci., 2: 287-293. Archer, D. et al. (1997) Myelination of the canine central nervous sytem by glial cell transplantation: a model for repair of human myelin disease. Nat. Med., 3: 54-59. Arsenijevic, Y. et al. (2001) Isolation of multipotent neural precursors residing in the cortex of the adult human brain. Exp. Neurol., 170(1): 48-62. Arsenijevic, Y. et al. (2001) Isolation of multipotent neural precursors residing in the cortex of the adult human brain. Exp. Neurol., 170: 48-62. Arvidsson, A. et al. (2002) Neuronal replacement from endogenous precursors in the adult brain after stroke. Nat. Med., 8: 963-970. Back, S. and Rivkees, S. (2004) Emerging concepts in periventricular white matter injury. Semin. PerinatoL, 6: 405-414. Back, S.A. et al. (2001) Late oligodendrocyte progenitors coincide with the developmental window of vulnerability for human perinatal white matter injury. J. Neurosci., 21(4): 1302-1312. Belachew, S. et al. (2003) Postnatal NG2 proteoglycan-expressing progenitor cells are intrinsically multipotent and generate functional neurons. J. Cell Biol., 161(1): 169-186. Benraiss, A. et al. (2001) Adenoviral brain-derived neurotrophic factor induces both neostriatal and olfactory neuronal recruitment from endogenous progenitor cells in the adult forebrain. J. Neurosci., 21(17): 6718-6731. Biffi, A. et al. (2004) Correction of metachromatic leukodystrophy in the mouse model by transplantation of genetically modified hematopoietic stem cells. J. Clin. Invest., 113:1118-1129. Brezun, J. and Daszuta, A. (2000) Serotonin may stimulate granule cell proliferation in the adult hippocampus, as observed in rats grafted with foetal raphe neurons. Eur. J. Neurosci., 12: 391-396.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

ACKNOWLEDGMENTS Brooks, A. et al. (2002) Functional correction of established CNS deficits in an animal model of lysosomal storage disease with feline immunodeficiency virus-based vectors. Proc. Natl. Acad. Sci., 99: 6216-6221. Cannella, B. et al. (1998) The neuregulin, glial growth factor-2, diminishes autoimmune demyelination and enhances remyelination in a chronic relapsing model for multiple sclerosis. Proc. Natl. Acad. Sci., 95:10100-10105. Cao, L. et al. (2004) VEGF links hippocampal activity with neurogenesis, learning and memory. Nat. Genet., 36: 827-835. Chen, H., et al. (1998) Gene transfer and expression in oligodendrocytes under the control of myelin basic protein transcriptional control region mediated by adeno-associated virus. Gene Ther., 5: 50-58. Chen, H. et al. (1999) Oligodendrocyte-specific gene expression in mouse brain: Use of a myelin-forming cell type-specific promoter in an adeno-asociated virus. J. Neurosci. Res., 55: 504-513. Chen, J., Magavi, S. and Macklis, J. (2004) Neurogenesis of corticospinal motor neurons extending spinal projections in adult mice. Proc. Natl. Acad. Sci., 101: 16357-16362. Cheng, A. et al. (2003) Nitric oxide acts in a positive feedback loop with BDNF to regulate neural progenitor cell proliferation and differentiation in the mammalian brain. Dev. Biol., 258: 319-333. Chmielnicki, E. and Goldman, S.A. (2002) Induced neurogenesis by endogenous progenitor cells in the adult mammalian brain. Prog. Brain Res., 138: 451-464. Chmielnicki, E. et al. (2004) Adenovirally expressed noggin and brain-derived neurotrophic factor cooperate to induce new medium spiny neurons from resident progenitor cells in the adult striatal ventricular zone. J. Neurosci., 24(9): 2133-2142. Cho, S.-R., Chmielnicki, E. and Goldman, S.A. (2004) Adenoviral codelivery of BDNF and noggin induces striatal neuronal replacement and delays motor impairment in a transgenic model of Huntington's Disease. Mol. Ther., 9: S86-S87. Craig, C , Tropepe, V., Morshead, C , Re)molds, B., Weiss, S., Vander Kooy, D., (1996) In vivo growth factor expansion of endogenous subependymal neural precursor cell populations in the adult mouse brain. J. Neuroscience, 16: 2649-2658. Curtis, M.A. et al. (2003) Increased cell proliferation and neurogenesis in the adult human Huntington's disease brain. Proc. Natl. Acad. Sci. USA, 100(15): 9023-9027. Di Pasquale, G. Davidson, B., Stein, C , Martins, I., Sevdiero, D., Monks, A. and Chiorri, J. (2003) Identification of PDGFR as a receptor for AAV-5 transduction. Nature med., 9(10): 1306-1312. Doetsch, F. et al. (1999) Subventricular zone astrocytes are neural stem cells in the adult mammalian brain. Cell, 97(6): 703-716. Doetsch, F. et al. (2002) EGF converts transit-amplifying neurogenic precursors in the adult brain into multipotent stem cells. Neuron, 36: 1021-1034. Duncan, I.D., Grever, W.E. and Zhang, S.C. (1997) Repair of myelin disease: strategies and progress in animal models. Mol. Med. Today 3(12): 554-561. Economides, A., Stahl, N.E. and Harland, R.M. (2000) Modified noggin polypeptide and compositions. Regeneron Pharmaceuticals, Inc., Tarrytown, NY; Regents of the University of California, Oakland, CA, USA. Eriksson, PS. et al. (1998) Neurogenesis in the adult human hippocampus. Nat. Med., 4(11): 1313-1317. Fallon, J. et al. (2000) In vivo induction of massive proliferation, directed migration, and differentiation of neural cells in the adult mammalian brain. Proc. Natl. Acad. Sci. USA, 97(26): 14686-14691.

63

FoUett, P. et al. (2004) Glutamate receptor-mediated oligodendrocyte toxicity in periventricular leukomalacia: a protective role for topiramate. J. Neurosci. Res., 24: 4412-4420. Franklin, R.J. et al. (2001) What roles do growth factors play in CNS remyelination? Prog. Brain Res., 132: 185-193. Franklin, R.J.M., Quick, M.M. and Haase, G. (1999) Adenoviral vectors for in vivo gene delivery to oligodendrocytes: transgene expression and cytopathic consequences. Gene Ther., 6:1360-1367. Gage, R (2000) Mammalian neural stem cells. Science, 287:1433-1438. Gage, F.H. et al. (1998) Multipotent progenitor cells in the adult dentate gyrus. J. Neurobiol., 36(2): 249-266. Garcia, A. et al. (2004) GFAP-expressing progenitors are the principal source of constitutive neurogenesis in adult mouse forebrain. Nat. Neurosci., 7:1233-1241. Garcia-Verdugo, J. et al. (1998) Architecture and cell types of the adult subventricular zone: in search of the stem cells. J. Neurobiol., 36: 234-248. Goldman, S. (1998) Adult neurogenesis: from canaries to the clinic. J. Neurobiol, 36: 267-286. Goldman, S. (2003) Glia as neural progenitor cells. Trends Neurosci., 26: 590-596. Goldman, S. (2004) Directed mobilization of endogenous neural progenitor cells: the intersection of stem cell biology and gene therapy. Curr. Opin. Mol. Ther., 6: 466^72. Goldman, S. and Sim, F. (2005) Neural progenitor cells of the adult brain. In: Gearhart, J. (Ed.), Stem Cells: Nuclear programming and therapeutic applications. Novartis Foundation Symposium 265, John Wiley, London, pp. 66-92. Goldman, S. et al. (1997) Neuronal precursor cells of the adult rat ventricular zone persist into senescence, with no change in spatial extent or BDNF response. J. Neurobio., 32: 554-566. Goldman, S. et al. (2002) Isolation and induction of adult neural progenitor cells. Clin. Neurosci. Res., 2: 70-79. Goldman, S.A. (1997) Comparative strategies of subependymal neurogenesis in the adult forebrain. In: Gage F. and Chris Y. (Eds.), Isolation, characterization and utilization of CNS stem cells. Springer-Verlag, Fondation IPSEN, New York and Berlin, pp. 43-65. Goldman, S.A. (2005a) Neural progenitor cells of the adult human forebrain. In: Rao, M. (Ed.), Stem Cells and CNS Development, 2d ed., Humana, New York City, pp. 267-298. Goldman, S.A. (2005b) Neural stem and progenitor cell-based therapy of the human central nervous sytem. Nat. Biotech., 23: 862-871. Goldman, S.A. and Nottebohm, F. (1983) Neuronal production, migration, and differentiation in a vocal control nucleus of the adult female canary brain. Proc. Natl. Acad. Sci., 80(8): 2390-2394. Gould, E. (1999) Serotonin and hippocampal neurogenesis. Neuropsychopharmacology, 21: 46S-51S. Gravel, M. et al. (1998) Four-kilobase sequence of the mouse CNP gene directs spatial and temporal expression of lacZ in transgenic mice. J. Neurosci. Res., 53(4): 393-404. Gross, R.E. et al. (1996) Bone morphogenetic proteins promote astroglial lineage commitment by mammalian subventricular zone progenitor cells. Neuron, 17(4): 595-606. Jin, K. et al. (2002a) Stem cell factor stimulates neurogenesis in vitro and in vivo. J. Clin. Invest, 110(3): 311-319. Jin, K. et al. (2002b) Vascular endothelial growth factor (VEGF) stimulates neurogenesis in vitro and in vivo. Proc. Natl. Acad. Sci., 99(18): 11946-11950. Jin, K. et al. (2003) Directed migration of neuronal precursors into the ischemic cerebral cortex and striatum. Mol. Cell Neurosci., 24(1): 171-189.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

64

5. TARGETED INDUCTION OF ENDOGENOUS NEURAL STEM AND PROGENITOR CELLS: A NEW STRATEGY FOR GENE THERAPY

Jin, K. et aL (2004) Increased hippocampal neurogenesis in Alzheimer's disease. Proc. NatL Acad. Sci. USA, 101(1): 343-347. Kaye, E. (2001) Update on genetic disorders affecting white matter. Pediatr. NeuroL, 24: 11-24. Keyoung, H.M. et aL (2001) High-yield selection and extraction of two promoter-defined phenotypes of neural stem cells from the fetal human brain. Nat. Biotech., 19(9): 843-850. Kirschenbaum, B. and Goldman, S.A. (1995) Brain-derived neurotrophic factor promotes the survival of neurons arising from the adult rat forebrain subependymal zone. Proc. Natl. Acad. Sci.,. 92: 210-214. Kirschenbaum, B. et al. (1994) In vitro neuronal production and differentiation by precursor ceUs derived from the adult human forebrain. Cereb. Cortex, 4(6): 576-589. Kondo, T. and Raff, M. (2000) OUgodendrocyte precursor cells reprogrammed to become multipotential CNS stem cells. Science, 289(5485): 1754-1757. Kuhn, H.G. et al. (1997) Epidermal growth factor and fibroblast growth factor-2 have different effects on neural progenitors in the adult rat brain. J. Neurosci., 17(15): 5820-5829. Kukekov, V. et al. (1999) Multipotent stem/progenitor cells with similar properties arise from two neurogenic regions of adult human brain. Exp. Neurol., 156: 333-344. Lai, K. et al. (2003) Sonic hedgehog regulates adult neural progenitor proliferation in vitro and in vivo. Nat. Neurosci., 6: 21-27. Levine, J.M., Reynolds, R. and Fawcett, J.W. (2001) The oligodendrocyte precursor cell in health and disease. Trends Neurosci., 24(1): 39-47. Li, M. et al. (1998) Generation of purified neural precursors from embryonic stem cells by lineage selection. Curr. Biol., 8: 971-974. Lie, D. et al. (2002) The adult substantia nigra contains progenitor cells with neurogenic potential. J. Neurosci., 22: 6639-6649. Lim, D. et al. (2000) Noggin antagonizes BMP signaling to create a niche for adult neurogenesis. Neuron, 28: 713-726. Lim, D. A. et al. (2000) Noggin antagonizes BMP signaling to create a niche for adult neurogenesis. Neuron, 28(3): 713-726. Lindvall, O., Kokaia, Z. and Martinez-Serrano, A. (2004) Stem cell therapy for human neurodegenerative disorders. Nat. Med., 10 (supplement): S42-S50. Liu, X. et al. (1997) Insulin-like growth factor-I treatment reduces immune cell responses in acute non-demyehnative experimental autoimmune encephalomyeUtis. J. Neurosci. Res., 47: 531-538. Liu, X. et al. (1998) Basic FGF and FGF receptor 1 are expressed in microglia during experimental autoimmune encephalomyeUtis: temporally distinct expression of midkine and pleiotrophin. Glia, 24: 390-397. Loeffler, M. and Potten, C.S. (1997) Stem cells and cellular pedigrees. In: Potten, C.S. (Ed.), Stem Cells, Academic Press, San Diego, pp. 1-28. Lois, C. and Alvarez-Buylla, A. (1993) Proliferating subventricular zone cells in the adult mammalian forebrain can differentiate into neurons and glia. Proc. Nati. Acad. Sci., 90(5): 2074-2077. Lois, C. and Alvarez-Buylla, A. (1994) Long-distance neuronal migration in the adult mammalian brain. Science, 264(5162): 1145-1148. Lois, C , Garcia-Verdugo, J.M. and Alvarez-Buylla, A. (1996) Chain migration of neuronal precursors. Science, 271(5251): 978-981. Louissaint, A. et al. (2002) Coordinated interaction of angiogenesis and neurogenesis in the adult songbird brain. Neuron, 34: 945-960. Louissaint, A., Jr. et al. (2002) Coordinated interaction of neurogenesis and angiogenesis in the adult songbird brain. Neuron, 34(6): 945-960.

Luskin, M.B. (1993) Restricted proliferation and migration of postnatally generated neurons derived from the forebrain subventricular zone. Neuron, 11(1): 173-189. Magavi, S., Leavitt, B. and Macklis, J. (2000) Induction of neurogenesis in the neocortex of adult mice. Nature, 405: 951-955. Malberg, J. et al. (2000) Chronic antidepressant treatment increases neurogenesis in adult rat hippocampus. J. Neurosci., 20: 9104-9110. Mallon, B.S. et al. (2002) Proteolipid promoter activity distinguishes two populations of NG2-positive cells throughout neonatal cortical development. J. Neurosci., 22(3): 876-885. Mangiarini, L. et al. (1996) Exon 1 of the HD gene with an expanded CAG repeat is sufficient to cause a progressive neurological phenotype in transgenic mice. Cell, 87: 493-506. McTigue, D.M. et al. (1998) Neurotrophin-3 and brain-derived neurotrophic factor induce oligodendrocyte proliferation and myelination of regenerating axons in the contused adult rat spinal cord. J. Neurosci., 18(14): 5354-5365. Menezes, J.R. et al. (1995) The division of neuronal progenitor cells during migration in the neonatal mammalian forebrain. Mol. CeU Neurosci., 6(6): 496-508. Morshead, C M . et al. (1994) Neural stem cells in the adult mammalian forebrain: a relatively quiescent subpopulation of subependymal cells. Neuron, 13(5): 1071-1082. Nakatomi, H. et al. (2002) Regeneration of hippocampal pyramidal neurons after ischemic brain injury by recruitment of endogenous progenitors. Cell, 110: 429-441. Niemann, C. and Watt, KM. (2002) Designer skin: lineage commitment in postnatal epidermis. Trends Cell Biol., 12: 185-192. Nilsson, M. et al. (1999) Enriched environment increases neurogenesis in the adult rat dentate gyrus and improves spatial memory. J. NeurobioL, 39: 569-578. Nottebohm, F. (1985) Neuronal replacement in adulthood. Ann. N. Y. Acad. Sci., 457:143-161. Nottebohm, F. (2002) Why are some neurons replaced in adult brain? J. Neuroscience, 22(3): 624-628. Nunes, M.C. et al. (2003) Identification and isolation of multipotential neural progenitor cells from the subcortical white matter of the adult human brain. Nat. Med., 9(4): 439-447. Packer, M. et al. (2003) Nitric oxide negatively regulates mammalian adult neurogenesis. Proc. Nati. Acad. Sci., 100: 9566-9571. Palmer, T. (2002) Adult neurogenesis and the vascular Nietzsche. Neuron, 34: 856-858. Palmer, T.D. et al. (1999) Fibroblast growth factor-2 activates a latent neurogenic program in neural stem cells from diverse regions of the adult CNS. J. Neurosci., 19(19): 8487-8497. Palmer, T.D., Ray, J. and Gage, FH. (1995) FGF-2-responsive neuronal progenitors reside in proliferative and quiescent regions of the adult rodent brain. Mol. Cell Neurosci., 6(5): 474-486. Palmer, T.D., Takahashi, J. and Gage, F.H. (1997) The adult rat hippocampus contains primordial neural stem cells. Mol. Cell Neurosci., 8(6): 389-404. Palmer, T.D., Willhoite, A.R. and Gage, FH. (2000) Vascular niche for adult hippocampal neurogenesis. J. Comp. Neurol., 425(4): 479-494. Parent, J. et al. (2002) Rat forebrain neurogenesis and striatal neuron replacement after focal stroke. Ann. Neurol., 52: 802-813. Pencea, V. et al. (2001) Infusion of brain-derived neurotrophic factor into the lateral ventricle of the adult rat leads to new neurons in the parenchyma of the striatum, septum, thalamus, and hypothalamus. J. Neurosci., 21(17): 6706-6717.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

ACKNOWLEDGMENTS Penderis, J., Shields, S.A. and Franklin, RJ. (2003) Impaired remyelination and depletion of oligodendrocyte progenitors does not occur following repeated episodes of focal demyelination in the rat central nervous system. Brain, 126(Pt 6): 1382-1391. Pincus, D.W. et al. (1997) In vitro neurogenesis by adult human epileptic temporal neocortex. Clin. Neurosurg., 44:17-25. Pincus, D.W. et al. (1998) Fibroblast growth factor-2/brain-derived neurotrophic factor-associated maturation of new neurons generated from adult human subependymal cells. Ann. Neurol., 43(5): 576-585. Potten, C.S. and Loeffler, M. (1990) Stem cells: attributes, cycles, spirals, pitfalls and uncertainties. Lessons for and from the crypt. Development, 110(4): 1001-1020. Powers, J. (2004) The leukodystrophies: overview and classification. In: Lazzarini, R.A. (Ed.), Myelin Biology and Disorders, Elsevier Academic Press, San Diego, pp. 663-690. Richards, L.J., Kilpatrick, T.J. and Bartlett, RR (1992) De novo generation of neuronal cells from the adult mouse brain. Proc. Natl. Acad. Sci., 89(18): 8591-8595. Robinson, S. et al. (2005) Developmental changes induced by graded prenatal systemic hypoxic-ischemic insults in rats. Neurobiol. Dis., 18: 568-581. Rowitch, D.H. et al. (1999) Sonic hedgehog regulates proliferation and inhibits differentiation of CNS precursor cells. J. Neurosci., 19(20): 8954-8965. Roy, N.S. et al. (1999) Identification, isolation, and promoter-defined separation of mitotic oligodendrocyte progenitor cells from the adult human subcortical white matter. J. Neurosci., 19(22): 9986-9995. Roy, N.S. et al. (2000a) In vitro neurogenesis by progenitor cells isolated from the adult human hippocampus. Nat. Med., 6(3): 271-277. Roy, N.S. et al. (2000b) Promoter-targeted selection and isolation of neural progenitor cells from the adult human ventricular zone. J. Neurosci. Res., 59(3): 321-331. Sanai, N. et al. (2004a) Unique astrocyte ribbon in adult human brain contains neural stem cells but lacks chain migration. Nature, 427(6976): 740-744. Sawamoto, K. et al. (2001) Generation of dopaminergic neurons in the adult brain from mesencephalic precursor cells labeled with a nestin-GFP transgene. J. Neurosci., 21: 3895-3903. Scolding, N. (1998) Glial precursor cells in the adult human brain. Neuroscientist, 4: 264-272. Scolding, N. et al. (1998) Oligodendrocyte progenitors are present in the normal adult human CNS and in the lesions of multiple sclerosis [see comments]. Brain, 121 (Pt 12): 2221-2228. Scolding, N.J., Rayner, P.J. and Compston, D.A. (1999) Identification of A2B5-positive putative oligodendrocyte progenitor cells and A2B5-positive astrocytes in adult human white matter. Neuroscience, 89(1): 1 ^ .

65

Seaberg, R.M. and van der Kooy, D. (2002) Adult rodent neurogenic regions: the ventricular subependyma contains neural stem cells, but the dentate gyrus contains restricted progenitors. J. Neurosci., 22:1784-1793. Song, H., Stevens, C.R and Gage, RH. (2002) Astroglia induce neurogenesis from adult neural stem cells. Nature, 417(6884): 39-44. Svendsen, C.N., Bhattacharyya, A. and Tai, Y.T. (2001) Neurons from stem cells: preventing an identity crisis. Nat. Rev. Neurosci., 2(11): 831-834. Urayama, A. et al. (2004) Developmentally regulated maimose 6phophate receptor-mediated transport of a lysosomal enzyme across the blood-brain barrier. Proc. Natl. Acad. Sci., 101: 12658-12663. van Praag, H., Kempermann, G. and Gage, R (1999) Runiung increases cell proliferation and neurogenesis in the adult mouse dentate gyrus. Nat. Neurosci., 2: 266-270. Wang, S. et al. (1998) Isolation of neuronal precursors by sorting embryonic forebrain transfected with GFP regulated by the T alpha 1 tubulin promoter. Nat. Biotech., 16(2): 196-201. Watt, RM. (2001) Stem cell fate and patterning in mammalian epidermis. Curr. Opin. Genet. Dev., 11: 410^17. Weiss, S. et al. (1996) Multipotent CNS stem cells are present in the adult mammalian spinal cord and ventricular neuroaxis. J. Neurosci., 16(23): 7599-7609. Windrem, M.S. et al. (2002) Progenitor cells derived from the adult human subcortical white matter disperse and differentiate as oligodendrocytes within demyelinated lesions of the rat brain. J. Neurosci. Res., 69(6): 966-975. Windrem, M.S. et al. (2003) Identification, selection and use of adult human oligodendrocyte progenitor cells. In: Zigova, T. and Snyder, E. (Ed.), Neural Stem Cells for Brain Repair, Humana, New York City pp. 69-88. Windrem, M.S. et al. (2004) Fetal and adult human oligodendrocyte progenitor cell isolates myelinate the congenitally dysmyelinated brain. Nat. Med., 10(1): 93-97. Yao, D. et al. (1995) Insulin-like growth factor I treatment reduces demyelination and up-regulates gene expression of myelinrelated proteins in experimental autoimmune encephalomyelitis. Proc. Natl. Acad. Sci., 92: 6190-6194. Zhao, C. et al. (2003) Lentiviral vectors for gene delivery to normal and demyelinated white matter. Glia, 42(1): 59-67. Zhao, M. et al. (2003) Evidence for neurogenesis in the adult mammalian substantia nigra. Proc. Natl. Acad. Sci., 100: 7925-7930. Zigova, T, Peneca, V., Wiegand, S. and Huskin, M. (1998) Intraventricular administration of BDNF increases the number of newly generated neurons in the adult olfactory bulb. Molec. Cell. Neuroscience, 11: 234-245. Zimmerman, L.B., De Jesus-Escobar, J.M. and Harland, R.M. (1996) The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell, 86(4): 599-606.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

CHAPTER

6 Neurosurgical Targeting, Delivery, and Infusion of Gene Therapy Agents in the Brain Justin R Fraser, Neal Luther, Michael G. Kaplitt

Abstract: Gene delivery to the brain has until recently focused largely upon the molecular factors necessary to permit efficient transfer of therapeutic genes into target cells. However, as the ability to safely alter cellular function in a variety of settings has advanced, it has become increasingly clear that the physical mechanism of delivering gene therapeutics to the brain can limit effective clinical translation. Delivery from the vascular system to the brain is very difficult, and when focal gene expression is desirable, current molecular methods for controlling transduction and gene expression remain crude. Therefore, direct surgical delivery into the brain has been the method of choice for gene delivery in every trial of gene therapy in the brain conducted to date. Many of the techniques used in current applications derive from operations designed to either lesion portions of the brain or implant devices such as deep brain-stimulating electrodes. Stereotactic methods permit precise three-dimensional targeting of even the deepest structures, and this has been aided by advanced imaging techniques and computer-assisted reconstruction and navigation. These can be performed with either a traditional stereotaxic frame, or with frameless methods. The type of catheter and infusion parameters can also significantly influence both the efficiency of gene delivery and the area of spread, while adjuvant molecules can be added to the gene therapy solution to further influence these parameters. Finally, experiences with human trials in which a small focal area of the brain is targeted, such as Parkinson's disease, has revealed very different surgical delivery requirements compared with diseases where global delivery may be desirable, such as the genetic disorder Batten disease. As gene therapy continues to move into clinical practice, continued evolution of surgical techniques and infusion devices will aid in the safe and effective translation of biologically promising agents. Keywords: stereotactic surgery; deep brain stimulation; infusion; catheter; computer navigation

As described below, current methodology limits the clinical utility of delivering gene therapy from a peripheral administration, so proper targeting of an infusion catheter is necessary so that the intended target cell population is in fact treated. Second, the method of delivery may differ according to the surgical goal. Finally, variables that alter infusion can modify the efficacy of the final product. It is important to understand these principles as they guide current practice and future research in the emerging field of gene delivery to the human brain.

In an era of increasing research in gene therapy for neurological diseases, the delivery methodology and surgical implantation protocol can profoundly influence the ability to successfully translate a promising strategy into clinical practice. Utilization of gene therapy in the brain necessitates increasing safety, efficiency, and accuracy in stereotactic neurosurgery. Effective gene delivery depends upon several important elements. Anatomical targeting represents the first element; the tools and methods must be reliable and precise in order to facilitate accurate surgical planning.

Gene Therapy of the Central Nervous System: From Bench to Bedside

67

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

68

6. NEUROSURGICAL TARGETING, DELIVERY, AND INFUSION OF GENE THERAPY AGENTS IN THE BRAIN

I. ANATOMICAL TARGETING I N STEREOTACTIC NEUROSURGERY Targeting methodology represents a vital component in stereotactic neurosurgery. As therapeutic targets become more specific and delineated, the need for more accurate targeting tools drives continuing advances in the field. Current tools in neuroradiology provide excellent resources for evaluating brain regions and nuclei, and for their localization within a stereotactic plane. A review of the current techniques in stereotactic planning emphasizes the role of both these radiographic tools and novel advances in computer technology that permit real-time utilization and optimization of radiographic information. Recent studies have demonstrated roles for new targeting methodologies that will provide additional tools to improve neurosurgical stereotaxy. Stereotactic neurosurgery is a field that began before the availability of magnetic resonance imaging. Armed with only a thorough understanding of gross neuroanatomy and radiographs from such tools as aircontrast ventriculography, neurosurgeons undertook procedures to approach and impact the deep-brain nuclei (Walter and Vitek, 2004). While an understanding of these subjects is important in neurosurgical stereotaxy, current targeting methods depend largely upon radiographic anatomy from CT and MRI, and electrophysiologic monitoring. While CT and ventriculography can still be utilized for stereotactic planning, MRI has emerged as the standard for assessing neuroanatomical landmarks as it provides a high-resolution view of specific neurosurgical targets. Improvements in MR technology such as stronger magnets, faster acquisition, and employment of adjimcts such as MR angiogram have bolstered the utility of MRI as a planning tool for stereotactic neurosurgery. Within the MR environment, stereotactic targeting of deep-brain nuclei, such as the thalamus, subthalamic nucleus (STN), and globus pallidus (GP), employs two distinct methods for planning surgical approaches. Aptly named the 'direct' and 'indirect' methods, the former represents a newer method in the MR-era that relies on visual selection of the target from MR imaging, while the latter utilizes a standard set of measures from a midline landmark. The 'direct' method may be ideal for MRI, as the high resolution enhances the ability to directly visualize specific intracranial structures. Vayssierre et al., (2002) compared direct MRI selection of targets to selection based upon the Schaltenbrand and Talairach atlases (Talairach and Toumoux, 1998; Schaltenbrand and Wahren, 2002). Using the directly selected targets

as a standard verified by postoperative clinical results, the investigators found significant differences between the atlas-based coordinates and the directly selected coordinates(Vayssiere et al., 2002). As such, the direct method of targeting represents an increasingly useful tool as MRI imaging technology continues to improve. In contrast, the indirect method of targeting utilizes the line connecting the anterior and posterior commissures as a zeroing standard from which deep-brain nuclei are targeted using a set of calculations. Ventriculography, CT, and MRI can all act as foundation radiographic studies for this method, although MRI may provide the best atlas. In evaluating these options for surgical planning, Cuny et al. investigated the use of the direct and indirect methods for targeting the subthalamic nucleus (STN) in 14 patients. Using electrophysiologic guidance and functional stimulation response as a baseline standard for optimal electrode placement in patients with advanced Parkinson's disease, the investigators found that, while the indirect method was more accurate than the direct method, within the indirect method, the use of 3D MR imaging was superior to ventriculography in determining accurate, reliable, and reproducible targets for electrode placement in the STN (Cuny et al., 2002). While the specificity of the target in this study represents a caveat to its generalization in stereotactic targeting, it underscores the trend in stereotactic research toward methods that express reliable and reproducible precision in head-to-head comparisons. Thus, while surgical planning should not rely solely upon direct visual selection of targets, particularly in deep-brain nuclei, MRI represents a vital tool that allows target selection and surgical planning for both the direct and indirect method. As a radiographic tool, early MR! technology was impeded by lower resolution, slow image acquisition, and less well-defined protocols for diagnosis-specific imaging. However, recent studies have verified that current technical improvements have corrected some of these earlier deficiencies to produce relatively small targeting error with relative efficiency. In a study of MRI error among 11 patients undergoing repeat deep-brain stimulation electrode placement surgery, Simon et al. quantified targeting error through three-dimensional assessment of the distance between electrode placement and targeted coordinates (Simon et al., 2005). The investigators found that direct targeting utilizing images from a 1.5-T MRI resulted in a mean lateral-medial error of 0.09 ± 0.34 mm, a mean anterior-posterior error of 0.01 ± 0.32 mm, and a mean superior-inferior error of -0.08 ±0.33 mm (Simon et al., 2005). Thus, MRI represents a tool with demonstrated accuracy in surgical

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

ANATOMICAL TARGETING IN STEREOTACTIC NEUROSURGERY

planning for access to deep-brain nuclei. In addition to progress in accuracy, MRI can now be performed in relatively short times given appropriate sequence protocol designation. Hariz et al., (2003) in a study among eight medical centers utilizing predominantly 1.5-T MRI, tested a protocol using single T2-weighted nonvolumetric MRI to preoperatively and postoperatively assess deep-brain stimulation electrode placement (Hariz et al., 2003). Image acquisition varied between 3 min, 5 s and 7 min, 48 s (Hariz et a l , 2003). From such throughput studies, it is clear that, while MRI remains slower than CT, image acquisition time using high-resolution scanners continues to improve, enhancing the applicability of MRI as a practical method for preoperative stereotactic imaging. In addition to advances in MRI acquisition, quality, and processing, progress in other radiographic methodologies provides opportunities for other instruments in stereotactic neurosurgical planning. Functional MRI (fMRI) utilizes blood oxygen level contrasts to radiographically differentiate task-specific cortical activity (Ogawa et al.,1990; Atlas et a l , 1996; Krishnan et al., 2004). Krishnan et al., (2004) in a study of 54 patients with intracranial tumors near the motor cortex, found that fMRI could be used to calculate distance of the lesion/resection from the primary motor cortex, and that increasing distance is correlated with better neurological outcome if resection. Pirotte et al., (2005) found that preoperative fMRI findings correlated to intraoperative cortical mapping as a functional targeting method for epidural motor cortex stimulation in 17 of 18 patients with neuropathic pain. Chavez et al., (2005) studied three-dimensional fast imaging employing steady-state acquisition (3-D FIESTA) MRI to plan surgical intervention for trigeminal neuralgia. In 14 of 15 patients with trigeminal neuralgia, FIESTA imaging demonstrated clear anatomy of the trigeminal complex including the root entry zone, the trigeminal ganglion, and vasculature (Chavez et al., 2005). As such, fMRI and FIESTA represent variations in image acquisition, and may clearly add new dimensions to surgical planning in functional stereotaxy. Despite such advances, MRI continues to have limits as a tool for stereotactic preoperative planning. Not all institutions can utilize fast acquisition protocols with high resolution. Additionally, scanning the patient prior to the day of the procedure provides an opportunity to plan the approach before the operation begins. However, the typical headframes utilized are not fit for outpatient application and use. Due to such limitations, practitioners now routinely employ image fusion software to bring preoperative imaging and

69

planning into the real-time operating environment. One method entails a preoperative MRI used to plan the surgery, an early CT on the day of the procedure with the headframe in place, followed by fusion of the two images immediately and preoperatively. The software overlays the high-resolution MRI images (containing the surgical target and proposed tract) onto the three-dimensional stereotactic grid created by the head CT. While fusion software has been used in the literature to validate targeting methods by fusion pre- and postoperative imaging, its incorporation into the operating room has created a tool for real-time targeting and coordinate adjustment (Ferroli et al., 2005; Hamid et al., 2005). Furthermore, fusion technology allows for a more efficient operative chronology: the patient is fitted with a headframe, taken to noncontrast head CT, and taken directly to the operating room with no loss of time performing preoperative targeting measurements on the day of surgery. As such, fusion technology represents an important tool in the current and future practice of functional neurosurgery. Despite recent advances in MRI accuracy, imaging protocols, and software processing, some radiographic questions remain. For example, while many stereotactic procedures such as biopsies and deep-brain stimulator implantations have emerged as frame-dependent operations, stereotactic targeting for neuro-oncology and tumor resection has thrived on frameless stereotaxy. Frameless stereotaxy involves placement of some identifying mark on the patient's head, either stickers or small screws, which can be seen on MRI or CT. These are then identified to the navigation machine, and when a sufficient number of such 'fiducials' are identified, the device can then track a pointer anywhere within a certain radius of three-dimensional space with acceptable accuracy. Recent studies have examined the true accuracy of frameless stereotaxy. Gralla et al, (2003) found a technique of frameless stereotactic biopsy for intracranial tumor to deliver accurate diagnoses in 96.5% of patients (N = 57) (Gralla et al., 2003). Future research and advancement will provide further data to improve the efficacy and accuracy of frameless and frame-based stereotaxy, particularly as frameless stereotaxy is increasingly utilized in functional procedures such as surgery for Parkinson's disease. However, each method has some clear advantages and limitations. While a headframe provides a rigid three-dimensional system that can incorporate a mechanized delivery system, frameless stereotaxy provides flexibility, freedom of movement, and adjustment in all planes. Also, while frame-based stereotaxy continues to have a slight advantage in accuracy over frameless systems, only one

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

70

6. NEUROSURGICAL TARGETING, DELIVERY, AND INFUSION OF GENE THERAPY AGENTS IN THE BRAIN

target can be accessed at a time using a frame, while multiple targets such as bilateral procedures can be performed simultaneously if desired using frameless systems. Future techniques may combine such concepts, utilizing 'fiducials' to orient the patient to a nonattached rigid frame which would provide the support needed for certain types of infusion systems. While controversies in imaging improvements and coordinate mapping exist, they are frequently compared to electrophysiological data as the gold standard for targeting. Despite the increasing availability of intraoperative MRI, many technical limits issues continue to limit MRI to use primarily as a preoperative tool. Electrophysiologic monitoring, however, allows real-time evaluation of positioning in stereotactic neurosurgery. Methods for monitoring deep-brain nuclei have been studied and validated (Macias et al., 1997; Molinuevo et al., 2003; Nowinski et al., 2004).Hamani et al, (2005) found excellent correlation between MRI and microelectrode physiology in 10 patients with Parkinson's disease. However, the electrophysiologically defined subthalamic nucleus sometimes extended more anteriorly than that appreciated on MRI . As such, electrophysiologic monitoring can be important for real-time intraoperative confirmation of a functionally relevant target, which may even differ slightly from the radiographic target. This of course depends upon understanding the physiology of the structure to be treated, and whUe this is well-understood for several deep-brain structures targeted for movement disorder or epilepsy surgery, many other areas of the human brain have not been as well studied and therefore the value of intraoperative electrophysiology may be limited by a lack of relevant information for some applications. Neurosurgical targeting has undergone exponential growth and advancement recently, and continued research and development in neuroradiological methods, imaging software, and electrophysiologic monitoring will provide better techniques for preoperative planning and perioperative targeting in stereotactic neurosurgery. These tools are vital to the implantation of gene therapy in the brain as complete accuracy must be the goal.

IL

METHODS OF ACCESSING THE CENTRAL NERVOUS SYSTEM

While target selection and surgical planning are vital steps to gene delivery, the method of delivery is as important. Several techniques for accessing the central

nervous system (CNS) have been studied, including retrograde translation via the olfactory tracts, intravascular injection, intraventricular injection, intracavitary placement, and direct intraparenchymal infusion. The technique most appropriate depends highly upon the overall goal and specific neuronal target. The olfactory route offers a direct and nonpenetrating method for CNS gene delivery. The olfactory nerve endings, penetrating through the olfactory mucosa, trace directly through the cribiform plate into the CNS. While it is unknown whether substances are taken up directly by the neurons and transported in a retrograde fashion or whether they move into the subarachnoid space via the olfactory mucosa, it is clear than many substances can access the CNS through this pathway (Begley, 2003, 2004; Davis et al., 2003; Ilium, 2003). However, most substances absorbed this way express effects throughout the CNS, suggesting a lack of defined endpoint targets of delivery. However, Jerusalmi et al.,(2003) indirectly studied intranasal injections of gene therapy for experimental autoimmune encephalomyelitis. The investigators utilized a Semliki Forest virus expression system to express IL10 and green florescent protein (GFP) in Balb/c mice, and found that protein expression was visually detected by fluorescence in the olfactory bulb (Jerusalmi et al., 2003). While this method has demonstrated some promise for treating diffuse neuropathology, much future research is required to further elucidate the capabilities and limitations of the olfactory route for delivery of gene therapy to the brain. Intravascular injection represents an important technique for delivering therapeutics in medicine. As with olfactory delivery, it is limited in its ubiquitous nature; all parts of the CNS would be exposed to the potential agent. The blood-brain barrier represents the most severe limitation for intravascular delivery of gene therapy to the brain. The blood-brain barrier severely limits penetration of particular molecules from the intravascular space, protecting the brain from systemically administered substances .(Kaplitt and Lozano, 2001). Larger molecules (>500kDa), charged substances, molecules with a high propensity to form hydrogren bonds, and molecules with significant polarity are less able to penetrate the blood-brain barrier (Bodor and Buchwald, 2003; Begley, 2004). However, transvascular gene therapy may be possible if one can access endogenous transport systems that are present within the blood-brain barrier and serve to permit selective transport under normal conditions (Pardridge, 2002; Schlachetzki et al., 2004). In a review of transvascular approaches to gene therapy, Schelachetzki et al.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

METHODS FOR ENHANCING INTRACRANIAL GENE TRANSFER

hypothesized that CNS-directed gene therapy could be successfully delivered by intravascular injection through the use of selective transporter targeting in conjunction with a vector containing a CNS-specific promoter (Shi et al., 2001; 2003; Schlachetzki et al., 2004). While these methods may be feasible and have excellent potential, they are not at a translational stage to be employed as active surgical techniques for applying gene therapy to the brain. Intraventricular injection of gene therapy, while more invasive, represents a previously well-studied method for delivery of drugs to the central nervous system. In particular, the Ommaya reservoir is a chronic implant that permits injection of chemotherapy directly into the cerebrospinal fluid. However, while the Ommaya reservoir permits long-term access to the CSF space, its purpose is directed at treatment of leptomeningeal and diffuse disease. Ooboshi et al. (1995) demonstrated that adenoviral injection into the cisterna magna results in infection of overlying major arteries, adventitial cells of large blood vessels, and some smooth-muscle of smaller vessels. This diffusion pattern is supported by recent studies, most notably by Sugiura et al., (2005) who demonstrated that intraventricular administration of recombinant adenovirus expressing heparin-binding epidermal growth factorlike growth factor significantly improved functional recovery, angiogenesis, and neurogenesis (as assessed by bromodeoxyuridine injection) in Wistar rats that underwent middle cerebral artery occlusion ischemic strokes. As such, intraventricular injection of gene therapy in the brain could play an important role in diffuse cerebrovascular diseases such as ischemic stroke, as well as leptomeningeal diseases. However, the brain-CSF barrier does present an important limitation to intraventricular injection, acting to restrict the applications of intraventricular gene therapy. In addition, the immune system within the ventricular lining is different and more robust than that in the normal brain parenchyma, so there may be concerns regarding a more profound immune reaction to gene therapy delivered via an intraventricular route. Despite such limitations, directed research to further define indications and applications for intraventricularly injected gene therapy is needed. Intraparenchymal infusion of gene therapy offers a more direct method for localized delivery. Such a method avoids the BBB altogether, limits potential for systemic toxicity of the infusate, and limits the amount of virus necessary to deliver the gene to the therapeutic target area (Tang and Chiocca, 1997). This method applies to both neuro-oncological surgery

71

and functional neurosurgery. Direct implantation of chemotherapeutics in tumor cavities has provided a demonstrated foundation for embedding of adjuncts in the treatment of intracranial tumors. For example, utilization of the carmustine-loaded 'Gliadel' wafer has demonstrated significant survival benefits in clinical trials for glioblastoma multiforme (Brem et al., 1995; Westphal et al., 2003; Chiocca et al., 2004). Through such pioneering innovation, a method for adjunctive gene therapy in neuro-oncology becomes clear. Retroviruses, adenovirus, and Herpes Simplex Virus have all acted as vectors in human clinical trials for gene therapy adjuncts in neuro-oncology. Implantation of such genes as thymidine kinase (often paired with gangiclovir), j8-galactosidase, p-53, and oncolytic adenovirus have been subjects of several phase I, II and even III trials (Shand et al., 1999; Rainov, 2000; Rampling et al., 2000; Chen et a l , 2001; Lang et al., 2003; Chiocca et al., 2004). Direct implantation of gene therapy has also entered a clinical phase in functional neurosurgery. We have been involved in two current trials, including a Phase I trial investigating the infusion into the subthalamic nucleus of the glutamic acid decarboxylase (GAD) gene via an adeno-associated virus (AAV) vector, as well as a trial of gene therapy for Batten's disease ((NIH) TNIoH, 2005). In the Parkinson's disease trial, the goal is to efficiently deliver vectors focally to the subthalamic nucleus, which is a structure of roughly 6mm X 5mm X 3mm in the human. Therefore, standard frame-based stereotactic techniques were used for targeting this area, similar to those used for traditional deep-brain stimulation. The vector was infused via a single injection of 50 ]x\ of solution over 100 min via a borosilicate catheter of only 140 i^m in diameter. By contrast. Batten's disease is a global pediatric neurogenetic degenerative disorder, with the goal of widely delivering the potentially corrective gene therapy to large areas of the brain. Therefore, this protocol utilized frameless stereotaxy to identify three areas on each side of the brain (six total injection sites), and then catheters were inserted into these sites along a trajectory and to a depth guided by the computerized frameless stereotactic system. This was essential in this disorder, since many of the patients have significant brain atrophy and therefore simply creating six random injection sites would likely result in infusion into the cerebrospinal fluid rather than into brain substance in most cases. Similar infusion parameters were utilized to the Parkinson's disease trial, although two infusions were performed per site, with each infusion performed at a different depth to optimize

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

72

6. NEUROSURGICAL TARGETING, DELIVERY, AND INFUSION OF GENE THERAPY AGENTS IN THE BRAIN

spread. These two trials reflect only a small sampling of the differences which may be faced in various trials depending upon the specific issues of gene delivery, target and disease state of the individual brain. Future clinical trials such as these will rely both upon previous stereotactic success with electrode implantation, such as deep-brain stimulation in the STN for Parkinson's disease, and upon animal data demonstrating reliable, accurate, and beneficial infusion of the therapeutic viral vector (Luo et al., 2002; Erola et al., 2005; Nilsson et al., 2005). Thus, intraparenchymal infusion represents a central method for delivery of gene therapy in functional and neuro-oncologic neurosurgery. Intraparenchymal infusion, intraventricular infusion, intravascular administration, and olfactory transportation combine to provide a virtual armamentarium to approach intracranial disease processes with gene transfer therapy. However, in considering these options, it is important to understand their limitations as well as variables that can alter their efficacy and applicability. III. METHODS FOR E N H A N C I N G INTRACRANIAL GENE TRANSFER While the different approaches to gene transfer in the brain have specific advantages and limitations, their value may be enhanced through alterations in technique and through the employment of specific adjuncts. Improvements in intraparenchymal injection techniques increase the potential volume of infusate, and potentiate improved dissemination of gene transfer. Coadministration of supplemental materials such as mannitol may augment absorption and delivery in several infusion methods. Finally, it is important to be mindful of the target environment at the time of gene delivery; pathological changes can alter the efficacy of gene transfer. These important elements exemplify the importance of research aimed not at comparisons of different delivery approaches, but at improving delivery techniques to maximize effect and therapeutic benefit. While different methods exist for intraparenchymal injection of gene-carrying vectors, one of the most efficacious methods is convection-enhanced delivery. Convection-enhanced delivery (CED), also known as interstitial infusion or intracerebral clysis, is a method of delivering therapeutic agents intracranially via a stereotactically positioned cannula. High-flow, continuous pressure gradients are utilized to drive the infusate through the interstitial compartment; this pressure gradient is theorized to improve overall

uniformity and volume of distribution of a therapeutic agent (Bobo et al., 1994; Morrison et al., 1994; Laske et al,. 1997). Through CED, the amount of agent injected directly affects distribution (KroU et al., 1996; Tang et al., 1997). Furthermore, High-flow microinfusion permits distribution in a mathematically predictable model (Morrison et al., 1999). Over some thresholds, rapid infusion may cause significant backflow along a catheter tract resulting in extra-target dissemination of viral vectors (Morrison et al., 1994). In one study of flow dynamics, Morrison et al. concluded that smaller flow rates (e.g.,
I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

CONCLUSION

injection at the time of intraventricular infusion of adenoviral vectors resulted in penetration of the vector into the subependymal layer, and increased j8-glucuronidase activity (Ghodsi et al., 1999). Bourgoin et al., (2003) demonstrated enhanced delivery of adenoviral vectors when co-injected locally with mannitol in intracerebral injections of the j6-subunit of jS-hexosaminidase. However, Burger et al. found that systemic administration of mannitol amplified striatal transduction of rAAV2 in a rat model by 400% and enhanced distribution volume by 200%, while local co-administration only enhanced transduction by 25% (Burger et al., 2005). As such, systemic administration of mannitol can be an efficacious adjunct in promoting gene transfer in local intraparenchymal infusion. While the specific mechanism of such enhancement still requires further study, the hyperosmolality would likely enhance vector transduction. Other adjunctive infusates may be specific to the viral vector. For example, heparin can induce increased distribution volumes when used as an adjunct co-infusate with AAV2 (Nguyen et al., 2001). As such, further research may provide simple adjuncts that will greatly augment gene transfer efficacy in the brain. While techniques as convection-enhanced delivery and co-administration of transfer adjuncts may enhance gene therapy in the brain, it is important to consider the target environment, and to understand what obstructions and obstacles may impair efficient gene transfer. The presence of brain edema may alter infusion rate and efficacy (Laske et al., 1997; Kaplitt and Lozano, 2001). Pathological cells such as astrocytomas may undergo transduction of viral vectors at different rates and with different characteristics than nonpathological brain tissue. As such. Further study is required to understand how pathological processes in the brain alter gene therapy as an efficacious and beneficial protocol. IV-

CONCLUSION

In considering gene therapy for the brain, arenas of research have erupted for different vectors, innumerable targets, and a plethora of neurological diseases. However, human gene transfer to the brain also depends upon the methodology of delivery and the mechanics of infusion. While current tools for stereotactic targeting provide an excellent base for determining extremely specific infusion coordinates, the field continues to progress as new tools become more clinically recognized and utilized. The methods

73

of gene delivery provide sharp contrasts; each method carries particular benefits and drawbacks. In addition, adjunctive methodologies can serve to augment gene transfer, and to increase efficiency and efficacy of gene delivery in the brain. Future research should focus upon these important principles, and should continue to guide the focus of gene therapy in the brain. References Abe, T., Wakimoto, H., Bookstein, R., Maneval, D.C., Chiocca, E.A. and Basilion, J.R (2002) Intra-arterial delivery of p53-containing adenoviral vector into experimental brain tumors. Cancer Gene Ther., 9: 228-235. Atlas, S.W., Howard II R.S., Maldjian, J., Alsop, D., Detre, J.A., Listerud, J., D'Esposito, M., Judy, K.D., Zager, E. and Stecker, M. (1996) Functional magnetic resonance imaging of regional brain activity in patients with intracerebral gliomas: findings and implicatioris for clinical management. Neurosurgery, 38:329-338. Begley, D.J. (2003) Understanding and circumventing the bloodbrain barrier. Acta. Paediatr. Suppl., 92: 83-91. Begley, D.J. (2004) Delivery of therapeutic agents to the central nervous system: the problems and the possibilities. Pharmacol. Ther., 104: 29-45. Bobo, R.H., Laske, D.W., Akbasak, A., Morrison, PR, Dedrick, R.L. and Oldfield, E.H. (1994) Convection-enhanced delivery of macromolecules in the brain. Proc. Natl. Acad. Sci. USA, 91: 2076-2080. Bodor, N. and Buchwald, P. (2003) Brain-targeted drug delivery: experiences to date. Am. J. Drug Targ., 1:13-26. Bourgoin, C , Emiliani, C , Kremer, E.J., Gelot, A., Tancini, B., Gravel, R.A., Drugan, C , Orlacchio, A., Poenaru, L. and Caillaud, C. (2003) Widespread distribution of beta-hexosaminidase activity in the brain of a Sandhoff mouse model after coinjection of adenoviral vector and mannitol. Gene Ther., 10: 1841-1849. Brem, H., Piantadosi, S., Burger, P C , Walker, M., Selker, R., Vick, N.A., Black, K., Sisti, M., Brem, S., Mohr, G., et al. Placebo-controlled trial of safety and efficacy of intraoperative controlled delivery by biodegradable polymers of chemotherapy for recurrent gliomas. The Polymer-brain Tumor Treatment Group. Lancet, 345: 1008-1012. Burger, C , Nguyen, F.N., Deng, J. and Mandel, R.J. (2005) Systemic marmitol-induced hyperosmolality amplifies rAAV2-mediated striatal transduction to a greater extent than local co-infusion. Mol. Ther. 11: 327-331. Chavez, G.D., De Salles, A.A., Solberg, T.D., Pedroso, A., Espinoza D. and Villablanca, P. (2005) Three-dimensional fast imaging employing steady-state acquisition magnetic resonance imaging for stereotactic radiosurgery of trigeminal neuralgia. Neurosurgery 56: E628; discussion E628. Chen, Y., DeWeese, T, Dilley J., Zhang, Y, Li, Y, Ramesh, N., Lee, J., Pennathur-Das, R., Radzyminski, J., Wypych, J., Brignetti, D., Scott, S., Stephens, J., Karpf, D.B., Henderson, D.R. and Yu D.C. (2001) CV706, a prostate cancer-specific adenovirus variant, in combination with radiotherapy produces synergistic antitumor efficacy without increasing toxicity. Cancer Res., 61: 5453-5460. Chiocca, E.A., Broaddus, W.C, Gillies, G.T., Visted, T and Lamfers, M.L. (2004) Neurosurgical delivery of chemotherapeutics, targeted toxins, genetic and viral therapies in neuro-oncology. J. Neurooncol., 69:101-117.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

74

6. NEUROSURGICAL TARGETING, DELIVERY, AND INFUSION OF GENE THERAPY AGENTS IN THE BRAIN

Cuny, E., Guehl, D., Burbaud, P., Gross, C , Dousset, V. and Rougier, A. (2002) Lack of agreement between direct magnetic resonance imaging and statistical determination of a subthalamic target: the role of electrophysiological guidance. J. Neurosurg., 97: 591-597. Davis, S.S.and Ilium, L. (2003) Absorption enhancers for nasal drug delivery. Clin. Pharmacokinet., 42: 1107-1128. Erola, T., Karinen, P., Heikkinen, E., Tuominen, J., Haapaniemi, T., Koivukangas, J. and Myllyla, V. (2005) Bilateral subthalamic nucleus stimulation improves health-related quality of Hfe in Parkinsonian patients. Parkinsonism Relat. Disord., 11: 89-94. Ferroli, P., Franzini, A., Marras, C , Maccagnano, E., D'Incerti, L. and Broggi, G. (2004) A simple method to assess accuracy of deep brain stimulation electrode placement: pre-operative stereotactic CT + postoperative MR image fusion. Stereotact. Funct. Neurosurg., 82: 14-19. Ghodsi, A., Stein, C., Derksen, T., Martins, I., Anderson, R.D. and Davidson, B.L. (1999) Systemic hyperosmolality improves betaglucuronidase distribution and pathology in murine MPS VII brain following intraventricular gene transfer. Exp. Neurol., 160: 109-116. Gralla, J., Nimsky, C , Buchfelder, M., Fahlbusch, R. and Ganslandt, O. (2003) Frameless stereotactic brain biopsy procedures using the Stealth Station: indications, accuracy and results. Zentralbl. Neurochir., 64:166-170. Hamani, C , Richter, E.O., Andrade-Souza, Y, Hutchison, W., Saint-Cyr., J.A. and Lozano, A.M. (2005) Correspondence of microelectrode mapping with magnetic resonance imaging for subthalamic nucleus procedures. Surg. Neurol., 63: 249-253; discussion 253. Hamid, N.A., Mitchell, R.D., Mocroft, P, Westby, G.W., Mihier, J. and Pall, H. (2005) Targeting the subthalamic nucleus for deep brain stimulation: technical approach and fusion of pre- and postoperative MR images to define accuracy of lead placement. J. Neurol. Neurosurg. Psychiatry, 76: 409-414. Hariz, M.I., Krack, P, Melvill, R., Jorgensen, J.V., Hamel, W., Hirabayashi, H., Lenders, M., Wesslen, N., Tengvar, M. and Yousry, T.A. (2003) A quick and universal method for stereotactic visualization of the subthalamic nucleus before and after implantation of deep brain stimulation electrodes. Stereotact. Funct. Neurosurg., 80: 96-101. Ilium, L. (2003) Nasal drug delivery—possibilities, problems and solutions. J Control Release, 87:187-198. Jerusalmi, A., Morris-Downes, M.M., Sheahan, B.J. and Atkins, G.J. (2003) Effect of intranasal administration of Semliki Forest virus recombinant particles expressing reporter and cytokine genes on the progression of experimental autoimmune encephalomyelitis. Mol. Ther., 8: 886-894. Kaplitt, M.G. and Lozano, A.M., (2001) Surgical drug delivery for neurodegenerative diseases. Clin. Neurosurg., 48: 127-144. Krishnan, R., Raabe, A., Hattingen, E., Szelenyi, A., Yahya, H., Hermann, E., Zimmermann, M. and Seifert, V. (2004) Functional magnetic resonance imaging-integrated neuronavigation: correlation between lesion-to-motor cortex distance and outcome. Neurosurgery, 55: 904-914; discussion 914-905. KroU, R.A., Pagel, M.A., Muldoon, L.L., Roman-Goldstein, S. and Neuwelt, E.A. (1996) Increasing volume of distribution to the brain with interstitial infusion: dose, rather than convection, might be the most important factor. Neurosurgery, 38: 746-752; discussion 752-744. Lang, F.R, Bruner, J.M., Fuller, G.N., Aldape, K., Prados, M.D., Chang, S., Berger, M.S., McDermott, M.W., Kunwar, S.M., Junck, L.R., Chandler, W., Zwiebel, J.A., Kaplan, R.S. and Yung, W.K. (2003) Phase I trial of adenovirus-mediated

p53 gene therapy for recurrent glioma: biological and clinical results. J. Clin. Oncol., 21: 2508-2518. Laske, D.W., Morrison, PR, Lieberman, D.M., Corthesy, M.E., Reynolds, J.C., Stewart-Henney, P.A., Koong, S.S., Cummins, A., Paik, C.H. and Oldfield, E.H. (1997) Chronic interstitial infusion of protein to primate brain: determination of drug distribution and clearance with single-photon emission computerized tomography imaging. J. Neurosurg., 87: 586-594. Luo, J., Kaplitt, M.G., Fitzsimons, H.L., Zuzga, D.S., Liu, Y, Oshinsky, M.L. and During, M.J. (2002) Subthalamic GAD gene therapy in a Parkinson's disease rat model. Science, 298: 425-429. Macias, R., Teijeiro, J., Torres, A. and Alvarez, L (1997) Electrophysiological targeting in stereotaxic surgery for Parkinson's disease. Adv Neurol., 74:175-182. Molinuevo, J.L., Valldeoriola, F. and Valls-Sole, J. (2003) Usefulness of neurophysiologic techniques in stereotactic subthalamic nucleus stimulation for advanced Parkinson's disease. Clin. Neurophysiol., 114: 1793-1799. Morrison, PR, Chen, M.Y, Chadwick, R.S., Lonser, R.R. and Oldfield, E.H. (1999) Focal delivery during direct infusion to brain: role of flow rate, catheter diameter, and tissue mechanics. Am. J. Physiol., 277: R1218-1229. Morrison, PR, Laske, D.W., Bobo, H., Oldfield, E.H. and Dedrick, R.L. (1994) High-flow microinfusion: tissue penetration and pharmacodynamics. Am. J. Physiol., 266: R292-305. Nguyen, J.B., Sanchez-Pemaute, R., Cunningham, J. and Bankiewicz, K.S. (2001) Convection-enhanced delivery of AAV-2 combined with heparin increases TK gene transfer in the rat brain. Neuroreport, 12: 1961-1964. (NIH) TNIoH: Human Gene Transfer Protocols. Last Updated: 03-08-05, in. The National Institutes of Health, 2005. Nilsson, M.H., Tomqvist, A.L. and Rehncrona, S. (2005) Deep-brain stimulation in the subthalamic nuclei improves balance performance in patients with Parkinson's disease, when tested without anti-parkinsonian medication. Acta Neurol. Scand., I l l : 301-308. Nowinski, W.L., Belov, D., PoUak, P and Benabid, A.L. (2004) A probabilistic functional atlas of the human subthalamic nucleus. Neuroinformatics, 2: 381-398. Ogawa, S., Lee, T.M., Kay, A.R. and Tank, D.W. (1990) Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc. Natl. Acad. Sci. USA, 87: 9868-9872. Ooboshi, H., Welsh, M.J., Rios, CD., Davidson, B.L. and Heistad, D.D. (1995) Adenovirus-mediated gene transfer in vivo to cerebral blood vessels and perivascular tissue. Circ. Res., 77: 7-13. Pardridge, W.M. (2002) Drug and gene targeting to the brain with molecular Trojan horses. Nat. Rev. Drug. Discov., 1: 131-139. Pirotte, B., Voordecker, P., Neugroschl, C , Baleriaux, D., Wilder, D., Metens, T., Denolin, V., Joffroy, A., Massager, N., Brotchi, J. and Levivier, M (2005) Combination of functional magnetic resonance imaging-guided neuronavigation and intraoperative cortical brain mapping improves targeting of motor cortex stimulation in neuropatiiic pain. Neurosurgery, 56: 344-359; discussion 344r-359. Rainov, N.G. (2000) 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., 11: 2389-2401. Rampling. R., Cruickshank, G., Papanastassiou, V., Nicoll, J., Hadley, D., Brennan, D., Petty, R., MacLean, A., Harland, J., McKie, E., Mabbs, R. and Brown, M. (2000) Toxicity evaluation of replicationcompetent herpes simplex virus (ICP 34.5 null mutant 1716) in patients with recurrent malignant glioma. Gene Ther., 7: 859-866.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

CONCLUSION Schaltenbrand, G. and Wahren, W. (1977) Atlas for Stereotaxy of the Human Brain. Stuttgart, Georg Thieme. Schlachetzki, R, Zhang, Y, Boado, R.J. and Pardridge, W.M. (2004) Gene therapy of the brain: the trans-vascular approach. Neurology, 62:1275-1281. Shand, N., Weber, R, Mariani, L., Bernstein, M., Gianella-Borradori, A., Long, Z., Sorensen, A.G. and Barbier, N. (1999) A phase 1-2 clinical trial of gene therapy for recurrent glioblastoma multiforme by tumor transduction with the herpes simplex thymidine kinase gene followed by ganciclovir. GLI328 EuropeanCanadian Study Group. Hum. Gene. Ther., 10: 2325-2335. Shi, N., Zhang, Y, Zhu, C , Boado, R.J. and Pardridge, W.M. (2001) Brain-specific expression of an exogenous gene after i.v. administration. Proc. Natl. Acad. Sci. USA, 98: 12754-12759. Simon, S.L., Douglas, P, Baltuch, G.H. and Jaggi, J.L. (2005) Error analysis of MRI and Leksell stereotactic frame target localization in deep brain stimulation surgery. Stereotact. Funct. Neurosurg., 83: 1-5. Sugiura, S., Kitagawa, K., Tanaka, S., Todo, K., Omura-Matsuoka. E., Sasaki, T., Mabuchi, T., Matsushita, K., Yagita, Y and Hori, M, (2005) Adenovirus-mediated gene transfer of heparinbinding epidermal growth factor-like growth factor enhances

75

neurogenesis and angiogenesis after focal cerebral ischemia in rats. Stroke, 36: 859-864. Talairach, J. and Tournoux, P. (1998) Co-Planar Stereotaxic Atlas of the Human Brain: 3-Dimensional Proportional System: An Approach to Cerebral Imaging. New York, Thieme Medical. Tang, G. and Chiocca, A. (1997) Gene transfer and delivery in central nervous system disease. Neurosurg. Rocus, 3: e2. Vayssiere, N., Hemm, S., Cif, L., Picot, M.C., Diakonova, N., El Pertit, H., Rrerebeau, P. and Coubes, P. (2002) Comparison of atlas- and magnetic resonance imaging-based stereotactic targeting of the globus pallidus intemus in the performance of deep brain stimulation for treatment of dystonia. J. Neurosurg., 96: 673-679. Walter, B.L. and Vitek, J.L. (2004) Surgical treatment for Parkinson's disease. Lancet Neurol., 3: 719-728. Westphal, M., Hilt, D.C., Bortey, E., Delavault, P , Olivares, R., Warnke, P C , Whittle, I.R., Jaaskelainen, J. and Ram, Z. (2003) A phase 3 trial of local chemotherapy with biodegradable carmustine (BCNU) wafers (Gliadel wafers) in patients with primary malignant glioma. Neuro-oncol, 5: 79-88. Zhang, Y, Schlachetzki, R, Li, J.Y, Boado, R.J. and Pardridge, W.M. (2003) Organ-specific gene expression in the rhesus monkey eye following intravenous non-viral gene transfer. Mol. Vis., 9:465-472.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

C H A P T E R

7 Gene Transfer for Neurological Disease: Agencies, Policies, and Process* Jerry R. Mendelly Andra Miller

Abstract: Advances in molecular biology have contributed to a growing interest in gene therapy as a form of treatment for neurological diseases. Implementation, however, requires knowledge of the regulatory policies governing this field of research, especially in view of the greater stringency imposed by the serious adverse events affecting some patients participating in gene therapy protocols. Educational resources accessible to basic and clinical scientists hoping to bring their work to the clinical forefront in the form of a gene therapy protocol for a phase I clinical trial are not available through any single source. The steps involved can be arduous and incredibly time-consuming, including laboring through many documents to discover the route to implementation of a clinical protocol for gene therapy. In this chapter, the authors have summarized the steps involved including: the regulatory agencies and their requirements, the phases of clinical development with emphasis on a phase I study, and specific steps leading to an Investigational New Drug (IND) application for a biological product to be used in a gene therapy clinical trial. Links are provided to all appropriate websites, which will facilitate the efforts for investigators. Keywords: gene therapy; neurological disease; regulatory policies

L THE LANDSCAPE OF GENE THERAPY A. Overview of Clinical Approaches to Gene Transfer The rapid expanse of understanding in molecular biology nurtures expectations that gene transfer can provide treatment for a wide range of disorders. To date, over 500 clinical gene therapy trials have been initiated in the United States (http://www4.od.nih.gov/ oba/rac/clinicaltrial.htm). Most have been in cancer but interest is rapidly developing for application of this methodology to neurological disorders. Gene therapy is not restricted to any single paradigm. Replacement of a defective gene with absent or reduced protein product. *This work was supported in part by funding from National Institutes of Health, National Institute of Arthritis, Musculoskeletal, and Skin Diseases RFA: AR-03-001

Gene Therapy of the Central Nervous System: From Bench to

as demonstrated in the mdx mouse model for Duchenne muscular dystrophy, represents one application of gene therapy (Wang et al., 2000; Gregorevic et al., 2004). Other strategies include the use of anti-sense oligonucleotides to induce exon skipping (van Deutekom et al., 2001; Bertoni and Rando, 2002), or DNA/RNA chimeric hybrid oligonucleotides to restore the normal reading frame (Smith et a l , 1999). Some diseases, inherited or acquired, are potentially treatable by strategically delivered neuroprotective agents that combat the underlying process. For example, neural-derived growth factors can provide advantages for diseases as diverse as Alzheimer's disease and amyotrophic lateral sclerosis (Kaspar et al., 2003). In primates cholinergic neurons in the basal forebrain bundle, known to be important in memory function, are protected by transfer of fibroblasts engineered to secrete nerve growth factor (Tuszynski, 2002) encouraging the initiation of a clinical phase I study in Alzheimer's. In a mouse model Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

78

7. GENE TRANSFER FOR NEUROLOGICAL DISEASE: AGENCIES, POLICIES, AND PROCESS

for amyotrophic lateral sclerosis, IGF-1 protects spinal motor neurons by retrograde axonal transport following injection of limb and intercostal muscles (Kaspar et al., 2003). Alternative strategies employ gene products to inhibit overactive neuronal pools. Excitable neurons in the subthalamic nucleus can acquire an inhibitory phenotype after delivery of the glutamic acid decarboxylase (GAD) gene by a viral vector which upregulates y-amino butyric acid (GABA) (Oransky, 2003). Discharging neurons resulting in uncontrollable seizures are quelled by viral vector-mediated delivery of galanin, a 29- to 30-amino acid peptide, known to antagonize excitatory glutamatergic neurotransmission (Lin et al., 2003). In some cases, protection can be achieved by targeting a molecular pathway to prevent the consequences of over-expressed genes, illustrated by transfer of cytokine genes to reduce tumor growth (Lichtor et al., 2003). In addition, neoplastic cells can be killed by transfer of a virus with retained lytic properties (Vecil and Lang, 2003). Confidence in the potential for gene therapy for neurological disorders is facilitated by a growing repertoire of viral vectors well suited for expression in nervous system tissues. Herpes simplex virus (HSV) is a neurotropic DNA virus with high affinity for peripheral sensory neurons, potentially useful for treatment of pain syndromes (Glorioso et al., 2003) and peripheral nerve disorders (Goss et al., 2002). Skeletal muscle is the natural host for the non-pathogenic, non-replicative adeno-associated virus (AAV), an agent that also exhibits efficient, stable gene expression in human neurons (Du et al., 1996). The studies cited here emphasize the potential for bringing gene therapy closer to the arena of clinical investigation for neurological disease. The lack of exposure of clinicians and scientists to the stepwise process involved in the implementation of a clinical gene transfer protocol is a weakness in our educational process. The usual continuing education resources for our field are not inclusive of such materials and there is no single resource pointing the way to websites and regulatory documents that are essential in developing clinical gene transfer protocols. This chapter is designed to facilitate the process. B.

Although the vector was primarily delivered to the liver, postmortem analysis found it in a variety of tissues and death was attributed to systemic Ad vector-induced shock syndrome, acute respiratory distress, and multiorgan system failure. This event had ramifications throughout the gene therapy "world." Anxiety intensified in the autumn of 2002 because of a report of T cell leukemia in a patient with X-linked severe combined immunodeficiency disease (X-SCID) treated by ex vivo, retroviral-mediated gene transfer of the IL2rg gene (Hacein-Bey-Abina et al., 2003). The risks for cancer are presumed to be related to insertional mutagenesis, supported by evidence for retroviral vector integration in proximity to LM02, a known human T cell oncogene (Dave et al., 2004). The issue, however, further intensified when two additional X-SCID patients (total of three) developed leukemia, and one of the three died of a cancer-related complication. The US Food and Drug Administration (FDA) reacted by halting all gene therapy trials for SCID (Check, 2005a). Subsequently, an advisory board met in March 2005 and permitted continuation of gene transfer for adenosine deaminase deficiency (ADA)-SCID in contrast to the X-linked variant (Check, 2005b). They further noted, however, that close follow up of the ADA patients will be necessary before final recommendations can be made. These events have clearly changed the environment for clinical gene transfer studies. For one thing, expression of transgene from an episomal vector location may no longer be considered as a disadvantage for gene therapy, and more research is likely to be directed toward preventing transcriptional silencing from episomal gene expression using strategies that include regulatory sequences, introns, and native promoter elements. From a regulatory perspective, informed consent documents will clearly need to state the risks of neoplasia and death following gene therapy clinical trials. Protocols employing viruses dependent on genome integration for gene expression, such as retroviruses (Nakamura, 2005) and lentiviruses (Wu and Burgess, 2004) will undergo closer scrutiny by regulatory agencies. Even presumably safe vectors, such as AAV will require long-term experience in human trials to unequivocally establish safety (Berns and Linden, 1995; Nakai et al., 2001; Schnepp et al., 2003).

Impact of Serious Adverse Events

Serious adverse events have changed the landscape for gene therapy. In September 1999, a death occurred during a gene therapy clinical trial that involved the transfer of an adenoviral (Ad) vector containing the ornithine transcarbamylase (OTC) gene (Verma, 2000).

IL OVERSIGHT OF GENE THERAPY PROTOCOLS The oversight of gene therapy falls under the purview of the Department of Health and Human Services

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

79

OVERSIGHT OF GENE THERAPY PROTOCOLS

(DHHS) (Cornetta and Smith, 2002). An organizational chart for the agencies within DHHS is illustrated in Fig. 1. Table 1 provides a listing of the key websites for agencies and committees involved in gene therapy. A.

Federal Agencies

related to protecting human subjects participating in research. OHRP also evaluates the effectiveness of the policies and programs of DHHS, and specifically promotes approaches that avoid unwarranted risks to human research subjects. OHRP provides, on its website, guidance on gene therapy.

1 • Office for Human Research Protections (OHRP)

2.

OHRP has responsibility for developing, monitoring, and implementing compliance for the protection of human subjects in research conducted or supported by any component of the DHHS. In addition, OHRP fulfills an important teaching role, developing programs and generating educational resource materials

In 1993, the FDA issued a statement published in the Federal Register (Taylor, 1993) defining statutory authorities governing therapeutic products that apply to human cell therapy and gene therapy according to the Public Health Service Act and the Federal Food, Drug, and Cosmetic Act. All gene therapy clinical trial protocols must be conducted under Investigational New Drug (IND) application. Regulations pertaiiung to this process appear in Title 21 of the Code of Federal Regulations (CFR), Part 312. It is the Center for Biologies Evaluation and Research (CBER) that regulates human gene therapy products in the Office of Cellular, Tissue, and Gene Therapies of the FDA. CBER provides a website for assistance in IND preparation for gene therapy.

DHSS

FDA

NIH

OHRP

CBER

OBA/RAC

IND Investigative site Investigator IBC

IRB Subject

FIGURE 1 The diagram shows the interaction of the regulatory agencies involved in the implementation of a gene therapy protocol. Oversight for the process falls to the Department of Health and Human Services (DHHS). The Center for Biologies Evaluation and Research (CBER) of the FDA reviews gene therapy IND applications submitted by the investigator/sponsor. The National Institutes of Health (NIH) establishes guidelines for genetic research and implements policy through the Office of Biotechnology Activities (OBA). The operation of the Recombinant DNA Advisory Committee (RAC), originally established to address concerns of recombinant DNA research, is managed by OBA. Al gene therapy protocols must be submitted to RAC for review but not all will be discussed. RAC may make suggestions regarding ethical, legal, and scientific issues. The Office for Human Research Protections (OHRP) is responsible for monitoring and implementing compliance, and supplying educational materials protecting human research subjects. At the investigative site the Institutional Biosafety Committee (IBC) ensures environmental and personnel safety and the Institutional Review Board (IRB) must approve the protocol and consent forms for the gene transfer trial. Enrollment of patients for the clinical study begins after attaining an IND, review by RAC, and approval by IBC and IRB. Reproduced with permission of Lippincott, Williams, and Wilkins from Neurology 2004: 2225-2232.

US Food and Drug Administration

(FDA)

3* ISIational Institutes of Health (NIH) and Office of Biotechnology Activities (OBA) The NIH establishes guidelines for genetic research and executes policy through OBA [originally called Office of Recombinant DNA Activities (ORDA)]. OBA is responsible for developing policies and procedures for the safe conduct of recombinant DNA activities and human gene transfer. In collaboration with CBER, OBA has developed a registry of activities related to recombinant DNA research and human gene transfer: the Genetic Modification Clinical Research Information System (GeMCRIS). OBA serves as a comprehensive information resource for scientists, research participants, sponsors, institutional oversight committees, federal officials, and others with an interest in human gene transfer research. OBA manages the operation of the Recombinant DNA Advisory Committee (RAC) established in 1974 as a result of public concerns over the potential risks of recombinant DNA research. NIH guidelines related to human gene transfer were first published in 1976 and have been periodically revised by the RAC. Committee members represent the disciplines of science, medicine, law and ethics, as well as representation from the public domain. RAC meetings are held quarterly and discussions take place in a public forum. All human gene transfer trials in which NIH funding is involved (either directly or indirectly) must be registered with the OBA/RAC. Appendix M of the NIH guidelines

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

80

7. GENE TRANSFER FOR NEUROLOGICAL DISEASE: AGENCIES, POLICIES, AND PROCESS TABLE 1.

Websites for Agencies and Committees Involved in Gene Therapy

Office for Human Research Protections (OHRP) • Responsibilities of OHRP for protection of human subjects http://www.hhs.gov/ohrp • OHRP guide for IRB reviewers and investigators for conduct of gene therapy: http://www.hhs.gov/ohrp/irb/irb_chapter5.htm Food and Drug Administration • Statutory authority invested in FDA regarding gene therapy: http://www.fda.gov/cber/genadmin/frl01493.pdf • Center for Biologies Evaluation and Research (CBER) (assistance for IND preparation): h t t p : / / w w w / f d a / g o v / c b e r / i n d / i n d . h t m • FDA Guidelines for ''Good Practices" (conduct of pre-clinical studies, product manufacture, and clinical trials): http://www.access.gpo. gov/cgi-bin/cfrassemble.cgi?title=199821, http://www.fda.gov/oc/gcp/regulations.html • FDA/CBER website policies, regulatory approach, and expectations for gene therapy: "Guidance for Human Somatic Cell Therapy and Gene Therapy": http://www.fda.gov/cber/gdlns/somegene.pdf • Four phases of clinical development for drugs and biologies (phases I-IV): http://www.fda.gov/cder/guidance/1857fnl.pdf • CBER steps for IND submission: http://www.fda.gov/cber/summaries/cberl01032204jf.pdf • Information on the pre-IND meeting: http://www.fda.gov/cber/gdlns/mtpdufa.htm • CBER Letter March 6, 2000 related to adverse events in gene therapy trials: http://www.fda.gov/cber/ltr/gt030600.htm International Conference on Harmonisation Tripartite Guideline on Good Clinical Practice • http: / / www.ich.org/MediaServer.jser?®JD=482&@_TYPE_MULTIMEDIA&@_TEMPLATE=616&@_MODE=GLB Office of Biotechnology Activities (OBA)/Recombinant DNA Advisory Committee (RAC) • OBA policies and procedures for gene therapy: http://www4.od.nih.gov/oba/ • Recombinant DNA Advisory Committee (RAC) human gene therapy guidelines: http://www4.od.nih.gov/oba/rac/guidelines/guidelines.html • NIH guidelines for submission of gene transfer protocols to the RAC, Appendix M: http:www4/od.nih.gov/oba/rac/guidelines_ 02APPENDD(_M.htm • Meeting dates and deadlines for RAC submission: http://wvvw4.od.nih.gov/oba/rac/meeting.html • Registry of activities related to recombinant DNA research and human gene transfer: Genetic Modification Clinical Research Information System (GeMCRIS): http://www4.od.nih.gov/oba/rac/gemcris/gemcris.htm Policies related to long-term follow up of participating subjects in clinical gene therapy trials • http: / /grants.nih.gov/grants/guide/notice-files/NOT-RR-04-005.html • Institutional Review Board • http: / / www.hhs.gov/ohrp/humansubjects/guidance/45cfr46.htm Institutional Biosafety Committee • http://www4.od.nih.gov/oba/IBC/IBCrole.htm

describes the requirements for submission of gene transfer protocols and related materials to the RAC (http://www4.od.nih.gov/oba/rac/guidelines_02/ APPENDIX_M.htm). Not all submitted protocols are discussed by RAC. Preference is reserved for protocols with: (i) new vectors/new gene delivery systems, (ii) new disease indications, (iii) unique applications of gene transfer, and (iv) other issues considered to require further public discussion, such as ethical issues. B.

Local Agencies at Clinical Site

Two committees are operative at the clinical site (may be public or private).

1 • The Institutional Review Board (IRB) All human research requires IRB approval and the functions, operations, and IRB committee membership follows policy defined in Title 45 of the CFR, Part 46. This includes a requirement that IRBs register with OHRP. 2,

The Institutional Biosafety Committee (IBC)

The IBC approves all experiments involving the transfer of recombinant DNA, or DNA or RNA derived from recombinant DNA into human participants. The IBC is charged with the obligation to determine the risks and ensure public and environmental safety in

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

CLINICAL PHASES FOR DEVELOPMENT OF BIOLOGICS

the locale where the research takes place. Originally, IBC oversight v^as restricted to recombinant DNA research, but over time responsibility has expanded to include a w^ide range of biohazardous materials (e.g., infectious agents and carcinogens). Members of the IBC committee must have expertise and training in recombinant DNA technology. Ad hoc consultants participate as required. The IBC files an annual report v^ith OBA.

IIL GOOD PRACTICE REQUIREMENTS FOR GENE THERAPY PROTOCOLS The FDA has established a set of "Good Practices" that provides the framework for all steps building toward and including implementation of the clinical gene therapy trial. These expectations for conducting pre-clinical studies, product manufacturing, and conduct of the clinical trial should be thoroughly understood by the sponsor of a gene therapy protocol (http://www/access.gpo.gov/cgi-bin/cfrassemble. cgi?title=199821). Title 21 of the CFR, Part 58 details the requirements for Good Laboratory Practice (GLP), a set of standards for the conduct of non-clinical laboratory studies used to support an IND application. Title 21of the CFR, Parts 210 and 211 describe the requirements for Good Manufacturing Practices (GMP) that apply to the manufacturing process and the manufacturing facility. While there is some flexibility at early IND phases in the level of adherence to GLP and GMP regulations, application of these principles in the initial stages of product development lays the foundation for establishing the appropriate controls and documentation necessary for the initiation of human clinical trials. Good Clinical Practice (GCP) guidelines apply to the conduct of the clinical trial and are described at the FDA website, h t t p : / / www.fda.gov/oc/gcp/regulations.html. Additional information can be assessed through the International Conference on Harmonisation Tripartite Guideline: Guideline on Good Clinical Practice (http://www.ich. org/MediaServer.jser?@_ID=482&@_TYPE_MULTIMEDIA&@_TEMPLATE=616&@_MODE=GLB). GCP guidelines implicitly require that clinical research be conducted for valid ethical and scientific reasons, be performed by qualified investigators, initiated only after IRB approval, and after valid informed consent has been obtained and documented. There must also be periodic monitoring of the clinical trial to assess the quality of the research and integrity of the data. A second set of tools published by the FDA is

81

available at the CBER w^ebsite describing specific policy, regulatory approach, and expectations for gene therapy: "Guidance for Human Somatic Cell Therapy and Gene Therapy (http://www.fda.gov/cber/gdlns/ somgene.pdf)."

IV. CLINICAL PHASES FOR DEVELOPMENT OF BIOLOGICS The FDA provides a detailed description of the four temporal phases (I-IV) of clinical development for biologics (http://www.fda.gov/cder/guidance/1857fnl. pdf). In general, the phases are performed serially to allow for results of prior studies to influence the plan of later phases. These phases were designed to principally address drug development but are also applied to gene therapy and other biologies. The initial administration into humans is a phase I trial (human pharmacology). Studies may be open, baseline-controlled, or randomized with blinding. For gene therapy, randomizing subjects is usually not appropriate, especially for a dose-escalation design, which is usually the paradigm, followed for phase I trials. Depending on study design, however, it is possible to randomize extremities for treatment (e.g., gene injection vs. vehicle-only into opposite limbs for gene therapy protocols). At the phase I level of development, non-therapeutic, pharmacologic objectives are paramount. Data collection focuses on (1) safety and tolerability; (2) distribution and clearance (pharmacokinetics) from local and remote sites, e.g., urine, semen, saliva; and (3) estimates of activity of recombinant agent (usually a secondary outcome measure). A phase II trial (therapeutic exploratory) assesses initial efficacy as its primary objective. Goals include determination of dose and regimen, and establishing endpoints. Target populations can be clarified, such as mild versus severe disease. Substantiation of efficacy is the goal of the phase III trial (therapeutic confirmatory) providing a basis for licensure and marketing. Post-marketing product modifications occur during a phase IV trial (therapeutic use). It is notable that no gene therapy product has yet to be licensed by the FDA. This is the result of a variety of factors. The time and resources needed to bring a product to market lead to excessive costs. Safety concerns, especially in light of recent gene therapy complications, further drive expenses. In addition, limited vector production capacity and technical issues of scale-up impose obstacles for some vector systems.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

82

7. GENE TRANSFER FOR NEUROLOGICAL DISEASE: AGENCIES, POLICIES, AND PROCESS

V. SEQUENTIAL STEPS IN IMPLEMENTATION OF A CLINICAL GENE TRANSFER TRIAL Seeking regulatory approval for a phase I gene therapy protocol is a stepwise process guided by the sponsor/investigator. Essentially, all of the agencies described above are involved. An important first step is a pre-IND meeting with CBER. A-

P r e 4 N D Meeting

1 • Scheduling the Pre-IND Meeting This meeting is arranged at the request of the sponsor/investigator through a formal written letter to the FDA. It is customary to fax the letter followed by a written hard copy. According to the requirements of the Prescription Drug User Fee Act II (PDUFA II), CBER will schedule the meeting within 60 days of the Agency's receipt of the written request for the meeting. The letter should provide a brief statement of the purpose of the meeting and succinct information regarding the vector, transgene, promoter, manufacturing process, and clinical trial. The sponsor should include a listing of the specific objectives and outcomes expected from the meeting. This must include targeted questions the investigators want to have addressed during the meeting. For example: is the patient population for the clinical trial acceptable? Are the proposed manufacturing process and testing procedures acceptable for release of vector? Are the toxicology and biodistribution plans adequate to support the titers of virus to be transferred? The sponsors should propose an agenda, including estimated times needed for each agenda item, a listing of planned external attendees, and a request to CBER for specific disciplines and individuals to be represented at the meeting. Within 14 calendar days of the Agency's receipt of a request for a formal meeting, CBER should notify the sponsor/investigator in writing (letter or fax) of the date, time, and place for the meeting, as well as CBER participants who will be present at the meeting. Information relating to the preIND meeting request and preparation of the pre-IND meeting package can be found at http://www.fda. gov/cber/gdlns/mtpdufa.htm. 2,

Preparing the Pre-IND Package

The pre-IND package prepared by the sponsor must be submitted 1 month before the meeting date. The package will have components similar to the IND application but with less details. A section on

background information establishes the premise and rationale for the clinical trial. The chemistry, manufacturing, and control (CMC) section summarizes vector derivation, the vector manufacturing process, and should specify procedures that assure consistency, sterility, and stability prior to product use. Information on cell banks or viral banks, gene modification of cells, as well as proposed specifications for product release should be included. The pre-clinical information should demonstrate proof-of-concept and feasibility using the proposed vector, transgene, and promoter. If possible, data should be presented in a disease animal model demonstrating gene expression and some indication of functional recovery following gene transfer. Plans for toxicology and biodistribution studies a n d / o r preliminary study data represent a critical component of the pre-IND package. This is often one of the most carefully examined aspects of the pre-IND package. The CBER reviewers will discuss inadequacies and the investigators are given the opportunity to defend their rationale for the proposed plan. The details of the clinical protocol in draft or synopsis are also presented, which include the study population, inclusion/exclusion criteria, plan for route of vector administration, dosing regimen, and safety monitoring. For a phase I trial, safety is the primary outcome measure, but efficacy can be included as a secondary measure (and is encouraged for most protocols). 3.

Conduct of the Pre-IND Meeting

In almost all cases, the meeting will be limited to 1 h and be conducted by teleconference. Only under exceptional circumstances will there be a face-to-face meeting. There is no need or time for the sponsor to make a presentation at the meeting. Instead, comments on the proposal by the FDA reviewers are the focus of the meeting. The FDA will provide written minutes documenting the discussion approximately 30 days after the meeting. This critical review process defines a clear path for IND preparation and strengthens the application for other regulatory submissions to RAC, IRB, and IBC B.

RAC Meeting

The meeting dates and deadlines for submission of materials are available through the OBA/RAC website: http://www4.od.nih.gov/oba/rac/meeting.html. Requirements for submission, format of materials, and all of the essential elements for the package are described at the Appendix M website: http://www4.od.nih.gov/ oba/rac/guidelines 02/APPENDIX__M.htm.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

SEQUENTIAL STEPS IN IMPLEMENTATION OF A CLINICAL GENE TRANSFER TRIAL

Although technically only NIH-supported protocols are subject to NIH/OBA review, this must be interpreted in the context of the language that extends authority to include all clinical trials conducted in institutions receiving NIH support for any recombinant DNA research. The RAC submission package determines if public discussion of a gene transfer protocol is required. NIH/OBA will notify the investigator of the need for RAC review within 15 working days of receipt of package. If in-depth review is required, three primary reviewers are assigned with the addition of ad hoc experts as needed. Applications submitted to RAC are available to the public and cannot be designated "confidential" in their entirety. If a sponsor determines that specific responses to one or more of the items described in the protocol should be considered proprietary or trade secret, these items must be clearly identified as such. OBA will make the final determination on confidentiality considering issues related to both the science and safety of the proposal (see Appendix M-I-C-5, Confidentiality). 1 • Components of the RAC Submission The prepared package for RAC will be submitted to the OBA, National Institutes of Health. NIH-OBA will confirm receipt within 3 working days. Investigators should contact OBA if they do not receive this confirmation. The package contains the following items. a. Cover letter A cover letter on institutional letterhead, signed by the Principal Investigator(s), that: (1) acknowledges that the documentation submitted to NIH-OBA complies with the requirements set forth in Appendix M-I-A, Requirements for Protocol Submission; (2) identifies the IBC and Institutional Review Board (IRB) at the proposed clinical trial site(s) responsible for local review and approval of the protocol; and (3) acknowledges that no research participant will be enrolled (i.e., obtaining informed consent from a potential research participant, or a designated legal guardian) until the RAC review process has been completed, and IBC and IRB approval have been obtained. b. Scientific abstract Summarizes the priniary (safety) and secondary (some measure of efficacy) end points, the vector and its functional capabilities in proof-of-concept studies, and the clinical protocol (subjects, vector and titers, route of administration). c. Non-technical abstract tion of study in lay terms.

Includes a brief descrip-

83

d. Clinical protocol This section describes the protocol in all of its facets. e* Responses to Appendices M'll through M-V This portion of Appendix M requires detailed answers to carefully guided questions and concerns intended to flush out the rationale and justification for the study and all of the safety features of the study design. The questions address the qualifications of the investigators, ethical and social issues related to the protocol, the informed consent document and the process of obtaining consent, and protection of subject privacy and confidentiality. f. Proposed informed consent document A draft of the proposed informed consent document is included. This follows the guidelines described in Appendix MIII. The RAC may suggest changes to be incorporated into the final consent form. g. Curriculum vitae of the principal investigator (s) NIH biographical sketches limited to no more than two pages are appropriate. C.

IBC Approval

The timing for submission to IBC is somewhat arbitrary but the IBC committee will want to consider issues and recommendations of the RAC. Therefore, it makes the most sense to wait until after the RAC presentation before seeking IBC approval. Information to IBC must include the source of the DNA, the nature of the inserted DNA sequences, the vectors to be used, the transgene and protein product to be produced, and the containment conditions that will be implemented. The IBC ensures that all aspects of Appendix M have been addressed. D. 1.

IRB Approval Timing of IRB Submission

IRB approval for the gene therapy protocol and the informed consent documents must take place prior to beginning a study. However, the timing for this request is left to the investigator keeping in mind that RAC will review and often modify the informed consent document to be used in the study. Therefore, the initial IRB approval can only be tentative until input from the FDA (IND), RAC, and IBC are received, demonstrating the highly integrated activities of these regulatory bodies in approval of gene therapy protocols.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

84 2.

7. GENE TRANSFER FOR NEUROLOGICAL DISEASE: AGENCIES, POLICIES, AND PROCESS

Major Points to he Addressed in the IRB

These issues are outlined in Appendix M-II and include the following: a. Description/purpose of the study The subjects should be provided with a detailed explanation in non-technical language of the purpose of the study and the procedures associated with the proposed study. The availability of alternative therapies and the possibility of other investigational interventions and approaches should be included. The subjects should be informed that participation in the study is voluntary and that failure to participate in the study or withdrawal of consent will not result in any penalty or loss of benefits to which the subjects are otherwise entitled. b. Benefits and risks The subjects should be provided with an accurate description of the possible benefits, if any, of participating in the proposed study. For studies that are not reasonably expected to provide a therapeutic benefit to subjects, the informed consent document should clearly state that no direct clinical benefit to subjects is expected to occur as a result of participation in the study, although knowledge may be gained that may benefit others. There should be clear itemization in the document of types of adverse experiences, their relative severity, and their expected frequencies. For consistency, the following definitions are suggested: side effects that are listed as mild should be ones which do not require a therapeutic intervention; moderate side effects require an intervention; and severe side effects are potentially fatal or life-threatening, disabling, or require prolonged hospitalization. If verbal descriptors (e.g., "rare," "uncommon," or "frequent") are used to express quantitative information regarding risk, these terms should be explained. The informed consent document should provide information regarding the approximate number of people who have previously received the genetic material under study It is necessary to warn potential subjects that, for genetic materials previously used in relatively few or no humans, unforeseen risks are possible, including ones that could be severe. Informed consent should indicate any possible adverse niedical consequences that may occur if the subjects withdraw from the study once it has started. The subjects should be provided with specific information about any financial costs associated with the protocol and in the long-term follow-up that are not covered by the investigators or the institution involved. Subjects should be provided an explanation about the extent to which they will

be responsible for any costs for medical treatment required as a result of research-related injury. c. Reproductive considerations To avoid the possibility that any of the reagents employed in the gene transfer research could cause harm to a fetus/child, subjects should be given information concerning possible risks and the need for contraception during the active phase of the study. The period of time for the use of contraception should be specified. The inclusion of pregnant or lactating women should be addressed. d. Long-term follow up To permit evaluation of long-term safety and efficacy of gene transfer, subjects should be informed that they are expected to cooperate in a long-term follow-up program. The exact period remains a topic for discussion by the FDA (and investigators should be alert for new recommendations). Currently, subjects are asked to have annual physical exams for the first 5 years following gene transfer. During a subsequent 10-year period, subjects are asked to complete questionnaire on health status. A list of persons who can be contacted in the event that questions arise during the follow-up period should be provided by the investigator. In the case of replication competent retroviruses, CBER recommends life-long monitoring on an annual basis with additional requirement for patients participating in studies involving hematopoietic stem cells transduced by retroviral vectors. In the latter case, subjects must have laboratory studies performed semi-annually for 5 years and then annually for the following 10 years to determine whether clonal cell populations have developed. To facilitate compliance with FDA requirements, the National Center for Research Resources (NCRR) announced on January 28,2004 that it will support the cost of subject visits at the General Clinical Research Center (GCRC) located either at study site or elsewhere if necessary (http://grants. nih.gov/grants/guide/notice-files/NOT-RR-04-005. html). Industry-sponsored projects are expected to reimburse costs encumbered by the GCRC. Storage of blood samples will be the responsibility of the NGVL at Indiana University. Investigators are compelled to make every effort to ensure that follow-up studies are carried out and that autopsy be requested at the time of death, no matter what the cause. Subjects should be asked to advise their families of the request and of its scientific and medical importance. The IRB document should also inform subjects that any significant findings resulting from their participation in a gene transfer study will be made known in a timely

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

SEQUENTIAL STEPS IN IMPLEMENTATION OF A CLINICAL GENE TRANSFER TRIAL

manner to them including new information about the experimental procedure, the harms and benefits experienced by other individuals involved in the study, and any long-term effects that have been observed. e. Privacy A statement should address the measures to be taken to protect the privacy of subjects and their families as well as maintain the confidentiality of research data. This will include provisions to honor the wishes of individual human subjects (and the parents or guardians of pediatric or mentally handicapped subjects) as to whether, when, or how the identity of a subject is publicly disclosed.

TABLE 2.

E.

85

I N D Submission

Under current regulations, any gene therapy product not previously authorized for marketing in the United States requires submission of an IND to the FDA. The specific requirements for content and format for a phase I gene transfer protocol are specified in Title 21of the CFR, Part 312.23, which lists the items that a sponsor should submit to the FDA. A summary of requirements for therapeutic biotechnology-derived products can be found at the following website: http://www.fda.gov/ cder/guidance/phasel.pdf. Specific requirements for a phase I study are enumerated in Table 2. CBER

Contents and Format for IND Submission for Phase I Gene Therapy Trial (FDA Form-lS?!, Title 21 of the CFR, Part 312,23):http://www.fda.gov/cder/guidance/phasel.pdf

Cover Sheet Form 1571 (a) Name, address, telephone number of sponsor (b) Phases of the clinical investigation to be conducted (c) A commitment to obtain appropriate IRB approval (d) Commitment to obtain IND approval before trial begins (e) Commitment to follow all regulatory requirements (f) Name of person responsible for monitoring conduct of trial including any contract research organization involved (g) Signature of sponsor Table of Contents Introductory statement and general investigational plan (a) Brief introductory statement with name of drug, active ingredients, pharmacologic class, structural formula, dosage, route of administration, broad objectives, planned duration of study (b) Summary of previous human experience with drug, reference to other INDs, investigational or marketing experience in other countries relevant to safety (c) Description of any prior withdrawal of product and reasons (d) Brief description for overall plan including rationale for product, indication for study, approach to evaluate product, clinical studies to be conducted, estimated number of patients, and serious risks anticipated from toxicological studies Investigator's

Brochure

Not required if sponsor is a single center investigator Protocols (a) Objectives and purpose of the study (b) Number of subjects, inclusion/exclusion criteria (c) Description of control subjects and methods to avoid bias (d) Dosing plan (duration, maximum dose, or method to determine dose) (e) Outcome measures (f) Details of plan to ensure subject safety: clinical and laboratory and toxicity-based stopping or dose adjustment rules (g) Description of investigators and sub-investigators (residents, fellows) Chemistry, manufacturing, and control information (a) Description of the agent, including physical and biological characteristics (b) Method of preparation (Continued)

I GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

86

7. GENE TRANSFER FOR NEUROLOGICAL DISEASE: AGENCIES, POLICIES, AND PROCESS

TABLE 2 {Continued) (c) Acceptable limits and analytical methods used to assure the identity, strength, quality, purity, and stability of the agent (d) Description of the composition and manufacture of placebo to be used in a controlled clinical trial (e) A copy of all labels and labeling to be provided to each investigator (f) Environmental analysis Pharmacology and toxicology information (a) Description of pharmacological effects and mechanism(s) of action of the product in animals, and information on absorption, distribution, metabolism, and excretion if known (b) Integrated summary of the toxicological effects (subacute and chronic); toxicity test related to the mode of administration or conditions of use and any in vitro studies to evaluate drug toxicity (c) Documentation that tests were performed in compliance with good laboratory practice regulations or brief statement of the reason it could not be done Previous human experience with the investigational drug Any information relevant to the safety of the study related to prior investigations (published) or if marketed outside the USA (provide a list of countries and any known places where product was withdrawn). (a) Special information possibly relevant to marketing product (b) Information previously submitted with identification of file, name, reference number, volume, and page number; if submitted by another person, a written/signed statement by an authorized person must accompany request to refer to material (c) Any part of IND in a foreign language must be accompanied by an English translation including any literature citations (d) Original and two copies of all submissions to IND file is required numbered serially using a single, three-digit serial number starting with initial IND to be numbered thousand (e) Sponsors must identify on the cover sheet any investigation exempt from informed consent FDA, Food and Drug Administration; IRB, Institutional Review Board; IND, Investigational New Drug.

presents a very succinct summary of the IND process in slide format that is worthy of review (http://www. fda.gov/cber/summaries/cberl01032204jf.pdf) and in addition, the Draft Guidance for FDA Review Staff and Sponsors published in November 2004 provides advice on the CMC information for gene therapy IND (http://www.fda.gov/cber/gdlns/gtindcmc.pdf). The basic elements of the IND include the following broad areas: (1) Pre-dinical in vitro and animal pharmacology establishing proof of concept, and toxicology studies assessing product safety for human trials. (2) Chemistry and manufacturing information pertaining to the composition, manufacture, stability, and controls used for manufacturing the drug substance and the drug product. This information is assessed to ensure consistency and control in batches of product to be used in the clinical trial. (3) Clinical protocols and investigator information are used to assess the risks of the initialphase trials and provide assurance that the clinical investigators are qualified to fulfill their clinical duties. Commitment must also be given to obtain informed consent, to obtain IRB approval, and to adhere to the IND regulations. The IND application addresses all issues identified during the pre-IND meeting plus specific points

recommended by FDA related to recent serious adverse events spelled out in the March 6, 2000 letter (http://www.fda.gov/cber/Itr/gt030600.htm). Most points will be familiar to the investigator and include the need for manufacturing summaries related to cell banks and viral banks, quality assurance and quality control programs for product manufacturing, the procedures ensuring compliance with GCP, a clinical monitoring plan, an organizational chart defining the role of individuals involved in the clinical trial, and the need for continued reporting of animal safety data that might raise awareness for clinical risk. The IND is submitted to CBER (regulatory management staff) in triplicate and upon receipt the sponsor will be issued an acknowledgment letter containing the date of receipt, the assigned IND number, and a reminder of the responsibility for submission to OBA / RAC. The point of contact within CBER is the Regulatory Manager, who is responsible for coordinating the review process. The IND review takes place over the next 30 calendar days after receipt of the application. The sponsor may be contacted during the review process for additional information or to discuss deficiencies. INDs automatically become effective 30 days after receipt unless FDA notifies the sponsor that the

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

LIMITATIONS OF THE SYSTEM

IND is subject to clinical hold, in which case the project is delayed until the concerns are resolved. This decision is communicated to the sponsor by phone, and is followed by a letter that provides the hold comments, and requests for additional information. For a phase I clinical study, the IND may be placed on hold for the following reasons, as detailed in Title 21 of the CFR, Part 312.42: human subjects are exposed to unreasonable and substantial risk of illness or injury; sufficient information is lacking to allow adequate assessment of risk; the information in the investigator's brochure is misleading, erroneous or materially incomplete; or the clinical investigators are not qualified to conduct the study In order to proceed with the clinical study, the sponsor must correct the hold deficiencies and submit a response as an amendment to the IND (Title 21 of the CFR, Part 312.30). This amendment will be reviewed at CBER within 30 calendar days and if satisfactory, the sponsor will be notified by phone that the clinical trial may proceed. A written letter will also be sent. All IND amendments are submitted in triplicate and are used to report protocol changes, new protocols or the addition of a new investigator or clinical site, as well as changes in the manufacturing process or new toxicology data. Annual reports (Title 21of the CFR, Part 312.33) are due within 60 days of the anniversary of the IND. The sponsor must provide IND safety reports at 15 days for any serious and unexpected adverse event or within 7 days for fatal or life threatening events (Title 21of the CFR, Part 312.32).

VI.

LIMITATIONS OF THE SYSTEM

The current system administered by DHHS is wellsuited to achieve the primary goal of patient safety for clinical gene therapy trials. Ethical issues are carefully scrutinized by RAC and IRB. The process usually assures proper safeguards for both patients and investigators. On the other hand, limitations to the current process include a tendency to err on the side of extreme caution in making recommendations for pre-clinical and clinical therapy protocols often resulting in exorbitant expenses for gene therapy studies. Such demands stretch the resources of many academic-based investigators. The affects on industry are also critical, influencing development toward those products with higher earning potential (e.g., heart disease and cancer). Thus, rare conditions, including many neurodegenerative disorders, may

87

be neglected. Finally, the redundancy in preparation of documents needed for approval of a gene therapy protocol results in a "mountain" of paper work that slows the process.

References Bems, K.L and Linden, R.M. (1995) The cryptic life style of adenoassociated virus. Bioessays, 17: 237-245. Bertoni, C. and Rando, T.A. (2002) Dystrophin gene repair in mdx muscle precursor cells in vitro and in vivo mediated by RNADNA chimeric oligonucleotides. Hum. Gene Ther., 13: 707-718. Check, E. (2005a) Gene therapy put on hold as third child develops cancer. Nature, 433: 561. Check, E. (2005b) Gene therapy trials to restart following cancer risk review. Nature, 434:137. Cornetta, K. and Smith, F.O. (2002) Regulatory issues for clinical gene therapy trials. Hum. Gene Ther., 13:1143-1149. Dave, U.R, Jenkins, N.A. and Copeland, N.G. (2004) Gene therapy insertional mutagenesis insights. Science, 303: 333. Du, B., Wu, R, Boldt-Houle, D.M. and Terwilliger, E.F. (1996) Efficient transduction of human neurons with an adeno-associated virus vector. Gene Ther., 3: 254-261. Glorioso, J.C, Mata, M. and Fink, D.J. (2003) Gene therapy for chronic pain. Curr. Opin. Mol. Ther., 5: 483-488. Goss, J.R., Goins, W.F., Lacomis, D., Mata, M., Glorioso, J.C. and Fink, D.J. (2002) Herpes simplex-mediated gene transfer of nerve growth factor protects against peripheral neuropathy in streptozotocin-induced diabetes in the mouse. Diabetes, 51: 2227-2232. Gregorevic, P., Blankinship, M.J., Allen, J.M., Crawford, R.W., Meuse, L., Miller, D.G., Russell, D.W. and Chamberlain, J.S. (2004) Systemic delivery of genes to striated muscles using adeno-associated viral vectors. Nat. Med., 10: 828-834. Hacein-Bey-Abina, S., Von Kalle, C , Schmidt, M., McCormack, M.R, Wulffraat, N., Leboulch, R et al. (2003) LM02-associated clonal T cell proliferation in two patients after gene therapy for SCID-Xl. Science, 302: 415-419. Kaspar, B.K., Llado, J., Sherkat, N., Rothstein, J.D. and Gage, RH. (2003) Retrograde viral delivery of IGF-1 prolongs survival in a mouse ALS model. Science, 301: 839-842. Lichtor, T. and Click, R.R (2003) Cytokine immuno-gene therapy for treatment of brain tumors. J. Neurooncol., 65: 247-259. Lin, E.J., Richichi, C , Young, D., Baer, K., Vezzani, A. and During, M.J. (2003) Recombinant AAV-mediated expression of galanin in rat hippocampus suppresses seizure development. Eur. J. Neurosci., 18: 2087-2092. Nakai, H., Yant, S.R., Storm, T.A., Fuess, S., Meuse, L. and Kay, M.A. (2001) Extrachromosomal recombinant adeno-associated virus vector genomes are primarily responsible for stable liver transduction in vivo. J. Virol., 75: 6969-6976. Nakamura, T. (2005) Retroviral insertional mutagenesis identifies oncogene cooperation. Cancer Sci., 96: 7-12. Oransky, I. (2003) Gene therapy trial for Parkinson's disease begins. Lancet, 362: 712. Schnepp, B.C., Clark, K.R., Klemanski, D.L., Pacak, C.A. and Johnson, P.R. (2003) Genetic fate of recombinant adeno-associated virus vector genomes in muscle. J. Virol., 77: 3495-3504. Smith, D.E., Roberts, J., Gage, FH. and Tuszynski, M.H. (1999) Ageassociated neuronal atrophy occurs in the primate brain and is

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

88

7. GENE TRANSFER FOR NEUROLOGICAL DISEASE: AGENCIES, POLICIES, AND PROCESS

reversible by growth factor gene therapy. Proc. NatL Acad. Sci. USA, 96: 10893-10898. Taylor, M.R. (1993) Application of current statutory authorities to human cell therapy products and gene therapy products. Notice. Federal Register, 58: 53248-53251. Tuszynski, M.H. (2002) Growth-factor gene therapy for neurodegenerative disorders. Lancet Neurol., 1:1-57. van Deutekom, J.C., Bremmer-Bout, M., Janson, A.A., Ginjaar, I.B., Baas, R, den Dunnen, J.T. and van Ommen, G.J. (2001) Antisense-rnduced exon skipping restores dystrophin expression in DMD patient derived muscle cells. Hum. Mol. Genet., 10: 1547-1554.

Vecil, G.G. and Lang, F.F. (2003) Clinical trials of adenoviruses in brain tumors: a review of Ad-p53 and oncolytic adenoviruses. J. NeurooncoL, 65: 237-246. Verma, I.M. (2000) A tumultuous year for gene therapy. Mol. Ther., 2: 415-416. Wang, B., Li, J. and Xiao, X. (2000) Adeno-associated virus vector carrying human minidystrophin genes effectively ameliorates muscular dystrophy in mdx mouse model. Proc. Natl. Acad. Sci. USA, 97: 13714-13719. Wu, X. and Burgess, S.M. (2004) Integration target site selection for retroviruses and transposable elements. Cell Mol. Life Sci., 61: 2588-2596.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

C H A P T E R

8 Gene Therapy for Parkinson's Disease Patricia A. Lawlor

Abstract: Parkinson's disease (PD) is a debilitating, neurodegenerative disorder resulting from loss of dopaminergic neurons in the substantia nigra pars compacta leading to distressing motor symptoms. The current standard pharmacological treatment for PD is direct replacement of dopamine by treatment with its precursor, levodopa (L-dopa). However, this does not significantly alter disease progression and becomes less efficacious over time. As a result, a significant amount of PD research over the last decade has been aimed at development of novel PD therapeutics such as gene therapy. The multi-factorial aetiology of sporadic PD and the expanding list of genetic mutations linked to familial forms of the disease means that it is unlikely any single gene will cure the disease. As a result, at least three distinct gene transfer strategies are currently being pursued — transfer of genes for enzymes involved in dopamine production, transfer of genes for growth factors involved in dopaminergic cell survival and regeneration, and use of genes to reset neuronal circuitry by switching cellular phenotype. The pre-clinical data and merits of each of these three strategies are discussed along with other potential PD gene therapy targets such as manipulation of the ubiquitin-proteasome system and use of anti-apoptotic factors. Also highlighted are remaining concerns that may impede transfer of gene therapy technology to the clinic as a PD treatment. Keywords: Parkinson's disease; glial-derived neurotrophic factor; glutamic acid decarboxylase; tyrosine hydroxylase; apoptosis; substantia nigra; striatum; adeno-associated virus

L

The current standard pharmacological treatment for this disease focuses on the direct replacement of dopamine by oral administration of its precursor, levodopa (L-dopa) (Cotzias et al., 1967; Kordower and Goetz, 1999). However, prolonged systemic use of L-dopa can itself lead to motor side-effects (Marsden and Parkes, 1977), and use of this drug does not alter disease progression and may itself contribute to the ongoing pathology (Fahn, 1996; Fahn et al., 2004). Certain features of PD make this disease particularly suited to a gene therapy-based approach to treatment. The major pathology is confined to a compact, localised neuronal population therefore global gene transfer is not required, and the anatomy of the basal ganglia circuitry means there are well-defined, multiple sites

INTRODUCTION

Parkinson's disease (PD) is arguably the most promising neurological target for clinical gene therapy. This debilitating neurodegenerative disorder affects at least 1% of the population over the age of 65 (Zigmond et al., 2002) and is characterised pathologically by the loss of dopaminergic neurons within the substantia nigra (reviewed in Lewis et al., 2003) and appearance of intracellular protein inclusions (Lewy bodies) within remaining cells. The depletion of dopamine results in disturbances in basal ganglia circuitry, leading to the characteristic and distressing motor abnormalities of rigidity, bradykinesia, tremor and gait disturbance.

Gene Therapy of the Central Nervous System: From Bench to

91

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

92

8. GENE THERAPY FOR PARKINSON'S DISEASE

for gene transfer. The identification of an active and preventable cell-death process occurring within the substantia nigra (SN) and the progression of the disease over a long time frame also suggests that this disease represents an achievable target for neurological gene therapy. Indeed, the world's first gene therapy protocol for PD was approved by the US Food and Drug Administration in August 2002, with the first patient treated in August 2003. However, the multifactorial aetiology of sporadic PD and the expanding list of genetic mutations linked to familial forms of the disease mean there is no single gene therapy treatment identified as the cure. As a result, there are several different gene-transfer strategies currently being pursued in an effort to combat this debilitating disease. The scope of this chapter is to outline the progress of research on these multiple gene therapy strategies and to highlight outstanding issues that must be resolved if gene therapy for PD is to proceed to the clinic as a viable treatment option. IL

PARKINSON^S DISEASE

The primary pathological hallmark of PD is death of dopaminergic neurons within the substantia nigra pars compacta (SNpc) and loss of their axonal projections to the striatum. This depletion of dopaminergic tone in the nigrostriatal projection causes a cascade of modifications to the functioning of the basal ganglia circuitry, resulting in the characteristic motor symptoms observed in PD — tremor, muscle rigidity and bradykinesia (Dunnett and Bjorklund, 1999). This cell-death process can take place over a time period of 20 or more years and the clinical symptoms of PD are not exhibited until a loss of -^-SO^o of striatal dopamine has occurred, representing a loss of --50% of the dopaminergic cell bodies within the SNpc (Feamley and Lees, 1991; Homykiewicz, 2001), and by the end stage of the disease as many as 90% of the dopaminergic neurons may have been lost. There is a normal age-dependent cell loss from the SN, with 4.5% of these cells lost with each decade of life (Anglade et al., 1997a) increasing up to 45% cell loss per decade in PD (Feamley and Lees, 1991), with a significant proportion of this cell death occurring before clinical diagnosis of the disease. Remaining neurons in the SNpc often contain cytoplasmic inclusions known as Lewy bodies — spherical cellular entities that contain ubiquitin (Andersen, 2000) and other proteins such as a-synuclein (Spillantini et al., 1997). Most (90-95%) cases of PD are idiopathic and several risk factors for PD have been identified, including

exposure to well water (Rajput et al., 1987; Smargiassi et al., 1998), pesticides (Butterfield et al., 1993; Sherer et a l , 2002), herbicides (Hubble et a l , 1993) and industrial chemicals (Chaturvedi et al., 1995); however, no specific toxin has consistently been found in the brain of PD patients. The remaining 5-10% of PD cases have a genetic basis with mutations in several genes putatively associated with familial PD having been identified in the last decade. These include mutations in a-synuclein (Polymeropoulos et al., 1997; Kruger et al., 1998), parkin (Kitada et al., 1998), ubiquitin carboxy terminal hydrolase (UCH-Ll) (Leroy et al., 1998a, b), DJ-I (Bonifati et al., 2003) and PINKl (Valente et a l , 2004). The exact mechanisms by which these mutations result in PD are still under investigation. Mutations in a-synuclein lead to accumulation of misfolded, toxic forms of this protein; mutations in parkin and UCHL-1 result in aberrations in protein degradation by the ubiquitin-proteasome system (UPS), leading to accumulation of toxic protein species such as a-synuclein (Shimura et al., 2001) and PaelR (Imai et al., 2001). Exact cellular roles of DJ-1 and PINK-1 have not been elucidated but evidence suggests they play a role in neuroprotection, with mutations in these genes affecting oxidative stress and mitochondrial function (Abou-Sleiman et al., 2004; Greenamyre and Hastings, 2004; Shen and Cookson, 2004). It is likely that several divergent (yet tightly linked) pathways — oxidative stress, mitochondrial dysfunction and protein mishandling — are initiated by a combination of environmental and genetic risk factors to converge and interact, resulting in the common outcome of dopaminergic neuronal dysfunction, atrophy and ultimately death. IIL C U R R E N T PHARMACOLOGICAL THERAPY VS> GENE THERAPY FOR P D Traditional pharmacological treatments for PD focus on direct replacement of dopamine by administration of the dopamine precursor L-dopa in combination with agents that act to prolong the action of dopamine at the synapse or prevent its breakdown. However, treatment with L-dopa becomes less efficacious over time and can lead to debilitating side effects such as dyskinesia, hallucinations and disorientation. Direct augmentation of dopamine in this way does nothing to halt the progress of the disease, and addition of exogenous dopamine to a compromised system may contribute to the ongoing pathology (Fahn, 1996). In addition, as the disease progresses and nigrostriatal innervation is lost, fewer striatal terminals are available to metabolise

II. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

STRATEGIES FOR GENE THERAPY INTERVENTION IN PD

the exogenously supplied L-dopa to dopamine. The side effects observed writh prolonged L-dopa treatment can be attributed to the availability of dopamine to neurons other then the nigrostriatal pathv^ay, and the fluctuating levels of dopamine produced foUov^ing oral administration of L-dopa (Chase et al., 1993; Obeso et al., 1994; Mandel et al., 1999). Of all the neurodegenerative disorders, PD is particularly amenable to treatment using a gene-therapy strategy. The observation that the symptoms of PD only become apparent w^hen an estimated 50% of nigral neurons have already been lost suggests the presence of a reserve and a capacity for compensation within the nigrostriatal system (Lloyd, 1977; Zigmond, 1997; Bezard et al., 2001a, b; Zigmond et al., 2002). It is possible that even a minimal increase in dopamine production via gene therapy might be sufficient for amelioration of symptoms. The initial pathology is confined to a v^ell-characterised subtype of neuron located writhin a small compact area in the brain (Moore, 2003) so a single injection of a high-titre viral vector could transduce a large proportion of the cells w^ithin the human SN, leading to an impact on disease phenotype. There are potentially three sites for gene transfer in PD, depending on the gene to be transferred — the dopaminergic cell bodies in the SNpc, their terminals v^ithin the striatum, or other nuclei such as the subthalamic nucleus (STN), v\rhich become overactive in PD. The disease is progressive, v^orsening over 10-20 years (Zigmond, 1997; Andersen, 2001) and so offers a large window of time for undertaking a therapeutic intervention. IV,

ANIMAL MODELS OF PD

PD is a well-characterised disease with many animal models available for evaluation of the benefits of any gene therapy strategy and as a result a wealth of pre-clinical data exists. Despite the fact that the initiating factors in prompting the cell death observed in PD have not been fully elucidated, the pathways implicated in this cell death, and its effects on the functioning and innervation of the nigrostriatal pathway, can still be mimicked in animal models of the disease. For instance, the neurotoxin l-methyl-4-phenyl-l,2,3, 6-tetrahydropyridine (MPTP) (Langston et al., 1983, 1999), the pesticide rotenone (Betarbet et al., 2000) and the toxin 6-hydroxydopamine (6-OHDA) (Ungerstedt, 1968; Sauer and Oertel, 1994) can evoke PD-like symptoms and neuropathological changes in rodents and primates, providing insights into oxidative stress.

93

excitotoxicity, necrosis and apoptotic cascades within the SN (reviewed in Dauer and Przedborski, 2003). In the last few years, the identification of genes for PD (mutations in a-s}muclein, parkin, etc.) has not only presented novel targets for development of treatments, but paved the way for generation of genetic models of PD. Gene-transfer technology can be used not only for the development of therapeutics but to generate more accurate animal models of PD such as localised overexpression of a-synuclein in rats (Kirik et al., 2002b) and primates (Kirik et al., 2003) or overexpression of CDCrell (Dong et al., 2003). The realisation that several of the gene mutations linked to PD involve components of the UPS has also led to the development of PD models based on proteasomal inhibition such as the use of lactacystin (McNaught et a l , 2002; Miwa et al, 2005) and epoxomicin (McNaught et al., 2004). There are a large number of standardised behavioural tests defined for both rodent and primate models of PD that can be used to gauge the effects of any gene therapy strategy (Ungerstedt and Arbuthnott, 1970; Barneoud et a l , 1995, 2000; Borlongan and Sanberg, 1995; Olsson et al., 1995; Chang et al., 1999; Henderson et al., 1999). However, the relevance of measuring behavioural parameters such as amphetamine- and apomorphine-induced rotation to gauge the efficacy of potential human PD treatments remains unclear. Screening of potential PD gene therapy treatments requires comprehensive studies encompassing use of clinically relevant animal models (mimicking both protein aggregation and prolonged cell-degeneration processes accurately) coupled with measurement of meaningful behavioural indices (e.g. spontaneous forepaw use or forepaw-adjusting steps as well as drug-induced rotation). Any attenuation of behavioural deficits needs to be correlated with parallel quantitative measures of dopamine metabolism and examination of nigrostriatal innervation. V-

STRATEGIES FOR GENE THERAPY INTERVENTION IN PD

The multi-factorial aetiology of PD means that intervention in the disease process via gene therapy is possible on many different levels, with several distinct strategies currently being investigated. First, transfer of genes involved in dopamine biosynthesis could help with the immediate motor symptoms of the disease by sustained local production of dopamine in the same way as systemic administration of oral L-dopa. Second, transfer of growth factor genes such

II. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

94

8. GENE THERAPY FOR PARKINSON'S DISEASE

as glial cell line-derived neurotrophic factor (GDNF) could restore atrophic neurons to a normal state and promote reinnervation of the damaged nigrostriatal system (neurorestoration) and prevent further dopaminergic cell death (neuroprotection). Third, transfer of genes involved in inhibitory neurotransmission could be used to dampen the activity of brain nuclei that become overactive in PD. These three main approaches have currently reached pre-clinical primate studies or Phase I human clinical trials, based largely on demonstration of efficacy in rodent and primate toxin models of PD such as 6-OHDA and MPTP. However, the identification of genetic components in PD pathogenesis has resulted not only in development of new PD animal models, but in identification of novel targets for PD gene therapy (e.g. parkin). Treatment options for a chronic neurodegenerative disease need to be long lasting or permanent, making PD particularly suited to treatment with viral vectors, where a single application of vector can result in prolonged, stable expression of the chosen transgene. Viral vector systems developed for use in the brain include adeno-associated virus (AAV) (Kaplitt et al., 1994; Klein et al., 1998), adenovirus (Ad) (AkU et al., 1993; Davidson et al., 1993), lentivirus (Naldini et al., 1996; Blomer et al., 1997) and herpes simplex virus (HSV) (During et al., 1994), all of which transduce neurons (postmitotic cells) efficiently. Each of these vector systems has advantages and disadvantages related to genome size, induction of immune response, transgene shut-off and integration into the host genome (Lu, 2004; Chen et al., 2005). Despite their limited genome size, the latest generation AAV vectors have emerged as a safe option for use in neurological gene therapy, with prolonged transgene expression and no significant immune response elicited following transduction of brain tissue (Mastakov et al., 2002). Of aU the AAV serotypes, AAV2 is the most widely used and evaluated in pre-clinical PD studies, its tropism for DA cells of the SN (Burger et al., 2004; Patema et al., 2004) making it ideal for use in PD gene therapy VL BIOCHEMICAL AUGMENTATION: GENE TRANSFER OF ENZYMES INVOLVED IN DOPAMINE SYNTHESIS Many of the side effects caused by the fluctuating levels of L-dopa obtained after oral administration can be abolished by the use of continuous intravenous infusion of L-dopa (Chase et al., 1989; Schuh and Bennett, 1993; Obeso et al., 1994). This suggests that

attaining steady-state levels of the drug by continual infusion may be of benefit clinically, with a reduction in the occurrence of phenomena such as 'wearing off or 'on-off motor fluctuations related to the pulsatile nature of dopamine production after oral L-dopa. However, it is not practical for patients to receive continuous intravenous infusions and this method of administration still does not overcome the problem of unwanted mental side effects caused by the availability of L-dopa to dopaminergic regions of the brain other than the nigrostriatal projection. Targeted delivery of L-dopa in continuous, physiological quantities specifically to the striatum, as could be achieved by localised gene transfer, would be an advance in PD treatment. A.

Dopamine Synthesis Pathway

The dopamine synthesis and storage pathway involves several enzymes and co-factors, any one of which could be manipulated genetically to yield increased dopamine levels. The rate-limiting enzyme in dopamine production is tyrosine hydroxylase (TH), which converts the amino acid tyrosine to L-dopa. L-dopa is then metabolised to dopamine by aromatic amino acid decarboxylase (AADC) (Nagatsu et al., 1964). Another factor that influences this pathway is the essential TH co-factor 6-tetrahydrabiopterin (BH4), the level of which is limited by availability of the enzyme GTP-cyclohydrolase I (GTPCHI) (Nagatsu et al., 1984; Nagatsu and Ichinose, 1999). Modification of the levels of any of these three key enzymes (TH, AADC or GTPCHI) through gene therapy could significantly impact on striatal dopamine levels and many studies have been published on the use of genes encoding these enzymes in both rodent and primate PD models. B.

Tyrosine Hydroxylase

The original studies in this area suggested beneficial increases in dopamine production could be achieved by use of ex vivo gene therapy approaches. Cells engineered to produce L-dopa by introduction of the TH gene were implanted into lesioned animals and correlated with increases in L-dopa levels (Horellou et a l , 1990a) and some recovery from behavioural deficits (Horellou et al., 1990b; Fisher et al., 1991). However, this indirect approach to increasing striatal dopamine levels means that there are two significant technical hurdles that must be overcome — cells must not only be modified to produce L-dopa but must also survive

11. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

BIOCHEMICAL AUGMENTATION: GENE TRANSFER OF ENZYMES INVOLVED IN DOPAMINE SYNTHESIS

transplantation into the host animal and demonstrate long-term stability and integration into host circuitry. Use of the more direct in vivo approach (injection of viral vectors encoding for TH into the denervated striatum) removes one of these technical variables. Initial experiments demonstrated that direct transfer of the TH gene to the striatum of parkinsonian rats using either AAV (Kaplitt et al., 1994), herpes simplex virus (HSV) (During et al., 1994) or adenoviral vectors (Horellou et al., 1994) resulted in long-term expression of TH and some phenotypic correction of behavioural deficits in these animals. Howrever, the use of the partial lesion model in some of these studies meant the presence of residual endogenous dopamine synthesis capacity, and as such determination of the amount of dopamine produced as the direct result of gene transfer v^as not possible, and the use of drug-induced rotational behaviour as the main indicator of transgene function may have given limited information. Nevertheless, these pioneering studies provided proof-of-principle that increasing TH levels in the striatum via gene transfer could lead to some behavioural recovery, but further studies using alternative behavioural tests and concurrent measurement of biochemical and histopathological markers was needed. Also the optimal gene combination that w^ould result in physiologically relevant levels of dopamine required further investigation. C.

GTP-Cyclohydrolase I

Reports on use of ex vivo gene therapy approaches to modify cells genetically prior to grafting in PD animal models suggested that co-transduction of target cells w^ith both TH and GTPCHI resulted in greater L-dopa production (Bencsics et al., 1996; Leff et al., 1998) than use of TH alone. Furthermore, direct striatal AAV-TH transduction v^ith addition of exogenous BH4 or co-transduction v^ith AAV-GTPCHI also resulted in increased L-dopa production (Mandel et al., 1998) suggesting that significant amounts of L-dopa are not produced foUow^ing transfer of the TH gene are not produced unless BH4 levels are also increased. Use of regulatable adenoviral vectors (Corti et al., 1999) demonstrated that residual levels of BH4 found in denervated animals may not allov\r production of the maximal amount of dopamine follov\ring TH gene transfer, and this may be achieved only by co-transduction w^ith GTPCHI or application of exogenous BH4. D-

AADC

Gene transfer of AADC to increase dopamine levels has been investigated in both rat and primate PD

95

models (Leff et al., 1999; Bankiewicz et al., 2000; Sanchez-Pernaute et a l , 2001). Transduction of striatal tissue v^rith AAV-AADC combined v^ith an exogenous dose of L-dopa in dopaminergically denervated animals resulted in the phenotypic correction of motor deficits. This was proposed to be due to increased decarboxylation of L-dopa to dopamine as a result of increased AADC levels, and suggests that gene transfer of AADC by itself may be beneficial in PD treatment by reducing the dose of L-dopa required to give relief from motor abnormalities, thereby avoiding or delaying development of side effects from L-dopa administration. E.

Gene Combinations

Other studies have reported recovery-promoting effects after gene transfer of various combinations of these three key enzymes, using different vector systems and various PD models. Co-transduction of rat striatum with AAV-TH and AAV-AADC resulted in increased dopamine production and greater behavioural recovery than that observed in rats receiving AAV-TH alone (Fan et a l , 1998). Use of this double enzyme combination has also been reported in MPTPtreated monkeys with the use of a bicistronic AAV vector, resulting in increased dopamine production (During et al., 1998). These findings were extended to triple transduction of striatal tissue with three separate AAV vectors expressing genes for TH, AADC and GTPCHI. In these studies in rat (Shen et al., 2000) and primate (Muramatsu et a l , 2002) models of PD, gene transfer of all three enzymes was achieved and resulted in increased behavioural modification above that seen in the double transduction study, suggesting that expression of all three of these factors is necessary to obtain biologically relevant levels of dopamine. However, direct comparison between studies is difficult, given the variety of models and transductions paradigms investigated — for example, the use of bi-cistronic vectors encoding for both TH and AADC compared with injection of two of three separate vectors, one encoding each enzyme. Use of a tri-cistronic lentiviral vector containing all three genes (TH, AADC and GTPCHI) increased striatal dopamine production and led to a reduction in behavioural deficits in 6OHDA-treated rats (Azzouz et al., 2002), and the inclusion of all genes into one vector meant that transduced cells produced all three enzymes in close proximity to each other, overcoming the need to transduce a single neuron with multiple vectors. However, studies that directly compare behavioural outcomes and dopamine

II. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

96

8. GENE THERAPY FOR PARKINSON'S DISEASE

levels following transduction with multiple vectors and bi- or tri-cistronic vectors are needed to determine the relevant importance of each enzyme to increasing dopamine concentration. A recent study using transduction of striatal tissue with AAV vectors containing genes to TH and GTPCHI in 6-OHDA-denervated rats (Kirik et al., 2002a) highlighted the importance of determining levels of transgene and dopamine produced after gene therapy treatment. Results from this study suggest a threshold level of 1.5 pmol L-dopa/mg striatal tissue must be produced in order to give a quantifiable effect on spontaneous and drug-induced behaviours in parkinsonian rats. Although significant recovery was observed in both complete and partially lesioned animals, this recovery was most pronounced in animals with partial preservation of striatal dopaminergic fibres, suggesting that spared fibres play an important role in storage and release of newly synthesised dopamine. In a follow-up study, this treatment regime also led to reversal of L-dopa-induced dyskinesias in L-dopa-primed parkinsonian rats, in addition to reversing 6-OHDA lesion-induced motor deficits (Carlsson et al., 2005). The observation that continuous production of L-dopa within the striatum of vector-treated animals does not worsen, and even reverses, dyskinesias suggests that continuous (as opposed to intermittent) stimulation of DA receptors and not the overall level of L-dopa within striatum is the critical factor in appearance of dyskinesias. R

Vesicular Monoamine Transporter-2 (VMAT2)

One other possible intervention point in the dopamine synthesis and storage pathway is use of gene therapy to modify the levels of the vesicular monoamine transporter-2 (VMAT2). VMAT2 plays a crucial role in storage of dopamine by packaging of the neurotransmitter into synaptic vesicles and regulating vesicle release. VMAT2 has been used in combination with AADC to increase striatal dopamine levels (Lee et al., 1999). In this study, fibroblasts that had been genetically modified to contain both VMAT2 and AADC were transplanted into parkinsonian rats. Following systemic administration of L-dopa, these animals contained higher levels of striatal dopamine, as measured by microdialysis, than those treated with AADC-modified cells alone. The impact of this treatment on restoration of behavioural deficits was not measured; however, this study underlines the fact that while increased dopamine production is required (perhaps via exogenous L-dopa), the ability to store

and release this neurotransmitter in a gradual fashion is also bene-ficial. Indeed, direct comparison of a tricistronic vector expressing TH, GTPCHI and AADC with a vector expressing VMAT2 in addition to TH, GTPCHI and AADC showed that the quadruple gene combination led to a more pronounced reduction in apomorphine-reduced rotation than the three-gene vector. This was reflected in the striatal biochemistry of transduced rats, with only the four-gene vector resulting in near-normal extracellular levels of dopamine and metabolites, and exhibiting high potassiumdependent release of dopamine (Sun et al., 2004). However, one point that should not be overlooked following transduction of striatal tissue with the aim of increasing DA production and release is the type of cell transduced — in this case striatal y-amino butyric acid (GABA)ergic neurons were transduced. It is unclear what effect this would have on the long-term functioning of the striatum and its projection areas. G, Dopamine Augmentation via Gene Transfer in Humans A Phase I clinical trial for gene transfer of human AADC (hAADC) is underway — in this dose escalation study, AAV-hAADC will be infused into the striatum of 15 PD patients (five patients at each of three doses) (Avigen, 2004). In addition to assessing the effects of AADC expression on clinical symptoms of PD and determining the dose of AAV-hAADC needed to achieve this, the trial will assess whether increasing AADC levels enhances the effectiveness of L-dopa, as reducing the requirement could reduce the side effects experienced after L-dopa treatment. H.

Summary

Intervention by gene therapy in the dopamine synthesis and storage pathway offers several possible strategies with the common aim of increasing dopamine production and release in a site-specific and continuous manner. Gene therapy to augment current pharmacological treatments (e.g. use of vectors for AADC a n d / o r VMAT2 to enhance decarboxylation and storage of exogenously supplied L-dopa) would be useful in early stages of the disease where some striatal dopaminergic innervation remains intact. In other cases where therapeutic efficacy of L-dopa has already been lost due to irreplaceable loss of nigrostriatal afferents, a gene therapy strategy involving replacement of multiple enzymes involved in dopamine synthesis and release may be beneficial. However, whatever the

11. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

GENE TRANSFER OF GROWTH FACTORS

approach used to increase dopamine levels, the fact remains that this strategy is palliative only — correcting the motor symptoms but not influencing disease pathogenesis.

VII. GENE TRANSFER OF GROWTH FACTORS The second main strategy being investigated in the pursuit of a gene therapy treatment for PD is the use of growth factors to promote recovery of the damaged nigrostriatal system. Various neurotrophic factors with putative effects on the nigrostriatal dopamine system have been evaluated for their therapeutic potential in PD. These include basic fibroblast growth factor (bFGF) (Ferrari et al., 1989), epidermal growth factor (EGF) (Knusel et al., 1990; Park and Mytilineou, 1992), brain-derived neurotrophic factor (BDNF) (Hyman et al., 1991) and sonic hedgehog (Miao et a l , 1997). These factors displayed promising attributes in dopaminergic in vitro systems such as enhancement of neurite outgrowth and protection from insults such as neurotoxin l-methyl-4-phenylpyridinium (MPP+), but have produced variable results when translated to animal models of PD (Ventrella, 1993; Pearce et al., 1996, 1999; Svendsen et a l , 1996; Zeng et al., 1996; Klein et al., 1999; Dass et al., 2002; Tsuboi and Shults, 2002; Hurtado-Lorenzo et al., 2004). A. The Effects of GDNF on the Nigrostriatal System Arguably, the most intensively studied trophic factor for treatment of PD to date is GDNF. Initial studies identified this molecule as a potent trophic factor for dopaminergic neurons of the SN following in vitro studies examining survival of mid-brain cultures (Lin et al., 1993). The in vivo potential of GDNF as a PD therapeutic agent was established following studies utilising the 6-OHDA rat model and infusion of GDNF recombinant protein, suggesting the dual abilities of GDNF to not only protect cells from death, but also to promote regeneration of an already damaged system. Injection of 6-OHDA into the striatal terminals causes axonal retraction and death of DA cell bodies in the SN to occur over a period of several weeks (Sauer and Oertel, 1994) offering the opportunity to study both the protection of degenerating neurons by GDNF during the acute cell death phase, and the regenerative and recovery-promoting effects of GDNF during the chronic phase of the lesion when spared dopaminergic

97

neurons persist in a dysfunctional state. Cell death was prevented by infusion of GDNF protein in the region of the SN throughout this time period (Sauer et al., 1995) and the neurorestorative power of this molecule was demonstrated in a study in which stably lesioned animals displaying behavioural deficits received a single infusion of GDNF into the SN, resulting in attenuation of apomorphine-induced rotation and an increase in dopaminergic cell number (Bowenkamp et al., 1995). Reports on the use of various GDNF protein infusion paradigms in this partial lesion model have confirmed both the neuroprotective (Rosenblad et al., 1999; Kirik et al., 2000a) and restorative powers (Winkler et a l , 1996; Rosenblad et al., 1998; Kirik et al., 2001) of GDNF Recovery-promoting effects of GDNF protein have also been demonstrated in primate PD models. Infusion of GDNF into the SN of rhesus monkeys stably lesioned with MPTP resulted in improvement of behavioural deficits (Gash et al., 1996). In addition, cells of the SN increased in mean cell size and density of fibres, and mid-brain dopamine levels increased (Gash et al., 1996). However, a later study observed that discontinuation of GDNF administration resulted in the slow decline of animals back to baseline deficit levels (Zhang et al., 1997). A more recent study on the action of GDNF in primate PD models examined the effect of prolonged infusion of GDNF directly into the striatum (rather than the SN) 3 months after MPTP treatment, and found a reduction in parkinsonian symptoms coupled with an increase in both striatal dopamine levels and in the number of cells in the SN expressing TH (Grondin et al., 2002). GDNF also has direct effects on TH and DA levels. In vitro, application of GDNF to mid-brain cultures leads to an increase in both the spontaneous firing rate and in the quantal size of terminal DA release (Pothos et al., 1998), as well as increased excitability of DA neurons mediated by A-type K+ channels (Yang et a l , 2001) and high voltage-activated Ca^^ channels (Wang et al., 2003). In vivo, administration of striatal GDNF to normal rats increases TH phosphorylation in both the striatum and SN, suggesting one effect of GDNp is on DA storage and somatodendritic release (Salvatpre et al., 2004). Overall, the data accumulated on the effects of GDNF protein in both 6-OHDA and MPTP models showed not only that GDNF protects dopaminergic cells from death, but promotes axonal sprouting and regeneration of lesioned neurons, depending on the site and regime of GDNF administration. Use of the factor in primates highlighted the need for an efficient

11. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

98

8. GENE THERAPY FOR PARKINSON'S DISEASE

route of GDNF delivery directly into the striatum (rather than the SN or cerebral ventricles) and suggested that sustained presence of GDNF would be necessary for maintaining any beneficial effects, and as such a viral vector-based gene therapy strategy resulting in prolonged, stable production of GDNF may be more applicable to the human clinical condition than long-term infusion of protein. The production of transgenic GDNF should also be focal as widespread distribution of GDNF to systems other than the nigrostriatal projection after intracerebroventricular (icv) administration has been observed (Lapchak et al., 1997a) and might account for some side effects. B-

Gene Transfer of G D N F

Various viral vector systems have been used to introduce GDNF into brain tissue — adenovirus, lentivirus and AAV — and efficacious effects on both behaviour and nigrostriatal innervation levels have been demonstrated in models designed to assess both the neuroprotective and neurorestorative properties of GDNF. Adenoviral-mediated transfer of GDNF resulted in protection from 6-OHDA-induced cell death following infusion of the vector adjacent to the cell bodies of the SN (Choi-Lundberg et al., 1997) and into the striatum (Bilang-Bleuel et al., 1997; Choi-Lundberg et al., 1998). A direct comparison between the benefits of infusing GDNF-encoding vectors into either the SN or striatum (Connor et al., 1999, 2001) showed that infusion into the striatum (the site of terminal degeneration) resulted in greater prevention of behavioural deficits and greater preservation of striatal innervation. Other studies have provided evidence of the ability of virally delivered GDNF to promote restoration of function and innervation in 6-OHDA-lesioned rats (Lapchak et al., 1997b; Kozlowski et al., 2000). A potential problem with the use of adenoviral vectors is production of an inflammatory response and consequent downregulation of transgene levels over time following injection into the brain. Other groups have reported on the use of AAV- or lentiviral-mediated delivery of GDNF to ensure longterm transgene expression and have demonstrated both neuroprotection (Mandel et al., 1997; Kirik et al., 2000b; Georgievska et al., 2002b; Eslamboli et al., 2003) and restoration of function and innervation in rat and primate models of PD (Wang et al., 2002), with transgenic GDNF still at maximal levels 6 months after gene transfer (Kirik et al., 2000b). Use of lentiviral GDNF gene transfer to both the striatum and SN prevented neurodegeneration in a

primate model of PD (Kordower et al., 2000). An additional effect of transgenic GDNF expression in the striatum of parkinsonian primates was an increase in the number of cells expressing TH endogenously present in the striatum (Palfi et al., 2002), suggesting that one effect of this growth factor is to convert striatal neurons to a dopaminergic phenotype; or this observation could represent an effect of GDNF on neurogenesis and migration of cells from the sub-ventricular zone, an area known to contain neuronal progenitor cells. However, the benefits of long-term striatal GDNF overexpression may not be clear-cut, with reports that expression of TH is down-regulated over time following lentiviral GDNF treatment, and aberrant sprouting of fibres in the globus pallidus and entopeduncular nucleus occurs (Georgievska et al., 2002a; Rosenblad et al., 2003). This down-regulation of TH following lenti-GDNF is time- and dose-dependent (Georgievska et al., 2004), possibly reflecting a compensatory effect in response to continuous GDNF stimulation of DA neurons. Latest work in this area defined the concentration of GDNF, following AAV-GDNF transduction, which would provide anatomical and behavioural protection against intrastriatal 6-OHDA in marmosets, but not affect normal DA neurons (Eslamboli et al., 2005). A high level of GDNF production (14 n g / m g tissue) affected dopamine transmission bilaterally whereas a low level of GDNF production (0.04 n g / m g tissue or three-fold above baseline) provided unilateral protection and correction of behavioural deficits, demonstrating that continuous low-level GDNF production in striatum provides the optimal functional outcome. Interestingly, the bulk of the data on the protective and regenerative effects of GDNF has been generated using neurotoxic lesion models of PD (6-OHDA, MPTP) that are based on oxidative damage. Use of lenti-GDNF in a genetic rat PD model (overexpression of mutant a-synuclein) did not prevent neurodegeneration (Lo Bianco et al., 2004a) suggesting that GDNF treatment cannot modulate cellular toxicity related to abnormal protein accumulation. This finding demonstrates that the ability to show efficacy of a particular therapy in a given model is dependent on the mechanism of action of the toxic insult, and highlights the necessity of developing relevant animal models. C.

Use of G D N F in Humans

These encouraging results in rodent and primate models augured well for the clinical use of this trophic factor. However, the first reported use of icv GDNF protein in a PD patient did not produce the expected

[I. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

GENE THERAPY TO RESET NEURONAL CIRCUITRY

beneficial effects (Kordower et al., 1999). The administration of GDNF via the ventricles did not result in any reduction of parkinsonian symptoms, but did cause several adverse side effects including nausea, loss of appetite, hallucinations and depression. Postmortem examination of brain tissue revealed no evidence for nigrostriatal regeneration. A large, multi-centre, double-blind trial on the use of recombinant GDNF also reported a large number of adverse effects and no impact on parkinsonian symptoms following icv infusion (Nutt et al., 2003), suggesting infusions into ventricles were not an ideal route of delivery and GDNF needs to be applied directly into brain parenchyma to be effective. A further trial on human patients (Gill et al., 2003) used chronic infusion of GDNF protein directly into the caudate putamen, resulting in improvements in motor symptoms, with positron emission tomography (PET) scans showing an increase in dopamine storage (Gill et al., 2003). A follow-up report (Patel et al., 2005) on patients in this Phase I study, after 2 years of continual GDNF infusion, documented improvements in off-medication motor and activities of daily living sub-scores of United Parkinson's Disease Rating Scale (UPDRS), and an increase in fluordopa uptake in whole putamen, maximal closest to cannula tip. No serious side effects or detrimental effects on cognition were reported. This was confirmed by another trial of intraputamenal GDNF in 10 patients with advanced PD (Slevin et al., 2005a) with unilateral administration of protein resulting in significant bilateral effects. However, a Phase II trial using GDNF has produced disappointing results, possibly due to use of lower GDNF doses and different infusion parameters than were used in both successful Phase I trials (Slevin et al., 2005b). D,

Summary

A large body of data on use of both GDNF recombinant protein and virally mediated GDNF gene transfer in animal models of PD demonstrates both the neuroprotective and neurorestorative effects of this growth factor, with intra-striatal GDNF providing protection of both striatal terminals and cell bodies. The beneficial effects of intra-striatal GDNF protein were confirmed by encouraging results in Phase I clinical trials. The less impressive effects of GDNF in a Phase II trial highlights crucial points such as the necessity of the continual presence of the growth factor at optimal concentrations in discrete striatal locations. Pre-clinical GDNF studies have also highlighted the potential problems associated with long-term over expression of

99

this growth factor such as aberrant sprouting, the magnitude of which is dose-dependent (Kirik et al., 2004). In addition, the biological action of the growth factor may not be restricted to directly transduced cells; for example, a study on the use of GDNF for protection of motor neurons resulted in survival of a greater number of cells than were actually transduced (Baumgartner and Shine, 1997). Any proposed human GDNF gene therapy treatment would need to be assessed closely to ensure attainment of optimal GDNF doses — high enough for beneficial effects such as increased striatal dopamine metabolism and improvements in motor symptoms, but low enough to avoid adverse outcomes such as uncontrolled sprouting. VIII.

GENE THERAPY TO RESET N E U R O N A L CIRCUITRY

A third potential strategy for gene therapy treatment of PD involves the use of genes encoding enzymes involved in neurotransmitter production to alter activity of selected brain regions. Loss of the dopaminergic nigrostriatal projections in PD results in profound disturbances in the circuitry of the basal ganglia. One consequence is that the excitatory neurons of the STN become disinhibited and overactive. These cells project to the major inhibitory output nuclei of the basal ganglia, the SN pars reticulata and the internal segment of the globus pallidus. The STN overdriving of these nuclei results in inhibition of the thalamus and downstream motor pathways, and ultimately in the motor disturbances seen in PD. This has led researchers to speculate that silencing of the STN in PD may improve the motor symptoms of the disease. Indeed, electrical inhibition (Limousin et a l , 1995a, b), ablation (Gill and Heywood, 1997; Alvarez et al., 2001; Patel et a l , 2003) or pharmacological silencing (Levy et al., 2001) of the STN have all been reported to control PD symptoms. A recent report (Luo et al., 2002) provided proof-of-principle that gene therapy could also be used to modulate STN activity with infusion of an AAV vector coding for glutamic acid decarboxylase (GAD), the enzyme involved in synthesis of the inhibitory neurotransmitter GAB A, into the overactive STN of parkinsonian rats. Expression of transgenic GAD resulted in increased GABA production levels in the SN pars reticulata following STN stimulation (as determined by microdialysis to measure GABA levels), and electrophysiological recordings from the brains of GAD-treated animals confirmed increased inhibitory responses in these animals. This translated

11. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

100

GENE THERAPY FOR PARKINSON'S DISEASE

to a decrease in behavioural deficits in unilaterally 6-OHDA-lesioned rats, and the phenotypic shift of STN neurons from excitatory to inhibitory following GAD gene transfer also protected against further cell death in the SN. An ongoing study in a primate model of PD confirmed the effectiveness of AAV-GAD in normalising motor symptoms seen in these animals, with no reported toxicity. This novel work is being pursued in the human clinical setting as an alternative to deepbrain stimulation of the STN, with a phase I clinical trial approved by the US Food and Drug Administration currently underway (During et al., 2001; Hsich et al., 2002; Burton et al., 2003). This trial involves a dose-escalation safety study with 12 patients (four patients at each of three doses), and parameters being measured are UPDRS scores, MRI and PET scans. IX-

PREVENTION OF APOPTOSIS

The relative contributions of necrotic (passive) and apoptotic (requiring gene activation) cell death pathways to the degenerative process in PD is still an area of active investigation. Apoptosis can be initiated by a variety of insults to the adult brain and, as such, might have a role in the pathogenesis of PD. Postmortem PD brains have shown evidence of apoptotic cell death in some studies (Mochizuki et al., 1996; Anglade et al., 1997b; Tompkins et al., 1997; Hirsch et al., 1999), although this is controversial and other groups have found no evidence of neuronal apoptosis (Kosel et al., 1997; Banati et al., 1998). The results of these studies rely heavily on the use of the terminal deoxynucleotidyl transferase-mediated dUTP nick end labelling (TUNEL) technique in which the free 3'-ends of DNA generated after endonuclease digestion of DNA during apoptosis are labelled and detected immunohistochemically. The absence of positive TUNEL staining in postmortem tissue does not automatically mean apoptosis is not involved in PD, considering that the cell death occurs over as many as 20 years. Further independent evidence for the occurrence of apoptotic cell death in PD has also been reported with detection of the pro-apoptotic factor caspase-3 in the SN of postmortem PD brain (Hartmann et al., 2000; Tatton, 2000), a key molecular player in programmed cell-death pathways (Andersen, 2001; Tatton et al., 2003). Importantly, the realisation that loss of cells during PD may be the result of an active cell-death programme opens up new avenues of therapeutic strategies for treatment of this disease. If the disease can be detected early enough then the use of gene therapy treatments

that inhibit initiation or propagation of this cell-death cascade may be of benefit in arresting cell death and slowing disease progression. Interference in events upstream of apoptosis such as inhibition of the c-Jun N-terminal kinase (JNK) pathway (Xia et a l , 2001; Kuan and Burke, 2005; Silva et al., 2005) or blockade of pro-apoptotic cascades involving caspases (Eberhardt et al., 2000; Mochizuki et al., 2001; Crocker et al., 2003; Ebert et al., 2005) can lead to prevention of DA cell death. Interference with apoptotic cascades and prevention of cell death in PD may be only half the story — neurons might be saved from death but that is only useful if normal physiological functions of that cell, such as dopamine production, are still intact. For example, use of Ad-XIAP (X-chromosomal inhibitor of apoptosis protein) in vitro protected 6-OHDA-challenged DA neurons from death, but did not prevent functional impairment as cell terminals were still lost (von Coelln et al., 2001); similarly use of Ad-XIAP in vivo protected SN neurons from MPTP-induced death but did not protect nigrostriatal terminals (Eberhardt et al., 2000). There are many distinct apoptotic cascades that can be initiated within the SN by a variety of genetic and environmental insults — a truly neuroprotective strategy that will protect cells regardless of the cause of PD may only become a reality if a common molecular mechanism leading to degeneration is identified. X.

OTHER TARGETS FOR PD GENE THERAPY: THE U P S

Several familial forms of PD are associated with mutations in genes that directly or indirectly influence the UPS, including parkin and UCHLl. The UPS targets misfolded or mutant proteins for degradation by the 26S proteosome by covalently linking multiple ubiquitin molecules to the target (McNaught et al., 2001). Parkin is an E3 ligase that mediates ubiquitination of damaged proteins and so facilitates their removal by the proteasome. Loss of parkin activity has been postulated to lead to the build-up of protein substrates such as a-synuclein (Shimura et al., 2001) and PaelR (Imai et al., 2001), leading to speculation that enhancement of parkin function would be beneficial in PD. Indeed, parkin expression suppressed PD-like symptoms in both PaelR (Yang et al., 2003) and a-synuclein (Haywood and Staveley, 2004) Drosophila models, and lentiviral delivery of wild-type parkin prevented DA cell degeneration in an a-synuclein rat model of PD (Lo Bianco et al., 2004b), confirming that expression

11. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

101

SUMMARY

of parkin (or other factors that promote protein breakdown or prevent its build-up) is an additional strategy to be pursued in PD gene therapy. XL

REMAINING HURDLES IN GENE THERAPY FOR P D

Although PD is considered by many researchers to be the closest neurological target for treatment via gene therapy, there are still several outstanding research issues that need to be resolved before this becomes a mainstream treatment option. These include development of improved animal models, optimisation of systems for control of transgene expression and implementation of standards to ensure that manufacture of vector preparations for human use is safe. Identification of suitable agents for gene therapy treatment of PD in humans relies upon testing of putative therapeutic agents in animal models that mimic the human condition. At this stage, it is unclear how relevant to human PD the models that make use of toxins (MPTP, 6-OHDA and rotenone) are. Development of models that better recreate the prolonged degeneration and secondary pathologies observed in PD would hasten transfer of successful gene therapy technology to the clinic. Identification of gene mutations associated with PD has led not only to new gene therapy strategies being identified, but also to development of more relevant animal models (e.g. overexpression of a-synuclein in the SN) (Kirik et al., 2002b, 2003). Problems such as regulation of transgene expression also need to be resolved. Uncontrolled transgene expression of enzymes involved in dopamine synthesis could result in side effects, in much the same way as continual excessive dopamine stimulation results in side effects in pharmacological treatment paradigms, or unregulated growth factor expression may lead to uncontrolled axonal outgrowth in brain. Regulatable gene expression systems are being developed (Gossen et al., 1995; Fitzsimons et al., 2001, 2002; Vigna et al., 2002) but remain leaky or poorly inducible in vivo and would require prolonged use of drugs like tetracycline, the long-term effects of which are unknown. If the introduced gene causes unforeseen problems, there is currently no reliable method for excising the transgene from brain tissue, short of ablating transduced cells. In the past few years, issues regarding manufacture and quality control of viral vectors have been largely addressed with the introduction of efficient and reproducible packaging systems for vectors such

as recombinant adeno-associated viral vector (rAAV) that eliminate the need for use of helper viruses and result in no detectable contamination of stocks with wild-type virus (Cao et al., 2000; During et al., 2003). Development of affinity-based purification methods enables production of large amounts of high titre, pure vector stocks (Clark et al., 1999). There are currently three Phase I human clinical trials underway utilising AAV for gene transfer — the two PD trials previously mentioned, along with a Canavan disease trial (Janson et al., 2002) — with no reported safety concerns arising from use of these AAV vectors. Once a gene therapy strategy has been selected for use in humans, careful trial design must be undertaken to ensure accurate measurement both of appropriate clinical outcomes and output of the vectors. The selection of patients could include both early- and late-stage PD patients, depending on the specific gene therapy being evaluated. Endpoints to be measured should include clinical assessment of motor symptoms and imaging — using PET to assess the impact of treatment on striatal dopamine metabolism and brain glucose utilisation. Ideally, some non-invasive measure of vector output would also be needed such as concentration of L-dopa produced or amount of dopamine metabolised, and correlated to the clinical data. Gene therapy for PD based on supplementation of dopamine production, prevention of cell death or prompting reinnervation of the nigrostriatal system is more likely to succeed in patients with a moderate number of nigrostriatal afferents remaining. Unfortunately, definitive diagnosis of PD occurs only after motor symptoms become apparent, meaning a large number of SN cells have already been lost. Successful gene therapy for PD will probably go hand-in-hand with improvements in imaging techniques leading to earlier diagnosis and the ability to implement neuroprotective strategies. XIL

SUMMARY

PD is a complex neurodegenerative disorder caused by loss of cells within the circuitry of the basal ganglia, resulting in debilitating motor symptoms. The exact cause of dopaminergic cell loss remains to be elucidated but probably involves a combination of both genetic and environmental factors. Current mainstay treatments for PD, such as administration of L-dopa, treat the motor symptoms but do little to alter the ongoing pathology. As a result, alternative, long-term treatment strategies such as gene therapy are being pursued.

11. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

102

8. GENE THERAPY FOR PARKINSON'S DISEASE

Gene transfer to alter the course of PD is possible on many different levels, and at least three distinct strategies are currently under evaluation in clinical and pre-clinical studies: gene transfer to increase local dopamine production in a physiological manner; gene transfer of growth factors to rejuvenate the damaged nigrostriatal pathway and prevent nigral cell death and gene transfer to reset altered neuronal circuitry. These genetic manipulations could be used alone or to augment current pharmacological treatments (e.g. use of AADC gene transfer to enhance metabolism of exogenously supplied L-dopa). Isolation of gene mutations linked to PD has resulted in the identification of new targets for gene therapy treatment such as parkin expression. Gene transfer could also be used to stimulate neurogenesis and to facilitate neuronal survival during cell transplantation procedures in neurodegenerative diseases. Viral vectors systems capable of efficiently transducing neurons, such as AAV, have been developed, along with protocols for manufacture of pure vector stocks suitable for use in human brain. However, problems regarding regulation of transgene expression and, if required, excision of transgene from transduced cells or other rescue procedures need to be addressed if gene therapy for PD is to become a viable mainstream treatment option. As our knowledge of the molecular basis of this devastating disease expands, further options for gene therapy of this disease will doubtless present themselves. References Abou-Sleiman, P.M., Healy, D.G. and Wood, N.W. (2004) Causes of Parkinson's disease: genetics of DJ-1. Cell Tissue Res., 318: 185-188. Akli, S., Caillaud, C , Vigne, E., Stratford-Perricaudet, L.D., Poenaru, L., Perricaudet, M., Kahn, A. and Peschanski, M.R. (1993) Transfer of a foreign gene into the brain using adenovirus vectors. Nat. Genet., 3: 224-228. Alvarez, L., Macias, R., Guridi, J., Lopez, G., Alvarez, E., Maragoto, C , Teijeiro, J., Torres, A., Pavon, N., Rodriguez-Oroz, M.C, Ochoa, L., Hetherington, H., Juncos, J., DeLong, M.R. and Obeso, J.A. (2001) Dorsal subthalamotomy for Parkinson's disease. Mov. Disord., 16: 72-78. Andersen, J.K. (2000) What causes the build-up of ubiquitin-containing inclusions in Parkinson's disease? Mech. Ageing Dev., 118: 15-22. Andersen, J.K. (2001) Does neuronal loss in Parkinson's disease involve programmed cell death? Bioessays, 23: 640-646. Anglade, P, Vyas, S., Hirsch, E.G. and Agid, Y. (1997a) Apoptosis in dopaminergic neurons of the human substantia nigra during normal aging. Histol. Histopathol, 12: 603-610. Anglade, P, Vyas, S., Javoy-Agid, E, Herrero, M.T., Michel, P P , Marquez, J., Mouatt-Prigent, A., Ruberg, M., Hirsch, E.G. and

Agid, Y. (1997b) Apoptosis and autophagy in nigral neurons of patients with Parkinson's disease. Histol. Histopathol., 12: 25-31. Avigen (2004) Avigen's AV201 for Parkinson's disease begins clinical trials, www.avigen.com. Azzouz, M., Martin-Rendon, E., Barber, R.D., Mitrophanous, K.A., Carter, E.E., Rohll, J.B., Kingsman, S.M., Kingsman, A.J. and Mazarakis ND (2002) 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., 22: 10302-10312. Banati, R.B., Daniel, S.E. and Blunt, S.B. (1998) Glial pathology but absence of apoptotic nigral neurons in long-standing Parkinson's disease. Mov. Disord. 13: 221-227. Bankiewicz, K.S., Eberling, J.L., Kohutnicka, M., Jagust, W , Pivirotto. P., Bringas, J., Cunningham, J., Budinger, T.F. and HarveyWhite, J (2000) 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., 164: 2-14. Bameoud, P, Descombris, E., Aubin, N. and Abrous, D.N. (2000) Evaluation of simple and complex sensorimotor behaviours in rats with a partial lesion of the dopaminergic nigrostriatal system. Eur. J. Neurosci., 12: 322-336. Bameoud, P., Parmentier, S., Mazadier, M., Miquet, J.M., Boireau, A., Dubedat, P. and Blanchard, J.C. (1995) Effects of complete and partial lesions of the dopaminergic mesotelencephalic system on skilled forelimb use in the rat. Neuroscience, 67: 837-848. Baumgartner, B.J. and Shine, H.D. (1997) Targeted transduction of CNS neurons with adenoviral vectors carrying neurotrophic factor genes confers neuroprotection that exceeds the transduced population. J. Neurosci., 17: 6504-6511. Bencsics, C , Wachtel, S.R., Milstien, S., Hatakeyama, K., Becker, J.B. and Kang, U.J. (1996) Double transduction with GTP cyclohydrolase I and tyrosine hydroxylase is necessary for spontaneous s)mthesis of L-DOPA by primary fibroblasts. J. Neurosci., 16: 4449^456. Betarbet, R., Sherer, T.B., MacKenzie, G., Garcia-Osuna, M., Panov, A.V. and Greenamyre, J.T. (2000) Chronic systemic pesticide exposure reproduces features of Parkinson's disease. Nat. Neurosci., 3:1301-1306. Bezard, E., Grossman, A.R., Gross, C.E. and Brotchie, J.M. (2001a) Structures outside the basal ganglia may compensate for dopamine loss in the presymptomatic stages of Parkinson's disease. Faseb J., 15: 1092-1094. Bezard, E., Dovero, S., Prunier, C , Ravenscroft, P., Chalon, S., Guilloteau, D., Grossman, A.R., Bioulac, B., Brotchie, J.M. and Gross, C.E. (2001b) Relationship between the appearance of symptoms and the level of nigrostriatal degeneration in a progressive 1methyl-4-phenyl-l,2,3,6-tetrahydropyridine-lesioned macaque model of Parkinson's disease. J. Neurosci., 21: 6853-6861. Bilang-Bleuel, A., Revah, E, Colin, P, Locquet, I., Robert, J.J., Mallet, J. and Horellou, P. (1997) Intrastriatal injection of an adenoviral vector expressing gUal-ceU-line-derived neurotrophic factor prevents dopaminergic neuron degeneration and behavioral impairment in a rat model of Parkinson disease. Proc. Natl. Acad. Sci. USA, 94: 8818-8823. Blomer, U., Naldini, L., Kafri, T, Trono, D., Verma, I.M. and Gage, EH. (1997) Highly efficient and sustained gene transfer in adult neurons with a lentivirus vector. J. Virol., 71: 6641-6649. Bonifati, V., Rizzu, P., van Baren, M.J., Schaap, O., Breedveld, G.J., Krieger, E., Dekker, M.C, Squitieri, R, Ibanez, P., Joosse, M., van

II. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

103

SUMMARY Dongen, J. W., Vanacore, N., van Swieten, J.C, Brice, A., Meco, G., van Duijn, CM., Oostra, B.A. and Heutink, P. (2003) Mutations in the DJ-1 gene associated with autosomal recessive early-onset parkinsonism. Science, 299: 256-259. Borlongan, C.V. and Sanberg, P.R. (1995) Elevated body swing test: a new behavioral parameter for rats with 6-hydroxydopamineinduced hemiparkinsonism. J. Neurosci., 15: 5372-5378. Bowenkamp, K.E., Hoffman, A.F., Gerhardt, G.A., Henry, M.A., Biddle, P.T., Hoffer, B.J., Granholm, A.C. (1995) Glial cell linederived neurotrophic factor supports survival of injured midbrain dopaminergic neurons. J. Comp. Neurol., 355: 479-489. Burger, C., Gorbatyuk, O.S., Velardo, M.J., Peden, C.S., Williams, R, Zolotukhin, S., Reier, RJ., Mandel, R.J. and Muzyczka, N. (2004) Recombinant AAV viral vectors pseudotyped with viral capsids from serotypes 1, 2, and 5 display differential efficiency and cell tropism after delivery to different regions of the central nervous system. Mol. Ther., 10: 302-317. Burton, E.A., Glorioso, J.C. and Fink, D.J. (2003) Gene therapy progress and prospects: Parkinson's disease. Gene Ther., 10: 1721-1727. Butterfield, RG., Valaius, B.G., Spencer, RS., Lindeman, C.A. and Nutt, J.G. (1993) Envirorm\ental antecedents of young-onset Parkinson's disease. Neurology, 43:1150-1158. Cao, L., Liu, Y, During, M.J. and Xiao, W. (2000) High-titer, wild-type free recombinant adeno-associated virus vector production using intron-contairung helper plasmids. J. Virol., 74:11456-11463. Carlsson, T., Winkler, C., Burger, C., Muzyczka, N., Mandel, R.J., Cenci, A., Bjorklund, A. and Kirik, D. (2005) Reversal of dyskinesias in an animal model of Parkinson's disease by continuous L-dopa delivery using rAAV vectors. Brain, 128: 559-569. Chang, J.W., Wachtel, S.R., Young, D. and Kang, U.J. (1999) Biochemical and anatomical characterization of forepaw adjusting steps in rat models of Parkinson's disease: studies on medial forebrain bimdle and striatal lesions. Neuroscience, 88: 617-628. Chase, T.N., Baronti, R, Fabbrini, G., Heuser, I.J., Juncos, J.L., Mouradian, M.M. (1989) Rationale for continuous dopaminomimetic therapy of Parkinson's disease. Neurology, 39: 7-10. Chase, T.N., Mouradian, M.M. and Engber, T.M. (1993) Motor response complications and the function of striatal efferent systems. Neurology 43: S23-S27. Chaturvedi, S., Ostbye, T., Stoessl, A.J., Merskey, H. and Hachinski, V. (1995) Environmental exposures in elderly Canadians with Parkinson's disease. Can. J. Neurol. Sci., 22: 232-234. Chen, Q., He, Y and Yang, K. (2005) Gene therapy for Parkinson's disease: progress and challenges. Curr. Gene Ther., 5: 71-80. Choi-Lundberg, D.L., Lin, Q., Chang, YN., Chiang, YL., Hay, CM., Mohajeri, H., Davidson, B.L. and Bohn, M.C (1997) Dopaminergic neurons protected from degeneration by GDNF gene therapy Science, 275: 838-841. Choi-Lundberg, D.L., Lin, Q., Schallert, T., Crippens, D., Davidson, B.L., Chang, Y.N., Chiang, Y.L., Qian, J., Bardwaj, L. and Bohn, M.C. (1998) Behavioral and cellular protection of rat dopaminergic neurons by an adenoviral vector encoding glial cell linederived neurotrophic factor. Exp. Neurol., 154: 261-275. Clark, K.R., Liu, X., McGrath, J.R and Johnson, PR. (1999) Highly purified recombinant adeno-associated virus vectors are biologically active and free of detectable helper and wild-type viruses. Hum. Gene Ther., 10: 1031-1039. Connor, B., Kozlowski, D.A., Schallert, T., Tillerson, J.L., Davidson, B.L. and Bohn, M.C (1999) Differential effects of glial cell linederived neurotrophic factor (GDNF) in the striatum and substantia nigra of the aged Parkinsonian rat. Gene Ther., 6:1936-1951.

Connor, B., Kozlowski, D.A., Unnerstall, J.R., Elsworth, J.D., Tillerson, J.L., Schallert, T. and Bohn, M.C. (2001) Glial cell linederived neurotrophic factor (GDNF) gene delivery protects dopaminergic terminals from degeneration. Exp. Neurol., 169: 83-95. Corti, O., Sanchez-Capelo, A., Colin, P., Hanoun, N., Hamon, M. and Mallet, J. (1999) Long-term doxycycline-controUed expression of human tyrosine hydroxylase after direct adenovirus-mediated gene transfer to a rat model of Parkinson's disease. Proc. Natl. Acad. Sci. USA, 96: 12120-12125. Cotzias, C.G., Van Woert, M.H. and Schiffer, L.M. (1967) Aromatic amino acids and modification of parkinsonism. N. Engl. J. Med., 276: 374-379. Crocker, S.J., Liston, R, Anisman, H., Lee, CJ., Smith, RD., Earl, N., Thompson, CS., Park, D.S., Komeluk, R.G. and Robertson, G.S. (2003) Attenuation of MPTP-induced neurotoxicity and behavioural impairment in NSE-XIAP transgenic mice. Neurobiol. Dis., 12: 150-161. Dass, B., Iravani, M., Jackson, M., Engber, T., Galdes. A. and Jenner, P. (2002) Behavioural and immunohistochemical changes following supranigral administration of sonic hedgehog in l-methyl-4-phenyl-l,2,3,6-tetrahydropyridine-treated common marmosets. Neuroscience, 114: 99. Dauer, W. and Przedborski, S. (2003) Parkinson's disease: mechanisms and models. Neuron, 39: 889-909. Davidson, B.L., Allen, E.D., Kozarsky, K.F., Wilson, J.M. and Roessler, B.J. (1993) A model system for in vivo gene transfer into the central nervous system using an adenoviral vector. Nat. Genet., 3: 219-223. Dong, Z., Ferger, B., Patema, J.C, Vogel, D., Purler, S., Osinde, M., Feldon, J. and Bueler, H. (2003) Dopamine-dependent neurodegeneration in rats induced by viral vector-mediated overexpression of the parkin target protein, CDCrel-1. Proc. Natl. Acad. Sci. USA, 100: 12438-12443. Dunnett, S.B. and Bjorklund, A. (1999) Prospects for new restorative and neuroprotective treatments in Parkinson's disease. Nature, 399: A32-A39. During, M.J., Kaplitt, M.G., Stem, M.B. and Eidelberg, D. (2001) Subthalamic GAD gene transfer in Parkinson disease patients who are candidates for deep brain stimulation. Hum. Gene Ther., 12: 1589-1591. During, M.J., Naegele, J.R., O'Malley, K.L. and Geller, A.L (1994) Long-term behavioral recovery in parkinsonian rats by an HSV vector expressing tyrosine hydroxylase. Science, 266:1399-1403. During, M.J., Samulski, R.J., Elsworth, J.D., Kaplitt, M.G., Leone, R, Xiao, X., Li, J., Freese, A., Taylor, J.R., Roth, R.H., Sladek, J.R. Jr., O'Malley, K.L., Redmond, D.E. Jr. (1998) In vivo expression of therapeutic human genes for dopamine production in the caudates of MPTP-treated monkeys using an AAV vector. Gene Ther., 5: 820-^27. During, M.J., Young, D., Baer, K., Lawlor, P. and Klugmann, M. (2003) Development and optimization of adeno-associated virus vector transfer into the central nervous system. Methods Mol. Med., 76: 221-236. Eberhardt, O., Coelln, R.V, Kugler, S., Lindenau, J., Rathke-Hartlieb, S., Gerhardt, E., Haid, S., Isenmann, S., Gravel, C , Srinivasan, A., Bahr, M., Weller, M., Dichgans, J. and Schulz, J.B. (2000) Protection by synergistic effects of adenovirus-mediated X-chromosome-linked inhibitor of apoptosis and glial cell line-derived neurotrophic factor gene transfer in the l-methyl-4-phenyl1,2,3,6-tetrahydropyridine model of Parkinson's disease. J. Neurosci., 20: 9126-9134.

11. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

104

GENE THERAPY FOR PARKINSON'S DISEASE

Ebert, A.D., Chen, R, He, X., Cryns, V.L. and Bohn, M.C. (2005) A tetracycline-regulated adenovirus encoding dominant-negative caspase-9 is regulated in rat brain and protects against neurotoxin-induced cell death in vitro, but not in vivo. Exp. Neurol., 191: S80-S94. Eslamboli, A., Cummings, R.M., Ridley, R.M., Baker, H.F., Muzyczka, N., Burger, C , Mandel, R.J., Kirik, D. and Annett, L.E. (2003) Recombinant adeno-associated viral vector (rAAV) delivery of GDNF provides protection against 6-OHDA lesion in the common marmoset monkey (Callithrix jacchus). Exp. Neurol., 184: 536-548. Eslamboli, A., Georgievska, B., Ridley, R.M., Baker, H.F., Muzyczka, N., Burger, C , Mandel, R.J., Annett, L. and Kirik, D. (2005) Continuous low-level gUal cell line-derived neurotrophic factor delivery using recombinant adeno-associated viral vectors provides neuroprotection and induces behavioral recovery in a primate model of Parkinson's disease. J. Neurosci., 25: 769-777. Fahn, S. (1996) Is levodopa toxic? Neurology, 47: S184-S195. Fahn, S., Oakes, D., Shoulson, I., Kieburtz, K., Rudolph, A., Lang, A., Olanow, C.W., Tanner, C. and Marek, K. (2004) Levodopa and the progression of Parkinson's disease. N Engl. J. Med., 351: 2498-2508. Fan, D.S., Ogawa, M., Fujimoto, K.I., Ikeguchi, K., Ogasawara, Y, Urabe, M., Nishizawa, M., Nakano, I., Yoshida, M., Nagatsu, I., Ichinose, H., Nagatsu, T., Kurtzman, G.J. and Ozawa, K. (1998) 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., 9: 2527-2535. Feamley, J.M. and Lees, A.J. (1991) Ageing and Parkinson's disease: substantia nigra regional selectivity. Brain, 114: 2283-2301. Ferrari, G., Minozzi, M.C, Toffano, G., Leon, A. and Skaper, S.D. (1989) Basic fibroblast growth factor promotes the survival and development of mesencephalic neurons in culture. Dev. Biol., 133: 140-147. Fisher, L.J., Jinnah, H.A., Kale, L.C., Higgins, G.A. and Gage, EH. (1991) Survival and function of intrastriatally grafted primary fibroblasts genetically modified to produce L-dopa. Neuron, 6: 371-380. Fitzsimons, H.L., Bland, R.J. and During, M.J. (2002) Promoters and regulatory elements that improve adeno-associated virus transgene expression in the brain. Methods, 28: 227-236. Fitzsimons, H.L., McKenzie, J.M. and During, M.J. (2001) Insulators coupled to a minimal bidirectional tet cassette for tight regulation of rAAV-mediated gene transfer in the mammalian brain. Gene Ther., 8: 1675-1681. Gash, D.M., Zhang, Z., Ovadia, A., Cass, W.A., Yi, A., Simmerman, L., Russell, D., Martin, D., Lapchak, PA., Collins, R, Hoffer, B.J. and Gerhardt, G.A. (1996) Functional recovery in parkinsonian monkeys treated with GDNR Nature, 380: 252-255. Georgievska, B., Kirik, D. and Bjorklund, A. (2002a) Aberrant sprouting and downregulation of tyrosine hydroxylase in lesioned nigrostriatal dopamine neurons induced by long-lasting overexpression of glial cell line derived neurotrophic factor in the striatum by lentiviral gene transfer. Exp. Neurol., 177: 461-474. Georgievska, B., Kirik, D. and Bjorklund, A. (2004) Overexpression of glial cell line-derived neurotrophic factor using a lentiviral vector induces time- and dose-dependent downregulation of tyrosine hydroxylase in the intact nigrostriatal dopamine system. J. Neuroscience, 24: 6437-6445. Georgievska, B., Kirik, D., Rosenblad, C , Lundberg, C. and Bjorklund, A. (2002b) Neuroprotection in the rat Parkinson

model by intrastriatal GDNF gene transfer using a lentiviral vector. Neuroreport, 13: 75-82. Gill, S.S. and Heywood, P (1997) Bilateral dorsolateral subthalamotomy for advanced Parkinson's disease. Lancet, 350:1224. Gill, S.S., Patel, N.K., Hotton, G.R., O'Sullivan, K., McCarter, R., Bunnage, M., Brooks, D.J., Svendsen, C.N. and Heywood, P (2003) Direct brain infusion of glial cell line-derived neurotrophic factor in Parkinson disease. Nat. Med., 9: 589-595. Gossen, M., Freundlieb, S., Bender, G., MuUer, G., Hillen, W. and Bujard, H. (1995) Transcriptional activation by tetracyclines in mammalian ceUs. Science, 268:1766-1769. Greenamyre, J.T. and Hastings, T.G. (2004) Biomedicine. Parkinson's-divergent causes, convergent mechanisms. Science, 304: 1120-1122. Grondin, R., Zhang, Z., Yi, A., Cass, W.A., Maswood, N., Andersen, A.H., Elsberry, D.D., Klein, M.C, Gerhardt, G.A. and Gash, D.M. (2002) Chronic, controlled GDNF infusion promotes structural and functional recovery in advanced parkinsonian monkeys. Brain, 125: 2191-2201. Hartmann, A., Hunot, S., Michel, PR, Muriel, M.P, Vyas, S., Paucheux, B.A., Mouatt-Prigent, A., Turmel, H., Srinivasan, A., Ruberg, M., Evan, G.I., Agid, Y and Hirsch, E.G. (2000) Caspase3: a vulnerability factor and final effector in apoptotic death of dopaminergic neurons in Parkinson's disease. Proc. Natl. Acad. Sci. USA, 97: 2875-2880. Haywood, A.F. and Staveley, B.E. (2004) Parkin counteracts symptoms in a Drosophila model of Parkinson's disease. BMC Neurosci., 5: 14. Henderson, J.M., Annett, L.E., Ryan, L.J., Chiang, W., Hidaka, S., Torres, E.M. and Dunnett, S.B. (1999) Subthalamic nucleus lesions induce deficits as well as benefits in the hemiparkinsonian rat. Eur. J. Neurosci., 11: 2749-2757. Hirsch, E.G., Hunot, S., Faucheux, B., Agid, Y, Mizuno, Y, Mochizuki, H., Tatton, W.G., Tatton, N. and Olanow, W.C (1999) Dopaminergic neurons degenerate by apoptosis in Parkinson's disease. Mov Disord., 14: 383-385. Horellou, P, Brundin, P , Kalen, P , Mallet, J. and Bjorklund, A. (1990a) In vivo release of dopa and dopamine from genetically engineered cells grafted to the denervated rat striatum. Neuron, 5: 393^02. Horellou, P, Marlier, L., Privat, A. and Mallet, J. (1990b) Behavioural Effect of Engineered Cells that Sjmthesize l-dopa or Dopamine after Grafting into the Rat Neostriatum. Eur. J. Neurosci., 2:116-119. Horellou, P., Vigne, E., Castel, M.N., Bameoud, P., Colin, P., Perricaudet, M., Delaere, P. and Mallet, J. (1994) Direct intracerebral gene transfer of an adenoviral vector expressing tyrosine hydroxylase in a rat model of Parkinson's disease. Neuroreport, 6: 49-53. Homykiewicz, O. (2001) Chemical neuroanatomy of the basal ganglia-normal and in Parkinson's disease. J. Chem. Neuroanat., 22: 3-12. Hsich, G., Sena-Esteves, M. and Breakefield, X.O. (2002) Critical issues in gene therapy for neurologic disease. Hum. Gene Ther., 13: 579-604. Hubble, J.P, Cao, T., Hassanein, R.E., Neuberger, J.S. and Koller, W.C. (1993) Risk factors for Parkinson's disease. Neurology, 43: 1693-1697. Hurtado-Lorenzo, A., Millan, E., Gonzalez-Nicolini, V., Suwelack, D., Castro, M.G. and Lowenstein, PR. (2004) Differentiation and transcription factor gene therapy in experimental parkinson's disease: sonic hedgehog and Gli-1, but not Nurr-1, protect nigrostriatal cell bodies from 6-OHDA-induced neurodegeneration. Mol. Ther., 10: 507-524.

11. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

105

SUMMARY Hyman, C , Hofer, M., Barde, Y.A., Juhasz, M., Yancopoulos, G.D., Squinto, S.R and Lindsay, R.M. (1991) BDNF is a neurotrophic factor for dopaminergic neurons of the substantia nigra. Nature, 350: 230-232. Imai, Y, Soda, M., Inoue, H., Hattori, N., Mizuno, Y. and Takahashi, R. (2001) An unfolded putative transmembrane polypeptide, which can lead to endoplasmic reticulum stress, is a substrate of Parkin. CeU, 105: 891-902. Janson, C , McPhee, S., Bilaniuk, L., Haselgrove, J., Testaiuti, M., Freese, A., Wang, D.J., Shera, D., Hurh, P., Rupin, J., Saslow, E., Goldfarb, O., Goldberg, M., Larijani, G., Sharrar, W., Liouterman, L., Camp, A., Kolodny, E., Samulski, J. and Leone, P (2002) Clinical protocol. Gene therapy of Canavan disease: AAV-2 vector for neurosurgical delivery of aspartoacylase gene (ASPA) to the human brain. Hum. Gene Ther., 13: 1391-1412. Kaplitt, M.G., Leone, P , Samulski, R.J., Xiao, X., Pfaff, D.W., O'Malley, K.L. and During, M.J. (1994) Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat. Genet., 8:148-154. Kirik, D., Annett, L.E., Burger, C , Muzyczka, N., Mandel, R.J. and Bjorklund, A. (2003) Nigrostriatal alpha-synucleinopathy induced by viral vector-mediated overexpression of human alpha-synuclein: a new primate model of Parkinson's disease. Proc. Natl. Acad. Sci. USA, 100: 2884-2889. Kirik, D., Georgievska, B. and Bjorklund, A. (2004) Localized striatal delivery of GDNF as a treatment for Parkinson disease. Nat. Neurosci., 7:105-110. Epub 2004 Jan 2027. Kirik, D., Georgievska, B., Burger, C , Wii\kler, C , Muzyczka, N., Mandel, R.J. and Bjorklund, A. (2002a) Reversal of motor impairments in parkinsonian rats by continuous intrastriatal delivery of L-dopa using rAAV-mediated gene transfer. Proc. Natl. Acad. Sci. USA, 99: 4708^713. Kirik, D., Georgievska, B., Rosenblad, C. and Bjorklund, A. (2001) Delayed infusion of GDNF promotes recovery of motor function in the partial lesion model of Parkinson's disease. Eur. J. Neurosci., 13:1589-1599. Kirik, D., Rosenblad, C. and Bjorklund, A. (2000a) Preservation of a functional nigrostriatal dopamine pathway by GDNF in the intrastriatal 6-OHDA lesion model depends on the site of administration of the trophic factor. Eur. J. Neurosci., 12: 3871-3882. Kirik, D., Rosenblad, C , Bjorklund, A. and Mandel, R.J. (2000b) 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., 20: 4686^700. Kirik, D., Rosenblad, C , Burger, C , Lundberg, C , Johansen, T.E., Muzyczka, N., Mandel, R.J. and Bjorklund, A. (2002b) Parkinson-like neurodegeneration induced by targeted overexpression of alpha-synuclein in the nigrostriatal system. J. Neurosci., 22: 2780-2791. Kitada, T., Asakawa, S., Hattori, N., Matsumine, H., Yamamura, Y, Minoshima, S., Yokochi, M., Mizuno, Y and Shimizu, N. (1998) Mutations in the parkin gene cause autosomal recessive juvenile parkinsonism. Nature, 392: 605-608. Klein, R.L., Lewis, M.H., Muzyczka, N. and Meyer, E.M. (1999) Prevention of 6-hydroxydopamine-induced rotational behavior by BDNF somatic gene transfer. Brain Res., 847: 314-320. Klein, R.L., Meyer, E.M., Peel, A.L., Zolotukhin, S., Meyers, C , Muzyczka, N. and King, M.A. (1998) Neuron-specific transduction in the rat septohippocampal or nigrostriatal pathway by recombinant adeno-associated virus vectors. Exp. Neurol., 150: 183-194.

Knusel, B., Michel, PR, Schwaber, J.S. and Hefti, R (1990) Selective and nonselective stimulation of central cholinergic and dopaminergic development in vitro by nerve growth factor, basic fibroblast growth factor, epidermal growth factor, insulin and the insulin-like growth factors I and II. J. Neurosci., 10: 558-570. Kordower, J.H., Emborg, M.E., Bloch, J., Ma, S.Y, Chu, Y, Leventhal, L., McBride, J., Chen, E.Y, Palfi, S., Roitberg, B.Z., Brown, W.D., Holden, J.E., Pyzalski, R., Taylor, M.D., Carvey, R, Ling, Z., Trono, D., Hantraye, P., Deglon, N. and Aebischer, P. (2000) Neurodegeneration prevented by lentiviral vector delivery of GDNF in primate models of Parkinson's disease. Science, 290: 767-773. Kordower, J.H. and Goetz, C.G. (1999) The first miracle in neurodegenerative disease: the discovery of oral levodopa. Brain Res. Bull., 50: 377-378. Kordower, J.H., Palfi, S., Chen, E.Y, Ma, S.Y, Sendera, T., Cochran, E.J., Mufson, E.J., Penn, R., Goetz, C.G. and Comella, C D . (1999) Clinicopathological findings following intraventricular glial-derived neurotrophic factor treatment in a patient with Parkinson's disease. Ann. Neurol., 46: 419-424. Kosel, S., Egensperger, R., von Eitzen, U., Mehraein, P. and Graeber, M.B. (1997) On the question of apoptosis in the parkinsonian substantia nigra. Acta. Neuropathol. (Berl), 93: 105-108. Kozlowski, D.A., Connor, B., Tillerson, J.L., Schallert, T. and Bohn, M.C. (2000) Delivery of a GDNF gene into the substantia nigra after a progressive 6-OHDA lesion maintains functional nigrostriatal connections. Exp. Neurol., 166:1-15. Kruger, R., Kuhn, W., MuUer, T., Woitalla, D., Graeber, M., Kosel, S., Przuntek, H., Epplen, J.T., Schols, L. and Riess, O. (1998) Ala30Pro mutation in the gene encoding alpha-synuclein in Parkinson's disease. Nat. Genet., 18: 106-108. Kuan, C.Y and Burke, R.E. (2005) Targeting the JNK signaling pathway for stroke and Parkinson's diseases therapy. Curr. Drug Targets CNS Neurol. Disord., 4: 63-67. Langston, J.W., Ballard, R, Tetrud, J.W. and Irwin, I. (1983) Chronic Parkinsonism in humans due to a product of meperidine-analog synthesis. Science, 219: 979-980. Langston, J.W., Fomo, L.S., Tetrud, J., Reeves, A.G., Kaplan, J.A., Karluk, D., DeLanney, L.E. and Irwin, I. (1999) Evidence of active nerve cell degeneration in the substantia nigra of humans years after l-methyl-4-phenyl-l,2,3,6-tetrahydropyridine exposure. Similarities and differences between MPTP-induced parkinsonsim and Parkinson's disease. Neuropathologic considerations. Ann. Neurol., 46: 598-605. Lapchak, PA., Araujo, D.M., Hilt, D.C., Sheng, J. and Jiao, S. (1997b) Adenoviral vector-mediated GDNF gene therapy in a rodent lesion model of late stage Parkinson's disease. Brain Res., 777\ 153-160. Lapchak, PA., Jiao, S., Collins, F and Miller, P.J. (1997a) Glial cell line-derived neurotrophic factor: distribution and pharmacology in the rat following a bolus intraventricular injection. Brain Res., 747: 92-102. Lee, W.Y, Chang, J.W., Nemeth, N.L. and Kang, U.J. (1999) Vesicular monoamine transporter-2 and aromatic L-amino acid decarboxylase enhance dopamine delivery after L-3, 4-dihydrox)rphenylalanine administration in Parkinsonian rats. J. Neurosci., 19: 3266-3274. Leff, S.E., Rendahl, K.G., Spratt, S.K., Kang, U.J. and Mandel, R.J. (1998) In vivo L-DOPA production by genetically modified primary rat fibroblast or 9L gliosarcoma cell grafts via coexpression of GTPcyclohydrolase I with tyrosine hydroxylase. Exp. Neurol., 151: 249-264. Leff, S.E., Spratt, S.K., Snyder, R.O. and Mandel, R.J. (1999) Longterm restoration of striatal L-aromatic amino acid decarboxylase activity using recombinant adeno-associated viral vector gene

II. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

106

GENE THERAPY FOR PARKINSON'S DISEASE

transfer in a rodent model of Parkinson's disease. Neuroscience, 92: 185-196. Leroy, E., Boyer, R., Auburger, G., Leube, B., Ulm, G., Mezey, E., Harta, G., Brownstein, M.J., Jonnalagada, S., Chemova, T., Dehejia. A., Lavedan, C , Gasser, T., Steinbach, P.J., Wilkinson, K.D. and Polymeropoulos, M.H. (1998b) The ubiquitin pathway in Parkinson's disease. Nature, 395: 451^52. Leroy, E., Boyer, R. and Polymeropoulos, M.H. (1998a) Intron-exon structure of ubiquitin c-terminal hydrolase-Ll. DNA Res., 5: 397-400. Levy, R., Lang, A.E., Dostrovsky, J.O., Pahapill, P., Romas, J., SaintCyr, J., Hutchison, W.D. and Lozano, A.M. (2001) Lidocaine and muscimol microinjections in subthalamic nucleus reverse Parkinsonian symptoms. Brain, 124: 2105-2118. Lewis, S.J.G., Caldwell, M.A. and Barker, R.A. (2003) Modem therapeutic approaches in Parkinson's disease. Expert Reviews in Molecular Medicine 5:DOI:10.1017/S1462399403006008. Limousin, P., Pollak, P., Benazzouz, A., Hoffmann, D., Broussolle, E., Perret, J.E. and Benabid, A.L. (1995a) Bilateral subthalamic nucleus stimulation for severe Parkinson's disease. Mov. Disord., 10: 672-674. Limousin, P., Pollak, P., Benazzouz, A., Hoffmann, D., Le Bas, J.F., Broussolle, E., Perret, J.E. and Benabid, A.L. (1995b) Effect of parkinsonian signs and symptoms of bilateral subthalamic nucleus stimulation. Lancet, 345: 91-95. Lin, L.R, Doherty, D.H., Lile, J.D., Bektesh, S. and Collins, R (1993) GDNF: a glial cell line-derived neurotrophic factor for midbrain dopaminergic neurons. Science, 260: 1130-1132. Lloyd, K.G. (1977) CNS compensation to dopamine neuron loss in Parkinson's disease. Adv. Exp. Med. Biol, 90: 255-266. Lo Bianco, C , Deglon, N., Pralong, W. and Aebischer, P. (2004a) Lentiviral nigral delivery of GDNF does not prevent neurodegeneration in a genetic rat model of Parkinson's disease. Neurobiol. Dis., 17: 283-289. Lo Bianco, C , Schneider, B.L., Bauer, M., Sajadi, A., Brice, A., Iwatsubo, T. and Aebischer, P. (2004b) Lentiviral vector delivery of parkin prevents dopaminergic degeneration in an alphasynuclein rat model of Parkinson's disease. Proc. Natl. Acad. Sci. USA, 101: 17510-17515. Lu, Y. (2004) Recombinant adeno-associated virus as delivery vector for gene therapy — a review. Stem Cells Dev, 13: 133-145. Luo, J., Kaplitt, M.G., Fitzsimons, H.L., Zuzga, D.S., Liu, Y, Oshinsky, M.L. and During, M.J. (2002) Subthalamic GAD gene therapy in a Parkinson's disease rat model. Science, 298: 425-429. Mandel, R.J., Rendahl, K.G., Snyder, R.O. and Leff, S.E. (1999) Progress in direct striatal delivery of L-dopa via gene therapy for treatment of Parkinson's disease using recombinant adenoassociated viral vectors. Exp. Neurol., 159: 47-64. Mandel, R.J., Rendahl, K.G., Spratt, S.K., Snyder, R.O., Cohen, L.K. and Leff, S.E. (1998) 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., 18: 4271-4284. Mandel, R.J., Spratt, S.K., Snyder, R.O. and Leff, S.E. (1997) 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, 94:14083-14088. Marsden, C D . and Parkes, J.D. (1977) Success and problems of long-term levodopa therapy in Parkinson's disease. Lancet, 1: 345-349.

Mastakov, M.Y, Baer, K., Symes, C.W., Leichtlein, C.B., Kotin, R.M. and During, M.J. (2002) Immunological aspects of recombinant adeno-associated virus delivery to the mammalian brain. J. Virol., 76: 8446-8454. McNaught, K.S., Bjorklund, L.M., Belizaire, R., Isacson, O., Jenner, P. and Olanow, C.W. (2002) Proteasome inhibition causes nigral degeneration with inclusion bodies in rats. Neuroreport, 13: 1437-1441. McNaught, K.S., Olanow, C.W., Halliwell, B., Isacson, O. and Jenner, P. (2001) Failure of the ubiquitin-proteasome system in Parkinson's disease. Nat. Rev. Neurosci., 2: 589-594. McNaught, K.S., Perl, D.P, Brownell, A.L. and Olanow, C.W. (2004) Systemic exposure to proteasome inhibitors causes a progressive model of Parkinson's disease. Arm. Neurol., 56:149-162. Miao, N., Wang, M., Ott, J.A., D'Alessandro, J.S., Woolf, T.M., Bumcrot, D.A., Mahanthappa, N.K. and Pang, K. (1997) Sonic hedgehog promotes the survival of specific CNS neuron populations and protects these cells from toxic insult In vitro.PG - 5891-9. J. Neurosci., 17: 5891-5899. Miwa, H., Kubo, T., Suzuki, A., Nishi, K. and Kondo, T. (2005) Retrograde dopaminergic neuron degeneration following intrastriatal proteasome inhibition. Neurosci. Lett., 380: 93-98. Epub 2005 Jan 2026. Mochizuki, H., Goto, K., Mori, H. and Mizuno, Y (1996) Histochemical detection of apoptosis in Parkinson's disease. J. Neurol. Sci., 137: 120-123. Mochizuki, H., Hayakawa, H., Migita, M., Shibata, M., Tanaka, R., Suzuki, A., Shimo-Nakanishi, Y, Urabe, T., Yamada, M., Tamayose, K., Shimada, T., Miura, M. and Mizuno, Y (2001) An AAVderived Apaf-1 dominant negative inhibitor prevents MPTP toxicity as antiapoptotic gene therapy for Parkinson's disease. Proc. Nati. Acad. Sci. USA, 98:10918-10923. Moore, R. Y (2003) Organization of midbrain dopamine systems and the pathophysiology of Parkinson's disease. Parkinsonism Relat. Disord., 9: S65-S71. Muramatsu, S., Fujimoto, K., Ikeguchi, K., Shizuma, N., Kawasaki, K., Ono, R, Shen, Y, Wang, L., Mizukami, H., Kume, A., Matsumura, M., Nagatsu, I., Urano, R, Ichinose, H., Nagatsu, T., Terao, K., Nakano, I. and Ozawa, K. (2002) 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 Then, 13: 345-354. Nagatsu, T. and Ichinose, H. (1999) Molecular biology of catecholamine-related enzymes in relation to Parkinson's disease. Cell Mol. Neurobiol., 19: 57-66. Nagatsu, T., Levitt, M. and Udenfriend, S. (1964) Conversion of L-tyrosine to 3,4-dihydroxyphenylalanine by cell-free preparations of brain and sympathetically innervated tissues. Biochem. Biophys. Res. Commun., 14: 543-549. Nagatsu, T., Yamaguchi, T., Rahman, M.K., Trocewicz, J., Oka, K., Hirata, Y, Nagatsu, I., Narabayashi, H., Kondo, T. and lizuka, R. (1984) Catecholamine-related enzymes and the biopterin cofactor in Parkinson's disease and related extrapyramidal diseases. Adv Neurol., 40: 467-473. Naldini, L., Blomer, U., Gage, RH., Trono, D. and Verma, I.M. (1996) Efficient transfer, integration, and sustained long-term expression of the transgene in adult rat brains injected with a lentiviral vector. Proc. Nati. Acad. Sci. USA, 93:11382-11388. Nutt, J.G., Burchiel, K.J., Comella, C.L., Jankovic, J., Lang, A.E., Laws, E.R. Jr., Lozano, A.M., Perm, R.D., Simpson, R.K. Jr., Stacy, M. and Wooten, G.F. (2003) Randomized, double-blind trial of

II. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

107

SUMMARY glial cell line-derived neurotrophic factor (GDNF) in PD. Neurology, 60: 69-73. Obeso, J.A., Grandas, R, Herrero, M.T. and Horowski, R. (1994) The role of pulsatile versus continuous dopamine receptor stimulation for functional recovery in Parkinson's disease. Eur. J. Neurosci., 6: 889-897. Olsson, M., Nikkhah, G., Bentlage, C. and Bjorklund, A. (1995) Forelimb akinesia in the rat Parkinson model: differential effects of dopamine agonists and nigral transplants as assessed by a new stepping test. J. Neurosci., 15: 3863-3875. Palfi, S., Leventhal, L., Chu, Y., Ma, S.Y., Emborg, M., Bakay R., Deglon, N., Hantraye, P., Aebischer, P. and Kordower, J.H. (2002) Lentivirally delivered glial cell line-derived neurotrophic factor increases the number of striatal dopaminergic neurons in primate models of nigrostriatal degeneration. J. Neurosci., 22: 4942-4954. Park, T.H. and Mytilineou, C. (1992) Protection from l-methyl-4phenylpyridinium (MPP+) toxicity and stimulation of regrowth of MPP(+)-damaged dopaminergic fibers by treatment of mesencephalic cultures with EGF and basic FGF. Brain Res., 599: 83-97. Patel, N.K., Bunnage, M., Plaha, P., Svendsen, C.N., Heywood, P. and Gill, S.S. (2005) Intraputamenal infusion of glial cell linederived neurotrophic factor in PD: a two-year outcome study. Ann. Neurol, 57: 298-302. Patel, N.K., Heywood, R, O'Sullivan, K., McCarter, R., Love, S. and Gill, S.S. (2003) Unilateral subthalamotomy in the treatment of Parkinson's disease. Brain, 126: 1136-1145. Patema, J.C., Feldon, J. and Bueler, H. (2004) Transduction profiles of recombinant adeno-associated virus vectors derived from serotypes 2 and 5 in the nigrostriatal system of rats. J. Virol., 78: 6808-6817. Pearce, R.K., Collins, P , Jenner, P., Emmett, C. and Marsden, C D . (1996) Intraventricular infusion of basic fibroblast growth factor (bFGF) in the MPTP-treated common marmoset. Synapse, 23: 192-200. Pearce, R.K., Costa, S., Jenner, P. and Marsden, C D . (1999) Chronic supranigral infusion of BDNF in normal and MPTP-treated common marmosets. J. Neural Transm., 106: 663-683. Polymeropoulos, M.H., Lavedan, C , Leroy E., Ide, S.E., Dehejia, A., Dutra, A., Pike, B., Root, H., Rubenstein, J., Boyer, R., Stenroos, E.S., Chandrasekharappa, S., Afhanassiadou, A., Papapetropoulos, T., Johnson, W.G., Lazzarini, A.M., Duvoisin, R.C., Di lorio, G., Golbe, L.I. and Nussbaum, R.L. (1997) Mutation in the alphasynuclein gene identified in families with Parkinson's disease. Science, 276: 2045-2047. Pothos, E.N., Davila, V. and Sulzer, D. (1998) Presynaptic recording of quanta from midbrain dopamine neurons and modulation of the quantal size. J. Neurosci., 18: 4106-4118. Rajput, A.H., Uitti, R.J., Stem, W., Laverty, W., O'Donnell, K., O'Donnell, D., Yuen, W.K. and Dua, A. (1987) Geography drinking water chemistry, pesticides and herbicides and the etiology of Parkinson's disease. Can. J. Neurol. Sci., 14: 414r-418. Rosenblad, C , Georgievska, B. and Kirik, D. (2003) Long-term striatal overexpression of GDNF selectively downregulates tyrosine hydroxylase in the intact nigrostriatal dopamine system. Eur. J. Neurosci., 17: 260-270. Rosenblad, C , Kirik, D., Devaux, B., Moffat, B., Phillips, H.S. and Bjorklund, A. (1999) Protection and regeneration of nigral dopaminergic neurons by neurturin or GDNF in a partial lesion model of Parkinson's disease after administration into the striatum or the lateral ventricle. Eur. J. Neurosci., 11:1554-1566.

Rosenblad, C , Martinez-Serrano, A. and Bjorklund, A. (1998) Intrastriatal glial cell line-derived neurotrophic factor promotes sprouting of spared nigrostriatal dopaminergic afferents and induces recovery of function in a rat model of Parkinson's disease. Neuroscience, 82:129-137. Salvatore, M.F, Zhang, J.L., Large, D.M., Wilson, RE., Gash, C.R., Thomas, T.C, Haycock, J.W., Bing, G., Stanford, J.A., Gash, D.M. and Gerhardt, G.A. (2004) Striatal GDNF administration increases tyrosine hydroxylase phosphorylation in the rat striatum and substantia nigra. J. Neurochem., 90: 245-254. Sanchez-Pernaute, R., Harvey-White, J., Cunningham, J. and Bankiewicz, K.S. (2001) Functional effect of adeno-associated virus mediated gene transfer of aromatic L-amino acid decarboxylase into the striatum of 6-OHDA-lesioned rats. Mol. Ther., 4: 324-330. Sauer, H. and Oertel, W.H. (1994) Progressive degeneration of nigrostriatal dopamine neurons following intrastriatal terminal lesions with 6-hydroxydopamine: a combined retrograde tracing and immunocytochemical study in the rat. Neuroscience, 59: 401-415. Sauer, H., Rosenblad, C and Bjorklund, A. (1995) Glial cell linederived neurotrophic factor but not transforming growth factor beta 3 prevents delayed degeneration of nigral dopaminergic neurons following striatal 6-hydroxydopamine lesion. Proc. Natl. Acad. Sci. USA, 92: 8935-8939. Schuh, L.A. and Bennett, J.P. Jr. (1993) Suppression of dyskinesias in advanced Parkinson's disease. I. Continuous intravenous levodopa shifts dose response for production of dyskinesias but not for relief of parkinsonism in patients with advanced Parkinson's disease. Neurology, 43:1545-1550. Shen, J. and Cookson, M.R. (2004) Mitochondria and dopamine: new insights into recessive parkinsonism. Neuron, 43: 301-304. Shen, Y, Muramatsu, S.I., Ikeguchi, K., Fujimoto, K.I., Fan, D.S., Ogawa, M., Mizukami, H., Urabe, M., Kume, A., Nagatsu, I., Urano, F, Suzuki, T., Ichinose, H., Nagatsu, T., Monahan, J., Nakano, I. and Ozawa, K. (2000) Triple transduction with adqnoassociated virus vectors expressing tyrosine hydroxylase, aromatic-L-amino-acid decarboxylase, and GTP cyclohydrolase I for gene therapy of Parkinson's disease. Hum. Gene Then, 11: 1509-1519. Sherer, T.B., Betarbet, R. and Greenamyre, J.T. (2002) Environment, mitochondria, and Parkinson's disease. Neuroscientist, 8: 192-197. Shimura, H., Schlossmacher, M.G., Hattori, N., Frosch, M.P., Trockenbacher. A., Schneider, R., Mizuno, Y, Kosik, K.S. and Selkoe, D.J. (2001) Ubiquitination of a new form of alpha-synuclein by parkin from human brain: implications for Parkinson's disease. Science, 293: 263-269. Silva, R.M., Kuan, C Y , Rakic, P and Burke, R.E. (2005) Mixed lineage kinase-c-jun N-terminal kinase signaling pathway: a new therapeutic target in Parkinson's disease. Mov Disord., 20: 653-664. Slevin, J.T., Gerhardt, G.A., Smith, C D . , Gash, D.M., Kryscio, R. and Young, B. (2005a) Improvement of bilateral motor functions in patients with Parkinson disease through the unilateral intraputaminal infusion of glial cell line-derived neurotrophic factor. J. Neurosurg., 102: 216-222. Slevin, J.T., Gerhardt, G.A., Smith, C D . , Gash, D.M., Kryscio, R. and Young, B. (2005b) Research on Parkinson disease. J. Neurosurg., 102: 401. Smargiassi, A., Mutti, A., De Rosa, A., De Palma, G., Negrotti, A. and Calzetti, S. (1998) A case-control study of occupational and

[I. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

108

8. GENE THERAPY FOR PARKINSON'S DISEASE

environmental risk factors for Parkinson's disease in the EmiliaRomagna region of Italy. Neurotoxicology, 19: 709-712. Spillantini, M.G., Schmidt, M.L., Lee, V.M., Trojanowski, J.Q., Jakes, R. and Goedert, M. (1997) Alpha-synuclein in Lewy bodies. Nature, 388: 839-840. Sun, M., Kong, L., Wang, X., Holmes, C , Gao, Q., Zhang, G.R., Pfeilschifter, J., Goldstein, D.S. and Geller, AJ. (2004) Coexpression of tyrosine hydroxylase, GTP cyclohydrolase I, aromatic amino acid decarboxylase, and vesicular monoamine transporter 2 from a helper virus-free herpes simplex virus type 1 vector supports high-level, long-term biochemical and behavioral correction of a rat model of Parkinson's disease. Hum. Gene Ther., 15:1177-1196. Svendsen, C.N., Clarke, D.J., Rosser, A.E. and Dunnett, S.B. (1996) Survival and differentiation of rat and human epidermal growth factor-responsive precursor cells following grafting into the lesioned adult central nervous system. Exp. Neurol., 137: 376-388. Tatton, N.A. (2000) Increased caspase 3 and Bax immunoreactivity accompany nuclear GAPDH translocation and neuronal apoptosis in Parkinson's disease. Exp. Neurol., 166: 2 9 ^ 3 . Tatton, W.G., Chalmers-Redman, R., Brown, D. and Tatton, N. (2003) Apoptosis in Parkinson's disease: signals for neuronal degradation. Ann. Neurol., 53: S61-S70; discussion S70-62. Tompkins, M.M., BasgaU, E.J., Zamrini, E. and HiU, W.D. (1997) Apoptotic-like changes in Lewy-body-associated disorders and normal aging in substantia nigral neurons. Am. J. Pathol., 150: 119-131. Tsuboi, K. and Shults, C.W. (2002) Intrastriatal injection of sonic hedgehog reduces behavioral impairment in a rat model of Parkinson's disease. Exp. Neurol., 173: 95-104. Ungerstedt, U. (1968) 6-Hydroxy-dopamine induced degeneration of central monoamine neurons. Eur. J. Pharmacol., 5: 107-110. Ungerstedt, U. and Arbuthnott, G.W. (1970) Quantitative recording of rotational behavior in rats after 6-hydroxy-dopamine lesions of the nigrostriatal dopamine system. Brain Res., 24: 485-493. Valente, E.M., Abou-Sleiman, PM., Caputo, V., Muqit, M.M., Harvey, K., Gispert, S., Ali, Z., Del Turco, D., Bentivoglio, A.R., Healy, D.G., Albanese, A., Nussbaum, R., Gonzalez-Maldonado, R., Deller, T., Salvi, S., CorteUi, P, Gilks, W.P, Latchman, D.S., Harvey, R.J., Dallapiccola, B., Auburger, G. and Wood, N.W. (2004) Hereditary early-onset Parkinson's disease caused by mutations in PlNKl. Science, 304: 1158-1160. Ventrella, L.L. (1993) Effect of intracerebroventricular infusion of epidermal growth factor in rats hemitransected in the nigrostriatal pathway. J. Neurosurg. Sci., 37: 1-8. Vigna, E., Cavalieri, S., Allies, L., Geuna, M., Loew, R., Bujard, H. and Naldini, L. (2002) Robust and efficient regulation of transgene

expression in vivo by improved tetracycline-dependent lentiviral vectors. Mol. Ther., 5: 252-261. von Coelln, R., Kugler, S., Bahr, M., Weller, M., Dichgans, J. and Schulz, J.B. (2001) Rescue from death but not from functional impairment: caspase inhibition protects dopaminergic cells against 6-hydroxydopamine-induced apoptosis but not against the loss of their terminals. J. Neurochem., 77\ 263-273. Wang, J., Chen, G., Lu, B. and Wu, C.P (2003) GDNF acutely potentiates Ca2+ channels and excitatory synaptic transmission in midbrain dopaminergic neurons. Neurosignals, 12: 78-88. Wang, L., Muramatsu, S., Lu, Y., Ikeguchi, K., Fujimoto, K., Okada, T., Mizukami, H., Hanazono, Y., Kume, A., Urano, R, Ichinose, H., Nagatsu, T., Nakano, I. and Ozawa, K. (2002) Delayed delivery of AAV-GDNF prevents rugral neurodegeneration and promotes functional recovery in a rat model of Parkinson's disease. Gene Ther., 9: 381-389. Winkler, C , Sauer, H., Lee, C.S. and Bjorklund, A. (1996) Short-term GDNF treatment provides long-term rescue of lesioned nigral dopaminergic neurons in a rat model of Parkinson's disease. J. Neurosci., 16: 7206-7215. Xia, X.G., Harding, T., Weller, M., Bieneman, A., Uney, J.B. and Schulz, J.B. (2001) Gene transfer of the JNK interacting protein-1 protects dopaminergic neurons in the MPTP model of Parkinson's disease. Proc. Natl. Acad. Sci. USA, 98:10433-10438. Yang, F, Feng, L., Zheng, F, Johnson, S.W, Du, J., Shen, L., Wu, C.P and Lu, B. (2001) GDNF acutely modulates excitability and A-type K(+) channels in midbrain dopaminergic neurons. Nat. Neurosci., 4:1071-1078. Yang, Y, Nishimura, I., Imai, Y, Takahashi, R. and Lu, B. (2003) Parkin suppresses dopaminergic neuron-selective neurotoxicity induced by Pael-R in Drosophila. Neuron, 37: 911-924. Zeng, B.Y, Jenner, P. and Marsden, C D . (1996) Altered motor function and graft survival produced by basic fibroblast growth factor in rats with 6-OHDA lesions and fetal ventral mesencephalic grafts are associated with glial proliferation. Exp. Neurol., 139: 214-226. Zhang, Z., Miyoshi, Y, Lapchak, PA., Collins, F , Hilt, D., Lebel, C , Kryscio, R. and Gash, D.M. (1997) Dose response to intraventricular glial cell line-derived neurotrophic factor administration in parkinsonian monkeys. J. Pharmacol. Exp. Ther., 282: 1396-1401. Zigmond, M., Hastings, T. and Perez, R. (2002) Increased dopamine turnover after partial loss of dopaminergic neurons: compensation or toxicity? Parkinsonism Relat. Disord., 8: 389. Zigmond, M.J. (1997) Do compensatory processes underlie the preclinical phase of neurodegenerative disease? Insights from an animal model of parkinsonism. Neurobiol. Dis., 4: 247-253.

II. GENE THERAPY FOR DEGENARATIVE AND FUNCTIONAL DISORDERS

C H A P T E R

9 Nonhuman Primate Models for Testing Gene Therapy for Neurodegenerative Disorders Jeffrey Moirano, Marina E. Emborg

Abstract: In the coming years, the rapid aging of the world's population will lead to a higher incidence of neurological diseases. Because of the dire consequences of the diseases, and general lack of effective treatments, neurological diseases are attractive candidates for novel treatments such as gene therapy. Characteristics such as the progressive nature and site specificity of neurological diseases make gene therapy a viable treatment strategy. Gene therapeutic techniques are still in the earlier stages of development, with many factors left to perfect, especially in the areas of vector design and delivery. Before treatments can be brought to the clinic, the safety and efficacy of the therapy must be evaluated experimentally. Because they are the closest phylogenic, morphological, and genetic relatives of humans, nonhuman primates are ideal candidates to test the capabilities of a neurological disease treatment. Several primate models of neurological disease have been developed, and are essential to both the understanding of the disease processes and the prospect of successful treatments. Keywords: nonhuman primates; gene therapy; neurodegenerative disorders; Parkinson's disease; neuroprotection; restoration

L

characterized by a progressive deterioration of a select, defined population of neurons of specific function over an extended period of time, and typically manifest abnormal proteins within the brain (Mata et al., 2003; Ross and Poirier, 2004). They are a heterogeneous group of disorders, and despite having distinct clinical phenotypes, remain difficult to diagnose (Lansbury, 2004). Although neurodegeneration is prevalent in the normal aging brain, it is accelerated in disease, leading to the distinguishing symptoms that are tremendously devastating to the patient and their families, and carry a great cost to society as a whole (Bohn and Choi-Lundberg, 1998). Neurodegenerative disorders are typically chronic, and patients can live with a markedly reduced quality of life for many years. The most common conditions include Alzheimer's disease (AD), Parkinson's disease (PD), Himtington's disease (HD), and amyotrophic lateral sclerosis (ALS). The disorders with the highest incidence are AD, which affects over

NEURODEGENERATIVE DISORDERS

The world's population is rapidly aging. In the coming decades, the number of elderly people worldwide is projected to increase from about 600 million in the year 2000 to 1.2 billion in 2025 and to 2 billion in 2050 (UNDSEA, 2002). Because age is a major risk factor in many neurodegenerative disorders, the number of people affected by these disorders will increase sharply as a consequence, and the societal and economic costs will be immense. This knowledge, coupled with the fact that current treatments are generally inadequate, has led to a recent focus on neurodegenerative diseases in the medical research community. Advances in our understanding of these disorders are linked to the development of animal models of disease, in particular in nonhuman primates (Sibal and Samson, 2001). Neurodegenerative disorders are

Gene Therapy of the Central Nervous System: From Bench to Bedside

109

Copyright© 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

no

NONHUMAN PRIMATE MODELS FOR TESTING GENE THERAPY FOR NEURODEGENERATIVE DISORDERS

5% of the over-65 population and PD, which affects 1% of the same group. These disorders also share the unfortunate characteristic that their etiologies remain relatively unknovsrn. Currently there are no exceptionally effective clinical therapies for neurodegenerative disorders. Most treatments are symptomatic in nature, and do not attempt to provide any long-term curative or protective benefits. In light of the dire consequences of the diseases and the impasse in successful treatments, neurodegenerative disorders are appealing targets for novel therapeutic strategies. Clearly, it v^ould be beneficial to provide neuroprotection and possibly restoration to cells susceptible to neurodegeneration. Because the diseases are chronic, a successful treatment necessitates a high degree of sustainability. Treatments via classical pharmacotherapy are limited by the presence of the blood-brain barrier, which restricts systemic delivery. Direct interparenchymal or interventricular delivery methods in humans have often produced complications and unwanted side effects (Eriksdotter Jonhagen et al., 1998; Kordower et al., 1999). Other delivery methods have been proposed, such as implantable pumps, but require additional intervention to manipulate, refill and service, and pose greater risks of infection. Gene therapy has emerged from the laboratory setting as an alternative method to provide relief for neurodegenerative disorders.

II.

WHY USE GENE THERAPY FOR TREATMENT OF NEURODEGENERATIVE DISORDERS? Gene therapy is fundamentally defined as the transfer of therapeutic genetic material into cells as a means to rectify disease (Baekelandt et al., 2000; Constantini et al., 2000). Gene therapy was originally postulated as a treatment to correct genetically determined disorders by replacing the defective gene with a properly functioning copy (Friedmann et al., 1972). This complete gene replacement modality has not met with much success, but as the field of gene therapy evolved, a different use of gene transfer emerged. It was realized that genetic material could be used to treat conditions beyond those genetically linked by generating the expression of corrective proteins (Mata et al., 2003). Therefore, gene therapy is an ideal candidate to provide therapeutic benefit in neurodegenerative disorders, as it can locally deliver products designed to ameliorate symptoms by inducing the

production of beneficial or necessary compounds that have dissipated in the course of the disease. The compartmental nature of the brain and the redundant use of a relatively small amount of neurotransmitters throughout the nervous system require the local delivery of therapeutic agents (Mata et al., 2003). However, several distinctive features of the nervous system present challenges to gene therapy that must be overcome, such as the post-mitotic state of neurons, wide variety of cell types and function, volume constraints, difficulty of access, and the delicate and critical nature of the tissue (Constantini et a l , 2000; Hsich et al., 2002).

III.

METHODS OF DELIVERY OF THERAPEUTIC GENES

There are two major approaches to the delivery and expression of therapeutic genes. In ex vivo gene therapy, cultured host cells are transfected in vitro to express the gene of interest, and then transplanted into the body. In vivo gene therapy is a direct method of inserting the genetic material into the targeted tissue, and transduction takes place within the patient's own cells. In either method, there are many important considerations in the design of the transgene. Promoters and regulatory elements, which will control the expression of the gene must be carefully selected and manipulated, as well as an appropriate vector for gene delivery (Gimenez y Ribotta, 2001; Federoff, 2003). Factors, which influence the choice of vector, include the size of the transcription unit, the target cell type, the extent and duration of expression, and the potential for toxicity and immune response (Federoff, 2003). The two main classes of vectors used to deliver genes to their target either in or ex vivo are nonviral (synthetic) and viral vectors. Synthetic, or nonviral vectors include 'naked' DNA plasmids, liposomes, synthetic macromolecules, and DNA-protein complexes. Although they are seen as safer, synthetic vectors are less efficient at gene delivery and expression than viral vectors, and therefore viral vectors are the most often used method of gene delivery in neurological disorders (Baekelandt, 2000). Viral vectors capitalize on the natural ability of viruses to infect target cells. The viral genome is modified to remove or render inoperable pathogenic and replication genes, and a gene cassette is inserted that contains the therapeutic gene of interest, along with promoters and regulators that aid in the expression of the transgene (Glorioso et al., 2003).

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

CONSIDERATIONS FOR A SUCCESSFUL GENE THERAPY STRATEGY

Five main classes of viral vectors have been tested for the treatment of neurodegenerative disorders. Retroviruses are enveloped RNA viruses, which integrate their genetic material into host cells. Retroviruses are limited to expression in dividing cells, and therefore have diminished utility in neurons (Costantini et al., 2000; Hsich et al., 2002). Herpes simplex viral (HSV) vectors are enveloped double-stranded DNA viruses and have a very large transgene capacity. They are neurotropic, very efficient and can infect many types of neurons. HSV do not integrate within the genome, which decreases their risk for insertional mutagenesis, but can have problems with latent expression and immunogenic issues (Hsich et al., 2002; Glorioso and Fink, 2004). Adenoviral (Ad) vectors are nonenveloped double-stranded DNA, which have a large capacity for transgenes and can infect many types of cells. Ad vectors are limited in capability due to natural immune responses in the host, however, 'gutless' Ad vectors have been described that reduce this difficulty (Morsy et al., 1998). Adeno-associated vectors (AAV) are similar to Ad vectors, but are single-stranded, are not pathogenic or immunoreactive in humans, have the ability to insert their genes into the host genome, and have a much lower capacity for transgene information (Baekelandt et al., 2000). Lentiviruses are complex vectors that are capable of integration in nondividing cells. They are typically derived from the humanimmunodeficiency virus (HIV), and have long-term expression in a variety of cell types without being toxic or immunoreactive (Naldini et al., 1996). Although none of the current vectors are perfect, lentivirus and in particular AAV seem to be the most promising vectors for gene therapy strategies for neurodegenerative disorders (Mandel et al., 2004). IV. CONSIDERATIONS FOR A SUCCESSFUL GENE THERAPY STRATEGY Depending on the objective of the therapy and other considerations, one method of gene delivery must be chosen as more appropriate than the others. In vivo gene therapy benefits from its simplicity and the reduced volume of material introduced into the brain (Freeman, 1997). These properties make it an ideal method when the target is a specific, small area of the brain. Disadvantages of in vivo gene transfer are the possibility that the introduction and expression of virus proteins may activate endogenous pathogenic viruses (Romano et al., 2000). It has also been suggested

111

that replication-competent recombinant viruses may form due to random mutagenesis. The possibility exists that viruses, which integrate themselves into the host genome, such as retroviruses, may cause accidental malignant transformation of the host genome. Inflammation and immune response are also of particular concern, even in an 'immune-privileged' organ such as the brain (Hsich et al., 2002). As we previously described, these risks are increased or decreased depending on the vector. The major advantage of ex vivo gene therapy is that the expression of the gene and health of the transfected cells can be verified before it is introduced to the patient. This method can be useful for therapies where compounds should be released or expressed by specific cell types that can be selected on the culture dish before transplant. At this time, drawbacks include the sparse availability of suitable cells for transplantation, low survival rates of grafted cells, potential for migration and proliferation (in particular for stem-cell lines), the possible need of immunosuppression, and the time-consuming and complex process of preparing the cells. An alternative are polymer-encapsulated cells genetically modified to deliver the product of interest, which stops the possibility of migration, immunological response to the vector or engineered cell (no immunosuppression is required), and provides a container that can be replaced or retrieved in case of adverse effects or low efficacy. Its limitations include the relative large volume size of the capsules, low cell survival, and restricted distribution of the gene product to the area surrounding the capsule. Before clinical trials can begin with any treatment, the methodology and efficacy must be substantiated experimentally. Gene therapies for neurodegenerative disorders present many imique challenges to determine the best treatment for a condition. Long-term gene survival and production of the beneficial protein must be ensured due to the chronic nature of the disorders, and also because the potentially risky neurosurgical interventions should be kept to a minimum. The transduction efficiency of the gene must be optimized in order to prevent unnecessarily large amounts of vector. Target areas for gene therapy will also affect the distribution of the product and in the case of in vivo gene therapy, potential migration of vectors. Unwanted expression may cause sprouting and redirection of neurites, formation of inclusions, and apoptosis (Takayama et al., 1995; Bredesen and Rabizadeh 1997; Tabrizi etal., 2000). Ideally, the systems will also have mechanisms for regulation of gene expression, which until now have not been very successful (Georgievska et al., 2004).

11. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

112

9. NONHUMAN PRIMATE MODELS FOR TESTING GENE THERAPY FOR NEURODEGENERATIVE DISORDERS

V. THE NEED FOR N O N H U M A N PRIMATE EXPERIMENTATION Progression for a gene therapy strategy toward the clinic Starts in the culture dish where efficient, nontoxic vectors are tested. Rodents are the traditional next step in which to first assess in vivo safety and efficacy of a method. Rodents present the advantage that rats and mice are widely available, they have high reproductive rates, require reduced living space, and simple feeding and drinking schedules. Wellestablished models of disease in rodents are essential to assess the potential of a given therapy. Yet, the final proof to predict if therapy can be useful in the clinic is the testing in nonhuman primate models. Methodological issues such as distribution of the compound or selection of target areas depend on brain size and complexity. Physiological response, toxic effects, and immunological reactions may be specie-dependent. Nonhuman primates possess the closest phylogenic and genetic relationship to hunians, which may allow the prediction of potential adverse side effects associated to physiological response to the treatment. Their brain size, morphology, and complexity facilitate the testing of delivery methods and compound distribution compared to rodent brains (Emborg, 2004). For example, in primates, as in humans, the striatum can be divided into the distinct caudate nucleus and putamen, while in rodents it cannot. The muscle anatomy range of movements in primates are much more similar to humans than in rodents, and allow for a more clear examination of the abnormal movements found in neurodegenerative diseases. Monkeys can be trained to perform complex cognitive and motor tasks and data can be obtained that it is comparable to outcome measures used in the clinic.

VI. N O N H U M A N PRIMATE MODELS OF NEURODEGENERATIVE DISORDERS A N D GENE THERAPY The biggest challenge while studying neurodegenerative disorders is the development of models that resemble the disease of interest. Part of the problem has been the lack of knowledge of what causes the disease and in particular for HD, where a gene defect has been clearly identified in all affected patients, how to replicate the genetic component. In the following section, we will briefly describe the monkey models of neurodegenerative disorders currently available

and the gene therapies that were tested in these models. PD models have been at the forefront for assessment of new gene therapy technologies and the results are shaping the way that other diseases are evaluated and treated. A.

Alzheimer^s Disease

AD, the most common neurodegenerative disease, is a progressive disorder of unknown cause characterized by neuronal loss in regions critical for memory and cognition, particularly the hippocampus, entorhinal cortex, and basal forebrain cholinergic system (Jinnah and Friedmann, 1995; Nussbaum and Ellis, 2003). Of particular interest are the affected cholinergic neurons, the loss of which is thought to be a leading factor in the decline of cognitive function (Tuszynski et al., 2004). AD is also characterized by the formation of 6 -amyloid plaques and neurofibrillary tangles, the formation of which are thought to initiate the disease process (Selkoe, 2000). Most of the gene therapy strategies for AD tested in monkeys have focused on the property of nerve growth factor (NGF) to support cholinergic function. Fornix transections have been used in rhesus monkeys to roughly replicate the loss of cholinergic neurons. Encapsulated baby kidney cells genetically modified to deliver NGF can induce robust sprouting of cholinergic fibers within the septum ipsilateral to the transplant (Emerich et al., 1994; Kordower et al., 1994). Another model of AD has taken advantage of the natural decline in cholinergic activity and cognitive function associated with aging observed in rhesus monkeys. Grafting of fibroblasts genetically modified to secrete the growth factor NGF have been shown to reduce the rate of b-amyloid plaque formation, and reverse cholinergic neuronal atrophy in primates as compared to aged controls (Tuszynski et al., 1998; Smith et al., 1999). On the basis of this research, a phase I dose-escalation clinical trial of primary autologous fibroblasts modified to produce NGF is underway (Tuszynski, 2003). B.

Huntington's Disease

HD is a genetic progressive neurodegenerative disorder caused by a repetition mutation in a single gene, IT15, and results in the loss of striatal GABAergic neurons (Kells et al., 2004). Symptoms include a progressive motor disorder including dyskinesia, cognitive decline, and psychiatric problems (Hogarth, 2003). The underlying mechanism, which causes neuronal death is unknown, and therefore gene therapy strategies are

11. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

NONHUMAN PRIMATE MODELS OF NEURODEGENERATIVE DISORDERS AND GENE THERAPY

focused on neuroprotection rather than correcting the known genetic mutation (Gimenez y Ribotta, 2001). Current primate HD models are neurotoxin based. Stereotaxic administration of excitotoxins produce striatal lesions similar to those seen in HD. Excitotoxins cause deleterious effects and ultimately cell death in neurons by over-stimulating glutamate receptors, which causes a massive influx of Na"^ and Ca^^ cations into the cell (Brouillet et al., 1999). Various glutamate receptor agonists, such as ibotenic and quinolinic acid (QA) have been used to model HD in nonhuman primates (Hantraye et al., 1990; Ferrante et al., 1993; Roitberg et al., 2002). Of the various excitotoxins, QA seems to be the best at faithfully mimicking the lesion seen in the disease (Hantraye et a l , 1990). In addition to the morphological similarities, QA-lesioned animals display diskinetic movements and motor deficits similar to those seen in the human form of the disease (Burns et al., 1995). The other primate model makes use of the mitochondrial complex II inhibitor 3-nitropropionic acid (3NP) administered systemically. Chronic 3NP-treated primates show a progressive bilateral cell loss in the caudate and putamen, cognative impairment, and spontaneous and apomorphine-induced dyskinesias, including the hallmark chorea movements of HD (Brouillet et al., 1995; Palfi et al., 1996; Roitberg et al., 2002). Some progress has been made in treating animal models of HD using gene therapy. In the QA model of HD, encapsulated baby hamster kidney fibroblasts were genetically modified to express ciliary neurotrophic factor (CNTF). Striatal projections were protected, and retrograde atrophy was prevented in the target regions (Emerich et al., 1997). In another study using the 3NP model a similar therapy, CNTF was shown to restore both cognitive and motor function (Mittoux et al., 2000). The results of a phase I study evaluating the safety of intracerebral administration of encapsulated BHK cell line engineered to synthesize CNTF into HD patients was recently released (Bloch et al., 2004). Six subjects with stage 1 or 2 HD had one capsule implanted into the right lateral ventricle; the capsule was retrieved and exchanged for a new one every 6 months, over a total period of 2 years. No sign of CNTF-induced toxicity was observed; however, depression occurred in three subjects after removal of the last capsule, which may have correlated with the lack of any future therapeutic option. All retrieved capsules were intact but contained variable numbers of surviving cells, and CNTF release was low in 13 of 24 cases. Improvements in electrophysiological results were observed.

113

and were correlated with capsules releasing the largest amount of CNTF. This phase I study shows the safety, feasibility, and tolerability of this gene therapy procedure. Heterogeneous cell survival, however, stresses the need for improving the technique. C.

Amyotrophic Lateral Sclerosis

ALS is a progressive, fatal neurodegenerative disorder targeting motor neurons. A mutation in the superoxide dismutase 1 (SOD-1) gene is thought to be responsible in 20% of inherited ALS cases. The loss of motor neurons leads to spasms, weakness, and respiratory failure (Miller and Cleveland, 2003). Trophic factors have been proposed as therapies to prevent the motor neuron cell loss and its consequences. The delivery modality plays an important role in ALS treatment, as systemic, subcutaneous, and direct cerebral spinal fluid delivery of neurotrophic factors have been plagued by discouraging results and inefficiency (Alisky and Davidson, 2000; Miller and Cleveland, 2003). Currently, there are no monkey models that specifically replicate ALS. CNTF has been shown in various rodent models to reduce the motor neuron cell death similar to that seen in ALS. On the basis of these results, a phase I study in which six ALS patients were implanted with polymer capsules containing genetically engineered baby hamster kidney cells releasing CNTF was performed (Aebischer et al., 1996). The CNTF-releasing implants were surgically placed within the lumbar intrathecal space. Nanogram levels of CNTF were measured within the patients' cerebrospinal fluid (CSF) for at least 17 weeks post-transplantation, whereas it was undetectable before implantation. Although intrathecal delivery of CNTF was not associated with the limiting side effects observed with systemic delivery, nonclinical improvement could be detected during the observation period. D.

Parkinson's Disease

PD is a progressive neurodegenerative disorder that affects 1% of the population over 55 years of age. The pathologic hallmark of the disease is the loss of dopaminergic neurons in the Substantia Nigra pars compacta (SNpc) and the presence of intracytoplasmic inclusions named Lewy bodies, formed mainly by a-synuclein and ubiquitin. The main symptoms of the disease are tremor, brady and hypokinesia, and balance and gait disturbances. Dopamine replacement by oral administration of L - D O P A is the mainstay of therapy and a tool to identify the disease.

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

114

NONHUMAN PRIMATE MODELS FOR TESTING GENE THERAPY FOR NEURODEGENERATIVE DISORDERS

Although we do not know what causes PD, we have learned that aging, exposure to neurotoxins and mutations in the parkin gene are risk factors associated with its development. On the basis of this information, monkey models of PD have been developed. There is a body of literature documenting the similitude in loss of dopaminergic function with age in humans and monkeys (Emborg and Kordower, 2000). Aged monkeys present many of the alterations observed with PD such as slight tremor, stooped posture, gait and balance disturbances, although they can be very subtle (Emborg et al., 1998). Aged monkeys also present many of the problems that older patients present such as heart dysfunction, diabetes, and arthritis. Yet aged monkeys are not parkinsonian monkeys, and their changes are very subtle. They are not responsive to L-DOPA treatment unless studied with very sensitive tests (Zhang et al., 2000). Neurotoxin-induced models in monkeys used mainly 6-hydrohydopamine (6-OHDA) and the mitochondrial complex I inhibitor l-methyl-4-phenyl1,2,3,6-tetrahyropyridine (MPTP). 6-OHDA requires intracerebral stereotaxic delivery to be effective and marmosets monkeys have been the main specie used. Recovery of the symptoms after a few months has been described (Eslamboli et al., 2003). MPTP is a known human dopaminergic neurotoxin that has the ability to induce a wide range of motor disturbance. The symptoms and efficacy of the treatment depend on the method of administration and age of the animals. Intracarotid administration can induce a unilateral model that is stable for many years and its predictability makes it ideal for neuroprotection studies. Systemic administration can induce a complex bilateral syndrome, and can develop L-DOPA-induced choreoathetosic dyskinesias. Autonomic dysfunction such as cardiac sympathetic denervation that has been associated to PD's orthostatic hypotension can be observed in systemic MPTP-treated monkeys (Goldstein et al., 2003) as well as alterations in the sleep pattern (Almirall et al., 1999) and circadian levels of prolactin and melatonin (Barcia et al., 2003). Combination of systemic and intracarotid MPTP induces an asymmetrical bilateral syndrome, and can be a shortcut to induce a bilateral parkinsonism (Oiwa et al., 2003) similar to bilateral intracarotid administration (Kurlan et al., 1991). MPTP models also have drawbacks. MPTP requires a facility prepared to handle potentially sick animals, as well as trained personnel to work with MPTP and its waste. Possibility of recovery has to be considered when designing experiments. The intracarotid method has less health risk for

monkeys and investigators, yet induces a unilateral syndrome, requires a skill surgical team to be successful and does not develop L-DOPA-induced choreoathetosic dyskinesias. Systemic administration induces a bilateral syndrome, yet, it may require several months of repeated administration and careful follow up to induce a stable syndrome. Short regimes injecting MPTP several days in a row, can be very hard on the monkeys' health, and due to its variable outcome it requires to inject large numbers of monkeys to obtain a final number of monkeys with comparable syndrome. Combination of systemic and intracarotid MPTP can induce severe symptoms although less than systemic administration, it requires repeated dosing and close follow up of the animal to define syndrome and stability. Bilateral intracarotid induces a very severe syndrome that is hard on the animals and difficult for evaluation. Developnient of a genetic monkey model by overexpression of a-synuclein driven by AAV was attempted in three marmoset monkeys (Kirik et al., 2003). The main symptom described was head bias in two animals following strategies for gene therapy in PD focus on two main areas: restoration of the neurotransmitter balance or neuroprotection of affected cells (Tables 1 and 2). For restoration, dopamine-enhancement approaches utilize genes, which encode for enzymes in the dopamine production pathway, such as tyrosine hydroxylase (TH), aromatic-amino-dopa-decarboxylase (AADC), and GTP cyclohydrolase (GCH). Several studies in monkeys have demonstrated functional and histological improvements using vectors encoding for one or more of these enzymes (Anton et al., 1994; During et al., 1998; Bankiewicz et al., 2000; Muramatsu et a l , 2002). A novel approach is the subthalamic delivery of the enzyme glutamic acid decarboxylase (GAD) in order to increase the levels of the inhibitory

TABLE 1 Gene Therapy Approaches for the Treatment of Neurodegenerative Disorders Neuroprotective strategy

Restorative strategy

Goal of preventing or stopping cell loss

Goal of restoring function

Intervention early in disease course, when there are more cells to be protected

Intervention late in disease course, when few cells remain

Examples: antiapoptotic agents, trophic factors

Examples: neurotransmitter replacement, trophic factors

[I. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

115

NONHUMAN PRIMATE MODELS OF NEURODEGENERATIVE DISORDERS AND GENE THERAPY

Q

s a

PL,

O

^3 O

+ a> 13 «1^

2 ^ cS +

^ § s

> 2

oJ 5 '^

-2 S

(^ .§ ^ o

OH

(0

S ^ o

o

I i O CI.

<

O

5C'2

3

iS

0)

c«

9^ T3 TO

•+-•

a is

a

^ .a §

6 0* nl

ri 0

C 0

c 3

ga ^ C N

o

HJ

K

N

II M ^

N

S

(C II HJ

e

0> TJ yi

p^

<

S

^II PE s;

ir^

u II K

5

OH

CO

II

0

^

U Q

P5 o

V 3

+

(U -T3 =J C/)

>-

9
^

^

^

in

OK

O

S

0

a>

*-* u + <

0

2 vO

tin vO

CO

^

Q II

II

O

O

HH

:^

Q 0

CO

II s;

PLH

LO

^

II s

0 0

^II

tlH

R

D 0

^

II K

TS O

45 a>

0)

s

> H

tin

PH

II

pL. PtH

0

0 PU

OS 1 0

II

PH

OS

a> Q

o -i^

P >

>

> <

C

g

0

>; '>

X 0 0)

0

0)

3

0

^

O

>

>

'0

a

0

0

<

• >

X PLJ

u U

gU

<

0^ O

B

a u

u

a en

5S

PH

s

^

f^r

0^

en

en

;3

:3 en

1

(X)

o"

^ ON ON r-l

•

a> u C

fC

a>

o

Si 0^

P^

o; C C

<

00

00

ON

ON

T—1

13 "S

ON rH

^

0 0

Ci-

^

d

1

u

o'

0

a S

cs

£S

13 "S

13 "S

13 QJ

;3 _en

•4->

VH

a;

13 0

a fd

CQ

0) en 0

CS

a>

0

1

CO" 0 0

(D

WD

0)

Id

CO^ 0 0

13

UH

;H

0)

"3

0 0

13 H-i

13 "S

1

cu

MD

s ;3

1 ^

>;

£

1

X c 0 _o 1

a u

CD

S

13

13

1

CD

C 0 C

0

0 0 PQ

0 0

13

>

<

^ 0 X 0

^i

13

PH

a^

0

<

0

a

3

1 C 0

f3 0

0

u

U

ir^ 0 0

in

Ci

0 0 CN^

13

13

'0

'0

In W

In

^

a

In

2

en

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

C _o '55

(d

^QJ_>

0)

1 PJ

0

<

0 K C 0 .2

y

Csl

OH

pq

en

II

116

9. NONHUMAN PRIMATE MODELS FOR TESTING GENE THERAPY FOR NEURODEGENERATIVE DISORDERS

neurotransmitter GABA and restore the neurotransmitter imbalance (During et al., 2001; Carbon-Correll et al., 2004). Neuroprotective strategies in PD attempt to use viral vectors to deliver proteins encoding for neurotrophic factors to prevent further progression of the disease. Several growth factors have been utilized, but the most effective for PD appears to be glial cell-linederived neurotrophic factor (GDNF). Monkey studies have shown functional improvements, as well as neuronal protection and restoration (Kordower et al., 2000; Eslamboli et al., 2003; Kishima et al., 2004). Compared to other neurodegenerative disorders, it seems that more gene therapy strategies have been tested in monkey models of PD than in any other model of disease. There are several reasons for this: (1) the availability of monkey models; (2) the localized nature of the disease that is a good match for gene therapy stereotaxic intracerebral approaches; (3) the higher standards for PD treatments due to the improvement of pharmacological treatments beyond L-DOPA and the development of deep brain stimulation treatments. Although these treatments do not cure or stop the progression of the disease, can be extremely effective to improve function for a period of years. Currently, two phase I trials for PD using AAV have been initiated (Mandel and Burger, 2004). One proposes to increase the enzyme GAD to restore neurotransmitter balance (During et al., 2001). The other proposes to increase the striatal levels of AADC to increase L - D O P A metabolization. VIL WHAT MAKES A GENE THERAPY STUDY IN MONKEYS RELEVANT? Any monkey study has the potential to provide relevant information toward a clinical trial. A pilot study in few animals even in normal, intact monkeys can provide essential data on gene expression, gene product distribution, potential toxicity, or elicited immune response before planning an extensive and expensive project. Yet, a clinical trial should be justified by safety and efficacy. Outcome measures similar to the ones used in the clinic will increase the value of an experiment toward predicting the success of a given therapy. Blind acquisition and analysis of the data is essential for unbiased interpretation of the results. The animal model should replicate as close as possible the behavioral, neurochemical, and neuropathology components of the disease. In order to be able to

assess the results of a protective or restorative therapy, an animal model of any neurodegenerative disorder should induce a replicable lesion, and the cell-loss be stable over time, without spontaneous recovery. Most important for neuroprotection is that the model should provide a window of opportunity in which the neuroprotective strategy can work. Hence, the study design is critical for the interpretation of the results and the assessment of their clinical relevance. The progressive nature of these diseases has emphasized the development and assessment of neuroprotective strategies. The initiation and duration of a neuroprotective treatment depends on the animal model, the neuroprotective agent and its method of delivery (Emborg and Kordower, 2000). Prevention of cell loss requires delivery of neuroprotective agents when there are still cells to be protected. Ideally, neuroprotective substances are administered before the onset of the signs. In the case of HD, the knowledge of the genetic defect allows to predict who will develop the disease and hypothetically start treatment before the onset of the signs. However, we currently lack the tools to preclinically diagnose AD, PD, and ALS, and our best alternative is to start neuroprotective interventions early after the diagnosis of the disease. A parallel design for animal models is to intervene soon after inducing the syndrome. Although it could be argued that additional cell death may occur months to years after neurotoxin exposure (Langston et al., 1999), most of the cell loss in animal models occurs in the first weeks after lesioning and that is when a neuroprotective intervention will be the most beneficial. When using aged animals, timing becomes unimportant. Post-menopausal age monkeys will ensure a certain degree of age-related brain dysfunction. Overall, models that induce a predictable lesion over a relatively long period of time provide a window of opportunity for neuroprotective strategies to succeed. Depending on the experimental design, animals can be selected after lesioning but before treatment and randomly assigned to the experimental groups. If that it is not possible due to the properties of the neuroprotective compound and/or the delivery method, higher numbers of animals should have to be used to account for the variability in the response to the neurotoxin treatment. Later interventions (after the core lesion is completed) should be regarded as regenerative or restorative strategies instead of neuroprotective. An advantage of testing restorative strategies is that the experimental design allows for additional time to evaluate the monkeys' syndrome, ensure the stability of the lesion or continue the lesioning until the syndrome is established.

11. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

ACKNOWLEDGMENTS

The termination date of the experiment is as critical in assessing the strategies as the starting date of the treatment. The experimental design has to take into consideration the amount of time required for the specific model to induce the maximum loss of cells. The appropriate treating time period will ensure that a complete lesion can be found in the control-treatment group and, if the therapy is successful, significant recovery in the neuroprotective-treatment group. Treatments that continue for longer periods of time may induce a combination of neuroprotective and regenerative activities. In those cases behavioral and in vivo imaging follow up may help provide a functional evaluation of the affected system.

VIIL

A FINAL COMMENT

The number of patients with neurodegenerative disorders is increasing, as well as the expectations on new treatments. Gene therapy is not impervious to this trend. It is essential to remember that no one specific study by its own merit qualifies as the ultimate proof of therapeutic benefit. Moreover, it is the replication of the positive results in different animal models by independent groups with multiple outcome measures that validates the success of a therapy and its chances of becoming part of a clinical trial. And the final steps, unbiased and in-depth nonhuman primate research, will ensure that only the treatments with lowest risk and highest benefits reach the patients.

ACKNOWLEDGMENTS The authors gratefully acknowledge the support by grants from the Kinetics Foundation, NIH NINDS RO1-NS40578 and by NIH Grant 5P51RR000167 to the Wisconsin National Primate Research Center, University of Wisconsin-Madison. References Aebischer, P., Schluep, M., Deglon, N., Joseph, J.M., Hirt, L., Heyd, B., Goddard, M., Hammang, J.P., Zum, A.D., Kato, A.C., Regli, R and Baetge, E.E. (1996) Intrathecal delivery of CNTF using encapsulated genetically modified xenogeneic cells in amyotrophic lateral sclerosis patients. Nat. Med., 2: 696-699. Alisky, J.M. and Davidson, B.L. (2000) Gene therapy for amyotrophic lateral sclerosis and other motor neuron diseases. Hum. Gene Ther., 11: 2315-2329. Almirall, H., Pigarev, I., de la Calzada, M.D., Pigareva, M., Herrero, M.T. and Sagales, T. (1999) Nocturnal sleep structure and

117

temperature slope in MPTP treated monkeys. J. Neural Transm., 106:1125-1134. Anton, R., Kordower, J.H., Maidment, N.T., Manaster, J.S., Kane, D.J., Rabizadeh, S., Schueller, S.B., Yang, J., Rabizadeh, S., Edwards, R.H., Markham, C.H. and Bredesen, D.E. (1994) Neural-targeted gene therapy for rodent and primate hemiparkinsonism. Exp. Neurol., 127: 207-218. Baekelandt, V., De Strooper, B., Nuttin, B. and Debyser, Z. (2000) Gene therapeutic strategies for neurodegenerative diseases. Curr. Opin. Mol. Ther., 2: 540-554. Bankiewicz, K.S., Eberling, J.L., Kohutnicka, M., Jagust, W., Pivirotto. P., Bringas, J., Curmingham, J., Budinger, T.F. and HarveyWhite, J. (2000) 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, 164: 2-14. Barcia C., Bautista V., Sanchez-Bahillo A., Fernandez-Villalba E., Navarro-Ruis J.M., Barreiro A.F., Poza, Y, Poza, M. and Herrero M.T. (2003) Circadian determinations of Cortisol, prolactin and melatonin in chronic methyl-phenyl-tetrahydropyridine-treated monkeys. Neuroendocrinology, 78:118-128. Bloch, J., Bachoud-Levi, A.C., Deglon, N., Lefaucheur, J.P., Winkel, L., Palfi, S., Nguyen, J.P, Bourdet, C., Gaura, V., Remy, P., Brugieres. P., Boisse, M.F., Baudic, S., Cesaro, P., Hantraye, P., Aebischer, P. and Peschanski, M. (2004) Neuroprotective gene therapy for Huntington's disease, using polymer-encapsulated cells engineered to secrete human ciliary neurotrophic factor: results of a phase I study Hum. Gene Ther., 15: 968-975. Bohn, M.C. and Choi-Lundberg, D.L. (1998) Gene therapies for Parkinson's disease. In: Chiocca E.A. and Breakfield X.O. (Eds.), Gene Therapy for Neurological Disorders and Brain Tumors. Humana Press, NJ, pp. 377-397. Bredesen, D.E. and Rabizadeh, S. (1997) p75NTR and apoptosis: Trk-dependent and Trk-independent effects. Trends Neurosci., 20: 287-290. Brouillet, E., Conde, F , Beal, M.F and Hantraye, P (1999) Replicating Huntington's disease phenotype in experimental animals. Prog. NeurobioL, 59: 427-468. Brouillet, E., Hantraye, P., Ferrante, R.J., Dolan, R., Leroy-Willig, A., Kowall, N.W. and Beal, M.F. (1995) Chronic mitochondrial energy impairment produces selective striatal degeneration and abnormal choreiform movements in primates. Proc. Natl. Acad. Sci., 92: 7105-7109. Burns, L.H., Pakzaban, P., Deacon, T.W., Brownell, A.L., Tatter, S.B., Jenkins, B.G. and Isacson, O. (1995) Selective putaminal excitotoxic lesions in non-human primates model the movement disorder of Huntington's disease. Neuroscience, 64:1007-1017. Carbon-Corell, M., Emborg, M., Ma, Y, Holden; J., Kordower, J., Feigin, A.S., During, M., KapUtt, M. and Eidelberg, D. (2004) In vivo study of metabolic brain function in parkinsonian macaques following GAD therapy. Abst. 18th Armual symposia on Etiology, Parthogenesis, and Treatment of Parkinson's Disease and other Movement Disorders. Costantini, L.C., Bakowska, J.C., Breakefield, X.O. and Isacson, O. (2000) Gene therapy in the CNS. Gene Ther., 7: 93-109. During, M.J., Samulski, R.J., Elsworth, J.D., Kaplitt, M.G., Leone, P , Xiao, X., Li, J., Freese, A., Taylor, J.R., Roth, R.H., Sladek, J.R. Jr., O'Malley, K.L. and Redmond, D.E. Jr. (1998) In vivo expression of therapeutic human genes for dopamine production in the caudates of MPTP-treated monkeys using an AAV vector. Gene Ther., 5: 820-827.

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

118

9. NONHUMAN PRIMATE MODELS FOR TESTING GENE THERAPY FOR NEURODEGENERATIVE DISORDERS

During, M.J., Kaplitt, M.G., Stem, M.B. and Eidelberg, D. (2001) Subthalamic GAD gene transfer in Parkinson disease patients who are candidates for deep brain stimulation. Hum. Gene Ther., 12: 1589-1591. Emborg, M.E. (2004) Evaluation of animal models of Parkinson's disease for neuroprotective strategies. J. Neurosci. Meth., 139: 121-143. Emborg, M.E. and Kordower, J.H. (2000) Delivery of therapeutic molecules into the CNS. Prog. Brain Res., 128: 323-332. Emborg, M.E., Ma, S.Y., Mufson, E.J., Levey, A.I., Taylor, M.D., Brown, W.D., Holden J.E. and Kordower, J.H. (1998) Age-related declines in nigral neuronal function correlate with motor impairments in rhesus monkeys. J. Comp. Neurol., 401: 253-265. Emerich, D.F., Winn, S.R., Hantraye, P.M., Peschanski, M., Chen, E.Y., Chu, Y, McDermott, P, Baetge, E.E. and Kordower, J.H. (1997) Protective effect of encapsulated cells producing neurotrophic factor CNTF in a monkey model of Huntington's disease. Nature, 386: 395-399. Emerich, D.F., Winn, S.R., Harper, J., Hammang, J.P, Baetge, E.E. and Kordower, J.H. (1994) Implants of polymer-encapsulated human NGF-secreting cells in the nonhuman primate: rescue and sprouting of degenerating cholinergic basal forebrain neurons. J. Comp. Neurol., 349: 148-164. Eriksdotter Jonhagen, M., Nordberg, A., Amberla, K., Backman, L., Ebendal, T., Meyerson, B., Olson, L., Seiger, S.M., Theodorsson, E., Viitanen, M., Winblad, B. and Wahlund, L.O. (1998) Intracerebroventricular infusion of nerve growth factor in three patients with Alzheimer's disease. Dement. Geriatr. Cogn. Disord., 9: 246-257. Eslamboli, A., Baker, H.E, mdley R.M. and Annett, L.E. (2003) Sensorimotor deficits in a unilateral intrastriatal 6-OHDA partial lesion model of Parkinson's disease in marmoset monkeys. Exp. Neurol., 183:418-429. EslamboU, A., Cummings, R.M., Ridley, R.M., Baker, H.E, Muzyczka, N., Burger, C , Mandel, R.J., Kirik, D. and Annett, L.E. (2003) Recombinant adeno-associated viral vector (rAAV) delivery of GDNF provides protection against 6-OHDA lesion in the common marmoset monkey (Callithrix jacchus). Exp. Neurol., 184: 536-548. Eslamboli, A., Georgievska, B., Ridley, R.M., Baker, H.R, Muzyczka, N., Burger, C , Mandel, R.J., Annett, L. and Kirik, D. (2005) Continuous low-level glial cell line-derived neurotrophic factor delivery using recombinant adeno-associated viral vectors provides neuroprotection and induces behavioral recovery in a primate model of Parkinson's disease. J. Neurosci., 25: 769-777. Federoff, H.J. (2003) CNS diseases amenable to gene therapy. Ernst Schering Res. Found. Workshop, 43: 117-158. Ferrante, R.J., Kowall, N.W, Cipolloni, PB., Storey E. and Beal, M.E (1993) Excitotoxin lesions in primates as a model for Huntington's disease: histopathologic and neurochemical characterization. Exp. Neurol., 119: 46-71. Freeman, T.B. (1997) From transplants to gene therapy for Parkinson's disease. Exp. Neurol., 144: 47-50. Friedmann, T. and Roblin, R. (1972) Gene therapy for human genetic disease? Science, 175: 949-955. Georgievska, B., Jakobsson, J., Persson, E., Ericson, C , Kirik, D. and Lundberg, C. (2004) Regulated delivery of glial cell line-derived neurotrophic factor into rat striatum, using a tetracycline-dependent lentiviral vector. Hum. Gene Ther., 15: 934-944. Gimenez y Ribotta, M. (2001) Gene therapy strategies in neurodegenerative diseases. Histol. Histopathol., 16: 883-893. Glorioso, J.C, Mata, M. and Fink, D.J. (2003) Therapeutic gene transfer to the nervous system using viral vectors. J. Neurovirol., 9: 165-172.

Glorioso, J.C. and Fink, D.J. (2004) Herpes vector-mediated gene transfer in treatment of diseases of the nervous system. Armu. Rev Microb., 58: 253-271. Goldstein, D.S., Li, S.T., Hohnes, C. and Bankiewicz, K. (2003) Sympathetic innervation in the l-methyl-4-phenyl-l,2,3,6-tetrahydropyridine primate model of Parkinson's disease. J. Pharmacol. Exp. Ther., 306:855-860. Hantraye, P., Riche, D., Maziere, M. and Isacson, O. (1990) A primate model of Huntington's disease: behavioral and anatomical studies of unilateral excitotoxic lesions of the caudate-putamen in the baboon. Prog. Neurobiol., 59: 427-468. Hogarth, P. (2003) Huntington's disease: a decade beyond gene discovery. Curr. Neurol. Neurosci. Rep., 3: 279-284. Hsich, G., Sena-Esteves, M. and Breakefield, X.O. (2002) Critical issues in gene therapy for neurologic disease. Hum. Gene Ther., 13: 579-604. Jinnah, H.A. and Friedmann, T. (1995) Gene therapy and the brain. Br. Med. BuU., 51: 138-148. Kells, A.P., Fong, D.M., Dragimow, M., During, M.J., Young, D. and Connor, B. (2004) AAV-mediated gene delivery of BDNF or GDNF is neuroprotective in a model of Huntington's disease. Mol. Ther., 9: 682-688. Kirik, D., Annett, L.E., Burger, C , Muzyczka, N., Mandel, R.J., Bjorklund, A. (2003) Nigrostriatal alpha-sjmucleinopathy induced by viral vector-mediated overexpression of human alpha-synuclein: a new primate model of Parkinson's disease. Proc. Natl. Acad. Sci. USA., 100: 2884-2889. Kishima, H., Poyot, T., Bloch, J., Dauguet, J., Conde, E, DoUe, E, Hinnen, E, Pralong, W, Palfi, S., Deglon, N., Aebischer, P. and Hantraye, P (2004) Encapsulated GDNF-producing C2C12 cells for Parkinson's disease: a pre-clinical study in chronic MPTPtreated baboons. Neurobiol. Dis., 16: 428-439. Kordower, J.H., Emborg, M.E., Bloch, J., Ma, S.Y, Chu, Y, Leventhal, L., McBride, J., Chen, E.Y, Palfi, S., Roitberg, B.Z., Brown, W.D., Holden, J.E., Pyzalski, R., Taylor, M.D., Carvey P , Ling, Z., Trono, D., Hantraye, P., Deglon, N. and Aebischer, P (2000) Neurodegeneration prevented by lentiviral vector delivery of GDNF in primate models of Parkinson's disease. Science, 290: 767-773. Kordower, J.H., Palfi, S., Chen, E.Y, Ma, S.Y, Sendera, T., Cochran, E.J., Mufson, E.J., Penn, R., Goetz, C.G. and Comella, C D . (1999) Clinicopathological findings following intraventricular glial derived neurotrophic factor treatment in a patient with Parkinson's disease. Ann. Neurol., 46: 419-424. Kordower, J.H., Winn, S.R., Liu, Y.T., Mufson E.J., Sladek, J.R. Jr., Hammang, J.P, Baetge, E.E. and Emerich, D.F. (1994). The aged monkey basal forebrain: rescue and sprouting of axotomized basal forebrain neurons after grafts of encapsulated cells secreting human nerve growth factor. Proc. Natl. Acad. Sci. USA, 91:10898-10902. Kurlan, R., Kim, M.H. and Gash, D.M. (1991) The time course and magnitude of spontaneous recovery of parkinsonism produced by intracarotid administration of l-methyl-4-phenyl-l,2,3,6-tetrahydropyridine to monkeys. Ann. Neurol., 29: 677-679. Langston, J.W., Fomo, L.S., Tetrud, J., Reeves, A.G., Kaplan, J.A. and Karluk, D. (1999) Evidence of active nerve cell degeneration in the substantia nigra of humans years after l-methyl-4-phenyl1,2,3,6-tetrahydropyridine exposure. Ann. Neurol., 46:598-605. Lansbury, PT. Jr. (2004) Back to the future: the 'old-fashioned' way to new medications for neurodegeneration. Nat. Med., 10: S51-S57. Mandel, R.J. and Burger, C. (2004) Clinical trials in neurological disorders using AAV vectors: promises and challenges. Curr. Opin. Mol. Ther., 6: 482-490.

11. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

ACKNOWLEDGMENTS Mata, M., Glorioso, J.C. and Fink, DJ. (2003) Gene transfer to the nervous system: prospects for novel treatments directed at diseases of the aging nervous system. J. Gerontol. A Biol. Sci. Med. Sci., 58: M1111-M1118. Miller, T.M. and Cleveland, D.W. (2003) Has gene therapy for ALS arrived? Nat. Med., 9:1256-1257. Mittoux, v., Joseph, J.M., Conde, R, Palfi, S., Dautry, C , Poyot, T., Bloch, J., Deglon, N., Ouary, S., Nimchinsky, E.A., Brouillet, E., Hof, P.R., Peschanski, M., Aebischer, P and Hantraye, R (2000) Restoration of cognitive and motor functions by ciliary neurotrophic factor in a primate model of Huntington's disease. Hum. Gene Ther., 11:1177-1187. Morsy M.A., Gu, M., Motzel, S., Zhao, J., Lin, J., Su, Q., Allen, H., Frar\lin, L., Parks, R.J., Graham, F.L., Kochanek, S., Bett, A.J. and Caskey, C.T. (1998) An adenoviral vector deleted for all viral coding sequences results in enhanced safety and extended expression of a leptin transgene. Proc. Natl. Acad. Sci. USA, 95: 7866-7871. Muramatsu, S., Fujimoto, K., Ikeguchi, K., Shizuma, N., Kawasaki, K., Ono, R, Shen, Y., Wang, L., Mizukami, H., Kume, A., Matsumura, M., Nagatsu, I., Urano, R, Ichinose, H., Nagatsu, T., Terao, K., Nakano, I. and Ozawa, K. (2002) Behavioral recovery in a primate model of Parkinson's disease by triple transduction of striatal ceUs with adeno-associated viral vectors expressing dopamine-synthesizing ei\zymes. Hum. Gene Ther., 13: 345-354. Naldiiu, L., Blomer, U., Gallay, R, Ory, D., Mulligan, R., Gage, RH., Verma, I.M. and Trono, D. (1996) In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science, 272: 263-267. Nussbaum, R.L. and Ellis, C.E. (2003) Alzheimer's disease and Parkinson's disease. New Engl. J. Med., 348:1356-1364. Oiwa, Y., Eberling, J.L., Nagy, D., Pivirotto, P., Emborg, M.E. and Bankiewicz, K.S. (2003) Overlesioned hemiparkinsonian nonhuman primate model: correlation between clinical, neurochemical and histochemical changes. Front Biosci., 8: al55-al66. Palfi, S., Ferrante, R.J., Brouillet, E., Beal, M.F, Dolan, R., Guyot, M.C., Peschanski, M. and Hantraye, P. (1996) Chronic 3-nitropropionic acid treatment in baboons replicates the cognitive and motor deficits of Huntington's disease. J. Neurosci., 16: 3019-3025. Roitberg, B.Z., Emborg, M.E., Sramek, J.G., Palfi, S. and Kordower, J.H. (2002) Behavioral and morphological comparison of two nonhuman primate models of Huntington's disease. Neurosurgery, 50:137-145 Romano, G., Michell, P., Pacilio, C. and Giordano, A. (2000) Latest developments in gene transfer technology: achievements, per-

119

spectives, and controversies over therapeutic applications. Stem Cells, 18: 19-39. Ross, C.A. and Poirier, M.A. (2004) Protein aggregation and neurodegenerative disease. Nat. Med., 10: S10-S17. Schneider, J.S. and Wade, T.V. (2003) Experimental parkinsonism is associated with increased pallidal GAD gene expression and is reversed by site-directed antisense gene therapy. Mov. Disord., 18: 32-40. Selkoe, D.J. (2000) Toward a comprehensive theory for Alzheimer's disease:hypothesis: Alzheimer's disease is caused by the cerebral accumulation and cytotoxicity of amyloid -protien. Ann. NY Acad. Sci., 924: 17-25. Sibal, L.R. and Samson, K.J. (2001) Nonhuman primates: a critical role in current disease research. ILAR J., 42: 74-84. Smith, D.E., Roberts, J., Gage, RH. and Tuszynski, M.H. (1999) Ageassociated neuronal atrophy occurs in the primate brain and is reversible by growth factor gene therapy. Proc. Natl. Acad. Sci. USA, 96:10893-10898. Tabrizi, S.J., Orth, M., Wilkinson, J.M., Taanman, J.W, Warner, T.T., Cooper, J.M. and Schapira, A.H. (2000) Expression of mutant alpha-synuclein causes increased susceptibility to dopamine toxicity. Hum. Mol. Genet., 9: 2683-2689. Takayama, H., Ray, J., Raymon, H.K., Baird, A., Hogg, J., Fisher, L.J. and Gage, RH. (1995) Basic fibroblast growth factor increases dopaminergic graft survival and function in a rat model of Parkinson's disease. Nat. Med., 1: 53-58. Tuszynski, M.H. (2003) Gene therapy for neurological disease. Expert Opin. Biol. Ther., 3: 815-828. Tuszynski, M.H. and Blesch, A. (2004) Nerve growth factor: from animal models of cholinergic neuronal degeneration to gene therapy in Alzheimer's disease. Prog. Brain Res., 146: 441-449. Tuszynski, M.H., Smith, D.E., Roberts, J., McKay, H. and Mufson, E. (1998) Targeted intraparenchymal delivery of human NGF by gene transfer to the primate basal forebrain for 3 months does not accelerate beta-amyloid plaque deposition. Exp. Neurol, 154: 573-582. United Nations Division of Societal and Economic Affairs (UNDSEA), Population Division (2002) World Population Ageing 1950-2050. United Nations Secretariat, New York. Zhang, Z., Andersen, A., Smith, C , Grondin, R., Gerhardt, G., Gash, D. (2000) Motor slowing and parkinsonian signs in aging rhesus monkeys mirror human aging. J. Gerontol A Biol. Sci. Med. Sci., 55: B473-480.

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

C H A P T E R

10 Delivery of Molecular Therapeutics into the CNS and their Distribution within the Brain Piotr Hadaczek, John Forsayeth, Krystof Bankiewicz

Abstract: Brain is a particularly challenging organ for gene therapy. The clinical failure of potentially effective molecular therapeutics is often due not so much to a lack of potency, but rather to shortcomings in the method by which they are delivered. The major problem in gene delivery to brain is the presence of the blood-brain barrier (BBB) that often impedes vector transfer to this organ and to spinal cord. Research has recently focused on the development of new strategies to deliver gene therapy vectors to the CNS more effectively. This review details some recent advances in the field of delivery of genes to CNS. It describes methods of disrupting and bypassing the BBB, as well as local methods of drug delivery, such as convection-enhanced delivery (CED), with additional enhancements for vector distribution within the brain. Applications of gene delivery methods for brain tumors are also discussed. Keywords: blood-brain barrier (BBB); convection-enhanced delivery (CED); transport; perivascular space; adeno-associated virus (AAV); liposomes

Despite progress in current treatment modalities, the clinical outcome of neurological diseases remains dismal. Contemporary medicine seeks novel strategies to treat illnesses of the CNS. Efficient gene transfer into brain cells is a major goal both for molecular biologists studying the nervous system and for those attempting to develop gene therapy for various human neurological disorders including brain tumors. Although construction of vectors has advanced rapidly, significant limitations remain unresolved when it comes to delivery of various vectors to the CNS. No matter how sophisticated molecular therapeutics become, the key issue is how to deliver the drug most efficiently to the brain without causing systemic side effects and toxicity to the brain. Improvements in brainspecific drug delivery systems must be devised to overcome the limited therapeutic efficacy that is often a result of inadequate distribution of gene therapy vectors in brain tissue. Improved therapy is likely to depend on much more detailed knowledge of the anatomical and

Gene Therapy of the Central Nervous System: From Bench to

physiological features of the brain that guide movement of vectors through target tissues.

I.

BYPASSING THE BLOOD-BRAIN BARRIER (BBB)

Systemic administration to brain of therapeutic agents, such as small molecules and especially high molecular weight (HMW) agents like monoclonal antibodies, liposomes, and viral vectors, is severely limited by the BBB and other physiological obstacles, e.g., increased interstitial pressure within brain tumors. The BBB is created by the tight apposition of endothelial cells lining blood vessels in the brain, forming a barrier between the circulation and the brain parenchyma (e.g. astrocytes, microglia). Blood-borne immune cells such as lymphocytes, monocytes, and neutrophils cannot normally penetrate this barrier. A thin basement

121

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

122

10. DELIVERY OF MOLECULAR THERAPEUTICS INTO THE CNS AND THEIR DISTRIBUTION WITHIN THE BRAIN

membrane, composed of lamin, fibronectin and other proteins, surrounds the endothelial cells, and associated pericytes, and provides both mechanical support and a barrier function (Rubin and Staddon, 1999). Thus, the BBB is crucial for preventing infiltration of pathogens and restricting adaptive immune responses in the central nervous system, as well as for preventing disorganization of the fragile neural network. When designing strategies for delivery of therapeutics to the brain, one must keep in mind the existence of this tightly protected and closed environment. For that reason, brain is probably one of the most challenging human organs to target for gene therapy approaches. Various strategies that have been used for manipulating the BBB for drug delivery to the brain include osmotic and chemical opening of the BBB, as well as the use of transport/carrier systems. The BBB to water-soluble drugs and macromolecules can be opened in vivo by infusion of a hypertonic solution of arabinose or mannitol into the carotid artery for 30 s (KroU and Neuwelt, 1998; Rapoport, 2001). Opening the BBB involves widening of tight junctions between endothelial cells of the cerebrovasculature, and is mediated by endothelial cell shrinkage, vascular dilatation associated with removal of water from brain, and modulation of the contractile state of the endothelial cytoskeleton and junctional proteins by increased intracellular calcium ions. A 10-fold increase in BBB permeability to intravascular substances, lasting about 10 min after osmotic exposure, reflects both increased diffusion and bulk fluid flow from blood into brain. Furthermore, recent evidence indicates that the duration of peak BBB opening can be extended beyond 30 min, by pre-treatment with a Na^+VCa^^^^ channel blocker. In experimental animals, the osmotic method has been used to grant wide access to brain of watersoluble drugs, peptides, antibodies, viral vectors for gene therapy, and enzymes (Rapoport, 2001). Ongoing clinical studies suggest that this method, when used with intra-arterially administered anti-cancer drugs, can prolong survival in patients with malignant brain tumors, with minimal morbidity (KroU and Neuwelt, 1998; Siegal et al., 2000; Rapoport, 2001). Another potentially useful alternative to osmotic disruption of the BBB is the use of bradykinin that seems to selectively increase permeability of capillaries in brain tumors. This endogenous oligopeptide, produced by activation of the kallikrein-kinin system, can stimulate B2 receptors present at the BBB, and initiate second messenger cascades that induce opening of the tight junctions. Histamine and leukotrienes act in a similar way. As the use of endogenous ligands is generally

compromised by both the high concentrations required to increase BBB permeability as well as by physiological side effects, one method of administration involves infusion of bradykinin directly into the internal carotid artery (Inamura and Black, 1994; Nomura et al., 1994). This approach has given some success in animal studies. However, bradykinin's very short half-life (several seconds), potent vasoactive metabolite, and poor therapeutic index limit its safety and usefulness. For this reason, the selective B2 bradykinin agonist Cereport (RPM-7 or Labradimil) was developed to improve therapeutic index via longer B2 receptor stimulation (Bartus et al., 1996; Boddy and Thomas, 1997). The timing of a 15-min infusion of Cereport with an overlapping 5-min infusion of the drug has been demonstrated as the optimal dosing regimen. Electron microscopy studies have demonstrated that intravenous co-administration of Cereport with the electron-dense marker. Lanthanum, caused disengagement of the tight junctions between the endothelial cells comprising the BBB (Sanovich et al., 1995). In brain tumors, modulation of cellular drug transporters, such as P-glycoprotein (Pgp), has been proposed as an optional strategy to open the BBB. Pgp is a transmembrane protein of 170 kDa, formed by two homologous subunits that form an ATP-dependent efflux pump, localized on the apical side of cells (Higgins, 1992). Although the general role of the Pgp system is to eliminate numerous xenobiotics by actively pumping these molecules out of the cell, and consequently decreasing their intracellular concentrations, Pgp is also implicated in the transport and regulation of endogenous molecules such as hormones (Ueda et al., 1992; Wolf and Horwitz, 1992) and phospholipids (Pohl et al., 2002). Pgp is detected in numerous organs including brain and peripheral nerves (capillary endothelial cells). Therefore, modulation of Pgp activity, especially its down-regulation, can help to distribute drugs to the CNS. Inhibition of Pgp increases the brain penetration of anti-cancer drugs. Global neuronal gene expression in the brain is also possible with trans-vascular delivery of the gene through non-viral gene transfer technology (Shi and Pardridge, 2000). Liposomes (phospholipid bilayers formed into spheres in the presence of water) are a commonly used vehicle for administering therapeutic agents, including drugs and genes, to specific target sites. In this approach, the non-viral plasmid is encapsulated in the interior of a neutral liposome that prevents degradation of the DNA by endonucleases. By conjugating 2 kD polyethylene glycol (PEG) to the surface of the liposome, reduced adsorption

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

LOCAL METHODS OF DRUG DELIVERY TO THE CNS

of serum proteins to the liposome is achieved (CuUis et al., 1998). This prolongs the blood residence time of the nanocontainer due to minimization of its uptake by the cells lining the reticulo-endothelial system. In addition to facilitate successful transit of the BBB, liposomes are conjugated to a receptor-specific targeting ligand, such as a monoclonal antibody (MAb). The antibody attaches to a receptor expressed on the BBB, w^hich enables sequential receptor-mediated endocytosis. The targeting MAb acts as a molecular Trojan horse to ferry the pegylated immunoliposome (PIL) across biological barriers in the brain via endogenous transport systems (Pardridge, 2001). Brain capillary endothelial cells do possess specific receptor-mediated transport mechanisms that can be exploited. Transferrin (Tf) and its receptor (TfR) play an important role as the brain drug transporter vector (Li et al., 2002). The TfRs are concentrated on the plasma membrane of brain endothelial cells. Coupling liposomes (terminal ends of PEG) with MAbs against those receptors helps PILs to cross the BBB. Endogenous genes w^ere delivered to mouse brain v^ith the rat 8D3 MAb to the mouse TfR (Shi et al., 2001), and to the rat brain with the murine OX26 MAb to the rat TfR (Shi and Pardridge, 2000; Shi et al., 2001). After intravenous administration of a 6-7 kb plasmid, encoding either jS-galactosidase or luciferase, widespread transgene expression was achieved. Although gene expression was also observed in peripheral tissues, such as liver and spleen, this can be eliminated by the con\bined use of this vector with a brain-specific promoter such as the glial fibrillary acidic protein (GFAP) promoter (Shi et al., 2001). A similar approach with PIL gene-targeting technology was used in the primate brain. Zhang and colleagues used PILs conjugated to the murine MAb against the human insulin receptor (HIR) (Zhang et al., 2003). Previously used anti-TfRMAbs were specific for mice and rats and are not active in primates or humans. Moreover, the rate of transport of the HIR-MAb across the primate BBB is nearly 10-fold greater than the rate of transport of an anti-human TfRMAb across the primate BBB in vivo (Pardridge, 2001). These studies have proven that the HIR-MAb enables the liposome carrying the exogenous gene to undergo transcytosis across the BBB, and then endocytosis across the neuronal plasma membrane after intravenous injection. The level of luciferase gene expression in the brain was 50-fold higher in the rhesus monkey than in the rat. Confocal microscopy confirmed neuronal expression of the j5-galactosidase gene (jS-gal). Histochemistry of primate peripheral organs demonstrated tissuespecific expression; diffuse gene expression was

123

observed in liver and spleen, but not in heart, skeletal muscle, or fat in the primate. The expression of the transgene in liver and spleen was detected most likely because those organs have a permeable microvasculature and the plasmids were driven by the universal SV40 promoter. Use of the GFAP or a neuron-specific promoter could have probably eliminated that problem. Both osmotic and enzymatic/chemical manipulation of the brain environment can help to increase efficacy of the CNS treatment hindered by the BBB. Conjugation of molecular vectors with specific antibodies can also improve their transport through the BBB in vivo. Nevertheless, methods for general opening of the BBB may not be very efficient, since they do not achieve anatomical targeting. Thus, the main advantage of gene transfer is not met. More localized and specific methods of delivery are needed.

IL

LOCAL METHODS OF D R U G DELIVERY TO THE CNS

Other strategies for drug delivery to the brain bypass the BBB. Direct administration of therapeutics involves entry into the CNS by devices and needles (intrathecal and intra-cerebroventricular delivery). Such strategies may also be advantageous because they consider not only the net delivery of the agent to the CNS, but also the ability of the agent to access the relevant target site within the CNS. A.

Convection-Enhanced Delivery (CED)

To achieve therapeutic concentrations of a drug within the interstitium of the brain, extremely high systemic doses are needed, which normally results in unacceptable toxicities. Traditional local delivery of most therapeutic agents into the brain (by biodegradable polymers, direct intraventricular injection) has relied on diffusion that is dependent on a concentration gradient, is inversely related to the size of the agent, and is usually slow relative to tissue clearance. Thus, diffusion results in non-homogeneous distribution of most agents restricted to a few millimeters from the source. In contrast to diffusion, CED uses a pressure gradient established at the tip of an infusion catheter that creates bulk flow, which 'pushes' the drug through the space between brain cells (throughout the extracellular space) (Bobo et al., 1994). As a result, the drug is distributed more evenly, and at higher concentrations over a larger area, than in the absence of bulk flow, i.e., by diffusion alone. CED involves the

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

124

10. DELIVERY OF MOLECULAR THERAPEUTICS INTO THE CNS AND THEIR DISTRIBUTION WITHIN THE BRAIN

highly accurate placement of catheters in the brain for microinfusion of the drug, which makes this technique a very precise system for brain drug administration. For a given agent, the volume of distribution (V^) depends on structural properties of the tissue (hydraulic conductivity, vascular volume fraction, and extracellular fluid fraction) and the parameters of the infusion procedure (cannula placement, cannula design, and infusion rate). To maximize delivery and minimize reflux, the infusion procedure must be adjusted according to the tissue properties. Our laboratory has been involved in studies designed to optimize the parameters crucial for successful use of CED. We found that the design of an infusive cannula is important for optimal drug distribution. Morrison et al. (1999) have shown in animal experiments that reflux can be reduced and CED enhanced by reducing the diameter of the needle from a 27- to a 32-gauge needle. However, even the 32-gauge needle, one of the smallest needles commercially available, must be used with a flow rate of 0.5 jiil/min to avoid reflux (Morrison et al., 1994; Chen et al., 1999; Lonser et al., 2002; Degen et al., 2003). Our group developed a new, reflux-free, step-design cannula that permits CED with a higher flow rate. The cannula consists of a 27gauge needle with glued-in silica tubing, with an outer diameter (OD) of 168 jim and an inner diameter (ID) of 102 |Lim. The needle with silica inside is glued to tubing attached to the CED infusion system. The silica tubing extends beyond the cannula needle tip by 0.5 to 1 mm. With this design, a rate of 5 |il/min can be achieved without noticeable reflux, and more importantly without tissue damage caused by a high build-up of pressure. The reflux-free CED can reduce infusion time and the volume of drug required to cover the target structure in the brain, which consequently provides greater safety and efficacy in drug delivery. We have shown that CED can homogeneously distribute large therapeutic molecules over considerable distances (over 3 cm from the point of infusion). Convection has also shown enhanced distribution of HMW molecules in the normal rat, cat, and primate brain by an order of magnitude relative to diffusion (Lieberman et al., 1995; Bankiewicz et al., 2000; Cunningham et al., 2000; Nguyen et al., 2001). Thus, based on our own experience, CED allows infusion of therapeutic macromolecules and small viral constructs (such as adeno-associated virus, AAV) to significant regions of the non-human primate sub-cortical regions (Lieberman et al., 1995; Bankiewicz et al., 2000). In a direct comparison of CED and simple injection of AAV in monkeys, we found a significant increase of gene

transfer with the CED method. These results indicate that efficient and optimal gene delivery can be translated to the size of a human brain (Fig. 1). As mentioned before, recently liposomes have also been introduced to gene therapy systems to increase the efficiency of brain drug delivery (Huwyler et al., 1996; Imaoka et al., 1998; Segovia et al., 1998; Zhang et al., 2003; Mamot et al., 2004; Saito et al., 2004). Their systemic administration may not achieve satisfactory penetration of the BBB, and may not be able to restrict transgene expression to specific brain substructures, essential in many instances. In contrast, local injection cannot achieve acceptable distribution. However, CED provides a very attractive option for liposomal usage. It provides a larger distribution of liposomes within the target sites, allows for a locally sustained release of drugs, and minimizes systemic exposure, thereby producing fewer side effects. We have studied CNS distribution produced by local-regional administration of liposomes containing various markers, including attached or encapsulated fluorochromes, encapsulated gold particles and encapsulated gadodiamide for MRI contrast (Mamot et a l , 2004; Saito et al., 2004). Our results indicated that use of CED with liposomes achieved extensive and efficient distribution within normal rodent brains, s.c. flank tumor xenografts, and intracranial tumor xenografts. Recent advances in MR neuroimaging techniques have given rise to several methods for indirectly assessing the tissue properties (Hoehn-Berlage et al., 1992; Kurki et al., 1995). Further, the advent of nanotechnology-based MR contrast modalities provides a means for monitoring the progress of the CED infusion in real time. Thus, the inclusion of MRI in the therapeutic planning and monitoring of CED may improve the ability to achieve an optimum V^, and thereby improve the effectiveness of the infused therapeutic agent. B, Enhancements of Macromolecule Distribution within the Brain In designing optimal conditions for delivering drugs and viral vectors to the CNS, different coinfusates were used to enhance the distribution of .therapeutics within the brain. We have shown previously that by blocking interaction with cell surface heparan sulfate proteoglycan (HSPG), heparin can increase the distribution of therapeutic agents such as glial cell derived neurotrophic factor (GDNF) in vivo (Hamilton et al., 2001). On the basis of this finding, we showed that AAV2, when co-infused with heparin (a soluble HSPG receptor analog that transiently

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

LOCAL METHODS OF DRUG DELIVERY TO THE CNS

125

FIGURE 1 Digital reconstruction (BrainLab^M, Heimstetten, Germany) of the transduced area of monkey brain infused with AAV2-TK. Areas expressing TK by immunohistochemistry were mapped onto MRI of the same primate brain. In the next step, areas were delineated with BrainLab^M software. (A) Digital 3D MRI reconstruction showing CED of AAV2-TK expression. (B) Sagittal plane of MRI with striatal distribution of AAV2-TK expression. (C) Coronal plane of MRI showing areas of AAV2-TK expression. (D) Same as C with depicted regions of right putamen (yellow) and caudate nucleus (blue).

binds to AAV2), transduces a significantly larger area of rat brain. Presumably, this occurs because receptor-binding sites on the AAV capsid are temporarily blocked, allowing the vector to move farther through tissue before transduction occurs (Nguyen et al., 2001). Mastakov and colleagues extended our findings and optimized the heparin concentration for AAV2 infusion. They showed that transduction efficiency in animals infused with AAV2 plus heparin (1000 U/ml) was approximately 4.3 times higher than in animals infused with AAV2 alone (Mastakov et al., 2002). In 1999, Qing et al. presented evidence that, in addition to HSPG as a cell surface receptor, fibroblast growth factor receptor (FGFR) is required as a co-receptor for efficient binding and successful cellular entry by AAV (Qing et al., 1999). Consequently, we designed another strategy to investigate whether co-infusion of AAV2 with basic fibroblast growth factor (bFGF) enhances

AAV-mediated gene delivery in rat striatum. Our results clearly indicated that simultaneous injection of bFGF with AAV2 greatly enhances the volume of the transduced tissue (more than fourfold), probably by way of a competitive block of AAV2-binding sites within the striatum (Hadaczek et al., 2004). As mentioned before, mannitol has been used to disrupt the BBB, thereby facilitating spread of molecular therapeutics within the CNS (KroU and Neuwelt, 1998; Rapoport, 2001). Systemic intravascular delivery of viral vectors in combination with hyperosmotic mannitol resulted in very limited brain parenchymal transduction in rodents (Doran et al, 1995; Muldoon et al., 1995). Mastakov extended previous findings showing that local co-infusion of mannitol with rAAV produced a twofold improvement in transduction (Mastakov et al., 2001). The mechanism by which mannitol facilitated both transduction efficiency and viral spread is suspected

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

126

10. DELIVERY OF MOLECULAR THERAPEUTICS INTO THE CNS AND THEIR DISTRIBUTION WITHIN THE BRAIN

to include a local osmotic shrinkage of cells, thereby increasing interstitial space and facilitating diffusion. Optimization of all these parameters and conditions will almost certainly improve gene delivery to the CNS. C.

Intranasal Delivery

Another route of drug administration that bypasses the BBB is intranasal delivery. It provides a practical, noninvasive method to deliver therapeutic agents to the brain and spinal cord. In a w^ay similar to that of CED, it allows rapid access of drugs that normally cannot cross the BBB, thereby reducing unwanted systemic side effects. Intranasal delivery does not require any modification of therapeutic agents and does not require that drugs be coupled to any carrier (Ilium, 2003). The olfactory pathway may provide both intraneuronal and extraneuronal access into the brain. The intraneuronal pathway involves axonal transport and requires hours to days for drugs to reach brain regions. The extraneuronal pathway probably relies on bulk flow transport through perineural channels, which deliver drug directly to the brain parenchymal tissue, to the cerebrospinal fluid (CSF), or to both. This pathway allows agents to reach the CNS within minutes (Graff and Pollack, 2003). There have been many reports showing the usefulness of the intranasal route of administration for various drugs. However, due to the fact that < 1% of the initial drug concentration enters the brain, and also that it does not provide for localized delivery in specific brain structures, there may be limited therapeutic advantages of intranasal delivery route. In gene therapy, delivery of viral vectors via olfactory pathway has also been proposed (Flotte et al., 1996). Kinoshita and colleagues reported that adenoviral vector encoding wheat germ agglutinin (WGA), when infused into a mouse nostril, was transported and distributed to other brain structures beyond the olfactory bulb, i.e., olfactory cortical areas, anterior olfactory nucleus, olfactory tubercle, piriform cortex, and lateral entorhinal cortex (Kinoshita et al., 2002). In addition, trans-synaptic retrograde labeling was observed in cholinergic neurons in the horizontal limb of diagonal band, serotonergic neurons in the median raphe nucleus, and noradrenergic neurons in the locus coeruleus, all of which project centrifugal fibers to the olfactory bulb. In a phase 1 trial, an AAV encoding full length human cystic fibrosis transmembrane receptor (CFTR) was evaluated for safety after intranasal and endo-bronchial instillation in 25 cystic fibrosis patients with mild-to-moderate lung disease (Flotte et al., 1996). Unfortunately, the nasal gene

transfer efficiency was much lower than bronchial gene transfer efficiency, as judged by PCR. However, the nasal epithelium may be a poor surrogate model for viral gene transfer to airway epithelial cells, which have generally resisted entry by viral vectors (Knowles et al., 1998). In another study by Lemiale and colleagues, biodistribution of recombinant adenovirus (rADV) vectors, administered intranasally, revealed very confined infection of the central nervous system, specifically in the olfactory bulb, which may limit the utility of this route of delivery of ADV (Lemiale et al., 2003). Evidently, the intranasal method of vector delivery, in comparison with CED, gives lower transduction efficiencies as well as lower precision in targeting specific brain structures.

IIL DISTRIBUTION OF MOLECULAR THERAPEUTICS WITHIN THE BRAIN Administration of viral vectors and other nanoparticles to the CNS is only the first step for brain drug delivery. Subsequently, those molecules have to be distributed within the brain to its target structures. Overcoming anatomical and physiological barriers within the brain itself is another challenge to efficiently distribute therapeutics, especially if it needs to be achieved in a specific manner. The brain environment is very heterogeneous. Therefore, microanatomy, the degree of myelination, cellular density, fluid dynamic of the brain are important factors governing distribution of macromolecules within the CNS. A.

Axonal Transport

After delivery of viral vector to targeted brain regions (or via intranasal delivery), the transgene protein product is often detected at a significant distance from the injection site — a finding usually interpreted as indicating transport of the product along specific pathways within the brain. Neuronal transport of AAV2 along neural projections has been reported. Recognized axonal pathways are believed to be responsible also for transporting viral particles like herpes simplex virus (HSV-1) (Sun et al., 1996), pseudorabies virus (Yang et al., 1999; Husak et al., 2000), and adenovirus-5 (Bohn et al., 1999) from one region of the brain to another. Usually, such transport takes a long time (up to several weeks). Kaspar et al. injected AAV containing the reporter gene green fluorescence protein (GPP) into rat hippocampus and striatum (Kaspar et al., 2002). After 2 weeks, they observed GPP

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

127

BRAIN TUMORS

expression in brain regions retrogradely connected to the sites of injection — cortical entorhinal layer II neurons for the hippocampus and substantia nigra for the striatum. They suggested that axonal retrograde transport of viral particles had occurred. In our own studies, after injecting AAV2-TK into rat striatum, the presence of AAV2 vector transgene product, thymidine kinase, was detected in neurons of the globus pallidus (GP), a region distant from the site of delivery (Hadaczek et al., 2004). This phenomenon in the rat brain was described in an earlier experiment by Chamberlin et al. (1998). Connections between the striatum and other brain structures form closed circuits called basal ganglia circuits (Gerfen and Wilson, 1996). After AAV transduction, TK enzyme is present only within transduced cells and cannot be secreted into the synaptic cleft. Positive staining in the GP indicates that, after striatal cell transduction, the AAV vector may be anterogradely transported from the site of injection. After carefully collecting tissue punches from GP and subsequently performing RT-PCR, we detected TK RNA transcripts in GP, which implies transport of the vector itself and not its transgene protein product. B.

"Perivascular Pump^'

In our experience with viral vector delivery to the CNS, we have often observed that cellular transduction is confined to the space around blood vessels in the brain. This observation raised the possibility that viral particles might be spread within the brain by a simple, non-synaptic pathway. In one of our experiments, when we infused AAV2 into rat striatum (unpublished data) and euthanized animals rapidly after injection, viral capsids were detected (antibody A20) in the GP immediately after infusion. This apparently immediate and distant spread of AAV directed our attention to a fluid mechanism and to the brain's perivascular spaces. It is known that CSF extends into the perivascular (Virchow-Robin) spaces. These spaces are an extension of the sub-arachnoid space that accompanies penetrating arteries into the brain down to the level of capillaries (Lennart, 1995). Within this space, metabolites and small solutes can diffuse quite freely between extracellular fluids and CSF. The concept of a perivascular space system has been studied in detail, and the existence of a pathway for rapid flow of CSF from the sub-arachnoid space into the perivascular spaces has been confirmed (Gregory et al., 1985; Rennels et a l , 1985, 1990). The movement of fluid is caused by arterial pulsation resulting from normal heart action causing agents delivered directly into the

brain to be propelled within the perivascular conduit by a peristaltic mechanism. A normal heartbeat generates a periodic traveling wave of wall deformation in the outer wall of the artery, with its amplitude and velocity contributing to the spread of infusates within the brain tissue. This finding further supports our concept of a "perivascular pump'' distributing and transporting viral particles within the CNS. When that pump was inactivated (heartbeat arrest), capsids were trapped by cellular receptors (HSPG) at the site of infusion. Thus, the pressure from CED alone was insufficient to propel them to the neighboring GP. It seems that fluid circulation through the CNS occurring through the perivascular space is the primary mechanism by which viral particles and other therapeutic agents administered by CED are spread within the brain. Cardiac contractions are necessary to power this process. It is conceivable that manipulation with drugs with cerebral vasomotor properties may offer another option for optimizing the delivery and distribution of therapeutic molecules within the CNS. It is also worth mentioning that, when discussing distribution of the final protein transgene product, it is important to know whether the protein will be secreted or confined strictly to the cell body. With reporter genes like TK, Lac-Z, or GFP, their transduction area overlaps the spread of viral vectors. In the case of secreted transgene products (glial cell-derived neurotrophic factor (GDNF), interleukins), it appears that their distribution surpasses the transduced region by diffusion within the extracellular matrix and may be propelled further by the perivascular pump. IV-

BRAIN TUMORS

Brain tumors may be considered neurological diseases insofar as gene therapy treatments are concerned. The various, specific anatomical locations of such tumors make their treatment especially challenging for modern oncology employing molecular therapeutics. Similarly to other CNS diseases, gene therapy for brain tumors faces anatomical and physio-logical barriers such as the BBB, Blood-Tumor Barrier (BTB), and high interstitial pressure within tumor tissue. Moreover, neuro-oncology clinical studies have not addressed major limiting factor for in vivo gene therapy, such as insufficient gene transfer rates to the tumor with the local delivery modalities. Consequently, gene transfer approaches may have resulted in inadequate cytotoxicity in the tumor of any given transgene-prodrug system. Critical evaluation of gene transfer and therapy studies

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

128

10. DELIVERY OF MOLECULAR THERAPEUTICS INTO THE CNS AND THEIR DISTRIBUTION WITHIN THE BRAIN

has led to the conclusion that, even with identical vectors, the anatomical route of vector administration can dramatically affect both the efficiency of tumor transduction and its spatial distribution, as well as the extent of intratumoral and intracerebral transgene expression. Different delivery methods for vector administration to malignant primary brain tumors in experimental or clinical settings were investigated and ranged from stereotactic or intratumoral injection or convectionenhanced interstitial delivery; intrathecal and intraventricular injection; and intravascular infusion, with or without modification of the BTB were explored. Therapeutic genes of choice have typically been either toxic genes such as thymidine kinase (Perez-Cruet et a l , 1994; Okada et al., 1996; Hadaczek et al., 2005), proapoptotic genes like Bax (Gridley et al., 2004; Kaliberov et al., 2004), or immunostimulatory genes (for instance, interleukin-12) (Parker et al., 2000; Liu et a l , 2002; Ren et a l , 2003). A. Systemic Vector Delivery in the Treatment of Brain Tumors The arterial delivery of genetically engineered replication-deficient adenovirus vector (Ad) and cationic liposome-plasmid DNA complexes (lipoDNA) to experimental brain tumors was investigated by Rainov and colleagues (Rainov et al., 1999). They compared the systemic delivery (internal carotid artery) of those two vectors to evaluate the extent of transduction within 9L gliosarcomas in F344 rats. Bradykinin was used to selectively permeabilize the BTB. Forty-eight hours after vector injection, brains and other internal organs were collected for evaluation of expression of the marker gene product, jS-galactosidase. Intracarotid delivery of Ad to rat gliosarcoma without BTB disruption resulted in transgene expression in 3-10% of tumor cells distributed throughout the tumor. In the brain parenchyma only a few endothelial cells expressed j5-gal owing to Ad-mediated gene transfer. Intracarotid delivery of lipoDNA carrying a cytoplasmic expression cassette rendered >30% of the tumor mass positive for the marker gene without BTB disruption. Bradykinin infusion was able to increase further the number of transduced tumor cells to >50%. Although lipoDNA-mediated transfer resulted in increased efficacy as compared to Ad-mediated gene transfer, it was less specific since larger numbers of endothelial and glial cells also expressed the transgene. Both Ad and lipoDNA injections, in the absence and presence of bradykinin, also resulted in transduction of peripheral organs. LipoDNA transduced

parenchymal organs such as liver, lung, testes, lymphatic nodes, and especially spleen, whereas AAV displayed only its known affinity for liver and lung. B. Local Vector Delivery in Treatment of Brain Tumors Our group has also tested the use of local CED of AAV2 for treatment of brain tumors (Hadaczek et al., 2005). We transduced tumors (athymic rats bearing 87MG-derived glioblastomas) with the HSV-TK gene, which activates the nucleoside analog prodrug ganciclovir (GCV). This is one of the most effective and most commonly explored gene therapy approaches for treatment of experimental brain tumors. This therapy has the potential to selectively kill dividing cancer cells, since AAV2 selectively infects tumor cells and neurons both expressing heparan sulfate proteoglycan binding sites, leaving other cell types such as astrocytes and endothelial cells uninfected. The effectiveness of the AAV2-TK/ GCV strategy depends critically on transduction of a sufficient number of tumor cells to achieve total eradication of the tumor mass. Despite a statistically significant difference in survival between GCV-treated and control animals (25.8 compared with 21.3 days, p < 0.05), we concluded that even if an extensive tumor area (39%) was transduced with AAV2-TK vector, we were not able to eradicate tumors. Through the CED technique, AAV2 particles were delivered into the central portion of a growing neoplastic mass, locally transducing only the core of the tumor and leaving its periphery unaffected. Even if the central mass was most likely eradicated by GCV treatment, peripheral cells were still dividing because they were isolated from immediate contact with the TK-positive cells (Fig. 2). Therefore, the well-known bystander effect could not exert its function and the unaffected tumor masses had eventually overgrown the space left by the cells killed by GCV That experiment demonstrated the considerable divergence between in vivo and in vitro bystander effects. Our in vitro studies showed that, when the cell culture consisted of only 10% of TK-positive cells, the inhibition of cell growth by GCV was substantial. Such powerful action is most likely due to the fact that TK-positive cells are evenly dispersed among other untransduced cells, and thus toxic molecules can be distributed. This was not the case in vivo, where we encounter more focal patterns of transduction. Anti-tumoral efficacy depends most likely on achieving a highly diffuse transduction pattern within a tumor mass. This again emphasizes how critical it is to design efficient vector delivery system tailored for individual diseases of the CNS.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

ACKNOWLEDGMENT

129

FIGURE 2 Thymidine kinase immunoreactivity of the transduced U87MG-derived rat brain tumors. AAV-TK vector was used for intratumoral infusion by CED. Tumors were collected 3 days after infusion with the vector, and the transduction area of TK (arrows) within the tumor mass was calculated. In 18-day-old tumors (A), 39% of the total tumor volume was transduced with AAV2-TK; in 22-day-old tumors (B), this transduction accounted for only 18%. Through CED, AAV2 particles were delivered into the central portion of a growing neoplastic mass, locally transducing only the core of the tumor leaving the periphery unaffected.

The gene therapy of CNS diseases is particularly challenging because the delivery of drugs to the brain is often precluded by a variety of anatomical and physiological obstacles that collectively comprise the BBB or the BTB. Drug delivery directly to the brain interstitium has recently been markedly enhanced through development of methods such as CED and the optimization of its parameters. Human brain is a heterogeneous organ; therefore, all aspects of its organization (anatomical, physiological, and biochemical) should be taken into consideration in designing techniques of drug delivery. Even the most potent drug will not be effective until administered properly. Thus, formulating new vectors and molecular therapeutics should be undertaken in parallel with devising and optimizing their delivery and distribution within the CNS.

ACKNOWLEDGMENT The authors would like to thank Dr. Michal Krauze for contributing digital reconstruction figure. References Bankiewicz, K.S. et al. (2000) Convection-enhanced delivery of AAV vector in parkinsonian monkeys; in vivo detection of gene expression and restoration of dopaminergic function using prodrug approach. Exp. Neurol., 164: 2-14.

Bartus, R.T. et al. (1996) Controlled modulation of BBB permeability using the bradykinin agonist, RMP-7. Exp. Neurol., 142: 14-28. Bobo, R.H. et al. (1994) Convection-enhanced delivery of macromolecules in the brain. Proc. Natl. Acad. Sci. USA, 91: 2076-2080. Boddy, A. and Thomas, H. (1997) RMP-7: potential as an adjuvant to the drug treatment of brain tumors. CNS Drugs, 7: 257-263. Bohn, M.C. et al. (1999) Adenovirus-mediated transgene expression in nonhuman primate brafin. Hum. Gene Ther., 10: 1175-1184. Chamberlin, N.L. et al. (1998) Recombinant adeno-associated virus vector: use for transgene expression and anterograde tract tracing in the CNS. Brain Res., 793:169-175. Chen, M.Y. et al. (1999) Variables affecting convection-enhanced delivery to the striatum: a systematic examination of rate of infusion, cannula size, infusate concentration, and tissue-cannula sealing time. I. Neurosurg., 90: 315-320. Cullis, PR. et al. (1998) Interactions of liposomes and lipid-based carrier systems with blood proteins: relation to clearance behaviour in vivo. Adv. Drug Deliv. Rev., 32: 3-17. Cunningham, ]. et al. (2000) Distribution of AAV-TK following intracranial convection-enhanced delivery into rats. Cell Transplant., 9: 585-594. Degen, J.W. et al. (2003) Safety and efficacy of convection-enhanced delivery of gemcitabine or carboplatin in a malignant glioma model in rats. J. Neurosurg., 99: 893-898. Doran, S.E. et al. (1995) Gene expression from recombinant viral vectors in the central nervous system after blood-brain barrier disruption. Neurosurgery, 36: 965-970. Flotte, T. et al. (1996) A phase I study of an adeno-associated virusCFTR gene vector in adult CF patients with mild lung disease. Hum. Gene Ther., 7: 1145-1159. Gerfen, C.R. and Wilson, C.J. (1996) The basal ganglia. In: Swanson L.W., Bjorklund A. and Hokfelt T. (Eds.), Handbook of Chemical Neuroanatomy, Vol. 12. Integrated Systems of the CNS, part III. Elsevier, Amsterdam, pp. 371-468.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

130

10. DELIVERY OF MOLECULAR THERAPEUTICS INTO THE CNS AND THEIR DISTRIBUTION WITHIN THE BRAIN

Graff, C.L. and Pollack, G.M. (2003) P-Glycoprotein attenuates brain uptake of substrates after nasal instillation. Pharm. Res., 20: 1225-1230. Gregory, T.F. et al. (1985) A method for microscopic studies of cerebral angioarchitecture and vascular-parenchymal relationships, based on the demonstration of 'paravascular' fluid pathways in the mammalian central nervous system. J. Neurosci. Methods, 14: 5-14. Gridley, D.S. et al. (2004) Proton radiation and TNF-alpha/Bax gene therapy for orthotopic C6 brain tumor in Wistar rats. Technol. Cancer Res. Treat., 3: 217-227. Hadaczek, P. et al. (2004) Basic fibroblast growth factor enhances transduction, distribution, and axonal transport of adeno-associated virus type 2 vector in rat brain. Hum. Gene Ther., 15: 469^79. Hadaczek, P. et al. (2005) Limited efficacy of gene transfer in herpes simplex virus-thymidine kinase/ganciclovir gene therapy for brain tumors. J. Neurosurg., 102: 328-335. Hamilton, J.F. et al. (2001) Heparin coinfusion during convection-enhanced delivery (CED) increases the distribution of the glial-derived neurotrophic factor (GDNF) ligand family in rat striatum and enhances the pharmacological activity of neurturin. Exp. Neurol., 168: 155-161. Higgins, C.F. (1992) ABC transporters: from microorganisms to man. Annu. Rev Cell Biol., 8: 67-113. Hoehn-Berlage, M. et al. (1992) In vivo NMR T2 relaxation of experimental brain tumors in the cat: a multiparameter tissue characterization. Magn. Reson. Imaging, 10: 935-947. Husak, P.J. et al. (2000) Pseudorabies virus membrane proteins gl and gE facilitate anterograde spread of infection in projectionspecific neurons in the rat. J. Virol., 74: 10975-10983. Huwyler, J. et al. (1996) Brain drug delivery of small molecules using immunoliposomes. Proc. Natl. Acad. Sci. USA, 93: 1416414169. Ilium, L. (2003) Nasal drug delivery - possibilities, problems and solutions. J. Control Release, 87: 187-198. Imaoka, T. et al. (1998) Significant behavioral recovery in Parkinson's disease model by direct intracerebral gene transfer using continuous injection of a plasmid DNA-liposome complex. Hum. Gene Then, 9: 1093-1102. Inamura, T. and Black, K.L. (1994) Bradykinin selectively opens blood-tumor barrier in experimental brain tumors. J. Cereb. Blood Flow Metab., 14: 862-870. Kaliberov, S. et al. (2004) Enhanced apoptosis following treatment with TRA-8 anti-human DR5 monoclonal antibody and overexpression of exogenous Bax in human glioma cells. Gene Then, 11: 658-667. Kaspar, B.K. et al. (2002) Targeted retrograde gene delivery for neuronal protection. Mol. Then, 5: 50-56. Kinoshita, N. et al. (2002) Adenovirus-mediated WGA gene delivery for transsynaptic labeling of mouse olfactory pathways. Chem. Senses, 27: 215-223. Knowles, M.R. et al. (1998) A double-blind, placebo controlled, dose ranging study to evaluate the safety and biological efficacy of the lipid-DNA complex GR213487B in the nasal epithelium of adult patients with cystic fibrosis. Hum. Gene Then, 9: 249-269. Kroll, R.A. and Neuwelt, E.A. (1998) Outwitting the blood-brain barrier for therapeutic purposes: osmotic opening and other means. Neurosurgery, 42: 1083-1099; discussion 1099-1100. Kurki, T. et al. (1995) MR classification of brain gliomas: value of magnetization transfer and conventional imaging. Magn. Reson. Imaging, 13: 501-511. Lemiale, F. et al. (2003) Enhanced mucosal immunoglobulin A response of intranasal adenoviral vector human immunodeficiency virus

vaccine and localization in the central nervous system. J. Virol., 71: 10078-10087. Lennart, H. (1995) The Human Brain and Spinal Cord: functional neuroanatomy and dissection guide. Springer, New York. Li, H. et al. (2002) The role of the transferrin-transferrin-receptor system in drug delivery and targeting. Trends Pharmacol. Sci., 23: 206-209. Lieberman, D.M. et al. (1995) Convection-enhanced distribution of large molecules in gray matter during interstitial drug infusion. J. Neurosurg., 82: 1021-1029. Liu, Y. et al. (2002) In situ adenoviral interleukin 12 gene transfer confers potent and long-lasting cytotoxic immunity in glioma. Cancer Gene Then, 9: 9-15. Lonser, R.R. et al. (2002) Successful and safe perfusion of the primate brainstem: in vivo magnetic resonance imaging of macro-molecular distribution during infusion. J. Neurosurg., 97: 905-913. Mamot, C. et al. (2004) Extensive distribution of liposomes in rodent brains and brain tumors following convection-enhanced delivery J. Neurooncol., 68: 1-9. Mastakov, M.Y et al. (2001) Combined injection of rAAV with mannitol enhances gene expression in the rat brain. Mol. Then, 3: 225-232. Mastakov, M.Y. et al. (2002) Recombinant adeno-associated virus serotypes 2- and 5-mediated gene transfer in the mammalian brain: quantitative analysis of heparin co-infusion. Mol. Then, 5: 371-380. Morrison, PF. et al. (1994) High-flow microinfusion: tissue penetration and pharmacodynamics. Am. J. Physiol., 266: R292-305. Morrison, PF. et al. (1999) Focal delivery during direct infusion to brain: role of flow rate, catheter diameter, and tissue mechanics. Am. J. Physiol., 277: R1218-1229. Muldoon, L.L. et al. (1995) Comparison of intracerebral inoculation and osmotic blood-brain barrier disruption for delivery of adenovirus, herpes virus, and iron oxide particles to normal rat brain. Am. J. Pathol., 147: 1840-1851. Nguyen, J.B. et al. (2001) Convection-enhanced delivery of AAV-2 combined with heparin increases TK gene transfer in the rat brain. Neuroreport, 12: 1961-1964. Nomura, T. et al. (1994) Intracarotid infusion of bradykinin selectively increases blood-tumor permeability in 9L and C6 brain tumors. Brain Res., 659: 62-66. Okada, H. et al. (1996) Gene therapy against an experimental glioma using adeno-associated virus vectors. Gene Then, 3: 957-964. Pardridge, W.M. (2001) Brain Drug Targeting: The Future of Brain Development. Cambridge, United Kingdom, Cambridge University Press. Parker, J.N. et al. (2000) Engineered herpes simplex virus expressing IL-12 in the treatment of experimental murine brain tumors. Proc. Natl. Acad. Sci., USA, 97: 2208-2213. Perez-Cruet, M.J., et al. (1994) Adenovirus-mediated gene therapy of experimental gliomas. J. Neurosci. Res., 39: 506-511. Pohl, A. et al. (2002) Transport of phosphatidylserine via MDRl (multidrug resistance l)P-glycoprotein in a human gastric carcinoma cell line. Biochem. J., 365: 259-268. Qing, K. et al. (1999) Human fibroblast growth factor receptor 1 is a co-receptor for infection by adeno-associated virus 2. Nat. Med., 5: 71-77. Rainov, N.G. et al. (1999) Intraarterial delivery of adenovirus vectors and liposome-DNA complexes to experimental brain neoplasms. Hum. Gene Then, 10: 311-318. Rapoport, S.I. (2001) Advances in osmotic opening of the bloodbrain barrier to enhance CNS chemotherapy. Expert Opin. Investig. Drugs, 10: 1809-1818.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

ACKNOWLEDGMENT Ren, H. et al. (2003) Immunogene therapy of recurrent glioblastoma multiforme with a liposomally encapsulated replication-incompetent Semliki forest virus vector carrying the human interleukin-12 gene - a phase I/II clinical protocol. J. NeurooncoL, 64: 147-154. Rennels, M.L. et al. (1985) Evidence for a 'paravascular' fluid circulation in the mammalian central nervous system, provided by the rapid distribution of tracer protein throughout the brain from the subarachnoid space. Brain Res., 326: 47-63. Rennels, M.L. et al. (1990) Rapid solute transport throughout the brain via paravascular fluid pathways. Adv. Neurol, 52: 431^39. Rubin, L.L. and Staddon, J.M. (1999) The cell biology of the bloodbrain barrier. Annu. Rev. Neurosci., 22: 11-28. Saito, R. et al. (2004) Distribution of liposomes into brain and rat brain tumor models by convection-enhanced delivery monitored with magnetic resonance imaging. Cancer Res., 64: 2572-2579. Sanovich, E. et al. (1995) Pathway across blood-brain barrier opened by the bradykinin agonist, RMP-7. Brain Res., 705: 125-135. Segovia, J. et al. (1998) Astrocyte-specific expression of tyrosine hydroxylase after intracerebral gene transfer induces behavioral recovery in experimental parkinsonism. Gene Ther., 5: 1650-1655. Shi, N. et al. (2001a) Receptor-mediated gene targeting to tissues in vivo following intravenous administration of pegylated immunoliposomes. Pharm. Res., 18: 1091-1095.

131

Shi, N. et al. (2001b) Brain-specific expression of an exogenous gene after i.v. administration. Proc. Natl. Acad. Sci. USA, 98: 12754-12759. Shi, N. and Pardridge, W.M. (2000) Noninvasive gene targeting to the brain. Proc. Natl. Acad. Sci. USA, 97: 75^7-7572. Siegal, T. et al. (2000) In vivo assessment of the window of barrier opening after osmotic blood-brain barrier disruption in humans. J. Neurosurg., 92: 599-605. Sun, N. et al. (1996) Anterograde, transneuronal transport of herpes simplex virus type 1 strain H129 in the murine visual system. J. Virol., 70: 5405-5413. Ueda, K. et al. (1992) Human P-glycoprotein transports Cortisol, aldosterone, and dexamethasone, but not progesterone. J. Biol. Chem., 267: 24248-24252. Wolf, D.C. and Horwitz, S.B. (1992) P-glycoprotein transports corticosterone and is photoaffinity-labeled by the steroid. Int. J. Cancer, 52: 141-146. Yang, M. et al. (1999) Retrograde, transneuronal spread of pseudorabies virus in defined neuronal circuitry of the rat brain is facilitated by gE mutations that reduce virulence. }. Virol., 73: 4350-4359. Zhang, Y. et al. (2003a) Intravenous non-viral gene therapy causes normalization of striatal tyrosine hydroxylase and reversal of motor impairment in experimental parkinsonism. Hum. Gene Ther., 14: 1-12. Zhang, Y. et al. (2003b) Global non-viral gene transfer to the primate brain following intravenous administration. Mol. Then, 7:11-18.

I. GENE TRANSFER TECHNOLOGY AND REGULATORY ISSUES

C H A P T E R

11 Gene Therapy for CNS Diseases Using Intrabodies Todd W, Millery Anne Messer

Abstract: Single-chain Fv and single-domain antibodies retain the binding specificity of full-length antibodies, but they can be expressed as single genes in phage or yeast surface-display libraries, thus allowing efficient in vitro selection from a naive human repertoire using standard molecular and cellular techniques. Candidate genes can then be expressed intracellularly as intrabodies, with the potential for alteration of the folding, interactions, modifications, or subcellular localization of their targets. These reagents have already been developed as therapeutics against cancer and HIV. The misfolded and accumulated proteins that characterize a wide range of neurodegenerative disorders provide a novel class of potential intrabody targets. Here, we review the extension of intrabody technology to the nervous system, where studies of Huntington's disease have been used to develop the approach, and antisynuclein, anti-jS-amyloid, and anti-prion strategies are under development. Research on several other neurodegenerations suggests that intrabodies directed against specific targets, or possibly against more common downstream targets, might be developed as novel genetic therapeutics, and as drug discovery tools, to further unravel disease pathways. Keywords: intrabody; antibody; scFv; DAB; therapeutics; Alzheimer's; Parkinson's; poly-glutamine; Huntington's; prion; tauopathy; synucleinopathy; amyloid; ALS

L

Alzheimer's disease (AD), and prion diseases, and we discuss potential targets in other prominent neurological disorders. Table 1 summarizes the current literature on intrabodies and antibodies for neurodegenerative protein targets as of early 2005.

INTRODUCTION

Intrabodies are intracellularly expressed antibodies (Abs) or Ab fragments that target intracellular antigens. As several neurological disorders are mediated by abnormal proteins, intrabodies may serve as genetic therapeutics to selectively target such proteins, and as drug discovery and validation tools. Intrabodies have already been investigated as treatments for a variety of conditions (Stocks, 2004), including HIV infection (Marasco et al., 1999; Tewari et al., 2003), tumor growth (Wheeler et al., 2003a, b), and tissue transplantation (Beyer et al., 2004). They are also being tested in clinical trials for cancer (Alvarez et a l , 2000; Leath et al., 2004). Here, we highlight the progress made with intrabodies for Huntington's disease (HD), synucleinopathies.

Gene Therapy of the Central Nervous System: From Bench to Bedside

A,

Intrabody Generation and Selection

Antibody engineering has facilitated the generation of small Ab fragments that can recapitulate the binding properties of a full Ab while using a much smaller gene coding sequence. A single-chain Fv (scFv; sometimes also abbreviated as sFv) is generated by cloning of the variable domains of an Ab, and then joining the single domain cDNA sequences with DNA encoding a flexible linker (Bird et al., 1988; Huston et a l , 1988), thereby allowing a scFv (--250 amino acids.

133

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

134

11. GENE THERAPY EOR CNS DISEASES USING INTRABODIES

TABLE 1

Current Literature on the Utilization of Intrabodies and Antibodies for Diseases of the Central Nervous System

Neurological disorder

Protein target

Publications

Parkinson's disease and synucleinopathies

a-synuclein

Emadi et al. (2004), Maguire-Zeiss et al. (2004), Zhou et al. (2004a)

Huntington's disease

Huntingtin

Marks et al. (1991), Lecerf et al. (2001), Khoshnan and Patterson (2002), Colby et al. (2004a, b). Murphy and Messer (2004), Miller and Messer (2005)

Alzheimer's disease Tauopathies"

j5-amyloid Tau

Rangan et al. (2003), Liu et al. (2004a, b), Paganetti et al. (2005) Visintin et al. (2002)

Prion disease

Prion diseases

Leclerc et al. (2000), Heppner et al. (2001), Wuertzer et al. (2004), Cardinale et al. (2005)

Amyloidogenic diseases^

Amyloid conformation

O'Nuallain and Wetzel (2002), Kayed et al. (2003) "Disorders for which scFvs or DABs have been selected, but for which efficacy in disease model systems has not been reported. Modified from Miller and Messer (2005). ''Amyloidogenic diseases include Alzheimer's, Huntington's, Parkinson's, and prion diseases.

29 kDa) to be expressed from a single gene. This technique may be performed using the cDNA from a single cell line to create an intrabody of known specificity (e.g., hybridoma, Orlandi et al., 1989), or from popu-lations of cells (e.g., naive spleens, peripheral blood lymphocytes) to generate phage or yeast surface-display libraries (Marks et a l , 1991). Single-variable domains of an Ab (DAB) can similarly be utilized (Tanaka et a l , 2003). Figure 1 summarizes the different selection approaches which are described in more detail below. The advantage of starting with a monoclonal Ab is that the capacity to target a defined epitope is retained, taking advantage of previous work, as shown in recent publications (Khoshnan et al., 2002; Cardinale et a l , 2005; Paganetti et al., 2005). The advantages of scFv or DAB libraries, which can represent the entire repertoire of potential Abs from a human pool, include a reduced immunogenic potential for clinical use (since they are derived from a human, rather than a mouse), and an enhanced probability of finding an initial candidate with acceptable intracellular folding (Lecerf et al., 2001; Colby et al., 2004b). Synthesis of an scFv or DAB is not guaranteed to yield a functional intrabody, primarily due to misfolding of the fragment intracellularly, with an additional contribution from unformed intrachain disulfide bonds in the reducing cytoplasmic environment (Worn et al., 2000; Mossner et a l , 2001). This is less of a problem if the target protein is within the endoplasmic reticulum, as discussed in Sections II.A.4 and II.D. An scFv or DAB library incorporated into a phage (Marks et a l , 1991) or yeast surface-display system (Boder and Wittrup, 1997; Kieke et al., 1997) for in vitro

biopanning against an antigen of interest can partially overcome this pitfall by initially providing several high-affinity candidates for functional intracellular screening. Yeast libraries provide a eukaryotic expression context, although these Ab fragments are generally not glycosylated, so a bacterial environment is sufficient in most cases. Typically, the antigen will consist of a peptide sequence found in the target protein; however, selection against discontinuous conformational epitopes may require biopanning against larger protein fragments, and/or altering selection conditions. Strategies for secondary screening for intrabody expression and stability, following initial selection from phage or yeast libraries, include in vivo selection in a yeast two-hybrid strategy (yeast ratracellular Ab capture) (Tse et a l , 2002; Visintin et a l , 2002), and direct phage to intrabody screening (DPIS) in mammalian cells, starting with pools of intrabodies selected from phage ELISA screens (Gennari et al., 2004). Antibody fragments may be further engineered to produce derivatives with higher affinity, lower dissociation rate, or improved stability through targeted mutagenesis of complementarity-determining regions (CDRs), light-chain shuffling (Osbourn et al., 1996), DNA shuffling for molecular evolution (Stemmer, 1994), BIAcore-driven selection (Schier and Marks, 1993), error-prone PCR (Colby et a l , 2004c), or grafting of CDRs onto a stable framework (Ewert et al., 2004). Stability improvement could be critical, since Zhu et al. (1999) strongly linked half-life to efficacy. The true intracellular affinity of an intrabody for its antigen cannot be absolutely determined by current methods, although it can be approximated using in vitro

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

135

INTRODUCTION

scFv

VH-DAB

RT/PCR and clone variable domains / yeast surface-display

/

yeast

\

VLDAB

yeast intracellular antibody capture

\ phage surface-display phag(

Functional testing in disease models FIGURE 1 Intrabody selection. Intrabodies may exist as single-chain Fvs (scFvs), where the DNA encoding both the VH (red) and VL (blue) domains of an immunoglobulin are cloned and joined with DNA encoding a flexible linker. Alternatively, single-variable domain antibodies (DABs) may also be used as intrabodies. ScFv and DAB genes are cloned and expressed on the surface of bacteriophage or yeast to create a surface display library. The library is then used for biopanning against an antigen of interest, where yeast or phage expressing an scFv/DAB specific for the antigen will bind more strongly. Candidate scFvs/DABs can then be further screened in vivo by the yeast intracellular antibody capture technique; this approach helps ensure that the selected scFvs/DABs will fold properly and bind antigen in an intracellular environment. The candidate intrabody genes are fused to a VP16 transcriptional activation domain (yellow). The antigen of interest (blue star) is fused to a Lex A DNA-binding domain (pink). The histidine-deficient yeast genome has a LexA binding element (green) just upstream of the His3 gene (orange). Therefore, binding of a specific intrabody to the antigen will bring the VP16 domain into close proximity with the DNA upstream of the His3 gene, allowing transcription and production of histidine, and hence survival on a histidine-deficient medium, only in yeast expressing an antigen-specific intrabody

binding or yeast intracellular Ab capture techniques. Primary selection against an antigen of interest is not currently being performed intracellularly in mammalian cells due to the large size of the libraries. Hence, selection of desirable candidates does not

guarantee the same antigen-binding properties when they are expressed as intrabodies in mammalian cells. However, iterative in situ testing of candidate intrabodies and further engineering can be readily accomplished.

IL GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

136 B,

11. GENE THERAPY FOR CNS DISEASES USING INTRABODIES

The Power of Intrabodies

Antibodies and Ab fragments possess some distinct advantages over other classes of therapeutics for neurodegenerative diseases, which include random peptides, small molecules, and RNA interference (RNAi). Random peptides are the most similar to an intrabody approach. They can be selected from libraries with biopanning similar to that for scFvs. Nagai et al. (2000) have shown that a peptide against the expanded polyglutamine poly-Q can rescue some toxic effects of ataxin3 in cultures and Drosophila models, and Kazantsev et al. (2002) have also isolated a peptide with anti-aggregation effects in cultures. Intrabodies may offer a greater degree of antigen-binding specificity and stability, and may also prove to be less immunogenic than other protein/peptide therapeutics, particularly when selected from a human library. Small molecule drugs, chosen for their established effects on processes in the pathogenic cascade, or from high-throughput screens using aggregation suppression as an assay (Bates, 2003; Marsh and Thompson, 2004; Zhang et al, 2005), are likely to be much less specific than intrabodies. Intrabody genes may also be delivered to and stably expressed by target cells, providing a one-time procedure with long-term effects; small molecules would likely require repeated administrations and hence be more expensive. RNAi, which includes the advantages of gene therapy, provides protection by reducing the amount of substrate available to form toxic protein species (Xia et al., 2004; Harper et al., 2005). However, the proteomic approach provided by intrabodies offers the potential to utilize conformational specificities, and blockage of post-translational modifications and proteolytic inhibition. Offtarget effects may also be reduced. Use of combination therapies of small molecules, which can utilize doses of individual compounds that are below the threshold for toxicity, has improved outcomes in HD models (Hersch and Ferrante, 2004; Agrawal et al, 2005). Combinations that include intrabodies which act on early stages of the degenerative process may prove especially effective when combined with low doses of agents that reduce pathology globally due to downstream parts of the pathogenic cascade.

IL

MECHANISMS OF INTRABODY APPROACHES TO NEURODEGENERATION

The binding of an intrabody to its target protein may elicit any of several possible effects. The intrabody

may sterically prevent interactions of the target with other protein partners. The intrabody may stabilize or destabilize the target, thus preventing or facilitating turnover, respectively. Intrabody binding may alter folding of the target, possibly leading to altered stability or interaction capacity. The intrabody may also act as a signal to facilitate degradation, should the cellular machinery recognize this intrabody-target complex as a misfolded or foreign protein. Intrabodies may also be labeled with cellular localization sequences to re-target antigens to select subcellular compartments (e.g., lysosome, endoplasmic reticulum). The genetic manipulability of intrabodies and the rich diversity of intrabody libraries, provide a convenient platform from which to generate candidate therapeutics. Several prominent neurological diseases involve aberrantly modified proteins that can serve as unique targets. Here, we discuss three major categories of intrabodies which may affect the target by (1) altering protein folding a n d / o r interactions, (2) altering post-translational modifications, or (3) preventing pathogenic proteolysis. The major pathways through which intrabodies may act are depicted in Fig. 2. Each of these intrabody classes may be applicable to a spectrum of neurological disorders, due to both common disease mechanisms and common targets, and there is some overlap. A, Influencing Target Protein Folding and Interactions Using Intrabodies Abnormally folded proteins are a major theme among neurological diseases, including HD, AD, Parkinson's disease (PD), and prion diseases. The proteins implicated in such disorders are thought to misfold and assume abnormal conformations, thus rendering them susceptible to aberrant protein-protein interactions and formation of oligomers and aggregates. Binding of an exogenous molecule to the vulnerable protein could prevent misfolding a n d / o r sterically prevent aberrant interactions, thus blocking subsequent pathogenesis. 1 • Modifying the Folding and Interactions of Huntingtin in Huntington^s Disease A growing group of neurodegenerative disorders, including HD, several spinocerebellar ataxias, and spinobulbar muscular atrophy are attributable to the expansion of a CAG repeat in the coding region of a gene, leading to extension of a poly Q tract in the disease protein (Michalik and Van Broeckhoven, 2003). Proteins containing an expanded poly-Q tract are

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

MECHANISMS OF INTRABODY APPROACHES TO NEURODEGENERATION

137

Extracellular matrix Cytoplasm pathogenic ^ proteolysis^

benign proteolysis

l l l | post-transiationa! ^ modification

i1 V ^ Intrabodies

ipftiiiSiiiif

Wf-PO. M

proteasomal degradation

scFv VH-DAB VL-D

'^^m^

ITIB US

•

pathogenic misfolding/ aggregation

FIGURE 2 Mechanisms of intrabody action. Intrabodies may alter target proteins through one or a combination of a variety of actions. Depicted here are several of the more likely possible mechanisms, and steps at which intervention with intrabodies could block a pathogenic cascade. Black arrows indicate transitions and modifications of a protein, including misfolding, proteolytic processing, proteasomal degradation, post-translational modification, and transport between organelles and subcellular compartments. Green arrows indicate pathways where stimulation by intrabodies could be therapeutic. Red perpendicular lines indicate stages where inhibitory intrabodies could be useful. Intrabodies may be beneficial by stimulating or inhibiting protein transport between subcellular compartments depending on the compartment involved, as depicted by the dual green/red arrows. Additionally, several of the protein modifications such as proteasomal degradation, here depicted in the cytoplasm, may occur within organelles such as the nucleus, so intrabody targeting to subcellular compartments to alter protein modifications could also be favorable.

prone to misfolding (Ross et a l , 2003), aggregation (Onodera et al., 1997; Scherzinger et a l , 1997), and aberrant protein-protein interactions. Based on the hypothesis that the intrabody technology already in clinical trials for cancer and HIV therapies can be applied to neurodegenerative diseases, a collaboration between the Messer and Huston labs used a large naive human spleen sFv phage-display library (Marks et al., 1991) to examine the behavior of intrabodies against both the expanded poly-Q region and the amino-terminal flanking region. Intrabodies selected against mutant poly-Q showed both a lack of specificity and a degree of generalized toxicity. However, one intrabody, anti-HD-C4, selected against

amino-terminal residues 1-17 adjacent to the poly-Q of huntingtin, successfully counteracted in situ lengthdependent huntingtin aggregation in three different cell lines (Lecerf et al., 2001), as well as in organotypic slice-culture models (Murphy and Messer, 2004). Functional protection against mutant huntingtin-specific malonate toxicity was also demonstrated in the latter study. Binding intrabodies on either side of the mutant poly-Q sequence appears to have beneficial effects. The Patterson lab showed that mouse intrabodies to the poly-proline region flanking the poly-Q on the carboxyl-terminal side (generated from mouse monoclonal Abs) also reduce aggregation and apoptosis, while intrabodies to the expanded poly-Q itself

11. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

138

11. GENE THERAPY FOR CNS DISEASES USING INTRABODIES

were toxic (Khoshnan et al., 2002). These results also suggest that it may be possible to flank the poly-Q with intrabodies, which could prevent misfolding that starts at a region distant from the binding site of a single intrabody. Recently, complete rescue of the eclosion deficit in a Drosophila model of HD, < 25 to 100% was achieved by transgenic expression of the anti-HD-C4 intrabody. Partial rescue of lifespan and delay of neuronal degeneration were also observed. This is the first demonstration of intrabody protection in an intact nervous system (Wolfgang et a l , 2005). The anti-HD-C4 intrabody has also been stably expressed in mouse brain using an equine infectious anemia virus (EIAV) vector (Mazarakis et a l , 2001). Further studies of functional protection continue to show promise, with no apparent toxicity (Fig. 3). Additional intrabodies were isolated using a human scFv yeast surface-display library. A scFv specific for the amino-terminal 20-amino acid residues of huntingtin, and a smaller version consisting of a single-variable light-chain domain ( V L - D A B ) , inhibited aggregation and decreased yeast and cell culture toxicity, demonstrating the potential of further antibody engineering (Colby et a l , 2004a, b). By removing the

cysteine residues using site-directed mutagenesis, and increasing affinity via rounds of random mutations, an intrabody with stronger cytoplasmic effects on huntingtin aggregation has been selected. Analysis of mutant huntingtin aggregation in the presence of severe overexpression of huntingtin exon 1 fragments may test both homo- and heterophilic interactions. Successful intrabodies may sterically block aberrant interactions a n d / o r aggregation, or alter the folding of expanded poly-Q; however, the mechanism of intrabody action remains undefined. The intrabodies described above can also bind to wild-type huntingtin protein. Evidence suggests that the soluble fraction of proteolytically cleaved mutant huntingtin is a toxic species (Kim et al., 1999, 2001; Wellington et a l , 2000). Huntingtin fragments from the wild-type protein are apparently normal breakdown products. Encouragingly, the intrabody that rescued phenotypes in a Drosophila model of HD selectively binds soluble huntingtin fragments, rather than full-length huntingtin protein (Miller et al., 2005). Neurons have the ability to slowly clear huntingtin inclusions in vivo, resulting in a reversal of behavioral dysfunction, if expression of the mutant protein is silenced (Yamamoto et al., 2000; Martin-Aparicio et al., 2001). Therefore, targeting the

FIGURE 3 In vivo expression of an anti-huntingtin scFv intrabody in mouse brain. Four-week old mice were stereotactically injected intrastriatally into the left hemisphere with 10^ lU (2 pL of 10^ lU/mL) of rabies virus glycoprotein-pseudotyped equine infectious anemia virus (Mitrophanous et al,, 1999) encoding a cytomegalovirus promoter-driven anti-huntingtin scFv, anti-HD-C4 (Lecerf et al., 2001). Brains were removed at 3 months postinjection for immunostaining. C4 scFv expression was robust in striatum and cortex in the left hemisphere, suggesting that both striatal and corticostriatal neurons were transduced. The right hemisphere, injected with control virus, did not show C4 immunoreactivity. Ctx- cortex; Str- striatum; V-ventricle. Inset (lower right) shows a C4-expressing cell at higher magnification (40X). Adapted from Miller and Messer (2005).

:L GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

MECHANISMS OF INTRABODY APPROACHES TO NEURODEGENERATION

soluble pathogenic fragment of mutant huntingtin may be beneficial. 2* Influencing cn-'Synuclein Protein Folding and Interactions in Parkinson^s Disease and Synucleinopathies A similar strategy was utilized in the development of anti-a-synuclein intrabodies for synucleinopathies, a group of neurodegenerative disorders characterized by the intracellular accumulation of a-synuclein-positive fibrillar aggregates (Lewy bodies). Synucleinopathies include PD, dementia with Lewy bodies, pure autonomic failure, multiple system atrophy, Lewy body variant of AD, and neurodegeneration with brain iron accumulation type 1 (Lee et al., 2004). The roles of a-synuclein and Lewy body formation in pathogenesis remain controversial, but evidence indicates that a soluble (possibly protofibril) oligomeric a-synuclein intermediate may be a toxic species (Uversky and Fink, 2002; Voiles and Lansbury, 2003). a-Synuclein is believed to exist in an equilibrium between monomeric, jS-sheet oligomeric, and aggregated forms (Maguire-Zeiss and Federoff, 2003), where dominantly transmitted pathogenic mutations (Vila and Przedborski, 2004), oxidative insults (Dawson and Dawson, 2003), and proteasomal inhibition (Giasson and Lee, 2003; McNaught et al., 2004; Zhou et al., 2004b) shift the balance toward toxic oligomer formation. Peptide mimics of a-synuclein have been shown to prevent oligomer/aggregate formation and toxicity in cell culture (El-Agnaf et al., 2004). However, such peptides may interact with other cellular proteins, and they are generally less stable than Ab fragments. Emadi et al. (2004) reported that an scFv selected against monomeric a-synuclein prevents the formation of high-molecularweight oligomers, protofibrils, and aggregates in vitro. Moving into mammalian cells, Zhou et al. (2004a) similarly demonstrated that an anti-monomeric-a-synuclein scFv preferentially binds to and stabilizes monomeric a-synuclein, preventing incorporation into a-synuclein oligomers and increasing the level of monomeric a-synuclein, while decreasing dimer and trimer formation. Such an approach may block the generation of toxic a-synuclein oligomers. The latter intrabody also rescued a cell-adhesion phenotype linked to a-synuclein overexpression in cultured cells, indicating functional correction. Maguire-Zeiss et al. (2004) have begun testing conformation-specific anti-a-synuclein scFvs, which could help to identify the toxic a-synuclein species. Should a-synuclein prove non-essential in adults, targeting of monomeric a-synuclein for degradation

139

may also be therapeutic. An alternative strategy could utilize intrabody-mediated targeting of the toxic aS3niuclein species for degradation. Parkinson's disease is a particularly attractive target for intrabody gene therapy, since the crucial cells are initially limited to a distinct brain region, which should be accessible for gene therapy either directly or via retrograde transport from the striatum (Mazarakis et al., 2001; Martinov et al., 2002). 3. Altering Protein Folding and Interactions in Alzheimer^s Disease and Tauopathies Alzheimer's disease is the most prominent neurodegenerative disorder associated with aging, with rare early-onset disease occurring in younger individuals. This disorder is characterized by the accumulation of extracellular and intracellular fibrils, respectively, containing jS-amyloid and tau (Citron, 2004). A fraction of early-onset familial AD cases are associated with dominantly inherited missense mutations in the gene encoding amyloid precursor proteins (APP) (Goldgaber et al., 1987). jS-Amyloid, a secreted 40- or 42-amino acid peptide, arises from sequential proteolytic processing of APP (Citron, 2004). The jS-amyloid^_42 species is more prone to formation of neurotoxic fibrils (Barrow and Zagorski, 1991; Pike et al., 1991; Lorenzo and Yankner, 1994), and it is a major component of amyloid plaques in AD brains (Glenner and Wong, 1984; Masters et al., 1985; Kang et a l , 1987). Supporting evidence that implicates jSamyloid in AD pathology comes from Down's syndrome cases due to trisomy 21, leading to triplication of the APP gene, AD-like pathology, and formation of j8-amyloid plaques (Wisniewski et al., 1985; Oliver and Holland, 1986). Therefore, intrabodies that block the formation of jS-amyloid plaques or prevent misfolding may be protective. Should j5-amyloid be proven non-essential, targeting it for degradation may be beneficial. Although jS-amyloid plaques accumulate extracellularly, )8-amyloid production likely occurs intracellularly (Shoji et al., 1992; Busciglio et al., 1993; Haass et al., 1993), so intrabody binding and re-targeting of jS-amyloid prior to secretion may be feasible. Indeed, recent findings by the Molinari group indicate that a scFv intrabody targeting an epitope adjacent to the j5-secretase cleavage site of APP can be combined with an endoplasmic reticulum retention signal (ER-retention signal) to prevent jS-amyloid secretion (Paganetti et al., 2005). (This is also a proteolysis inhibition strategy, see Section ILD) Liu et al. (2004a, b) have also demonstrated that an anti-j8-amyloid scFv can prevent

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

140

11. GENE THERAPY FOR CNS DISEASES USING INTRABODIES

aggregation in a cell-free system, inhibiting subsequent toxicity in neuronal cells. Manoutcharian et al. (2004) have recently reported synthetic anti-jS-amyloidi_42 peptides based on the V^-CDRS sequences, which can bind and offer protection when delivered extracellularly to primary rat hippocampal cultures. Binding of scFv or DAB protein or fragments to extracellular ^-amyloid, so as to prevent deposition, plaque formation, and pathogenesis, may also be beneficial, although such a strategy lies outside the realm of intracellular Abs. A 35-amino acid fragment of a-synuclein, the nonamyloid component (NAC), is found in the j5-amyloid plaques of AD brains (Ueda et al., 1993), suggesting that a-synuclein is proteolytically cleaved to release the NAC fragment. NAC may stimulate j5-amyloid aggregation, and vice versa (Han et al., 1995; Yoshimoto et al., 1995). Fibrils of both NAC and a-synuclein are neurotoxic (Conway et al., 1998; El-Agnaf et al., 1998). Bodies et al. (2001) proposed that a-synuclein residues 68-76 are critical for NAC aggregation. Therefore, an intrabody that targets this region of NAC could be more broadly useful, serving to reduce NAC, a-synuclein, and jS-amyloid fibril assembly. Since many neurological disorders show common markers, they may also have common mechanisms of pathogenesis. Such commonalities could permit the development of widely applicable therapeutics. The most frequently occurring feature in neurological diseases is, arguably, neurofibrillary tangles composed primarily of filaments of the microtubule-associated protein, tau (Lewis et al., 2000; Zhang et al., 2004); such tangles have been correlated with neurodegeneration (Braak and Braak, 1991). However, it has not been explained whether tangle formation is itself pathogenic, or whether the tangles appear concomitantly with neuropathology caused by another mechanism. Should the former be the case, blockage of tangle formation could be protective for an array of conditions. Several tauopathies have been linked to dominant missense mutations in tau, including frontotemporal dementia with Parkinsonism linked to chromosome 17, Pick's disease, progressive supranuclear palsy, and corticobasal degeneration (Lee et a l , 2001). Once the domains of tau that are implicated in fibril formation have been thoroughly elucidated, intrabodies could theoretically be selected against these regions and would bind to prevent the regions' self-interactions. There are six isoforms of tau in brain, generated by alternate mRNA splicing (Lee et al., 2001). These isoforms have differing microtubule-binding abilities, and possibly differing roles in disease, so intrabody

selection against particular isoforms could be valuable to an understanding of fibril assembly and the disease process in general. Anti-tau intrabodies have been selected by yeast intracellular Ab capture technology (Visintin et al., 2002), but their efficacy in disease applications remains unknown. 4» Influencing Prion Protein Folding and Interactions In contrast to the disorders described above, prion diseases are caused by an infectious protein. There are several forms of prion diseases that occur in humans, including Creutzfeld-Jakob disease, Gerstmann-Straussler disease, familial fatal insomnia, and the transmissible forms, kuru and new variant Creutzfeld-Jakob disease. Most prion disease cases are idiopathic, but approximately 15% are caused by dominantly inherited mutations in the PRNP gene, encoding prion protein (Prusiner, 1998). The disease mechanism put forth by S.B. Prusiner identifies an abnormal conformation of the prion protein (PrP^^) as the toxic species, where this infectious protein can induce normal prion protein (PrP^) to change conformation and adopt the same abnormal features, eliciting a cascade effect and accumulation of PrP^^ (Prusiner, 1982). The altered conformation shows increased j8-sheet structure and protease resistance (Caughey et al., 1991; Prusiner, 1991; Pan et al., 1993), and it has been detected in both intracellular and extracellular prion amyloid (Kitamoto et al., 1991; Tagliavini et al., 1994; Ma and Lindquist, 2002; Ma et al., 2002). PrP^^ causes neurotoxicity through an undefined mechanism, although roles in proteasomal dysfunction, endoplasmic reticulum stress (Castilla et al., 2004), and astrocyte-mediated damage (Jeffrey et al., 2004) due to altered copper homeostasis (Brown, 2004) have been proposed. In a model of passive immunization, Heppner et al. (2001) demonstrated that transgenic expression of secreted anti-PrP Abs protected mice against PrP^^ infection. This mode of rescue likely utilized extracellular Ab-PrP interaction to prevent disease. Since prion amyloid has also been detected in the cytosol (Ma and Lindquist, 2002; Ma et a l , 2002), and since the disease mechanism may involve intracellular pathways (Brown, 2004; Castilla et al., 2004; Jeffrey et al., 2004), intracellular Ab expression may also be effective. Additionally, an scFv version of the therapeutic Ab was generated (Heppner et al., 2001), although the efficacy of this scFv is unknown. Cardinale et al. (2005) have generated two anti-PrP scFvs from monoclonal Abs, and fused them to secretory leader or ER-retention signals. Retention in the endoplasmic reticulum

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

MECHANISMS OF INTRABODY APPROACHES TO NEURODEGENERATION

of PC12 neuronal cells prevented the appearance of the PrP^ on the cell surface, and also prevented PrP^'^ accumulation. This provides evidence that the toxic misfolding occurs on the cell surface, thus illustrating the use of an intrabody as a tool to study cellular trafficking and molecular pathology, as well as a potential therapy. Other anti-PrP scFvs that may bind PrP and block the pathogenic PrP^'^ cascade have been selected from phage-display libraries (Leclerc et al., 2000; Wuertzer et a l , 2004), and could also prove useful. PrP may be an essential protein. Knock-out mice appear to develop normally, but they show subclinical demyelination with age (Weissmann and Flechsig, 2003). Therefore, anti-PrP'^ intrabody therapy to block Pj.pc ^ pj-psc conversion could require selection of an agent that binds PrP^ but does not interfere with essential functions. Alternatively, a PrP^^-specific intrabody, which would not bind PrP^, could remove PrP^^ or block its interaction with PrP"^. PrP^^ is known to be protease-resistant (Prusiner, 1998); however, Luhr et al. (2004) have recently demonstrated that neuronal cells can degrade PrP^^ when endogenous PrP expression is silenced. This suggests that the pathogenic cellular pathway requires a continuous supply of PrP^, and that interference with PrP^ -> PrP^^ recruitment could similarly enable cells to clear themselves of PrP^^. B.

Targeting Common Amyloid Structures

There appear to be common amyloid protein conformations within inclusions observed in AD, PD, poly-Q, and prion diseases; these conformations share structural features, including j8-sheets, j8-strands, and j8-turns (Ross and Poirier, 2004). The Wetzel group has found that Abs specific for such amyloid conformations can bind inclusions composed of various disease proteins (O'Nuallain and Wetzel, 2002), suggesting shared structural characteristics among disorders. Another Ab described by the Glabe lab recognizes a common conformation-dependent structure that appears to be unique to soluble oligomeric forms of APP, a-synuclein, prion, poly-Q, and some non-neuronal prefibrils, regardless of amino acid sequence. This latter Ab can also inhibit the in vitro toxicity of soluble oligomers (Kayed et al., 2003). Such structural similarities may permit the development of a generic, conformation-specific intrabody for misfolded, pathogenic proteins. However, such conformation-specific intrabodies may also stabilize the abnormal conformations; therefore, extensive testing under multiple conditions will be required. Conformation-dependent intrabodies could be selected from scFv libraries using

141

the colloidal gold bead-coupled-Ab molecular mimic of soluble toxic oligomers, as originally described (Kayed et al., 2003). The candidate intrabodies might then be expressed intracellularly to identify common features of toxic species of a series of proteins, thus making them useful in target identification and validation (Glabe, 2004). C.

Altering Post-Translational Modifications

Proteins implicated in several neurological diseases have been found to be abnormally phosphorylated or oxidatively damaged. Whether these modifications are a cause or effect of the disease process remains a matter of debate. However, substantial evidence suggests that the aberrant modifications of a-synuclein (Fujiwara et al., 2002; Ischiropoulos, 2003), tau (Geschwind, 2003; Horiguchi et al., 2003), superoxide dismutase-1 (Rakhit et al., 2004), neurofilaments (Ackerley et al., 2004), and ataxin-1 (Emamian et al., 2003) contribute to pathogenesis. Intrabody-mediated alteration of the post-translational modifications of disease proteins, either by binding unaltered proteins and directly blocking modifications, or by targeting modified proteins and removing them, could theoretically prevent disease and should be considered an avenue for therapeutic development. !• Intrabody Targeting of Post-Translationally Modified Tau A fraction of tauopathy cases are caused by mutations in tau (Lee et al., 2001). Such mutations can result in altered phosphorylation states of tau (Crowther and Goedert, 2000), decreased microtubule-binding and assembly abilities, and susceptibility to insoluble filament formation (Lee et al., 2004). However, most tauopathies are not related to tau mutations, and hyperphosphorylated, non-mutated tau has been found in neurofibrillary tangles (Drewes, 2004), suggesting that abnormal phosphorylation is linked to pathogenesis. Hyperphosphorylated tau exhibits decreased microtubule-binding ability and is arguably involved in insoluble fibril formation (Lee et al., 2004). Tau may play a role in axonal microtubule organization, although knock-out mice appear phenotypically normal (Harada et al., 1994). Haploinsufficiency has not been reported to result in human disease, suggesting that a single normal allele is adequate. Therefore, therapeutic antitau intrabodies could theoretically target abnormally phosphorylated tau for degradation, or to prevent its incorporation into fibrils, without adversely affecting essential levels or functions of tau.

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

142

11. GENE THERAPY FOR CNS DISEASES USING INTRABODIES

2. Targeting of Post'Translationally Modi^ed (X-Synuclein Using Intrabodies Synucleinopathy lesions similarly contain a-synuclein selectively phosphorylated at S129, and phosphorylation at this residue accelerates fibril formation in vitro (Fujiwara et al., 2002; Kahle et a l , 2002). This phosphorylated species is ubiquitinated in humans (Hasegawa et a l , 2002), implying that it has been targeted for proteasomal degradation. Targeting of S129-phosphorylated a-synuclein using intrabodies may prevent accumulation and fibril formation. Alternatively, an anti-a-synuclein intrabody that blocks phosphorylation could prevent pathogenesis; however, the normal role for the phosphorylated species must first be established. Increased nitration of a-synuclein has also been observed in synucleinopathies (Duda et al., 2000; Giasson et a l , 2000), although in inverse correlation to phosphorylation in mutant mice (Papay et a l , 2002). Nitrotyrosine residues stabilize a-synuclein oligomers (Souza et a l , 2000; Takahashi et al., 2002), suggesting a mechanism for oxidative damage-induced Lewy body formation. Intrabodies that selectively target nitrotyrosinated a-synuclein could remove this damaged species from the cell. 3. Potential for Intrabody Therapeutics in Amyotrophic Lateral Sclerosis Familial amyotrophic lateral sclerosis type 1 (ALS) is caused by mutations in SODl, the gene encoding superoxide dismutase-1. Mutant SODl is prone to aggregation (Bruijn et al., 1997,1998; Elam et al., 2003) and may prevent the formation of functional homodimers (Fridovich, 1986). It is theoretically possible to select an intrabody that will specifically target the mutant form for degradation, thus preventing aggregation, and possibly facilitating nornial dimer formation. However, given the large number of different mutations involved in this disorder (see www.alsod. org), it may be more reasonable to carefully examine the SODl structural data to determine the sites of interaction, and then try to block these sites. In the more common sporadic ALS, however, it is less clear that the accumulating material is SODl itself. Rather, there appears to be a more general breakdown of cellular pathways, leading to deposition of a range of other proteins, which may represent markers rather than primary pathogenic species. If further studies continue to support hypotheses that oxidative stress is a causal agent in ALS (Agar and Durham, 2003), it may be possible to intervene with intrabodies designed to identify abnormally modified proteins (nitrated

or phosphorylated), and to establish a more general approach to easing the toxic burden on neurons or support cells. 4* Intrabody Targeting of Post-Translationally Modified Ataxin-1 Phosphorylation of S776 of ataxin-1, the poly-Qcontaining protein involved in spinocerebellar ataxia type 1, was shown to be important for pathogenesis in transgenic mice (Emamian et a l , 2003). Accordingly, an anti-ataxin-1 intrabody targeting this phosphorylation site could block modification and prevent disease. Alternatively, intrabody-mediated removal of S776-phosphorylated ataxin-1 may be another therapeutic approach. Interestingly, this carboxyl-terminal region implicated in phosphorylation-dependent pathogenesis is also essential for ataxin-1 interaction with an ubiquitin-specific protease, USP7 (Hong et al., 2002); this link further supports the role of aberrant proteolytic processing in the disease pathway. D . Preventing Pathogenic Proteolysis of Disease Proteins A series of neurological diseases are believed to involve abnormal proteolytic cleavage, resulting in the liberation of a pathogenic protein fragment; notable among these are the cleavage of APP to release j5-amyloidi_42 in AD (Goldgaber et al., 1987; Kang et al., 1987; Konig et a l , 1992), and the cleavage of proteins to release the expanded poly-Q-containing fragments that exacerbate disease (Tarlac and Storey, 2003). Proteolytic processing of a-synuclein has also been hypothesized as a factor in PD and other synucleinopathies (Mishizen-Eberz et al., 2003; Li et al., 2005; Liu et al., 2005). Such sites of proteolysis present attractive targets for therapeutic intrabody development, as blockage of the cleavage of disease proteins may prevent the generation of pathogenic fragments and subsequent disease. While blockage of the proteolytic enzymes has been proposed as a treatment for such diseases, an alternative strategy, the use of engineered Ab fragments, could theoretically offer specificity against a sequence that is restricted to the target protein, while leaving the enzymes themselves free to participate in other crucial reactions in the cell. The intrabodies in this application would need to have higher affinities a n d / o r lower off-rates than those in the aforementioned paradigms, since the enzymatic modification of the substrate is an irreversible process. They could also be valuable as part of a multiplex strategy to reduce the toxic load.

. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

POTENTIAL INTRABODY TOXICITY

Paganetti et al. (2005) recently described inhibiting the generation of j8-amyloid using scFvs derived from monoclonal Abs to an epitope adjacent to the APP j8-secretase cleavage site. The intrabodies were targeted to the endoplasmic reticulum, where they attached to the newly formed APP chains, and prevented the abnormal cleavage. One construct also retained the protein in the endoplasmic reticulum, which has the advantage that it also deals with the form of jS-amyloid that escapes from intrabody shielding against cleavage, but at the cost of forcing endoplasmic reticular disposal of excess APP, which could lead to eventual toxicity itself. A potentially beneficial proteolytic cleavage event unique to the treatment of AD is a-secretase-mediated cleavage of APP within the j8-amyloid sequence. Interestingly, an anti-jS-amyloid DAB has shown a-secretase-like activity through internal cleavage of jSamyloid and prevention of cytotoxicity (Rangan et al., 2003; Liu et al., 2004a, b), suggesting that this strategy may also be advantageous in select disease contexts. IIL IMPROVING INTRABODY GENE DELIVERY, EXPRESSION, A N D FUNCTION One of the most powerful aspects of the combination of genetic and proteomic approaches to protein deposition is the capability to further engineer intrabodies via directed or random mutagenesis. As noted in Section LA, the environment used for selection does not currently mimic the intracellular environment in which the intrabody protein must fold and interact. The affinity and stability of effective intrabodies would be relative to the disease or condition under study, and such parameters are only moderately predictable with current selection techniques. The data from several studies have confirmed that while binding to the correct epitopes is valuable for partial correction of a disease, further improvements will be necessary to optimize these reagents. One recent study used a strategy of site-directed elimination of the cysteine residues in a DAB selected against the amino-terminal 20 residues of huntingtin, followed by random mutagenesis to revitalize high-binding affinity. This generated an intrabody whose efficacy has been increased by a factor of 5- to 10-fold in an aggregation assay, with significant protection against mutant huntingtin toxicity in two functional assays (Colby et a l , 2004a). These data suggest that the use of such an iterative mutagenesis

143

and testing strategy with future intrabodies that show promising preliminary properties in cellular assays could be worthwhile, since it would facilitate the more efficient utilization of intrabodies in vivo. However, such mutagenesis techniques may also run the risk of increasing the immunogenicity of an intrabody. Delivery of exogenous genes to the central nervous system poses perhaps the greatest obstacle to intrabody therapeutics for neurological disorders. We will briefly discuss the potential limitations of such an approach, as the reader can find detailed reviews of this subject elsewhere in this book. There are three foreseeable options for intrabody expression in the nervous system: (1) intrabody gene delivery to neurons in vivo using viral vectors, (2) such delivery using intrabody genes tagged with a protein transduction domain (PTD) (Dietz and Bahr, 2004), and (3) ex vivo transduction of cells with PTD-tagged intrabody transgenes followed by cell implantation into the nervous system. Option (1) is currently the most viable, since the ability of PTD-tagged proteins to leave a producer cell and then transduce neighboring cells' cytoplasm in a properly folded configuration is controversial. However, should a reliable PTD system be created, options (2) and (3) might allow intrabodies to treat larger populations of cells, and hence provide more widespread protection. Diseases that are most readily amenable to intrabody therapy will be those in which correction of neurons in a discrete region will be beneficial. It is possible that in diseases such as HD, where there is both a dramatic focus of initial damage in the striatum, and evidence of significant pathology elsewhere, combinations of intrabody therapy in the most affected regions, plus small molecules that can act more ubiquitously, would be most efficacious. IV

POTENTIAL INTRABODY TOXICITY

Many of the intrabodies already developed or proposed for the above studies target sequences that are shared by normal and mutant proteins; this is particularly true for those that have been designed to prevent misfolding or abnormal interactions, rather than to remove altered proteins. The full-length forms of many such proteins may be essential. Therefore, an intrabody targeting a proteolytic fragment may be more therapeutically effective than one specific for a much larger full-length protein in a setting in which both species exist. This situation is especially advantageous in HD, where proteolytic cleavage of poly-Q

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

144

11. GENE THERAPY FOR CNS DISEASES USING INTRABODIES

disease proteins is thought to result in the release of toxic, expanded poly-Q-containing fragments with no known essential functions (Tarlac and Storey, 2003). Expression of virally delivered anti-HD-C4 intrabody (Lecerf et a l , 2001; Murphy and Messer, 2004; Miller et al., 2005; Wolfgang et al., 2005) in mouse brain over several months did not elicit any obvious toxicity (Miller and Messer, 2005). Careful toxicity testing will be required prior to clinical use, and levels of intrabody that reduce — rather than eliminate — the abnormal protein load may strike a balance between preservation of adequate normal function and removal of the offending species. It is also possible that short-term, or pulsed/periodic expression of intrabody genes could act to clear neurons of toxic material that appears to take many years to accumulate, thereby "resetting the clock." This would require regulatable delivery systems, which are under development.

V-

PERSPECTIVE

These new classes of intrabody reagents that can be highly engineered, both as direct therapeutics and as tools for further drug discovery, are theoretically applicable to a wide range of neurological diseases, most of which currently have very few viable treatment options. The obstacles to intrabody applications in neurological disorders reside in (1) the delivery of intrabody-encoding genes to sufficient numbers of target cells in the nervous system, and (2) the need for further understanding of disease mechanisms to enable the optimal targeting of antigens. Engineered Abs may themselves be very valuable in elucidating disease mechanisms. As fusions with capsid or envelope proteins, Ab fragments may also help to direct delivery of viral vectors to cells with specific surface properties. Recombinant viral vectors are obviously promising vehicles for intrabody gene delivery to the nervous system, and the crucial regulatable delivery systems are the subject of intense investigation. The rapidly advancing fields of virus-mediated gene therapy and neurological diseases hold great promise for intrabody therapy for the nervous system in the near future.

ACKNOWLEDGMENTS We thank the current and past members of the Messer lab (Thomas Shirley, William Wolfgang, Jack

Webster, Kevin Manley, Robert Murphy, and Chun Zhou), as well as Drs. James Huston, Dane Wittrup, David Colby, Valerie Bolivar, Kyri Mitrophanous, and Michael Sierks for many helpful discussions of aspects of the work reported here. The EIAV vector used in Fig. 3 was generously provided by Dr. Nicholas Mazarakis of Oxford Biomedica, Ltd. Support for intrabody work in the Messer lab has been provided by NIH, Hereditary Disease Foundation, Cure HD Initiative, Huntington's Disease Society of America, High Q Foundation, and National Parkinson Foundation. References Ackerley, S., Grierson, A.J., Banner, S., Perkinton, M.S., Brownlees, J., Byers, H.L., Ward, M., Thomhill, P., Hussain, K., Waby, J.S., Anderton, B.H., Cooper, J.D., Dingwall, C , Leigh, PN., Shaw, C.E. and Miller, C.C. (2004) p38alpha stress-activated protein kinase phosphorylates neurofilaments and is associated with neurofilament pathology in amyotrophic lateral sclerosis. Mol. Cell. Neurosci., 26: 354-364. Agar, J. and Durham, H. (2003) Relevance of oxidative injury in the pathogenesis of motor neuron diseases. Amyotroph. Lateral Scler. Other Motor Neuron Disord., 4: 232-242. Agrawal, N., Pallos, J., Slepko, N., Apostol, B.L., Bodai, L., Chang, L.W., Chiang, A.S., Thompson, L.M. and Marsh, J.L. (2005) Identification of combinatorial drug regimens for treatment of Huntington's disease using Drosophila. Proc. Natl. Acad. Sci. USA, 102: 3777-3781. Alvarez, R.D., Barnes, M.N., Gomez-Navarro, J., Wang, M., Strong, T.V., Arafat, W., Arani, R.B., Johnson, M.R., Roberts, B.L., Siegal, G.P. and Curiel, D.T. (2000) A cancer gene therapy approach utilizing an anti-erbB-2 single-chain antibody-encoding adenovirus (AD21): a phase I trial. Clin. Cancer Res., 6: 3081-3087. Barrow, C.J. and Zagorski, M.G. (1991) Solution structures of beta peptide and its constituent fragments: relation to amyloid deposition. Science, 253: 179-182. Bates, G. (2003) Huntington aggregation and toxicity in Huntington's disease. Lancet, 361:1642-1644. Beyer, W.E., Palache, A.M., Luchters, G., Nauta, J. and Osterhaus, A.D. (2004) Seroprotection rate, mean fold increase, seroconversion rate: which parameter adequately expresses seroresponse to influenza vaccination? Virus Res., 103:125-132. Bird, R.E., Hardman, K.D., Jacobson, J.W., Johnson, S., Kaufman, B.M., Lee, S.M., Lee, T., Pope, S.H., Riordan, G.S. and Whitlow, M. (1988) Single-chain antigen-binding proteins [erratum appears in Science, 1989 Apr 28; 244(4903): 409]. Science, 242: 423-426. Boder, E.T. and Wittrup, K.D. (1997) Yeast surface display for screening combinatorial polypeptide libraries. Nat. BiotechnoL, 15: 553-557. Bodies, A.M., Guthrie, D.J., Greer, B. and Irvine, G.B. (2001) Identification of the region of non-Abeta component (NAC) of Alzheimer's disease amyloid responsible for its aggregation and toxicity. J. Neurochem., 78: 384-395. Braak, H. and Braak, E. (1991) Neuropathological stageing of Alzheimer-related changes. Acta NeuropathoL, 82: 239-259. Brown, D.R. (2004) Role of the prion protein in copper turnover in astrocytes. Neurobiol. Dis., 15: 534-543. Bruijn, L.I., Becher, M.W, Lee, M.K., Anderson, K.L., Jenkins, N.A., Copeland, N.G., Sisodia, S.S., Rothstein, J.D., Borchelt, D.R.,

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

ACKNOWLEDGMENTS Price, D.L. and Cleveland, D.W. (1997) ALS-linked SODl mutant G85R mediates damage to astrocytes and promotes rapidly progressive disease with SODl-contairung inclusions. Neuron, 18: 327-338. Bruijn, L.L, Houseweart, M.K., Kato, S., Anderson, K.L., Anderson, S.D., Ohama, E., Reaume, A.G., Scott, R.W. and Cleveland, D.W. (1998) Aggregation and motor neuron toxicity of an ALS-linked SODl mutant independent from wild-type SODl. Science, 281: 1851-1854. Busciglio, J., Gabuzda, D.H., Matsudaira, R and Yankner, B.A. (1993). Generation of beta-amyloid in the secretory pathway in neuronal and nonneuronal cells. Proc. Natl. Acad. Sci. USA, 90: 2092-2096. Cardinale, A., Filesi, I., Vetrugno, V., Pocchiari, M., Sy, M.S. and Biocca, S. (2005) Trapping prion protein in the endoplasmic reticulum impairs PrP^ maturation and prevents PrP^'^ accumulation. J. Biol. Chem., 280: 685-694. Castilla, J., Hetz, C. and Soto, C. (2004) Molecular mechanisms of neurotoxicity of pathological prion protein. Curr. Mol. Med., 4: 397-403. Caughey, B.W., Dong, A., Bhat, K.S., Ernst, D., Hayes, S.R and Caughey, W.S. (1991) Secondary structure analysis of the scrapie-associated protein PrP 27-30 in water by infrared spectroscopy[erratum appears in Biochemistry, 1991 Oct 29; 30(43): 10600]. Biochemistry, 30: 7672-7680. Citron, M. (2004) Strategies for disease modification in Alzheimer's disease. Nat. Rev. Neurosci., 5: 677-685. Colby, D.W., Chu, Y, Cassady, J.P, Duennwald, M., Zazulak, H., Webster, J.M., Messer, A., Lindquist, S., Ingram, V.M. and Wittrup, K.D. (2004a) Potent inhibition of huntingtin aggregation and cytotoxicity by a disulfide bond-free single-domain intracellular antibody Proc. Natl. Acad. Sci., 101:17616-17621. Colby, D.W, Garg, P , Holden, T., Chao, G., Webster, J.M., Messer, A., Ingram, V.M. and Wittrup, K.D. (2004b) Development of a human light chain variable domain (V(L)) intracellular antibody specific for the amino terminus of huntingtin via yeast surface display. J. Mol. Biol., 342: 901-912. Colby, D.W, Kellogg, B.A., Graff, C.R, Yeung, YA., Swers, J.S. and Wittrup, K.D. (2004c) Engineering antibody affinity by yeast surface display Method. EnzymoL, 388: 348-358. Conway, K.A., Harper, J.D. and Lansbury, PT. (1998) Accelerated in vitro fibril formation by a mutant alpha-s5niuclein linked to early-onset Parkinson disease. Nat. Med., 4:1318-1320. Crowther, R.A. and Goedert, M. (2000) Abnormal tau-containing filaments in neurodegenerative diseases. J. Struct. Biol., 130, 271-279. Dawson, T.M. and Dawson, V.L. (2003) Molecular pathways of neurodegeneration in Parkinson's disease. Science, 302: 819-822. Dietz, G.P. and Bahr, M. (2004) Delivery of bioactive molecules into the cell: the Trojan horse approach. Mol. Cell. Neurosci., 27: 85-131. Drewes, G. (2004) MARKing tau for tangles and toxicity. Trends Biochem. Sci., 29: 548-555. Duda, J.E., Giasson, B.I., Chen, Q., Gur, T.L., Hurtig, H.I., Stern, M.B., GoUomp, S.M., Ischiropoulos, H., Lee, V.M. and Trojanowski, J.Q. (2000) Widespread nitration of pathological inclusions in neurodegenerative synucleinopathies. Am. J. Pathol., 157:1439-1445. El-Agnaf, O.M., Jakes, R., Curran, M.D., Middleton, D., Ingenito, R., Bianchi, E., Pessi, A., Neill, D. and Wallace, A. (1998) Aggregates from mutant and wild-type alpha-synuclein proteins and NAC peptide induce apoptotic cell death in human neuroblastoma cells by formation of beta-sheet and amyloid-like filaments. FEBS Lett., 440: 71-75.

145

El-Agnaf, O.M., Paleologou, K.E., Greer, B., Abogrein, A.M., King, J.E., Salem, S.A., Fullwood, N.J., Benson, RE., Hewitt, R., Ford, K.J., Martin, FL., Harriott, P., Cookson, M.R. and Allsop, D. (2004) A strategy for designing inhibitors of alpha-synuclein aggregation and toxicity as a novel treatment for Parkinson's disease and related disorders. FASEB J., 18:1315-1317. Elam, J.S., Taylor, A.B., Strange, R., Antonyuk, S., Doucette, PA., Rodriguez, J.A., Hasnain, S.S., Hayward, L.J., Valentine, J.S., Yeates, T.O. and Hart, P.J. (2003) Amyloid-like filaments and water-filled nanotubes formed by SODl mutant proteins linked to familial ALS. Nat. Struct. Biol., 10: 461-467. Emadi, S., Liu, R., Yuan, B., Schulz, P, McAllister, C , Lyubchenko, Y, Messer, A. and Sierks, M.R. (2004) Inhibiting aggregation of alpha-synuclein with human single chain antibody fragments. Biochemistry 43: 2871-2878. Emamian, E.S., Kaytor, M.D., Duvick, L.A., Zu, T., Tousey, S.K., Zoghbi, H.Y, Clark, H.B. and Orr, H.T. (2003) Serine 776 of ataxin-1 is critical for polyglutamine-induced disease in SCAl transgenic mice. Neuron, 38: 375-387. Ewert, S., Honegger, A. and Pluckthun, A. (2004) Stability improvement of antibodies for extracellular and intracellular applications: CDR grafting to stable frameworks and structure-based framework engineering. Methods, 34: 184-199. Fridovich, I. (1986) Superoxide dismutases. Adv. EnzymoL Relat. Areas Mol. Biol., 58: 61-97. Fujiwara, H., Hasegawa, M., Dohmae, N., Kawashima, A., Masliah, E., Goldberg, M.S., Shen, J., Takio, K. and Iwatsubo, T. (2002) AlphaSynuclein is phosphorylated in synucleinopathy lesions. Nat. Cell Biol., 4:160-164. Gennari, R, Mehta, S., Wang, Y, St Clair Tallarico, A., Palu, G. and Marasco, WA. (2004) Direct phage to intrabody screening (DPIS): demonstration by isolation of cytosolic intrabodies against the TESl site of Epstein Barr virus latent membrane protein 1 (LMPl) that block NF-kappaB transactivation. J. Mol. Biol., 335: 193-207. Geschwind, D.H. (2003) Tau phosphorylation, tangles, and neurodegeneration: the chicken or the egg? Neuron, 40: 457-460. Giasson, B.I., Duda, J.E., Murray, I.V, Chen, Q., Souza, J.M., Hurtig, H.I., Ischiropoulos, H., Trojanowski, J.Q. and Lee, V.M. (2000) Oxidative damage linked to neurodegeneration by selective alpha-synuclein nitration in synucleinopathy lesions, [see comment]. Science, 290: 985-989. Giasson, B.I. and Lee, V.M. (2003) Are ubiquitination pathways central to Parkinson's disease? Cell, 114:1-8. Glabe, C.G. (2004) Conformation-dependent antibodies target diseases of protein misfolding. Trends Biochem. Sci., 29: 542-547. Glenner, G.G. and Wong, C.W. (1984) Alzheimer's disease: initial report of the purification and characterization of a novel cerebrovascular amyloid protein. Biochem. Biophys. Res. Commun., 120: 885-890. Goldgaber, D., Lerman, M.I., McBride, W.O., Saffiotti, U. and Gajdusek, D.C. (1987) Isolation, characterization, and chromosomal localization of human brain cDNA clones coding for the precursor of the amyloid of brain in Alzheimer's disease, Down's syndrome and aging. J. Neural Transm. Suppl., 24: 23-28. Haass, C., Hung, A.Y, Schlossmacher, M.G., Oltersdorf, T., Teplow, D.B. and Selkoe, D.J. (1993) Normal cellular processing of the beta-amyloid precursor protein results in the secretion of the amyloid beta peptide and related molecules. Ann. NY Acad. Sci., 695:109-116. Han, H., Weinreb, PH. and Lansbury, PT., Jr. (1995) The core Alzheimer's peptide NAC forms amyloid fibrils which seed and

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

146

11. GENE THERAPY FOR CNS DISEASES USING INTRABODIES

are seeded by beta-amyloid: is NAC a common trigger or target in neurodegenerative disease? Chem. Biol., 2: 163-169. Harada, A., Oguchi, K., Okabe, S., Kuno, J., Terada, S., Ohshima, T., Sato-Yoshitake, R., Takei, Y, Noda, T. and Hirokawa, N. (1994) Altered microtubule organization in small-calibre axons of mice lacking tau protein. Nature, 369: 488-491. Harper, S.Q., Staber, P.D., He, X., Eliason, S.L., Martins, I.H., Mao, Q., Yang, L., Kotin, R.M., Paulson, H.L. and Davidson, B.L. (2005) From the cover: RNA interference improves motor and neuropathological abnormalities in a Huntington's disease mouse model. Proc. Natl. Acad. Sci. USA, 102: 5820-5825. Hasegawa, M., Fujiwara, H., Nonaka, T., Wakabayashi, K., Takahashi, H., Lee, V.M., Trojanowski, J.Q., Mann, D. and Iwatsubo, T. (2002) Phosphorylated alpha-synuclein is ubiquitinated in alphasynucleinopathy lesions. J. Biol. Chem., 277: 49071-49076. Heppner, F.L., Musahl, C , Arrighi, I., Klein, M.A., Rulicke, T., Oesch, B., Zinkemagel, R.M., Kalinke, U. and Aguzzi, A. (2001) Prevention of scrapie pathogenesis by transgenic expression of anti-prion protein antibodies. Science, 294: 178-182. Hersch, S.M. and Ferrante, R.J. (2004) Translating therapies for Huntington's disease from genetic animal models to clinical trials. Neurorx, 1: 298-306. Hong, S., Kim, S.J., Ka, S., Choi, I. and Kang, S. (2002) USP7, a ubiquitin-specific protease, interacts with ataxin-1, the SCAl gene product. Mol. Cell. Neurosci., 20: 298-306. Horiguchi, T., Uryu, K., Giasson, B.I., Ischiropoulos, H., LightFoot, R., Bellmann, C , Richter-Landsberg, C , Lee, V.M. and Trojanowski, J.Q. (2003) Nitration of tau protein is linked to neurodegeneration in tauopathies [erratum appears in Am J Pathol, 2003 Dec; 163(6): 2645]. Am. J. Pathol., 163:1021-1031. Huston, J.S., Levinson, D., Mudgett-Hunter, M., Tai, M.S., Novotny, J., MargoUes, M.N., Ridge, R.J., Bruccoleri, R.E., Haber, E. and Crea, R. (1988) Protein engineering of antibody binding sites: recovery of specific activity in an anti-digoxin single-chain Fv analogue produced in Escherichia coli. Proc. Natl. Acad. Sci. USA, 85: 5879-5883. Ischiropoulos, H. (2003) Oxidative modifications of alpha-synuclein. Ann. NY Acad. Sci., 991: 93-100. Jeffrey M., Goodsir, CM., Race, R.E. and Chesebro, B. (2004) Scrapie-specific neuronal lesions are independent of neuronal PrP expression. Ann. Neurol., 55: 781-792. Kahle, P.J., Neumann, M., Ozmen, L., Muller, V., Jacobsen, H., Spooren, W., Fuss, B., Mallon, B., Macklin, W.B., Fujiwara, H., Hasegawa, M., Iwatsubo, T., Kretzschmar, H.A. and Haass, C. (2002) Hyperphosphorylation and insolubility of alphasynuclein in transgenic mouse oligodendrocytes. EMBO Rep., 3: 583-588. Kang, J., Lemaire, H.G., Unterbeck, A., Salbaum, J.M., Masters, C.L., Grzeschik, K.H., Multhaup, G., Beyreuther, K. and MuUer-Hill, B. (1987) The precursor of Alzheimer's disease amyloid A4 protein resembles a cell-surface receptor. Nature, 325: 733-736. Kayed, R., Head, E., Thompson, J.L., Mclntire, T.M., Milton, S.C, Cotman, C.W. and Glabe, C.G. (2003) Common structure of soluble amyloid oligomers implies common mechanism of pathogenesis. Science, 300: 486-489. Kazantsev, A., Walker, H.A., Slepko, N., Bear, J.E., Preisinger, E., Steffan, J.S., Zhu, Y.Z., Gertler, KB., Housman, D.E., Marsh, J.L. and Thompson, L.M. (2002) A bivalent Huntingtin binding peptide suppresses polyglutamine aggregation and pathogenesis in Drosophila. Nat. Genet., 30: 367-376. Khoshnan, A., Ko, J. and Patterson, PH. (2002) Effects of intracellular expression of anti-huntingtin antibodies of various specificities

on mutant huntingtin aggregation and toxicity. Proc. Natl. Acad. Sci. USA, 99:1002-1007. Kieke, M.C., Cho, B.K., Boder, E.T., Kranz, D.M. and Wittrup, K.D. (1997) Isolation of anti-T cell receptor scFv mutants by yeast surface display. Protein Eng., 10:1303-1310. Kim, M., Lee, H.S., LaForet, G., Mclntyre, C , Martin, E.J., Chang, R, Kim, T.W., WiUiams, M., Reddy, PH., Tagle, D., Boyce, KM., Won, L., Heller, A., Aronin, N. and DiFiglia, M. (1999) Mutant huntingtin expression in clonal striatal cells: dissociation of inclusion formation and neuronal survival by caspase inhibition. J. Neurosci., 19: 964-973. Kim, YJ., Yi, Y, Sapp, E., Wang, Y, Cuiffo, B., Kegel, K.B., Qin, Z.H., Aronin, N. and DiFiglia, M. (2001) Caspase 3-cleaved N-terminal fragments of wild-type and mutant huntingtin are present in normal and Huntington's disease brains, associate with membranes, and undergo calpain-dependent proteolysis. Proc. Natl. Acad. Sci. USA, 98: 12784-12789. Kitamoto, T, Muramoto, T., Mohri, S., Doh-Ura, K. and Tateishi, J. (1991) Abnormal isoform of prion protein accumulates in follicular dendritic cells in mice with Creutzfeldt-Jakob disease. J. Virol., 65: 6292-6295. Konig, G., Monning, U., Czech, C , Prior, R., Banati, R., SchreiterGasser, U., Bauer, J., Masters, C.L. and Beyreuther, K. (1992) Identification and differential expression of a novel alternative splice isoform of the beta A4 amyloid precursor protein (APP) mRNA in leukocytes and brain microglial cells. J. Biol. Chem., 267: 10804-10809. Leath, C.A., III, Douglas, J.T., Curiel, D.T. and Alvarez, R.D. (2004) Single-chain antibodies: a therapeutic modality for cancer gene therapy (review). Int. J. Oncol., 24: 765-771. Lecerf, J.M., Shirley, T.L., Zhu, Q., Kazantsev, A., Amersdorfer, P., Housman, D.E., Messer, A. and Huston, J.S. (2001) Human singlechain Fv intrabodies counteract in situ huntingtin aggregation in cellular models of Huntington's disease. Proc. Natl. Acad. Sci. USA, 98: 4764^769. Leclerc, E., Liemann, S., Wildegger, G., Vetter, S.W and Nilsson, F. (2000) Selection and characterization of single chain Fv fragments against murine recombinant prion protein from a synthetic human antibody phage display library. Hum. Antibodies, 9: 207-214. Lee, M.K., Stirling, W , Xu, Y, Xu, X., Qui, D., Mandir, A.S., Dawson, T.M., Copeland, N.G., Jenkins, N.A. and Price, D.L. (2002) Human alpha-synuclein-harboring familial Parkinson's diseaselinked Ala-53 ~> Thr mutation causes neurodegenerative disease with alpha-synuclein aggregation in transgenic mice. Proc. Natl. Acad. Sci. USA, 99: 8968-8973. Lee, V.M., Giasson, B.I. and Trojanowski, J.Q. (2004) More than just two peas in a pod: common amyloidogenic properties of tau and alpha-synuclein in neurodegenerative diseases. Trends Neurosci., 27,129-134. Lee, V.M., Goedert, M. and Trojanowski, J.Q. (2001) Neurodegenerative tauopathies. Annu. Rev. Neurosci., 24: 1121-1159. Lewis, J., McGowan, E., Rockwood, J., Melrose, H., Nacharaju, P., Van Slegtenhorst, M., Gwinn-Hardy, K., Paul Murphy, M., Baker, M., Yu, X., Duff, K., Hardy, J., Corral, A., Lin, W.L., Yen, S.H., Dickson, D.W, Davies, P and Hutton, M. (2000) Neurofibrillary tangles, amyotrophy and progressive motor disturbance in mice expressing mutant (P301L) tau protein. Nat. Genet., 25: 402-405. Li, W , West, N., Colla, E., Pletnikova, O., Troncoso, J.C, Marsh, L., Dawson, T.M., Jakala, P., Hartmann, T., Price, D.L. and Lee, M.K. (2005) Aggregation promoting C-terminal truncation of alphasynuclein is a normal cellular process and is enhanced by the

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

ACKNOWLEDGMENTS familial Parkinson's disease-linked mutations. Proc. Natl. Acad. Sci., 102: 2162-2167. Liu, C.W., Giasson, B.I., Lewis, K.A., Lee, V.M., Demartino, G.N. and Thomas, P.J. (2005) A precipitating role for truncated alphasynuclein and the proteasome in alpha-synuclein aggregation: implications for pathogenesis of Parkinson's disease. J. Biol. Chem., 280: 22670-22678. Liu, R., McAllister, C., Lyubchenko, Y and Sierks, M.R. (2004a). Proteolytic antibody light chains alter beta-amyloid aggregation and prevent cytotoxicity. Biochemistry, 43: 9999-10007. Liu, R., Yuan, B., Emadi, S., Zameer, A., Schulz, R, McAllister, C., Lyubchenko, Y, Goud, G. and Sierks, M.R. (2004b) Single chain variable fragments against beta-amyloid (Abeta) can inhibit Abeta aggregation and prevent abeta-induced neurotoxicity. Biochemistry 43: 6959-6967. Lorenzo, A. and Yankner, B.A. (1994) Beta-amyloid neurotoxicity requires fibril formation and is inhibited by congo red. Proc. Natl. Acad. Sci. USA, 91:12243-12247. Luhr, K.M., Nordstrom, E.K., Low, P., Ljunggren, H.G., Taraboulos, A. and Kristensson, K. (2004) Scrapie protein degradation by cysteine proteases in CDllc+ dendritic cells and GTl-1 neuronal cells. J. Virol., 78: 4776-4782. Ma, J. and Lindquist, S. (2002) Conversion of PrP to a self-perpetuating PrP^^-like conformation in the cytosol.[see comment]. Science, 298:1785-1788. Ma, J., Wollmaim, R. and Lindquist, S. (2002) Neurotoxicity and neurodegeneration when PrP accumulates in the cytosol.[see comment]. Science, 298:1781-1785. Maguire-Zeiss, K.A. and Federoff, H.J. (2003) Convergent pathobiologic model of Parkinson's disease. Ann. NY Acad. Sci., 991: 152-66. Maguire-Zeiss, K.A., Yehling, E., Giuliano, R., Sullivan, M. and Federoff, H.J. (2004) HSV amplicon expression of single-chain antibodies directed against alpha-synuclein conformers. Mol. Ther., 9: S86. Manoutcharian, K., Acero, G., Munguia, M.E., Becerril, B., Massieu, L., Govezensky T., Ortiz, E., Marks, J.D., Cao, C , Ugen, K. and Gevorkian, G. (2004) Human single chain Fv antibodies and a complementarity determirung region-derived peptide binding to amyloid-beta 1-42. Neurobiol. Dis., 17: 114-121. Marasco, W.A., LaVecchio, J. and Winkler, A. (1999) Human antiHIV-1 tat sFv intrabodies for gene therapy of advanced HIV-1infection and AIDS. J. Immunol. Methods, 231: 223-238. Marks, J.D., Hoogenboom, H.R., Bonnert, T.P., McCafferty, J., Griffiths, A.D. and Winter, G. (1991) By-passing immunization: human antibodies from V-gene libraries displayed on phage. J. Mol. Biol., 222: 581-597. Marsh, J.L. and Thompson, L.M. (2004) Can flies help humans treat neurodegenerative diseases? Bioessays, 26: 485-496. Martin-Aparicio, E., Yamamoto, A., Hernandez, R, Hen, R., Avila, J. and Lucas, J.J. (2001) Proteasomal-dependent aggregate reversal and absence of cell death in a conditional mouse model of Huntington's disease. J. Neurosci., 21: 8772-8781. Martinov, V.N., Sefland, I., Walaas, S.I., Lomo, T., Nja, A, and Hoover, F. (2002) Targeting functional subtypes of spinal motoneurons and skeletal muscle fibers in vivo by intramuscular injection of adenoviral and adeno-associated viral vectors. Anat. Embryol. (Berl.), 205: 215-221. Masters, C.L., Simms, G., Weinman, N.A., Multhaup, G., McDonald, B.L. and Beyreuther, K. (1985) Amyloid plaque core protein in Alzheimer disease and Down syndrome. Proc. Natl. Acad. Sci., 82: 4245-4249.

147

Mazarakis, N.D., Azzouz, M., Rohll, J.B., EUard, P.M., Wilkes, RJ., Olsen, A.L., Carter, E.E., Barber, R.D., Baban, D.R, Kingsman, S.M., Kingsman, A.J., O'Malley, K. and Mitrophanous, K.A. (2001) Rabies virus glycoprotein pseudotyping of lentiviral vectors enables retrograde axonal transport and access to the nervous system after peripheral delivery. Hum. Mol. Genet., 10: 2109-2121. McNaught, K.S., Perl, D.R, Brownell, A.L. and Olanow, C.W. (2004) Systemic exposure to proteasome inhibitors causes a progressive model of Parkinson's disease. Arm. Neurol., 56:149-162. Michalik, A. and Van Broeckhoven, C. (2003) Pathogenesis of polyglutamine disorders: aggregation revisited. Hum. Mol. Genet., 12 (Spec No 2): R173—R186. Miller, T.W. and Messer, A. (2005) Intrabody applications in neurological disorders: progress and future prospects. Mol. Ther., 12: 394-401. Miller, T.W, Zhou, C , Gines, S., MacDonald, M.E., Mazarakis, N.D., Bates, G.R and Messer, A. (2005) A human single-chain Fv intrabody selectively targets amino-terminal himtingtin fragments in striatal models of Huntington's disease. Neurobiol. Dis., 19:47-56. Mishizen-Eberz, A.J., Guttmann, R.P., Giasson, B.L, Day, G.A., III, Hodara, R., Ischiropoulos, H., Lee, V.M., Trojanowski, J.Q. and Lynch, D.R. (2003) Distinct cleavage patterns of normal and pathologic forms of alpha-s)muclein by calpain I in vitro. J. Neurochem., 86: 836-847. Mitrophanous, K., Yoon, S., Rohll, J., Patil, D., Wilkes, R, Kim, V, Kingsman, S., Kingsman, A. and Mazarakis, N. (1999) Stable gene transfer to the nervous system using a non-primate lentiviral vector. Gene Ther., 6: 1808-1818. Mossner, E., Koch, H. and Pluckthun, A. (2001) Fast selection of antibodies without antigen purification: adaptation of the protein fragment complementation assay to select antigen-antibody pairs. J. Mol. Biol., 308:115-122. Murphy, R.C. and Messer, A. (2004) A single-chain Fv intrabody provides functional protection against the effects of mutant protein in an organotypic slice culture model of Huntington's disease. Brain Res. Mol. Brain Res., 121:141-145. Nagai, Y, Tucker, T., Ren, H., Kenan, D.J., Henderson, B.S., Keene, J.D., Strittmatter, W.J. and Burke, J.R. (2000) Inhibition of polyglutamine protein aggregation and cell death by novel peptides identified by phage display screening. J. Biol. Chem., 275: 10437-10442. Oliver, C. and Holland, A.J. (1986) Down's syndrome and Alzheimer's disease: a review. Psychol. Med., 16: 307-322. Onodera, O., Burke, J.R., Miller, S.E., Hester, S., Tsuji, S., Roses, A.D. and Strittmatter, W.J. (1997) Oligomerization of expanded-polyglutamine domain fluorescent fusion proteins in cultured mammalian cells. Biochem. Biophys. Res. Commun.,. 238: 599-605, O'Nuallain, B. and Wetzel, R. (2002) Conformational Abs recognizing a generic amyloid fibril epitope. Proc. Natl. Acad. Sci. USA, 99:1485-1490. Orlandi, R., Gussow, D.H., Jones, RT. and Winter, G. (1989) Cloning immunoglobulin variable domains for expression by the polymerase chain reaction. Proc. Natl. Acad. Sci. USA, 86: 3833-3837. Osboum, J.K., Field, A., Wilton, J., Derbyshire, E., Eamshaw, |.C., Jones, RT., Allen, D. and McCafferty J. (1996) Generation of a panel of related human scFv antibodies with high affinities for human CEA. Immunotechnology 2:181-196. . Paganetti, R, Calanca, V, GalU, C , Stefani, M. and Molinari, M. (2005) j5-site specific intrabodies to decrease and prevent generation of Alzheimer's Aj5 peptide. J. CeU Biol, 168: 863-868.

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

148

11. GENE THERAPY FOR CNS DISEASES USING INTRABODIES

Pan, K.M., Baldwin, M., Nguyen, J., Gasset, M., Serban, A., Groth, D., Mehlhorn, L, Huang, Z., Fletterick, R.J. and Cohen, RE. (1993) Conversion of alpha-helices into beta-sheets features in the formation of the scrapie prion proteins. Proc. Natl. Acad. Sci. USA, 90: 10962-10966. Papay, R., Zuscik, M.J., Ross, S.A., Yun, J., McCune, D.F., Gonzalez-Cabrera, P., Gaivin, R., Drazba, J. and Perez, D.M. (2002) Mice expressing the alpha(lB)-adrenergic receptor induces a synucleinopathy with excessive tyrosine nitration but decreased phosphorylation. J. Neurochem., 83: 623-634. Pike, C.J., Walencewicz, A.J., Glabe, C.G. and Cotman, C.W. (1991) In vitro aging of beta-amyloid protein causes peptide aggregation and neurotoxicity. Brain Res., 563: 311-314. Prusiner, S.B. (1982) Novel proteinaceous infectious particles cause scrapie. Science, 216:136-144. Prusiner, S.B. (1991) Molecular biology of prion diseases. Science, 252:1515-1522. Prusiner, S.B. (1998) Prions. Proc. Natl. Acad. Sci. USA, 95: 1336313383. Rakhit, R., Crow, J.P, Lepock, J.R., Kondejewski, L.H., Cashman, N.R. and Chakrabartty, A. (2004) Monomeric Cu, Zn-superoxide dismutase is a common misfolding intermediate in the oxidation models of sporadic and familial amyotrophic lateral sclerosis. J. Biol. Chem., 279: 15499-15504. Rangan, S.K., Liu, R., Brune, D., Planque, S., Paul, S. and Sierks, M.R. (2003) Degradation of beta-amyloid by proteolytic antibody Ught chains. Biochemistry, 42: 14328-14334. Ross, C.A. and Poirier, M.A. (2004) Protein aggregation and neurodegenerative disease. Nat. Med., lO(Suppl): SIO—S17. Ross, C.A., Poirier, M.A., Wanker, E.E. and Amzel, M. (2003) Polyglutamine fibrillogenesis: the pathway unfolds. Proc. Natl. Acad. Sci. USA, 100: 1-3. Scherzinger, E., Lurz, R., Turmaine, M., Mangiarini, L., Hollenbach, B., Hasenbank, R., Bates, G.P, Davies, S.W., Lehrach, H. and Wanker, E.E. (1997) Huntingtin-encoded polyglutamine expansions form amyloid-like protein aggregates in vitro and in vivo. Cell, 90: 549-558. Schier, R. and Marks, J.D. (1993) Efficient in vitro affinity maturation of phage antibodies using BIAcore guided selections. Hum. Antibodies Hybridomas, 7: 97-105. Shoji, M., Golde, T.E., Ghiso, J., Cheung, T.T., Eshis, S., Shaffer, L.M., Cai, X.D., McKay D.M., Tintner, R. and Frangione, B. (1992) Production of the Alzheimer amyloid beta protein by normal proteolytic processing. Science, 258: 126-129. Souza, J.M., Giasson, B.I., Chen, Q., Lee, V.M. and Ischiropoulos, H. (2000) Dityrosine cross-linking promotes formation of stable alpha -synuclein polymers; implication of nitrative and oxidative stress in the pathogenesis of neurodegenerative synucleinopathies. J. Biol. Chem., 275:18344-18349. Stemmer, WP. (1994) DNA shuffling by random fragmentation and reassembly: in vitro recombination for molecular evolution. Proc. Natl. Acad. Sci. USA, 91: 10747-10751. Stocks, M.R. (2004) Intrabodies: production and promise. Drug Discov. Today, 9: 960-966. Tagliavini, R, Prelli, R, Porro, M., Rossi, G., Giaccone, G., Farlow, M.R., Dlouhy, S.R., Ghetti, B., Bugiani, O. and Frangione, B. (1994). Amyloid fibrils in Gerstmann-Straussler-Scheinker disease (Indiana and Swedish kindreds) express only PrP peptides encoded by the mutant allele. Cell, 79: 695-703. Takahashi, T., Yamashita, H., Nakamura, T., Nagano, Y and Nakamura, S. (2002) Tyrosine 125 of alpha-synuclein plays a critical role for dimerization following nitrative stress. Brain Res., 938: 73-80.

Tanaka, T., Chung, G.T., Forster, A., Lobato, M.N. and Rabbitts, T.H. (2003) De novo production of diverse intracellular antibody libraries. Nucleic Acids Res., 31: E23. Tarlac, V. and Storey, E. (2003) Role of proteolysis in polyglutamine disorders. J. Neurosci. Res., 74, 406-416. Tewari, D., Notkins, A.L. and Zhou, P (2003) Inhibition of HIV-1 replication in primary human T cells transduced with an intracellular anti-HIV-1 p l 7 antibody gene. J. Gene Med., 5:182-189. Tse, E., Lobato, M.N., Forster, A., Tanaka, T., Chung, G.T. and Rabbitts, T.H. (2002) Intracellular antibody capture technology: application to selection of intracellular antibodies recognising the BCR-ABL oncogenic protein. J. Mol. Biol., 317: 85-94. Ueda, K., Fukushima, H., Masliah, E., Xia, Y, Iwai, A., Yoshimoto, M., Otero, D.A., Kondo, J., Ihara, Y and Saitoh, T. (1993) Molecular cloning of cDNA encoding an unrecognized component of amyloid in Alzheimer disease. Proc. Natl. Acad. Sci. USA, 90:11282-11286. Uversky, V.N. and Fink, A.L. (2002) Amino acid determinants of alpha-synuclein aggregation: putting together pieces of the puzzle. FEBS Lett., 522: 9-13. Vila, M. and Przedborski, S. (2004) Genetic clues to the pathogenesis of Parkinson's disease. Nat. Med., 10 (Suppl): S58-S62. Visintin, M., Settanni, G., Maritan, A., Graziosi, S., Marks, J.D. and Cattaneo, A. (2002) The intracellular antibody capture technology (IACT): towards a consensus sequence for intracellular antibodies. J. Mol. Biol., 317: 73-83. Voiles, M.J. and Lansbury, P.T., Jr. (2003) Zeroing in on the pathogenic form of alpha-synuclein and its mechanism of neurotoxicity in Parkinson's disease. Biochemistry, 42: 7871-7878. Weissmann, C. and Flechsig, E. (2003) PrP knock-out and PrP transgenic mice in prion research. Brit. Med. Bull., 66: 43-60. Wellington, C.L., Singaraja, R., EUerby, L., Savill, J., Roy, S., Leavitt, B., Cattaneo, E., Hackam, A., Sharp, A., Thomberry, N., Nicholson, D.W, Bredesen, D.E. and Hayden, M.R. (2000) Inhibiting caspase cleavage of huntingtin reduces toxicity and aggregate formation in neuronal and nonneuronal cells. J. Biol. Chem., 275: 19831-19838. Wheeler, YY, Chen, S.Y and Sane, D.C. (2003a) Intrabody and intrakine strategies for molecular therapy. Mol. Ther., 8: 355-366. Wheeler, YY, Kute, T.E., Willingham, M.C., Chen, S.Y. and Sane, D.C. (2003b) Intrabody-based strategies for inhibition of vascular endothelial growth factor receptor-2: effects on apoptosis, cell growth, and angiogenesis. FASEB J., 17:1733-1735. Wisniewski, K.E., Wisniewski, H.M. and Wen, G.Y (1985) Occurrence of neuropathological changes and dementia of Alzheimer's disease in Down's syndrome. Ann. Neurol., 17: 278-282. Worn, A., Auf der Maur, A., Escher, D., Honegger, A., Barberis, A. and Pluckthun, A. (2000) Correlation between in vitro stability and in vivo performance of anti-GCN4 intrabodies as cytoplasmic inhibitors. J. Biol. Chem., 275: 2795-2803. Wuertzer, C.A., Sullivan, M. and Federoff, H.J. (2004) Single-chain antibodies for viral-based passive immuno-therapy for prion disease: linear epitope mapping and clearance from thalamic injection. Mol. Ther., 9: S208. Xia, H., Mao, Q., Eliason, S.L., Harper, S.Q., Martins, I.H., Orr, H.T., Paulson, H.L., Yang, L., Kotin, R.M. and Davidson, B.L. (2004) RNAi suppresses polyglutamine-induced neurodegeneration in a model of spinocerebellar ataxia. Nat. Med., 10: 816-820. Yamamoto, A., Lucas, J.J. and Hen, R. (2000) Reversal of neuropathology and motor dysfunction in a conditional model of Huntington's disease.[see comment]. Cell, 101: 57-66. Yoshimoto, M., Iwai, A., Kang, D., Otero, D.A., Xia, Y and Saitoh, T. (1995) NACP, the precursor protein of the non-amyloid beta/A4

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

ACKNOWLEDGMENTS protein (Abeta) component of Alzheimer disease amyloid, binds Abeta and stimulates Abeta aggregation. Proc. Natl. Acad. Sci. USA, 92: 9141-9145. Zhang, J.Y., Liu, S.J., Li, H.L. and Wang, J.Z. (2004) Microtubuleassociated protein tau is a substrate of ATP/Mg(2+)-dependent proteasome protease system. J. Neural Transm., 112: 547-555. Zhang, X., Smith, D.L., Merlin, A.B., Engemann, S., Russel, D.E., Roark, M., Washington, S.L., Maxwell, M.M., Marsh, J.L., Thompson, L.M., Wanker, E.E., Young, A.B., Housman, D.E., Bates, G.P., Sherman, M.Y. and Kazantsev, A.G. (2005) A potent small molecule inhibits polyglutamine aggregation in Huntington's disease neurons and suppresses neurodegeneration in vivo. Proc. Natl. Acad. Sci. USA, 102: 892-897.

149

Zhou, C , Emadi, S., Sierks, M.R. and Messer, A. (2004a) A human single-chain Fv intrabody blocks aberrant cellular effects of overexpressed a-synuclein. Mol. Ther., 10:1023-1031. Zhou, Y, Shie, F.S., Piccardo, R, Montine, T.J. and Zhang, J. (2004b) Proteasomal inhibition induced by manganese ethylene-bisdithiocarbamate: relevance to Parkinson's disease. Neuroscience, 128: 281-291. Zhu, Q., Zeng, C , Huhalov, A., Yao, J., Turi, T.G., Danley, D., Hynes, T, Cong, Y, DiMattia, D., Kennedy S., Daumy, G., Schaeffer, E., Marasco, W.A. and Huston, J.S. (1999) Extended half-life and elevated steady-state level of a single-chain Fv intrabody are critical for specific intracellular retargeting of its antigen, caspase-7. J. Immunol. Methods, 231: 207-222.

11. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

CHAPTER

12 Gene Therapy for Epilepsy Francesco Noe, Matthew J. During, Annamaria Vezzctni

Abstract: Gene therapy techniques provide a realistic therapeutic approach for intractable focal epilepsies not responding to conventional antiepileptic drugs. These techniques involve the transfer and expression of a "therapeutic" gene into the ictogenic brain area(s), thus permitting long-term central nervous system expression of neuromodulatory molecules with potential anticonvulsive and antiepileptogenic properties. This chapter will review the selection of the "therapeutic" genes delivered into the rodent brain for studying their ability to inhibit seizures and delay epileptogenesis in vivo experimental models of epilepsy. Important aspects that contribute to determine the success or failure of a gene therapy approach are also described, such as the methods of gene delivery, strategies for improving cell transfection and neuronal expression, regulation of gene expression, and possible host tissue reactions to the transgene. Preclinical studies focused on the antiepileptic efficacy of gene therapy in pathological brain tissue, and on its possible side-effects are instrumental for establishing a proofof-principle of the applicability of gene transfer technologies in epilepsy. Keywords: antiepileptic; cell transplantation; viral vectors; neuroprotection; neuropeptides; glutamate receptors; seizures; non viral delivery

I.

properties. In this context, gene therapy techniques, involving the transfer and expression of a therapeutic gene, offer the possibility of delivering specific genes directly into the brain area where seizures originate, thus permitting long-term CNS expression of specific proteins with neuromodulatory actions. Other forms of generalized epilepsies, for example, those of genetic origin, such as the channelopathies, are less suitable for this therapeutic approach because of the need of transfecting cells all over the brain. Ultimately, once global gene transfer technology becomes further developed, such genetic epilepsies will be a primary target for gene therapy, but ironically, today focal epilepsies are a better indication by nature of the localized pathology. Clinical applications for gene transfer to the CNS have been developed so far for some CNS neurodegenerative disorders like Parkinson's disease and

INTRODUCTION

Epilepsy affects about 1% of the population and the current medical therapy is largely symptomatic, thus it is aimed at controlling seizures in affected individuals. However, about 30% of patients are not responsive to available antiepileptic drugs in spite of appropriate treatment (Perucca, 1998). If antiepileptic drugs fail, surgical resection of epileptogenic tissue provides an alternative treatment; however, this choice is amenable only for patients where the epileptogenic area can be adequately defined and removed without major functional impairment (Foldvary et al., 2001). In principle, patients with intractable seizures of focal onset may benefit from therapeutic strategies that are an alternative to resective surgery, such as the delivery of molecules into the seizure focus with potential anticonvulsive and antiepileptogenic

Gene Therapy of the Central Nervous System: From Bench to Bedside

151

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

152

12. GENE THERAPY FOR EPILEPSY

Alzheimer's disease, in cancer, inherited monogenic disorders (cystic fibrosis) and genetic diseases such as Canvan and Batten's disease or infectious diseases (HIV). The selection of the "therapeutic" gene and the method of its delivery are two crucial aspects that determine the success or failure of a gene therapy approach, and their choice depends on the specific pathology under investigation.

11.

THERAPEUTIC TARGETS I N EPILEPSY

At least two nonmutually exclusive endpoints can be considered when devising a gene therapy approach to epilepsy, namely to suppress seizures and to spare or rescue neurons from otherwise irreversible damage (Table 1). These two therapeutic outcomes should lead to sparing of function at the level of synaptic physiology and plasticity, as well as at the level of behavior and cognition.

As discussed in more detail in Section IV, gene therapy has been successful in suppressing seizures both in rat models of acutely induced focal seizures with or without secondary generalization, and in rats with congenital audiogenic seizures or showing primarily generalized absence-like seizures and tonic convulsions. It is important to note that seizure control in models of focal vs. generalized seizures was achieved using different delivery methods (viral vs. nonviral delivery) and route of administration (intraparenchymal vs. intraventricular). Limited information is available on the efficacy of gene therapy approaches in delaying epileptogenesis, thus preventing the occurrence of spontaneous seizures by early intervention after an epileptogenic insult. Two pieces of evidence so far have been provided on the effect of transgenes on the recurrence of spontaneous seizures in rats in which epilepsy has been already established (Thompson and Suchomelova, 2004; Noe' et al., 2005). B.

A.

Anticonvulsant Activity

Therapeutic strategies using conventional antiepileptic drugs have shown that a reduction of excitatory glutamatergic neurotransmission, enhancement of gamma-aminobutyric acid (GABA)-mediated inhibitory effects and blockade of Na^ and Ca^^ channels can provide effective seizure control. The mechanims of action of antiepileptic drugs and their molecular targets are compatible with the widely supported hypothesis that neuronal hyperexcitability underlying the epileptic state depends on an imbalance in the excitatory and inhibitory transmission in CNS (Rogawski and Loscher, 2004). In addition, the fortuitous occurrence of seizures in both mutant and transgenic mice has enabled the discovery of a large array of genes that can directly or indirectly affect neuronal excitability (Noebels, 1996). Experimental approaches to gene therapy in rodent models of seizures have characterized several possible targets to suppress seizures in epilepsy, namely some neuropeptides (galanin, cholecystokinin (CCK) and neuropeptide Y (NPY)), GABA and adenosine (see Section IV of this review). Furthermore, the delivery of an antisense sequence against the N-methyl-Daspartate (NMDA) receptor has been proven successful in suppressing seizures of focal onset (Haberman et al., 2002). A proof-of-principle for the possibility of sparing cognitive function and rescue neurons post seizure has also been provided using the glucose transporter or antiapoptotic genes (McLaughlin et al., 2000).

Neuroprotection

Neuronal damage is at least in part due to ongoing and uncontrolled seizure activity, therefore, suppression of seizures may in turn stop progression of cell death. However, cell damage may precede in some instances the onset of seizures and contribute to epileptogenesis, as it has been proposed for those cases of symptomatic epilepsies where brain damage can evolve to active epilepsy. Rescue of neurons at the morphological and functional level is more likely at the early stages of damage and the latency between the onset of damage and the gene delivery probably impacts the likelihood of ultimate therapeutic success. One possible solution highlighted in the current literature is to develop vectors that can be introduced into the CNS where they remain transcriptionally quiescent until a proper injurydependent stimulus is provided. In this respect, neuroprotection has been achieved in vitro against an acute necrotic insult using vectors containing a synthetic glucocorticoid-responsive promoter, which would exploit the high levels of adrenal stress hormones secreted in response to this insult (Ozawa et al., 2000). Overexpression of antinecrotic and antiapoptotic genes, neurotrophins, Glut-1 glucose transporter and neuropeptides are among the candidate targets studied in gene therapy strategies using in vitro and in vivo models of cell death (Table 1). Transgene expres-sion has been shown to reduce neurodegeneration and, in some instances, also provides recovery from physiological and behavioral dysfunction (Dumas and Sapolsky, 2001).

11. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

153

DELIVERY METHODS FOR GENE THERAPY

TABLE 1 Target

Route

Molecular Targets for Gene Therapy in Epilepsy

Method of delivery

Eperimental model

Functional effect

Jlllerences

Anticonvulsive Adenosine GAD-65

Anti-NMDA

ICV

Grafting of fibroblasts

Hippocampal kindling

Suppression of generalized seizures

Huber et al. (2001)

Anterior SN

Immortalized mouse cortical neurons and glia

Entorhinal kindling

Delayed rate,of kindling

Thompson et al. (2000)

Posterior SN

Entorhinal kindling

Faster rate of kindling

Thompson et al. (2000)

Anterior SN

Status epilepticus

Reduced spontaneous seizures

Thompson and Suchomelova (2004)

Piriform cortex

Amygdala kindling

Increased threshold seizures

Gemert et al. (2002)

Focal electrical stimulation

Increased seizure threshold

Haberman et al. (2002)

Decreased seizure threshold

Haberman et al. (2002)

Increased seizure threshold

Haberman et al. (2003)

Increased seizure threshold

Haberman et al. (2003)

Collicular cortex

AAV-CMV vector AAV-TET-off vector

Galanin

Collicular cortex

AAV-CMV vector

Focal electrical stimulation

AAV-TET-off vector Hippocampus

Neuropeptide Y

Hippocampus

AAV-TET-off vector

Status epilepticus

Neuroprotection

Haberman et al. (2003)

AAV-NSE vector

Ictal activity

Seizure inhibition

Lin et al. (2003)

AAV-NSE vector

Ictal activity. Status epilecticus Hippocampal kindling

Seizure inhibition

Richichi et al. (2004)

Increased threshold and delayed rate of kindling

Richichi et al. (2004)

ASPA

ICV

AV-CAG vector

Spontaneous seizure in SER

Reduced incidence of tonic seizure

Seki et al. (2004)

Cholecystokinin

ICV

Lipofectin

Audiogenic seizures

Seizure inhibition

Zhang et al. (1997)

Hippocampus Cell cultures

Viral vectors

Necrotic or apoptotic injury

Neuronal survival

Dumas and Sapolsky (2001), Haberman et al. (2003), McLaughlin et al. (2000), Ozawa et al. (2000)

Neuroprotective Glut-1 Anti-apoptotic Neuropeptides Neurotrophins Calbindin

AAV, adeno-associated viral; AV, adenoviral; GAG; cytomegalovirus enhancer, chicken j&-actin promoter; GMV, cytomegalovirus; GAD-65, glutamic acid decarboxylase; Glut-1, glucose transporter; IGV, intracerebroventricular; NMDA, N-methyl-D-aspartate; NSE, neuron-specific enolase ; SER, spontaneously epileptic rats; TET off, tetracycline-off regulatable promoter.

III.

DELIVERY METHODS FOR GENE THERAPY

There are two main types of gene delivery methods, which consist of in vivo gene transfer using viral vectors, naked DNA or cation-lipid DNA complexes and

ex vivo gene transfer using cells previously transfected in vitro with the transgene of interest (Costantini et al., 2000; Mountain, 2000). Some of these strategies have been used to study the therapeutic effect of gene transfer in experimental models of seizures. At the present stage, viral vectors

11. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

154

12. GENE THERAPY FOR EPILEPSY

generally give the most efficient transfection in vivo, although their main disadvantage concerns size-limitation of the transgene and the potential immunogenicity. Among the viral vector-delivery systems, adeno-associated virus (AAV) vectors, lentivirus and herpes simplex virus (HSV) vectors have specific tropism for post-mitotic neurons of the CNS, with AAV having the best safety profile. Viral vectors derived from retrovirus are not adequate for gene transduction in neurons since they transfect only proliferating cells (Janson et al., 2001). Physical methods of gene delivery, such as lysosomes carrying plasmids have the potential advantage over viral vectors to allow systemic, thus noninvasive, delivery of genes, which then successfully penetrate into the brain parenchyma, providing widespread gene expression in the brain. This method of delivery is impaired when using vector-mediated gene transfer since the blood-brain barrier prevents the vectors to get into the brain parenchyma. However, noninvasive gene targeting to the brain is extremely inefficient and allows a very short expression of the transgene limited to a few days; in addition, gene transfer will occur also in peripheral tissues, thus requiring that the specific brain expression is regulated at the promoter level. In the context of ex vivo gene transfer, the production and release of a protein from a transduced gene using cell transplantation and grafting of in vitro engineered cells, will strongly depend on the survival of these cells in the transplanted tissue. A. Vector^Mediated Gene Delivery: The Focus on AAV Vectors Neurotropic AAV represent the most often used tool for gene delivery in experimental models of epilepsy. They present many advantages over the other available delivery methods both for brain functional studies and the brain expression of transgenes for therapeutic purposes. Thus, they can efficiently express single or multiple transgenes together with a wide range of regulatory elements; they permit long-term gene expression and can be engineered at the capsid and promoter level to preferentially target specific populations of neurons in a controllable manner, and very importantly they are nonpathogenic and appear to be innocuous on normal brain physiology (Monahan and Samulski, 2000). Figure 1 depicts the expression of green fluorescent protein (GFP) used as a reported gene in AAV vectors engineered with the neuron-specific enolase (NSE)

promoter. The type of transduced neurons and the spread of the transgene expression around the injection site is determined by the serotype of the viral capsid (Davidson et al., 2000). After intrahippocampal delivery, the AAV-2-mediated GFP expression is mainly observed in intemeurons for about 1.5 mm around the site of injection, while AAV-1/2 (a chimeric with 50/50 capsid proteins from both serotypes 1 and 2) mediates the transgene expression also in granule cells and in pyramidal neurons for about 2.5 mm around the injection site (Xu et al., 2001). The mechanisms determining these differences are still unresolved, but it is clear that the choice of the serotype affects the type of cell populations, which will express the transgene (Davidson et al., 2000). This aspect has functional relevance since the neuronal population preferentially transfected will determine the subsequent changes in synaptic transmission and, at a more general level, the physio-logy of the brain area targeted by the vector. Table 2 reports the relative transduction efficiency and the duration of transgene expression of AAV vectors with different promoters and post-transcriptional regulatory elements, as exemplified using AAV2-mediated gene transfer into the hippocampus. Transgene expression driven by the hybrid cytomegalovirus (CMV)-<:hicken j5-actin (CBA) promoter, appears to be the most efficient and it mediates gene expression for at least 25 months after its delivery (Tenenbaum et al., 2004). Regardless of the promoter used, the transduction efficiency can be enhanced using post-transcriptional regulatory elements such as the woodchuck hepatitis virus post-transcriptional regulatory element (WPRE) which appears to increase the steady-state levels of mRNA, thus resulting in higher gene expression. An effective way for regulating exogenous gene expression in mammalian cells is offered by the tetracycline-responsive system. The small size of the tetracycline-responsive cis element and repressor is particularly suitable for the limit of 4.5 kb of the AAV cassette. Importantly, gene expression from the tetracycline-responsive promoter is dose-dependent and tetracycline derivatives can cross the bloodbrain barrier, which is a prerequisite for regulating transgene expression in the brain (Haberman et al., 1998). This system was tested for regulating long-term gene expression into the rodent brain using the presence or absence of tetracycline or its derivatives in the drinking water (see Fig. 4). Long-term regulated gene expression in the brain may have a significant application in humans when the control of the therapeutic gene expression is required.

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

STUDIES IN EXPERIMENTAL MODELS OF SEIZURES

155

FIGURE 1 Schematic representation of the transduction sequelae of a viral vector-mediated gene transfer into a target cell, and transduction efficiency of a trangene by AAV with different serotypes. Panel (A) depicts the vector particle containing the trangene, which generally binds to a cell receptor to permit the entry of the genetic material into the nucleus of the target cell. The double-stranded DNA deriving from the vector genome can persist in an episomal form or become integrated into the host genome. Panels (C) shows the expression of green fluorescent protein (GFP) in interneurons of the injected rat hippocampus (schematic representation of the injected area is reported in panel (B) and injection site marked by an asterisk), 8 weeks after serotype 2 AAV-NSE-GFP infusion. Chimeric serotype 1/2 AAV-NSE-GFP provides an additional expression of GFP in granule cells and their mossy fibers (panel D). This picture has been adapted from Kay et al. (2001) and from Richichi et al. (2004).

TABLE 2 Serotype 2 AAV^Mediated Gene Transfer in the Rat Hippocampus: Effect of Promoter and Regulatory Elements Days of expression

IV.

CMV

NSE

NSE+WPRE

CBA+WPRE

3

-

+

+

++

21

+++

++++

90

++++

++++++ ++++++ ++++++

+++++++

++ -

270

++++

+++++++ +++++++

CBA, cytomegalovirus promoter-chicken j5-actin promoter; CMV, cytomegalovirus promoter; NSE, neuron-specific enolase promoter; WPRE, woodchuck hepatitis virus post-transcriptional regulatory element. — no expression; + moderate expression; + + + + + + maximum expression.

A.

STUDIES IN EXPERIMENTAL MODELS OF SEIZURES Transplantation of Genetically Modified Cells

Cells engineered in vitro to produce and release a therapeutic molecule have been transplanted in specific brain areas involved in the onset or propagation of seizures using models of focally induced epileptic activity in rats. Two molecules with neurotransmitter or neuromodulator activity, namely GABA and adenosine, have been used as therapeutic targets in cell transplantation approaches because of their well-established inhibitory actions on seizures in various models of epilepsy.

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

156

12. GENE THERAPY FOR EPILEPSY

FIGURE 2 Grafting of cells engineered in vitro to produce adenosine (A,B) or GABA (C,D). Panel (A) depicts viable adenosine-releasing fibroblasts encapsulated into semipermeable hollow fibers. Collagen matrix within the capsule is highlighted within the boxed area. Panel (B) shows the encapsulated adenosine-releasing cells stained by toluidine blue, which survived for 24 days after their intraventricular implantation. Note that only scattered viable cells can be found, which are surrounded by cellular debris. Panels (C, D) show immortalized cortical neurons producing GABA grafted into the rat piriform cortex. In panel (C), nissl staining shows densely packed cells in the brain parenchyma surrounding the injection track, which appear healthy. In panel (D), cells engineered in vitro to express GAD65 showed enzyme immunoreactivity after their in vivo grafting into the piriform cortex. Pictures were adapted from Huber et al. (2001) and Gemert et al. (2002).

1.

Adenosine

Huber et al. (2001) engineered fibroblasts to secrete adenosine by inactivating the adenosine-metabolizing enzymes, adenosine kinase and adenosine deaminase. These conditionally immortalized fibroblasts were encapsulated into semipermeable polymers and subsequently grafted into the brain ventricle of electrically kindled rats (Figs. 2A and B). Four to five days after grafting, kindled rats containing grafts of control cells or matrix material only, displayed generalized stage 5 seizures consistently after the application of a test stimulus. During the same period, generalized seizures were completely suppressed in rats receiving the adenosine-releasing grafts. An important aspect of this approach was the lack of peripheral side effects (sedation, ataxia) for 24 days after grafting which are otherwise observed after systemic delivery of adenosine or its analogs. However, protection from seizures

was diminished between 12 and 24 days from grafting which was apparently due to a decrease in cell viability within the graft. Thus, the future challenge will be that of devising alternative strategies for the generation of cells with enhanced viability. In this regard, the authors suggested the possibility of introducing into the cells survival-enhancing genes, such as constitutive active immortalizing oncogenes or apoptosis inhibitors. 2.

QABA

Genetically engineered GABA-producing cells have been obtained from conditionally immortalized mouse cortical neurons and glia (Thompson et al., 2000). These cells were generated by using a plasmid carrying the cDNA for the GABA-synthesizing enzyme glutamate decarboxylase (GAD65) under the control of doxycycline-sensitive minimal human CMV

. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

STUDIES IN EXPERIMENTAL MODELS OF SEIZURES

promoter. In vitro assays showed that GAD enzymatic activity increased two to three fold in these cells upon addition of cofactors, and that GABA was produced and released. a. Kindling These cells were initially implanted in the anterior or posterior substantia nigra of rats and 10 days later, kindling was induced by electrical stimulation of the entorhinal cortex. Rats receiving jS-galactosidase-producing cells were used as controls. The rats receiving neuron or glia-derived cell transplants producing GABA in the anterior substantia nigra, showed a delay in the progression of kindling while acceleration of kindling was observed if these cells were transplanted in the posterior substantia nigra. These effects were consistent with those elicited by exogenous application of GABA agonists to the same brain areas. In a subsequent work, the same cells from immortalized cortical neurons were implanted in the piriform cortex and their effects of kindling were evaluated. Although these rats displayed a higher threshold for the induction of behavioral seizures after the electrical stimulation of the amygdala, they did not show any alteration in the development of kindling (Gernert et al., 2002). In this study, the grafted cells survived for at least 5 weeks after grafting (Figs. 2C and D). These findings clearly indicate that genetically engineered GABA-producing cells, when transplanted into discrete brain regions, can significantly affect seizure threshold or the propagation pathways of seizures, and this action depends on the specific brain area targeted by the cell graft. b. Status Epilepticus In a separate study, the effect of GABA-producing cells was assessed in a model of status epilepticus evolving to spontaneous seizures. Cells were transplanted into the substantia nigra of rats, 2 weeks after the induction of pilocarpineinduced status epilepticus (Thompson and Suchomelova, 2004). The rats that received GABA-producing cells have significantly fewer spontaneous seizures and suppression of cortical epileptiform spikes than rats receiving control cells, or rats receiving GABAproducing cells but administered with doxocycline in the drinking water. The engineered cells showed evidence of integration with the host but ongoing cell degeneration was observed within 10 days. In summary, although substantial seizure suppression can be obtained with engineered cell grafting, a caveat of this approach is that the therapeutic effects of the implanted cells are highly dependent on their

157

survival. So far, long-term viability of transplanted cells in vivo has not been achieved, and this aspect needs to be improved since the clinical applicability in epilepsy requires that the therapeutic gene product is expressed over a long period of time. B.

Viral Vectors for Gene Transfer

Therapeutic strategies using viral vector delivery of therapeutic genes have been focused mainly on modulating signaling mediated by excitatory neurotransmitters, or enhancing the release of neuroactive peptides. These studies have largely been carried out using experimental models of focally induced seizures, with or without secondary generalization, and in some instances models of primary generalized, genetically determined seizures as well. 1.

NMD A Receptor

This receptor represents an obvious target for inhibition of seizures since a reduction in the number of functional NMDA receptors should impair glutamatemediated excitation in CNS. Focusing on this target, Haberman et al. (2002) cloned an antisense cDNA fragment of the constitutive NMDA receptor subunit (NMDARl) into an AAV vector, under the control of a CMV promoter with or without a tetracycline (TET)off cassette. These two vectors were separately injected into the inferior coUiculus, a well-characterized site of focal seizure genesis which is sensitive to local manipulation of NMDA receptor function (McCown et al., 1987). Three weeks after vector delivery, a significant reduction was found in the NMDA-like immunoreactivity as compared to the contralateral uninjected side. The threshold current for inducing wild-running seizures by electrical stimulation of the coUicular cortex, decreased significantly over a 4-week post-treatment period but only in rats receiving the AAV vectors containing the TET-off cassette. Conversely, the rats that received the AAV vector with the CMV-promoter only, showed a significant increase in seizure threshold (Fig. 3). Thus, a modest change in the promoter design resulted in opposite effects on focal seizure sensitivity. More in-depth investigations of the effects of the two vectors showed that the AAV vector carrying the TET-off controllable promoter, primarily transduced inhibitory GABA interneurons; instead, when using the CMV promoter only, the vector transduced the primary output neurons. Since NMDA receptors located on interneurons drive the GABA-mediated inhibition on excitatory output neurons, the removal of these receptors by the vector-carried antisense sequence, led

11. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

158

12. GENE THERAPY FOR EPILEPSY

m

• AAV-CMV-NMDAantisense • AAV-tTAK-NMDAantisense • AAV-tTAK-rGFP

100

o (A

50 H

N


11 c CD JC

O

oJ -50 H -100

0 1 2 3 4 5 Time after AAV infusion (weeks)

FIGURE 3 Transduction pattern of AAV-GFP or AAV-LacZ in neurons of the rat inferior colliculus as induced by vectors engineered with different promoters (A-E), and functional effects on seizure threshold of vectors carrying the NMDARl antisense sequence (F). Panels (A, B) depict GFP (green, panel A) or LacZ (red, panel B)-positive cells transduced in the rat collicular cortex with AAV-TET-off controllable promoter, or AAV-CMV promoter, respectively. The two images are merged in panel (C) to show the different patterns of transgene expression as shown using a dual pass filter. Several neurons express only one of the two gene products: in particular, inhibitory interneurons (panels A, E) were preferentially transduced by AAV-TET-off controllable promoter while output excitatory primary neurons (panels B, F) were preferentially transduced by AAV-CMV promoter. Panel (D) shows seizure threshold sensitivity to electrical stimulation of the inferior colliculus in rats previously injected locally with AAV-TET-off controllable promoter or AAV-CMV promoter carrying the NMDARl antisense sequence. An increase in seizure threshold was achieved using the AAV-CMV promoter while proconvulsant effects were obtained with the AAV-TET-off controllable promoter. Rats injected with AAV-TET-off-GFP were used as controls. This picture was adapted from Haberman et al. (2002) and McCown (2004).

to hyperexcitability, thus increasing seizure susceptibility in the rats. These findings raise concerns about the identification of neurotransmitter receptors as good candidates for gene therapy. Thus, without a detailed knowledge of the transduction pattern in humans, manipulation of neurotransmitter receptors, or by inference, ion channels, may lead to a paradoxical increase in seizure sensitivity. 2.

Neuropeptides

The preferential release of neuropeptides under conditions of increased neuronal activity, i.e. during seizures, has encouraged investigation of their potential role in seizure modulation. Both galanin, a 29/30-amino acid peptide and NPY, a 36-amino acid polypeptide, have been shown to antagonize excitatory glutamatergic neurotransmission in the hippocampus (Mazarati et al., 2001). Compelling evidence supports an anticonvulsant role of these peptides in various experimental models of seizures either when exogenously applied or when endogenously released. Neuroprotection against excitotoxic cell death (Silva et al., 2003) and involvement in seizure-induced neurogenesis (Howell et al., 2003,2005) are two additional aspects of peptide actions in the CNS, which are relevant for epilepsy.

These findings led to the hypothesis that augmentation of local inhibitory tone, resulting from overexpression of these neuroactive peptides in specific brain areas, may be an effective strategy for inhibition of seizures and epileptogenesis. a» Qalanin Haberman et al. (2003) engineered an AAV vector carrying the galanin coding sequence preceded by the fibronectin secretory signal sequence. Since fibronectin is a constitutively released laminar protein, the presence of its secretory sequence results in facilitation of galanin secretion from in vitro transduced cells. To modulate the secretion of galanin, the gene was under the control of a TET-off CMV promoter. Four days after galanin gene transduction in vivo, the constitutively released peptide was able to augment the threshold to seizure induced by stimulation of the inferior colliculus over a 4-week-testing period. When doxycycline was given to rats into the drinking water, then the threshold for seizures genesis decreased and reached baseline level within 1 week (Fig. 4). This effect was reversible upon doxycycline removal. Different from the experiments aimed at decreasing the NMDA receptors, the type of promoter

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

STUDIES IN EXPERIMENTAL MODELS OF SEIZURES

4 7

14

21

28

35

42

49

56

63

159

70

Days after AAV vector infusion

FIGURE 4 Attenuation of neuronal death and seizures by AAV-FIB vector overexpressing galanin under the control of TET-off regulatable CMV promoter. Panels (A, B) depict sparing of hilar interneurons and CA3-4 pyramidal cells (arrowheads) in the AAV-FIB-galanin (GAL) infused hippocampus of rats injected systemically with kainic acid (panel B) as compared to the contralateral uninjected hippocampus (panel A). Arrow in panel (A) depict the injection site. Panel (C) shows the protective effect of locally applied GAL on focal seizure sensitivity to electrical stimulation of the inferior colliculus. Note that protection from seizures was lost when adding doxycycline to the drinking water and this effect was rescued after withdrawal of doxycycline. Control rats were injected with AAV-FIB-GFP. FIB represents the secretory sequence signal of fibronectin, and it was added to induce a constitutive release of galanin. This picture was adapted from Haberman et al. (2003).

used (CMV with or without TET-off cassette) did not affect the final outcome of galanin overexpression on seizures since inhibition was found in both circumstances. Thus, choosing neuropeptides as gene targets may overcome the problems raised by the manipulation of neurotransmitters receptors. The AAV-galanin vector also led to neuroprotection in the injected hippocampus of rats following status epilepticus induced by kainic acid (Haberman et al., 2003) (Fig. 4). A similar approach was taken by Lin et al. (2003) using an AAV vector carrying the galanin gene under the control of NSE promoter together with WPRE. This AAV-NSE-galanin serotype 2 vector was bilaterally injected into the rat dorsal hippocampus and its effect was assessed after 2.5 months on hippocampal seizures induced by local application of kainic acid. After vector delivery, galanin immunoreactivity

was strongly enhanced in fibers in the dentate hilus and the molecular layer of the hippocampus in rats, which experienced a lower number of seizures and a decreased time spent in seizures, as compared to control rats injected with AAV vectors lacking the transgene (Figs. 5A and C). An important finding of this study is the demonstration that the peptide can be efficiently translated in vivo in interneurons and granule cells and transported along their axons to their terminal projection fields. h* ISJeuropeptide Y (NPY) This peptide also exerts antiictal activity in vivo. NPY is expressed in GABA and somatostatin-containing interneurons in the normal rat hippocampus; after seizures, this peptide is greatly increased in these inhibitory interneurons as well as ectopically in granule cells and their mossy fibers.

11. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

160

12. GENE THERAPY FOR EPILEPSY

fi"

30

If" 0 1

* AAV-GAL

AAV-Emotv

g Onset

3001

J

:i::fc

1 200J

Q

No Of SeiZUfeS

IT

1=4

1 ioo|

QA, serotype 2

serotype 1/2 , rAAV-NSE-NPY

FIGURE 5 AAV-mediated overexpression of galanin (panels A, B) and neuropeptide Y (panels D, E) 2.5 months after intrahippocampal injection in rats, and their functional effects on seizures (panels C, F). Panels (A, B) depict serotype 2 AAV-NSE vector-mediated galanin overexpression in the hippocampus proper and the dentate hilus, respectively. Galanin immunoreactivity was observed predominantly in fibers, likely representing the peptide transported by the transduced hilar and granule neurons into the synaptic terminals. Panels (D, E) depict the overexpression of NPY induced by AAV-NSE vectors with serotype 2 or chimeric serotype 1/2, respectively. In panel (D) fiber staining was mainly observed in hilus and in the outer molecular layer (arrows); in panel (E), projection fiber tracts were intensively immunostained throughout the injected hippocampus, including the mossy fibers, and pyramidal cells were strongly immunoreactive. Panels (C, F) show that the hippocampal overexpression of these two transgenes was effective in reducing seizures caused by intrahippocampal injection of kainic acid. Note that AAV-vector with chimeric serotype 1/2 carrying the NPY gene (panel F) provided the best protection from seizures (*;7<0.05; **p<0.01 vs. respective rats injected with AAV-empty (without the transgene)). Mol, molecular layer; CAS, pyramidal cells; h, hilus. Pictures were adapted from Lin et al. (2003) and Richichi et al. (2004).

NPY has a predominant inhibitory action on excitatory glutamate-mediated neurotransmission and displays anticonvulsant properties in a large variety of acute models of seizures in vitro and in vivo (Vezzani et al., 1999; Vezzani and Sperk, 2004). Thus, its endogenous overexpression in epileptic conditions may represent a defensive attempt of the brain to counteract the noxious effects of seizures. This consideration prompted Richichi et al. (2004) to study the antiictal effects of long-lasting NPY overexpression in the rat hippocampus. AAV-mediated expression of human prepro-NPY under the control of the NSE promoter reduces by 50-75% (depending on whether serotype 2 or chimeric serotype 1/2 were used, respectively), the EEG seizures induced by intrahippocampal administration of kainic acid and the onset of ictal activity was also markedly delayed (Figs. 5D-F). Differences in the viral serotypes could therefore influence significantly the functional effect of the transgene. Immunohistochemical analysis of the transgene expression revealed that serotype 2 vector increased NPY expression mostly in hilar inhibitory interneurons, while the chimeric

serotype 1/2 vector induced an additional expression of this neuropeptide in the mossy fibers, pyramidal cells and in the subiculum (Figs. 5D-F). The higher degree of seizure protection achieved with the chimeric serotype vector was likely due to higher efficiency, and larger spread, of gene transduction in neuronal populations, which are relevant for seizure control. Additional experiments in rats injected with the chimeric serotype 1/2 vector showed that AAV-NSE-NPY abolished the occurrence of status epilepticus, significantly retarded the acquisition of kindling and enhanced the threshold current for inducing an afterdischarge in the hippocampus (Richichi et al., 2004). The inhibitory effects on status epilepticus and the associated generalized behavioral seizures, and on kindling progression indicate that the predominant result of the NPY overexpression was a selective inhibition of the generalization of seizures from their site of onset. c. Enzyme Replacement The aspartoacylase (ASPA) gene is involved in the hydrolyzation of N-acetyl-Laspartate (NAA) that has neuroexcitatory properties.

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

THERAPEUTIC IMPLICATIONS AND CONCLUSIONS

In humans, a deficiency of ASPA due to a mutation in the gene leads to Canavan Disease. It has been reported that 63% of Canavan Disease patients exhibit epileptic seizures (Traeger and Rapin, 1998), suggesting a role for ASPA (and/or NAA) in the etiology of human epilepsy. This gene, which is deleted in tremor rats, a strain developing spontaneous absence-like seizures, has been inserted into a defective recombinant adenoviral vector under the control of a chimeric promoter (CAG promoter) (Seki et a l , 2004). Seven days after the intracerebroventricular injection of this vector in spontaneously epileptic rats (SER), which were obtained from crossbreeding of the tremor rats, the occurrence of tonic seizures was significantly reduced without affecting the duration of each single convulsive event. Control rats were injected with a vector carrying the LacZ gene. This protective effect on genetically determined spontaneous seizures was transient since it was lost within 2 weeks after the vector injection. Prevention of the long-term expression of the transgene in vivo may be due to the induction of an immune response to this first-generation adenoviral vectors. A very recent study conducted in the laboratory of one of us (M.J.D.) has shown that delivery of the ASPA gene using AAV in tremor rats significantly attenuated the seizure (but not the myelin) phenotype and this effect persisted throughout the duration of the study (Klugmarm et al., 2005). C,

Nonviral Delivery Systems

Although extensive progress has been made in viral vector-mediated gene delivery systems, these techniques still raise concerns about the possibility of inducing immunological host responses, cellular toxicity or because of variable transfection efficiencies. These concerns fostered the development of alternative strategies focused on nonviral DNA delivery systems. 1 • Lipofectin Facilitated-Qene Transfer Zhang et al. (1997) used lipofectin to transfer the CCK gene into the ventricles of rats with congenital audiogenic seizures. The choice of this peptide was suggested by the evidence that low-brain levels of CCK were measured in this strain of rats and this defect has been supposed to be the neurochemical basis of their congenital audiogenic seizures (Zhang et al., 1997). The CCK gene was inserted into a plasmid under the control of SV40 promoter and mixed with a lipofectin reagent solution. The second day after the intracerebroventricular injection of this mixture.

161

audiogenic seizures were markedly decreased with a dramatic suppression observed at day 4. After 1 week, however, the vulnerability to audiogenic seizures returned to basal level likely because the plasmid was degraded by endogenous cellular nucleases. A promising alternative to lipofectamin is represented by pegylated immunoliposomes (Shi et al., 2001). This technique consists of a double-stranded supercoiled plasmid DNA encapsulated into pegylated liposomes. In addition, polyethylene glycol (PEG) strands are conjugated with a peptidomimetic monoclonal antibody, which binds to a transporting receptor on the blood-brain barrier. The great advantage of pegylated immunoliposomes is that their net negative charge and the internalized DNA into the liposome protect the plasmid DNA from endogenous nucleases. In their elegant work, Shi et al. (2001) used a monoclonal antibody (mAb) to the rat transferrin receptor, which is abundant on blood-brain barrier. After intravenous administration in rats, the mAb targeted the exogenous gene through the blood-brain barrier into the brain parenchyma. Although this approach provides gene targeting to the brain using noninvasive procedures, gene expression is generalized to the whole brain and it is also achieved in various peripheral tissue. The use of a brain-specific promoter is needed to drive gene expression exclusively in CNS. Once again, one major limitation of this approach concerns the very restricted time of expression of the transgene, and limited transduction efficiency, as compared to viral vector-mediated gene delivery. V-

THERAPEUTIC IMPLICATIONS A N D CONCLUSIONS

Pharmacologically intractable seizures of focal onset likely represent the best clinical target for gene therapy among the various types of human epilepsies, although a few experimental studies in models of genetically determined seizures open the perspective of using gene therapy approach also for primarily generalized seizures. The choice of therapeutic genes is a crucial aspect, which should take into account the molecular basis of the CNS disorder. Several potentially suitable targets for epilepsy can be envisaged based on the available knowledge of the genes and related proteins involved in the susceptibility to seizure, their onset and generalization. Studies in experimental models have highlighted the therapeutic liability of gene therapy targeting ion channels or neurotransmitter

II. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

162

12. GENE THERAPY FOR EPILEPSY

receptors while supporting the anticonvulsant efficacy using GAB A or neuroactive peptides. Neuroprotection from the damaging effects of seizures or prevention of progressive cell loss that may predispose to epileptogenesis are also important aspects that may be taken into consideration in the context of a gene therapy strategy in epilepsy Although the studies in experimental models of seizures have established a proof-of-principle for the applicability of gene therapy to epilepsy these studies have focused so far on the prevention of seizures. An important issue of clinical significance is the requirement of studies characterizing the efficacy of gene therapy in epileptic tissue using models of spontaneous and recurrent seizures. Chronic epileptic tissue is in fact often characterized by the loss of neurons, synaptic and molecular rearrangements and phenotypic changes in parenchymal cells, which may affect the targets of the "therapeutic'' genes and impair their efficacy One recent report showed that the incidence and duration of spontaneous seizures in chronically epileptic rats is significantly reduced by hippocampal application of an AAV carrying the human NPY gene (Noe' et al., 2005). Other important considerations include the immune reactions that may be triggered by the gene transfer modalities, such as production of high levels of circulating antibodies against the vector serotypes, particularly if more than one injection may be required, promoter silencing, instability of the vector genome in the absence of integration and loss of transduced cells. All these aspects are under investigation in experimental models of seizures and some of them can apparently be circumvented. Regulation of gene expression is probably one of the more intense areas of research in gene therapy, since it will provide a tool to both dial up or down gene expression as well as silence the therapeutic gene when, and if ever, needed. To move preclinical research to clinical applications will also require the choice of a population of epileptic patients which represents the best candidates for a clinical trial. One possibility would be to select patients affected by temporal lobe epilepsy since they represent a population elective for temporal-lobe surgery and it is likely that surgical removal of the pathological tissue will improve seizure outcome. At some stage prior to resective surgery, these patients could have vector infusion into the focus and the efficacy of the vector could be evaluated over a period of months before tissue resection. This approach would also give important insights into the level of expression of the transgene in the surgically removed epileptic tissue specimen.

Direct gene transfer into human epileptogenic hippocampal tissue has been done so far in vitro using an AAV vector to transfer and express a lacZ marker gene in brain slices obtained from patients undergoing temporal lobectomy to control intractable seizures (Freese et al., 1997). Expression of the transgene was observed within 5 h and morphological analysis indicated that neurons were preferentially transfected with no evidence of toxicity. Further evaluation of the functional correlates of vector-mediated gene transfer in human brain slices may provide an additional incentive for possible applications of gene therapies to patients with intractable seizures. In conclusion, once gene therapy effectiveness is demonstrated in epileptic tissues, and the safety concerns are adequately addressed in primate models, then the therapeutic potential of this novel and alternative approach, may be applied to the clinic for controlling otherwise pharmacologically intractable seizures. References Costantini, L.C., Bakowska, J.C., Breakefield, X.O. and Isacson, O. (2000) Gene therapy in the CNS. Gene Ther., 7: 93-109. Davidson, B.L., Stein, C.S., Heth, J.A., Martins, L, Kotin, R.M., Derksen, T.A., Zabner, J., Ghodsi, A. and Chiorini, J.A. (2000) Recombinant adeno-associated virus type 2,4, and 5 vectors: transduction of variant cell types and regions in the mammalian central nervous system. Proc. Natl. Acad. Sci. USA, 97: 3428-3432. Dumas, T.C. and Sapolsky, R.M. (2001) Gene therapy against neurological insults: sparing neurons versus sparing function. Trends Neurosci., 24: 695-700. Foldvary, N., Bingaman, W.E. and Wyllie, E. (2001) Surgical treatment of epilepsy. Neurol. Clin., 19: 491-515. Freese, A., Kaplitt, M.G., O'Connor, W.M., Abbey, M., Langer, D., Leone, R, O'Connor, M.J. and During, M.J. (1997) Direct gene transfer into human epileptogenic hippocampal tissue with an adeno-associated virus vector: implications for a gene therapy approach to epilepsy. Epilepsia, 38: 759-766. Gernert, M., Thompson, K.W., Loscher, W. and Tobin, A.J. (2002) Genetically engineered GABA-producing cells demonstrate anticonvulsant effects and long-term transgene expression when transplanted into the central piriform cortex of rats. Exp. Neurol, 176: 183-192. Haberman, R., Criswell, H., Snowdy, S., Ming, Z., Breese, G., Samulski, R. and McCown, T (2002) Therapeutic liabilities of in vivo viral vector tropism: adeno-associated virus vectors, NMDARl antisense, and focal seizure sensitivity. Mol. Ther., 6: 495-500. Haberman, R.R, McCown, T.J. and Samulski, R.J. (1998) Inducible long-term gene expression in brain with adeno-associated virus gene transfer. Gene Then, 5: 1604-1611. Haberman, R.F., Samulski, R.J. and McCown, T.J. (2003) Attenuation of seizures and neuronal death by adeno-associated virus vector galanin expression and secretion. Nat. Med., 9: 1076-1080. Howell, O.W., Doyle, K., Goodman, J.H., Scharfman, H.E., Herzog, H., Pringle, A., Beck-Sickinger, A.G. and Gray, W.R (2005) Neuropeptide Y stimulates neuronal precursor proliferation in the post-natal and adult dentate gyrus. J. Neurochem., 93: 560-570.

11. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

THERAPEUTIC IMPLICATIONS AND CONCLUSIONS Howell, O.W., Scharfman, H.E., Herzog, H., Sundstrom, L.E., Beck-Sickinger, A. and Gray, W.P. (2003) Neuropeptide Y is neuroproliferative for post-natal hippocampal precursor cells. J. Neurochem., 86: 646-659. Ruber, A., Padrun, V., Deglon, N., Aebischer, P., Mohler, H. and Boison, D. (2001) Grafts of adenosine-releasing cells suppress seizures in kindling epilepsy. Proc. Natl. Acad. Sci. USA, 98: 7611-7616. Janson, C.G., McPhee, S.W., Leone, P., Freese, A. and During, M.J. (2001) Viral-based gene transfer to the mammalian CNS for functional genomic studies. Trends Neurosci., 24: 706-712. Kay, M.A., Glorioso, J.C. and Naldini, I. (2001) Vectors for gene therapy: the art of turning infectious agents into vehicles of therapeutics. Nat. Med., 7(1): 3 3 ^ 0 . Klugmann, M., Leichtlein, C.B., Symes, C.W., Serikawa, T, Young, D. and During, M.J. (2005) Restoration of aspartoacylase activity in CNS neurons does not ameliorate motor deficits and demyelination in a model of Canavan disease. Mol. Ther., 11: 745-753. Lin, E.J., Richichi, C , Young, D., Baer, K., Vezzani, A. and During, M.J. (2003) Recombinant AAV-mediated expression of galanin in rat hippocampus suppresses seizure development. Eur. J. Neurosci., 18(7): 2087-2092. Mazarati, A., Langel, U. and Bartfai, T. (2001) Galanin: an endogenous anticonvulsant? Neuroscientist, 7: 506-517. McCown, T.J., Givens, B.S. and Breese, G.R. (1987) Amino acid influences on seizures elicited within the inferior coUiculus. J. Pharmacol. Exp. Ther., 243: 603-608. McCown, T.I. (2004) The clinical potential of antiepileptic gene therapy Expert Qpin. Biol. Ther. 4(11): 1/71-76. McLaughlin, J., Roozendaal, B., Dumas, T, Gupta, A., Ajilore, O., Hsieh, J., Ho, D., Lawrence, M., McGaugh, J.L. and Sapolsky, R. (2000) Sparing of neuronal function postseizure with gene therapy. Proc. Natl. Acad. Sci. USA, 97:12804-12809. Monahan, RE. and Samulski, R.J. (2000) Adeno-associated virus vectors for gene therapy: more pros than cons? Mol. Med. Today, 6: 433-440. Mountain, A. (2000) Gene therapy: the first decade. Trends Biotechnol., 18: 119-128. Noe', R, Nissinen, J., Filippi, R, During, M.J., Pitkanen, A. and Vezzani, A., rAAV-mediated neuropeptide Y gene expression in the hippocampus of chronically epileptic rats reduces spontaneous seizures ensuing after electrically-induced status epilepticus. American Epilepsy Society and American Clinical Neurophysiology Society Joint Annual Meeting 2005, Washington, DC, USA. Noebels, J.L. (1996) Targeting epilepsy genes. Neuron, 16: 241-244. Ozawa, C.R., Ho, J.J., Tsai, D.J., Ho, D.Y. and Sapolsky, R.M. (2000) Neuroprotective potential of a viral vector system induced by a neurological insuU. Proc. Natl. Acad. Sci. USA, 97: 9270-9275.

163

Perucca, E. (1998) Pharmacoresistance in epilepsy: how should it be defined? CNS Drugs, 10: 171-179. Richichi, C , Lin, E.J., Stefanin, D., Colella, D., Ravizza, T, Grignaschi, G., Veglianese, P., Sperk, G., During, M.J. and Vezzani, A. (2004) Anticonvulsant and antiepileptogenic effects mediated by adeno-associated virus vector neuropeptide Y expression in the rat hippocampus. J. Neurosci., 24: 3051-3059. Rogawski, M.A. and Loscher, W. (2004) The neurobiology of antiepileptic drugs. Nat. Rev. Neurosci., 5: 553-564. Seki, T., Matsubayashi, H., Amano, T., Kitada, K., Serikawa, T., Sasa, M. and Sakai, N. (2004) Adenoviral gene transfer of aspartoacylase ameliorates tonic convulsions of spontaneously epileptic rats. Neurochem. Int., 45: 171-178. Shi, N., Zhang, Y, Zhu, C , Boado, R.J. and Pardridge, W.M. (2001) Brain-specific expression of an exogenous gene after i.v. administration. Proc. Natl. Acad. Sci. USA, 98: 12754-12759. Silva, A.P., Pinheiro, PS., Carvalho, A.P, Carvalho, CM., Jakobsen, B., Zimmer, J. and Malva, J.O. (2003) Activation of neuropeptide Y receptors is neuroprotective against excitotoxicity in organotypic hippocampal slice cultures. Faseb J., 17:1118-1120. Tenenbaum, L., Chtarto, A., Lehtonen, E., Velu, T., Brotchi, J. and Levivier, M. (2004) Recombinant AAV-mediated gene delivery to the central nervous system. J. Gene Med., 6(Suppl 1): S212—S222. Thompson, K., Anantharam, V, Behrstock, S., Bongarzone, E., Campagnoni, A. and Tobin, A.J. (2000) Conditionally immortalized cell lines, engineered to produce and release GAB A, modulate the development of behavioral seizures. Exp. Neurol., 161: 481^89. Thompson, K.W. and Suchomelova, L.M. (2004) Transplants of cells engineered to produce GABA suppress spontaneous seizures. Epilepsia, 45: 4-12. Traeger, E.G. and Rapin, I. (1998) The clinical course of Canavan disease. Pediatr. Neurol., 18: 207-212. Vezzaru, A. and Sperk, G. (2004) Overexpression of NPY and Y2 receptors in epileptic brain tissue: an endogenous neuroprotective mechanism in temporal lobe epilepsy? Neuropeptides, 38: 245-252. Vezzani, A., Sperk, G. and Colmers, W.R (1999) Neuropeptide Y: emerging evidence for a functional role in seizure modulation. Trends Neurosci., 22: 25-30. Xu, R., Janson, C.G., Mastakov, M., Lawlor, P., Young, D., Mouravlev, A., Fitzsimons, H., Choi, K.L., Ma, H., Dragunow, M., Leone, P, Chen, Q., Dicker, B. and During, M.J. (2001) Quantitative comparison of expression with adeno-associated virus (AAV-2) brain-specific gene cassettes. Gene Ther., 8:1323-1332. Zhang, L.X., Li, X.L., Smith, M.A., Post, R.M. and Han, J.S. (1997) Lipofectin-facilitated transfer of cholecystokinin gene corrects behavioral abnormalities of rats with audiogenic seizures. Neuroscience, T7\ 15-22.

11. GENE THERAPY FOR DEGENERATIVE AND FUNCTIONAL DISORDERS

CHAPTER

13 Genetic Manipulation of Learning and Memory Jane Dunning

Abstract: Genetic techniques have been integral in elucidating the molecular basis of learning and memory. This chapter presents an overview of the progress to date, and challenges facing the field. While much of the work serves to confirm preexisting pharmacological data, it does so with a spatial and temporal precision not previously possible. In addition to identifying the involvement of individual genes, the selectivity of this approach has facilitated the development of key concepts of learning and memory processes. Exploration of the molecules underlying the discrete temporal phases of memory storage, and an understanding of the restriction of plasticity to the individual synapse, would not have been achieved without the specificity conferred by genetic manipulation.The accumulated body of research has revealed memory to be influenced by a diverse range of genes, remarkably conserved across species, brain regions, and memory subtypes. Precisely how these molecules interact in the formation, maintenance, and retrieval of the memory trace remains to be determined.Vector-mediated gene transfer has been relatively underutilized as an investigative tool in studies of learning and memory, despite possessing a number of advantages over currently favored techniques. Increased adoption of this technology will provide a powerful approach for addressing the questions that remain. Keywords: memory; long-term potentiation (LTP); synaptic plasticity; cyclic AMP response element binding protein (CREB); tetracycline transactivator I,

though its precise relationship with memory remains unclear (Martin et al., 2000). Synaptic strength can potentially be influenced by a wide range of both pre- and postsynaptic mechanisms, such as the rate of neurotransmitter release, and expression of postsynaptic receptors. Manipulation of the genes underlying these processes can provide insight into the mechanisms of memory.

INTRODUCTION

Learning is defined as the process by which information is acquired, whereas memory is the retention of this information. Both are critical to our understanding of the world, and differences in our recall and interpretation of events form in large part the biological basis of our individuality. At the cellular level, an increase in the strength of synaptic connections due to coincident pre- and postsynaptic activities is considered the primary mechanism by which memory storage is achieved (Hebb, 1949). A form of long-lasting synaptic strength referred to as LTP possesses many of the properties required of a memory storage system. As such, it is considered the preeminent model for memory encoding and storage.

Gene Therapy of the Central Nervous System: From Bench to Bedside

IL

TECHNIQUES FOR STUDYING, LEARNING, AND MEMORY

Genetic manipulation has revolutionized the field of learning and memory. While pharmacological studies have provided initial evidence for the involvement

167

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

168

13. GENETIC MANIPULATION OF LEARNING AND MEMORY

of many genes, they are frequently limited by issues of nonspecificity. These include actions at sites other than the target molecule, drug toxicity, and compensatory alterations in gene expression. Additionally, drugs can affect brain regions not under direct study, potentially confounding the results of memory tasks. In other cases, drugs are not available to target specific molecules, or distinguish between closely related genes. In contrast, genetic techniques can not only counter many of these problems, but can also be used to target specific regions of a gene, for example, in the creation of nonphosphorylable mutants or dominant-negative constructs lacking key activation domains. The majority of such studies to date have involved the use of transgenic mice. While first generation transgenics employed global gene knockouts, this method is subject to interference from developmental effects, including compensatory changes in gene expression, as well as effects on nonmnemonic functions that may impair the ability of the animal to perform memory tasks. At the extreme end of the spectrum, the absence of some genes is lethal. Later studies devised means to eliminate these problems by restricting transgene expression in both spatial and temporal domains. A.

Spatial Restriction of Gene Expression

Many aspects of memory can be compartmentalized into distinct brain regions, enabling selective targeting of discrete aspects of cognitive function. The majority of region-specific transgenics express the gene of interest under the control of the calmodulin kinase Ila (CaMKIIa) promoter (Mayford et a l , 1996). This limits expression to the forebrain, an area with strong involvement in learning and memory processes. Spatially restricted knockouts can also be generated by crossing mice expressing CaMKIIa-mediated ere recombinase with those expressing the gene of interest flanked by loxP sites (Tsien et al., 1996). Gene therapy-based techniques can also be used to achieve spatial selectivity. They do so with greater precision than current transgenic methods, as well as eliminating the need to identify region-specific promoters. This is particularly important as transgenic spatial restriction is currently limited to the forebrain, although its subregions differ greatly with respect to their involvement in memory processes. Gene therapy also eliminates the need to produce cumbersome transgenics, allowing candidate genes to be screened quickly. Furthermore, rats can be used, which are advantageous as they are considered more amenable than mice for behavioral studies.

B.

Temporal Restriction of Gene Expression

Temporal control of gene expression is of particular significance for studies of memory, which both conceptually and at the molecular level consists of a number of temporally distinct phases. One way in which this can be achieved is by the use of the tetracycline transactivator (tTA) system. Fusion of the E. coli tetracycline repressor to a VP16 activating domain created a construct capable of activating transcription from the tetO promoter (Gossen and Bujard, 1992). Transcription can be reversibly blocked by application of the tetracycline analogue doxycycline. Mutation of the tetracycline repressor created the reverse tTA system (rtTA), allowing transgene expression only in the presence of doxycycline (Gossen et al., 1995). This circumvents the requirement for continuous doxycycline application, which can have detrimental effects on learning and memory (Mayford et al., 1996). A subsequent innovation was to place the tTA gene under the control of a CaMKIIa promoter, providing both spatial and temporal restriction (Mayford et al., 1996). While typically used in conjunction with transgenic techniques, a number of studies have demonstrated that this system can be used successfully with viral vectors (Fitzsimons et al., 2001), though to date this has not been utilized in a learning and memory paradigm.

IIL

ASPECTS OF MEMORY FORMATION

The following sections provide an overview of current knowledge regarding the molecular basis of memory, with examples of the contribution of genetic manipulation to our understanding of this field. A. Memory Acquisition The majority of excitatory synapses in the central nervous system use glutamate as their neurotransmitter. Glutamate binds receptors located in the postsynaptic membrane, a subset of which mediates ionic conductance through receptor channels. Ionic glutamatergic receptors can be further subdivided into a-amino-3-hydroxy-5-methyl-4-isoxazoleproprionic acid (AMPA), N-methyl-D-aspartate (NMDA), or kainate, on the basis of differing affinity for these ligands (Watkins et al., 1990). Basal synaptic (rapid excitatory) transmission is mediated by the AMPA receptors. The influx of Na^ and K^ ions through its channel causes depolarization

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

ASPECTS OF MEMORY FORMATION

of the postsynaptic membrane. As such, AMPA receptors serve as an initial gateway for memory encoding. 1.

NMD A Receptor

Conductance through the NMDA receptor channel is absent at the resting potential, due to the presence of a voltage-dependent magnesium block. NMDA receptor activation requires coincident glutamate binding and membrane depolarization induced by repeated high-frequency stimulation, properties consistent with a role in memory encoding (Bourne and NicoU, 1993). The NMDA receptor is a heteromultimer, comprising one NRl subunit, and one or more NR2 subunits (either A or B in the adult forebrain). The presence of different NR2 subunits confers distinct properties on the receptor channel (Nakanishi, 1992). Forebrain-restricted knockout of the NRl subunit led to complete abolition of the NMDA receptor in this region. Accordingly, mice exhibited an impairment of spatial memory (Tsien et al., 1996) and trace fear conditioning (Huerta et al., 2000), as well as impaired LTP (Tsien et al., 1996). In contrast, forebrain-specific overexpression of NR2B, which confers increased receptor channel open time, produced mice with enhanced mnemonic ability on a number of hippocampal-dependent tasks, a phenomenon postulated to be due to the increased interval over which coincident activity could be detected (Tang et al., 1999). The NMDA receptor's crucial role in gating memory formation has led to the study of region-specific NMDA receptor knockouts to elucidate the mnemonic functions of distinct brain regions. In this manner, hippocampal CAl has been linked to transverse patterning (RondiReig et al., 2001), whereas CA3 (targeted by expressing ere recombinase under the control of the kainate receptor promoter) has been implicated in associative memory recall (Nakazawa et al., 2002) and the rapid acquisition of novel experiences (Nakazawa et al., 2003). In addition, a number of other postsynaptic receptors interact to enhance the signal in downstream pathways, among them glucagon-like peptide 1 receptor (GLPIR), coupled to adenylate cyclase, PKC, and mitogen-activated protein kinase (MAPK) pathways. If the GLPIR knockout mice were impaired then the heterozygotes possessed an intermediate phenotype. Hippocampal administration of rAAV2-GLPlR was able to restore the contextual memory ability of GLPIR knockout mice. Furthermore, rAAV2-GLPlR overexpression in the rat hippocampus improved performance on spatial memory and contextual fear conditioning tasks (During et al., 2003).

169

Similarly, rAAV2-mediated overexpression of the vasopressin receptor VIAR in the rat septum facilitated social discrimination memory among juveniles, although it remains to be seen if this finding can be extended to other forms of memory (Landgraf et al., 2003). B.

Postsynaptic Receptor Clustering

A key factor influencing the ability of synapses to rapidly alter synaptic strength is the interaction of components of the signaling apparatus at a macromolecular cluster referred to as the postsynaptic density. This structure comprises receptor and second messenger proteins, as well as the scaffolding proteins that facilitate signal transduction by maintaining components in close proximity (reviewed in Kennedy, 1993). Several studies have demonstrated the importance of scaffolding proteins in facilitating learning and memory. A global knockout of PSD-95 produced impaired memory in conjunction with enhancement of LTP for a given stimulus frequency. This was particularly interesting due to the dissociation of a direct relationship between LTP and memory, instead highlighting the importance of bidirectional plasticity for optimal memory storage (Migaud et al., 1998). The role of homer proteins in maintaining the postsynaptic density was recently demonstrated by rAAV-mediated overexpression in the rodent hippocampus. All homer isoforms possess a ligand-binding domain, and are thus capable of binding a number of substrates localized in the postsynaptic membrane. While long isoforms also have a coiled domain, allowing the creation of multimeric chains that physically link receptor proteins with intracellular calcium stores, short isoforms lack this domain, and thus act in a dominant-negative fashion. Accordingly, different isoforms exerted differing effects on learning and memory, with the short isoform homer l a producing an impairment of recognition and spatial memory, while the long isoforms Ic and g produced a slight enhancement (Klugmann et al., 2005). C.

Memory Molecules

1.

CaMKIl

Subsequent to NMDA receptor activation, calcium influx to the postsynaptic neuron activates a variety of downstream targets, either directly or indirectly via activation of calmodulin. The frequency of synaptic stimulation is critical in determining the duration of calcium influx, and hence the persistence of the memory encoded (Frey et al., 1993).

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

170

13. GENETIC MANIPULATION OF LEARNING AND MEMORY

One of the key targets of calcium/calmodulin signaling is CaMKII, a dodecameric holoenzyme consisting of a and ^ subunits. Each subunit comprises 4 domains, designated catalytic, autoinhibitory, variable segment, and self-association (Soderling, 1996). The catalytic domain possesses ATP- and substrate-binding sites, and interacts with anchoring proteins within the postsynaptic density. Under resting conditions, the pseudosubstrate region of the autoinhibitory domain binds to and inhibits catalytic domain activity. When calcium enters the postsynaptic neuron and binds calmodulin, it enables calmodulin to bind a region of CaMKII overlapping the pseudosubstrate domain, causing its dissociation from the catalytic domain. This exposes Thr286, allowing it to be phosphorylated by the catalytic domain of a neighboring subunit (Rich and Schulman, 1998). Once phosphorylated, CaMKQ's activity can persist in the absence of Ca^^, and may provide a means by which it can act as a mechanism of memory storage (Lisman and Goldring, 1988). CaMKII phosphorylates target proteins, among them the AMPA receptor subunit GluRl, facilitating AMPAmediated ionic conductance as well as the insertion of further AMPA receptors into the postsynaptic membrane. This leads to an increase in synaptic efficacy, and contributes to the formation of early LTP (E-LTP). Global deletion of CaMKIIa is significant for being the first use of genetic manipulation to identify the role of a protein in memory, as the animals produced exhibited a profound impairment of spatial learning and memory (Silva et al., 1992). However, they were also shown to possess an abnormal fear response and aggressive behavior (Chen et al., 1994), necessitating the development of more specific methods of genetic manipulation. In particular, much has been discovered by the creation of transgenic animals with point mutations at key phosphorylation sites. For example, mutation of Thr286 to a nonphosphorylable Ala is detrimental to memory performance (Giese et al., 1998). However, overexpression of CaMKIIa T286D, which mimics phosphorylation and causes constitutive, phosphatase-resistant activation of the kinase, impairs LTP at certain frequencies, as well as spatial memory (Bach et al., 1995; Mayford et al., 1996). Interestingly, contextual memory was unaffected in these animals, suggesting that in certain cases distinct forms of memory differ in their molecular requirements, despite sharing a similar anatomical locus. Furthermore, this demonstrates that CaMKII activation is not sufficient on its own, and must be coupled with the appropriate signals to produce optimal memory storage.

A more subtle memory phenotype is obtained by mutation of Thr305 and 306 to a nonphosphorylable Val and Ala, respectively. This prevents inhibitory phosphorylation, which is believed to reduce the affinity of CaMKII for the postsynaptic density. While these animals were unimpaired in spatial memory tasks, they exhibited a subtle impairment in tasks requiring flexibility such as learning a new platform position or distinguishing between similar chambers. Additionally, mimicking constitutive autophosphorylation of ThrSOS by mutation to Asp blocks activation of the kinase by calmodulin and produces an impairment of both spatial memory and contextual conditioning (Elgersma et al., 2002). These results are also interesting in light of studies linking CaMKII activity to Angelman's syndrome, a form of mental retardation characterized by profound cognitive dysfunction and learning deficits. Animal models of this disorder have increased phosphorylation at Thr286 and 305, coupled with reduced CaMKII enzymatic activity and reduced association with the postsynaptic density. However, it is currently unclear how perturbation of the UBE3A gene implicated in Angelman's syndrome causes this effect (Elgersma et a l , 2004). 2* Ccdcineurin E-LTP decays after an hour, approximately the length of time that CaMKII can remain active in the absence of Ca^^. Dephosphorylation of CaMKII is mediated by protein phosphatase (PPl); a protein phosphatase that is active only when the activity of the inhibitory protein II is suppressed by calcineurin (Lisman, 1989). Calcineurin, also known as PP2B, exhibits a high affinity for calcium, and hence is activated by the low level Ca^^ influx arising from low-frequency postsynaptic stimulation. It is a heterodimer, comprising a catalytic subunit (CNAa or CNA^), and a regulatory subunit (CNBl). Calcineurin is capable of dephosphorylating a wide variety of substrates, including II, AMPA and NMDA receptors, and components of the endocytic machinery, facilitating removal of AMPA receptors from the synapse and leading to development of long-term depression (LTD), the functional inverse of LTP (Beattie et al., 2000). Calcineurin is a prime example of a memory suppressor gene, a key concept in studies of memory. It is believed that there are both positive and negative regulatory molecules, the balance of which determines the strength of memory encoded. At many points along the signaling pathway, positive and negative

HI. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

ASPECTS OF MEMORY FORMATION

regulators act in concert to 'gate' memory formation (Abel and Kandel, 1998). Although LTD is typically considered to have a negative effect on memory storage, a forebrain-specific knockout of CNBl selectively impaired both LTD and certain forms of memory, once again highlighting the importance of bidirectional plasticity in memory systems (Zeng et al., 2001). Spatially and temporally restricted overexpression of a constitutively active, truncated calcineurin (lacking autoinhibitory and calmodulin-binding domains of CNAa) impaired long-term spatial memory and LTP. Interestingly, this deficit could be rescued by increasing the number of training trials per day, suggesting it was not absolute, but could be compensated for, perhaps by the increased expression of positive memory modulators (Mansuy et al., 1998). In contrast, overexpression of the CNAa autoinhibitory domain under the control of rtTA facilitated LLTP, enhanced learning, and strengthened short- and long-term memory for object recognition and spatial navigation tasks (Malleret et al., 2001). To date, eliminating inhibitory constraints on memory storage has been one of the most successful strategies for improving memory in 'normal' rodents. A similar result was obtained by overexpression of a constitutively active II, leading to improved learning and retention of spatial tasks (Genoux et al., 2002). 3.

Protein Kinase A

Continued influx of Ca^+ activates further enzymes, some of which are capable of removing inhibitory constraints on memory storage. Among these are adenylate cyclase isoforms 1 and 8, which catalyze the conversion of ATP to cAMP, allowing it to subsequently bind and activate protein kinase A (PKA). A double knockout of these genes produced an impairment of memory (Wong et al., 1999), whereas conversely, adenylate cyclase 1 overexpression facilitated memory (Wang et al., 2004). PKA consists of two regulatory and two catalytic subunits, and acts as a molecular switch between short- and long-term memory By phosphorylating II, it is able to inhibit PPl, thereby preventing the dephosphorylation of key proteins involved in memory formation, and can thus be considered to act in direct apposition to calcineurin (Blitzer et a l , 1995). Mice expressing a dominant-negative PKA (dnPKA) in the forebrain express normal E-LTP, but are deficient in both L-LTP and long-term memory. This deficit can be rescued by pharmacological inhibitors of PPl and PP2A, an interesting demonstration of the combined

171

use of genetics and pharmacology to delineate the pathways involved in memory formation, as well as the ability to compensate for genetic deficits (Woo et al., 2002). This approach was further explored by several studies utilizing drug administration to expose the memory phenot5^e of otherwise silent genetic mutations (Frankland et al., 2003). Persistent activation of PKA causes the catalytic subunits to translocate to the nucleus, where they are one of several molecules capable of phosphorylating the transcription factor CREB (Huang and Kandel, 1994). 4-

CREB

It has been known for many years that while shortterm memory involves the modification of preexisting proteins, long-term memory storage requires new protein synthesis (Flexner and Flexner, 1966). One means by which this can be achieved is via the transcription factor CREB, which, once phosphorylated at Serl33, dimerizes with other family members and binds the CRE element located in the promoter region of target genes (Gonzalez and Montimny, 1989). A number of proteins are capable of phosphorylating CREB, allowing it to act as a convergence point for a number of pathways. Of these, calmodulin kinase IV (CaMKIV) is linked to a rapidly induced, early phosphorylation event, although there is some debate over its precise role in memory, as differing transgenic techniques have produced conflicting results (Ho et al., 2000, Kang et al., 2001, Wei et al., 2002). ^ CREB's role in memory was initially identified by use of a global knockout mouse (Bourtchuladze et al., 1994). While these animals exhibited a memory impairment, it was subsequently discovered that a novel p isoform, transcribed from a different start site, had not been eliminated, and in fact had been compensatorily upregulated, along with several other related genes (Blendy et al., 1996). The problem of compensatory upregulation was later addressed in a study using reversible inJiibition of CREB, and related genes CREM and ATFl by overexpressing a dominant-negative KCREB in the forebrain (dorsal hippocampus) under control of rtTA. These animals were impaired in spatial memory and object recognition, but not contextual fear conditioning, which does not rely on the dorsal hippocampus (Pittenger et al., 2002). The CREB gene family includes both positive and negative regulators of memory. Forebrain-specific overexpression of a tetracycline-regulated dominantnegative C/EBP was shown to specifically inhibit all negative C/EBP and ATF4 genes, as well as produce a reversible enhancement of memory (Chen et al., 2003).

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

172

13. GENETIC MANIPULATION OF LEARNING AND MEMORY

The role of CREB at distinct time points in memory formation was investigated by the use of an inducible construct consisting of a mutant nonphosphorylable CREB fused to the ligand-binding domain of a human estrogen receptor, under the control of a CaMKIIa promoter. Expression can be rapidly and reversibly induced by injection of tamoxifen. This was used to demonstrate CREB's involvement in both new and reactivated fear memories (Kida et a l , 2002), as well as in conditioned taste aversion (Josselyn et al., 2004). CREB is one of the few memory-related genes for which gene transfer techniques have been used. In part, this may be due to questions raised about the validity of prior studies in transgenic animals (Balschun et al., 2003). The HSV-mediated overexpression of CREB in the basolateral amygdala produced animals capable of forming memories after massed training sessions in the fear-potentiated startle paradigm, whereas normal animals require multiple, spaced training sessions to achieve the same strength of memory (Josselyn et al., 2001). This phenomenon is believed to result from the equal induction of CREB activating and inhibiting isoforms after training, with inhibitors decaying faster relative to activators during rest intervals. CREB animals are able to circumvent the requirement for spaced trials, as they are already biased toward increased levels of activators compared to inhibitors. Gene therapy is particularly useful in this study, as it is not possible to selectively target the basolateral amygdala using current transgenic techniques. Additionally, the anatomical specificity of the effect could be confirmed by injections that missed the basolateral amygdala, as these animals were not memory enhanced. Though HSV type 1-derived vectors only produce short-term gene expression, this was advantageous in dissecting the temporal requirements for CREB in memory, as rats infused 3 days before training exhibited a memory enhancement, whereas rats infused 14 days before training (so CREB expression would have decayed by training time) did not (Josselyn, 2001). 5-

CREB Binding Protein

In order to initiate transcription, CREB must interact with CREB binding protein (CBP), which acts as a conduit for recruiting proteins to the promoter region (Chrivia et al., 1993), as well as a histone acetyltransferase, capable of altering chromatin structure and hence controlling gene expression (Ogryzko et al., 1996). In humans, mutations in the CBP locus are associated with Rubenstein-Taybi syndrome, characterized

by severe mental retardation. Mice overexpressing a truncated CBP have a long-term memory deficit on a number of tasks (Oike et al., 1999; Bourtchouladze et al., 2003). However, the interpretation of these results is confounded by the severe phenotype of these mice. C B P + / - mice also exhibit long-term memory deficits in contextual and cued fear and object recognition tasks, though interestingly are unimpaired in the water maze (Alarcon et al., 2004). Overexpression of a tetracycline-inducible dominant-negative CBP transgene designed to selectively block histone acetyltransferase activity demonstrated an impairment of several forms of long-term memory (Korzus et al., 2004). This is an important example of an emerging field of research, the influence of epigenetic mechanisms on memory formation (Levenson and Sweatt, 2005). The phenotype could be rescued with a histone deacetylase inhibitor, suggesting an intervention point at which treatments for this disorder could be targeted. 6.

CREB Targets

The targets of CREB number in the hundreds, and perform a diverse range of cellular functions. While several of them have been shown to be involved in learning and memory processes via genetic manipulation, it is not clear if there are selective mechanisms by which CREB or other factors can activate a subset of genes. Use of genetic manipulations to decrease expression of several transcription factors activated downstream of CREB have been shown to impair memory, including c-fos (Fleischmann et al., 2003), C/EBPj8 (Taubenfeld et al., 2001), and Krox24 (Jones et al., 2001). Furthermore, overexpression of TrkB, the receptor for brain-derived neuronal factor (BDNF), both of which are activated downstream of CREB, facilitated learning in a variety of tasks including spatial and contextual learning, and conditioned taste aversion (Koponen et al., 2004). In contrast, use of a lentiviral system to overexpress dominant-negative TrKB in the amygdala produced an impairment of fear-potentiated startle. Once again, gene therapy was particularly advantageous in allowing the demonstration of anatomical specificity, as animals with missed injections were unimpaired, and temporal specificity, as infusion of vector before, but not after conditioning produced an impairment (Rattiner et al., 2004). D.

Synaptic Tagging

The identification of a transcription factor as a key memory molecule raised the question of how changes

. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

ASPECTS OF MEMORY FORMATION

in synaptic efficacy can be restricted to one synapse, when transcription is induced in the nucleus. One potential explanation is the 'synaptic tagging' hypothesis. This model suggests that initial weak synaptic stimulation leading to E-LTP leaves a protein tag at the stimulated synapse. Further stimulation at this synapse will lead to late LTP (L-LTP), requiring protein synthesis in the nucleus. While the newly synthesized proteins are transported to all synapse of the neuron, L-LTP will only be expressed in the stimulated synapse, due to the presence of the protein tag. An interesting prediction of this theory is that the presence of newly synthesized proteins at all synapses of a neuron allows L-LTP to be generated by a weak stimulus that would usually only produce E-LTP (Prey and Morris, 1997). While electrophysiological studies with transgenic animals expressing constitutively active CREB have suggested that this tag may be a protein product from a CRE-driven gene (Barco et al., 2002), recent work has examined the effect on memory of disrupting synaptic tagging. Neuronal activity-induced translation of mRNA localized in dendrites is thought to be a control point in neuronal plasticity and is a potential means by which spatial restriction of synaptic plasticity could be achieved. One study devised a clever technique to test the importance of local protein synthesis in dendrites by disrupting the dendritic localization signal located in the 3' untranslated region of endogenous CaMKIIa. While protein could still be produced, it was restricted to the soma. This reduced L-LTP, and impaired performance on a number of behavioral assays, including object recognition, spatial memory, and associative cued and contextual fear conditioning (Miller et al., 2002). One candidate for translational regulation is MAPK (also referred to as ERK), which, interestingly, is also capable of phosphorylating CREB (Wu et al., 2001). Previous studies investigating the role of MAPK in memory have produced unexpected results due to gene compensation between the p42 and p44 (ERKl and 2) isoforms, such as an ERKl knockout with no impairment of emotional learning (Selcher et al., 2001) or an enhancement of memory on striatum-dependent tasks (Mazzucchelli et al, 2002). However, forebrain-restricted expression of a dominant-negative MEKl construct was able to produce a memory phenotype consistent with the results of pharmacological studies. MEKl acts upstream of both ERK isoforms, and a K to M substitution in its ATP binding site was able to abolish kinase activity without preventing binding to ERKl and 2. Memory was

173

impaired on spatial, contextual, and cued fear conditioning tasks. In addition, ERK inhibition impaired the translation-dependent, transcription-independent phase of LTP, and blocked phosphorylation of a number of translation factors (Kelleher et al., 2004). To date, many components of pathways acting upstream of MAPK have been studied and shown to be involved in memory formation. MAPK is activated by two separate pathways downstream of cAMP. Intriguingly, many genes in these pathways are involved in aspects of cell proliferation, suggesting a potentially fruitful avenue for research into a number of other molecules with similar function but as yet undetermined role in post-mitotic neurons. Both pathways involve small GTPases, Rapl, and Ras. These proteins are activated by distinct GTP/ GDP guanine nucleotide exchange factors and exert differing effects on downstream kinases. While Ras activates Raf 1 and b-Raf, which in turn activate MEK, Rapl can either inhibit Rafl or activate b-Raf, as well as directly antagonize the function of Ras. They also exhibit different subcellular localization, with Ras anchored to the plasma membrane, whereas Rapl is found in the membrane of the endosomal compartment (Kim et al., 1990). Forebrain-specific expression of a tetracyclineregulated Rapl dominant-negative impaired spatial memory and context discrimination but not cued or contextual fear conditioning (Morozov et al., 2003), suggesting it is essential for hippocampal but not amygdala-dependent learning. In contrast, mice lacking the guanine nucleotide exchange factor Ras-GRF exhibited a selective dysfunction of amygdala-dependent emotional memory and LTP, while hippocampal memory and LTP remained unaffected (Brambilla et al., 1997). Similarly, global disruption of the Ras effector Rinl led to enhancement of amygdala-dependent LTP and memory but preservation of hippocampal LTP and learning, perhaps due to factors such as differential expression of splice variants, or presence of region-specific signaling pathways (Dhaka et al., 2003). However, another study using Ras-GRF deficient mice showed normal performance on amygdaladependent learning and memory, but impairment on hippocampal-dependent tasks. The differing results of these studies highlight problems inherent in the gene knockout approach, and has been suggested to be attributable to factors including the location of the neo gene, which can affect expression of nearby genes, differences in genetic background, or in behavioral procedures (Giese et al., 2001).

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

174

13. GENETIC MANIPULATION OF LEARNING AND MEMORY

A heterozygous null mutation of the NFl gene, which codes for a GTPase-activating protein capable of inactivating ras, produced animals with spatial memory deficits. Interestingly, this could be rescued by crossing these mice with K-Ras+/- mice, countering the Ras increase induced by the NFl mutation, and demonstrating a powerful use of genetic techniques to delineate a molecular pathway (Costa et al., 2002). E.

Memory Consolidation

Previous work has established that although memories are initially formed in the hippocampus, they are transferred to areas of the cortex for long-term storage (e.g. Bontempi et al., 1999). Several of the proteins implicated in earlier phases of memory may also play a role in consolidation. For example, mice heterozygous for a null mutation of CaMKIIa display normal hippocampal-dependent memory at 1-4 days post training, but are impaired at extended delays (greater than 10 days). In conjunction with this deficit, mice exhibited impaired cortical but not hippocampal LTP, suggesting that CaMKII is required during the time window in which memory is transferred from the hippocampus to the cortex. The lack of effect on hippocampally mediated memory is postulated to be due to the relatively higher basal expression of CaMKII in this region, such that a knockdown would have less effect (Frankland et al., 2001). CaMKIFs role in consolidation has also been investigated by overexpressing CaMKII aF89G in the forebrain. This silent mutation creates a unique cavity in the ATP-binding pocket of the catalytic domain, allowing selective binding by a synthetic kinase. This compound reaches peak expression in the brain 20 min after i.p. injection, and declines to baseline after 45 min, allowing for more rapid modulation of gene expression than can be achieved at the level of DNA. While transgenic mice had altered bidirectional plasticity (Wang et al., 2003), and were impaired on a number of memory tasks, this could be reversed by kinase administration. Pre-treating animals with the kinase, then withdrawing treatment at different time points after fear conditioning demonstrated that normal CaMKII expression is critical at 1 week post training for memory retention 1 month later. Changes in synaptic morphology are also associated with long-term cortical storage of memory, and are predominantly mediated by remodeling of actin filaments. Forebrain-specific overexpression of a dominant-negative p21-activated kinase (PAK), a serine threonine kinase that acts as a regulator of this process.

produced mice with alterations in synaptic morphology. Cortical neurons from these animals possessed fewer dendritic spines, coupled with an increased proportion of larger synapses. In addition, they exhibited an impairment of consolidation on several hippocampal-dependent memory tasks (Hayashi et al., 2004). Selectively knocking out NMDA receptors in CAl during the initial post training weeks also disrupts the formation of long-lasting memory. Based on this finding, a synaptic reentry reinforcement hypothesis (Shimizu et al., 2000) has been proposed as a mechanism for transferring memories from the hippocampus to the cortex. This suggests that multiple rounds of post learning synaptic changes, perhaps initiated by periodic reactivation of the NMDA receptor, are required to reinforce the synaptic changes initiated during memory acquisition and counter the drift in synaptic efficacy arising from receptor turnover (Wittenberg and Tsien, 2002). A later study developed a technique for reversible elinunation of NMDA receptor expression in the forebrain in order to further examine this hypothesis. A forebrain-restricted NMDA receptor knockout was rescued by expression of NMDA receptor under the control of tetracycline. Switching off forebrain NMDA receptor expression 6 months after training and at least 2 months prior to memory retrieval impaired the retrieval of both contextual and cued fear memories (Cui et al., 2004). While this implicates the NMDA receptor in the later phases of memory storage, the precise nature of this involvement remains to be defined. Another gene implicated in information transfer between hippocampal and cortical regions is the M l muscarinic receptor, as M l knockout mice exhibit a selective enhancement of hippocampal tasks, coupled with an impairment of cortical tasks including consolidation (Anagnostaras et al., 2003). F.

Interaction between Genes and Environment

In addition to anecdotal evidence, a number of scientific studies have shown that exposure to an enriched environment has beneficial effects on multiple brain functions, including memory (e.g. Greenough et a l , 1972; Kempermann et al., 1997). This has been associated with upregulation of a number of genes. One such gene is vascular endothelial growth factor (VEGF), a hypoxia-inducible protein, which promotes angiogenesis via receptor tyrosine kinases Fltl and Kdr, and neuropilin 1 and 2, located on endothelial cells. VEGF also acts as a neurotrophic factor, and

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

ASPECTS OF MEMORY FORMATION

exerts neurogenic effects on neuronal progenitors. The rAAV2-mediated overexpression of VEGF in the hippocampus leads to expression in all principal cell groups, coupled with improved performance on several hippocampally mediated tasks. In contrast, rAAV expressing placental growth factor, which signals through Fltl, and rAAV-dnKDR led to impaired performance, suggesting that KDR, n p l and np2 are the receptors responsible for mediating the effects of VEGF on cognition. The VEGF-treated rats also demonstrated enhanced neurogenesis, suggesting that VEGF may be a means of linking the effects of enrichment to memory (Cao et al., 2004). Another gene that has been linked to the role of environment on memory is presenilin 1 (PSl). While PSl knockout mice did not show any memory impairment as assessed by fear conditioning, they exhibited decreased neurogenesis after exposure to an enriched environment. When coupled with increased retention of a fear memory acquired prior to enrichment, this suggested that neurogenesis may be required to clear outdated memory traces, and that the absence of PSl prevented this from occurring (Feng et al., 2001). A subsequent study showed that a knockout of both PSl and 2 isoforms, both of which are implicated in familial Alzheimer disease, creates an impairment of hippocampal-dependent memory that progressively worsens with age, paralleling the changes observed in the human condition (Saura et al., 2004). G.

Extrasynaptic Modulation of Memory Storage

Although the majority of memory research reflects the dogma that memory is encoded by changes at the synapse, a number of recent studies have investigated the contribution of postsynaptic mechanisms at extrasynaptic sites. A recent study examined the effect of a forebrainrestricted knockout of a hyperpolarization-activated, cyclic-nucleotide gated, cation nonselective channel (HCNl). These channels exhibit an expression gradient along the apical dendrites of pyramidal cells, increasing with distance from the soma. H C N l generates hyperpolarization-activated inward currents, which serve to decrease the amplitude and duration of distal synaptic potentials. In doing so, they act as a brake to reduce dendritic depolarization, and are able to integrate multiple inputs arriving at the postsynaptic pyramidal cell. H C N l forebrain knockouts exhibited an improvement of both short- and long-term components of spatial memory, without affecting contextual or cued fear

175

memory performance. Furthermore, they displayed an increased power of theta frequency oscillations and enhanced LTP at the direct perforant path inputs arising from entorhinal cortex layer III (required for the representation of spatial information by CAl neurons; Brun et al., 2002), while Schaffer collateral LTP remained unaffected. The specificity of this effect is likely due to the termination of these paths at different regions of the apical dendrite — while the direct perforant path innervates distal regions where HCNl expression is greatest, Schaffer collaterals terminate in more proximal locations (Nolan et al., 2004). Interestingly, this effect differs from that observed with a global H C N l knockout, which exhibits deficits in motor learning due to absence of H C N l from cerebellar Purkinje cells (Nolan et al., 2003). This is postulated to be due to differences in electrophysiological properties of these cells. In contrast to Purkinje neurons, pyramidal cells are not spontaneously active in the absence of synaptic input, allowing a significant fraction of the hyperpolarization-activated inward current to be active at rest, and thus contribute to the integration of postsynaptic potentials. A number of studies have examined the role of neuronal excitability in influencing synaptic plasticity, and hence memory. Decreased neuronal excitability has been linked to the memory impairment associated with aging, and is believed to be due to an increase in intracellular free Ca^+ in response to synaptic stimulation, resulting in an increase in Ca^+-activated potassium currents. These currents mediate the slow afterhyperpolarization (sAHP) that follows action potential bursts in pyramidal neurons. It is thought that the age-related increase in induction threshold for synaptic plasticity may reflect the changes in sAHP. To examine this hypothesis, mice were developed with a deletion of the KvjSl.l subunit of voltage-gated K+ channels (Giese et al., 1998). Presence of this subunit causes the channel to exhibit fast (A-type) inactivation, where it would otherwise be noninactivating. Due to the relatively restricted distribution of this subunit, it was possible to use a global knockout yet cause only specific alterations in phenotype, an important consideration as ablation of A-type K+ channels can lead to epilepsy (Gandolfo et al., 1989). As well as leading to an increase in neuronal excitability (as measured by a reduction in sAHP) in the absence of an alteration in synaptic plasticity, impaired performance was observed in two hippocampal tasks — the Morris water maze and social transmission of food preference. In contrast, a later analysis of aged animals demonstrated a more ready induction

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

176

13. GENETIC MANIPULATION OF LEARNING AND MEMORY

of LTP, coupled with improved performance on the Morris water maze (Murphy et al., 2004). It is important to note that this difference was only observed in a genetic background where the mutants exhibited increased excitability, highlighting the importance of careful examination of manipulations in multiple backgrounds before drawing conclusions regarding the role of specific genes in memory

IV

CONCLUSION

The use of genetic manipulations has revealed much regarding the molecular nature of learning and memory However, the focus in the field remains on transgenics, while gene therapy remains a relatively under-utilized approach. The success of gene transfer studies conducted to date show that this is a viable method by which to continue investigations. While transgenic animals are important in facilitating our understanding of memory processes, gene therapy is the only viable strategy for clinical treatment of memory dysfunction. Our understanding of learning and memory processes is by no means complete. Research to date has focused on the hippocampally mediated forms of memory, whereas the involvement of other regions such as the prefrontal cortex remains under-explored. Furthermore, additional work is required to improve spatial and temporal control of gene expression, as well as develop neuron and circuit-specific targeting techniques. The recently developed rAAV/CRE loxP system, with the ability to selectively target preand postsynaptic terminals is a powerful approach with the potential to yield many interesting findings (Scammell et al., 2003). Translation of these findings to the clinic will require a number of issues to be addressed. The most successful strategies for memory enhancement have arisen from situations in which memory is already compromised, and initial treatments will likely be for the aged or impaired brain. However, the ability to enhance mnemonic function in phenotypically normal animals is intriguing from both evolutionary and ethical perspectives, and will provide the subject for much future debate. References Abel, T. and Kandel, E. (1998) Positive and negative regulatory mechanisms that mediate long-term memory storage. Brain Res. Brain Res. Rev, 26: 360-378.

Alarcon, J.M., Malleret, G., Touzani, K., Vronskaya, S., Ishii, S., Kandel, E.R. and Barco, A. (2004) Chromatin acetylation, memory, and LTP are impaired in CBP+ / - mice: a model for the cognitive deficit in Rubinstein-Taybi s)mdrome and its amelioration. Neuron, 42: 947-959. Anagnostaras, S.G., Murphy G.G., Hamilton, S.E., Mitchell, S.L., Rahnama, N.P., Nathanson, N.M. and Silva, A.J. (2003) Selective cognitive dysfunction in acetylcholine Ml muscarinic receptor mutant mice. Nat. Neurosci., 6: 51-58. Bach, M.E., Hawkins, R.D., Osman, M., Kandel, E.R. and Mayford, M. (1995) Impairment of spatial but not contextual memory in CaMKII mutant mice with a selective loss of hippocampal LTP in the range of the theta frequency Cell, 81: 905-915. Balschun, D., Wolfer, D.R, Gass, R, Mantamadiotis, T., Welzl, H., Schutz, G., Prey, J.U. and Lipp, H.R (2003) Does cAMP response element-binding protein have a pivotal role in hippocampal synaptic plasticity and hippocampus-dependent memory? J. Neurosci., 23: 6304-6314. Barco, A., Alarcon, J.M. and Kandel, E.R. (2002) Expression of constitutively active CREB protein facilitates the late phase of long-term potentiation by enhancing synaptic capture. Cell, 108: 689-703. Beattie, E.G., Carroll, R.C., Yu, X., Morishita, W. and Yasuda, H. (2000) Regulation of AMPA receptor endocytosis by a signaling mechanism shared with LTD. Nat. Neurosci., 3: 1291-1300. Blendy J.A., Kaestner, K.H., Schmid, W., Gass, R and Schutz, G. (1996) Targeting of the CREB gene leads to up-regulation of a novel CREB mRNA isoform. Embo J., 15:1098-1106. Blitzer, R.D., Wong, T, Nouranifar, R., Iyengar, R. and Landau, E.M. (1995) Postsynaptic cAMP pathway gates early LTP in hippocampal CAl region. Neuron, 15: 1403-1414. Bontempi, B., Laurent-Demir, C , Destrade, C. and Jaffard, R. (1999) Time-dependent reorganization of brain circuitry underlying long-term memory storage. Nature, 400: 671-675. Bourne, H.R. and NicoU, R. (1993) Molecular machines integrate coincident synaptic signals. Cell, 72: 65-75. Bourtchouladze, R., Lidge, R., Catapano, R., Stanley, J., Gossweiler, S., Romashko, D., Scott, R. and Tully, T. (2003) A mouse model of Rubinstein-Taybi syndrome: defective long-term memory is ameliorated by inhibitors of phosphodiesterase 4. Proc. Natl. Acad. Sci. USA, 100:10518-10522. Bourtchuladze, R., Frenguelli, B., Blendy J., Cioffi, D., Schutz, G. and Silva, A.J. (1994) Deficient long-term memory in mice with a targeted mutation of the cAMP- responsive element-binding protein. Cell, 79: 59-68. Brambilla, R., Gnesutta, N., Minichiello, L., White, G., Roylance, A.J., Herron, C.E., Ramsey, M., Wolfer, D.R, Cestari, V, Rossi-Arnaud, C , Grant, S.G., Chapman, PR, Lipp, H.R, Sturani, E. and Klein, R. (1997) A role for the Ras signalling pathway in synaptic transmission and long-term memory. Nature, 390: 281-286. Brun, V.H., Otnass, M.K., Molden, S., Steffenach, H.A., Witter, M.R, Moser, M.B. and Moser, E.I. (2002) Place cells and place recognition maintained by direct entorhinal-hippocampal circuitry Science, 296: 2243-2246. Cao, L., Jiao, X., Zuzga, D.S., Liu, Y, Fong, D.M., Young, D. and During, M.J. (2004) VEGF links hippocampal activity with neurogenesis, learning and memory. Nat. Genet., 36: 827-835. Chen, A., Muzzio, LA., Malleret, G., Bartsch, D., Verbitsky, M., Pavlidis, P., Yonan, A.L., Vronskaya, S., Grody, M.B., Cepeda, I., Gilliam, T.C. and T.C., Kandel, E.R. (2003) Inducible enhancement of memory storage and synaptic plasticity in transgenic mice expressing an inhibitor of ATF4 (CREB-2) and C/EBP proteins. Neuron, 39: 655-669.

[I. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

CONCLUSION Chen, C , Rainnie, D.G., Greene, R.W. and Tonegawa, S. (1994) Abnormal fear response and aggressive behavior in mutant mice deficient for alpha-calcium-calmodulin kinase 11. Science, 266: 291-294. Chrivia, J.C, Kwok, R.R, Lamb, N., Hagiwara, M., Montminy, M.R. and Goodman, R.H. (1993) Phosphorylated CREB binds specifically to the nuclear protein CBR Nature, 365: 855-859. Costa, R.M., Federov, N.B., Kogan, J.H., Murphy, G.G., Stern, J., Ohno, M., Kucherlapati, R., Jacks, T. and Silva, A.J. (2002) Mechanism for the learning deficits in a mouse model of neurofibromatosis type 1. Nature, 415: 526-530. Cui, Z., Wang, H., Tan, Y, Zaia, K.A., Zhang, S. and Tsien, J.Z. (2004) Inducible and reversible NRl knockout reveals crucial role of the NMDA receptor in preserving remote memories in the brain. Neuron, 41: 781-793. Dhaka, A., Costa, R.M., Hu, H., Irvin, D.K., Patel, A., Komblum, H.I., Silva, A.J., O'Dell, T.J. and Colicelli, J. (2003) The RAS effector RINl modulates the formation of aversive memories. J. Neurosci., 23: 748-757. During, M.J., Cao, L., Zuzga, D.S., Francis, J.S., Fitzsimons, H.L., Jiao, X., Bland, R.J., Klugmann, M., Banks, W.A., Drucker, D.J. and Haile, C.N. (2003) Glucagon-like peptide-1 receptor is involved in learning and neuroprotection. Nat. Med., 9:1173-1179. Elgersma, Y., Fedorov, N.B., Ikonen, S., Choi, E.S., Elgersma, M., Carvalho, O.M., Giese, K.R and Silva, A.J. (2002) Inhibitory autophosphorylation of CaMKII controls PSD association, plasticity, and learning. Neuron, 36: 493-505. Elgersma, Y, Sweatt, J.D. and Giese, K.R (2004) Mouse genetic approaches to investigating calcium/calmodulin-dependent protein kinase II function in plasticity and cognition. J. Neurosci., 24: 8410-8415. Feng, R., Rampon, C , Tang, YR, Shrom, D., Jin, J., Kyin, M., Sopher, B., Miller, M.W., Ware, C.B., Martin, G.M., Kim, S.H., Langdon, R.B., Sisodia, S.S. and Tsien, J.Z. (2001) Deficient neurogenesis in forebrain-specific presenilin-1 knockout mice is associated with reduced clearance of hippocampal memory traces. Neuron, 32: 911-926. Fitzsimons, H.L., McKenzie, J.M. and During, M.J. (2001) Insulators coupled to a minimal bidirectional tet cassette for tight regulation of rAAV-mediated gene transfer in the mammalian brain. Gene Then, 8:1675-1681. Fleischmann, A., Hvalby, O., Jensen, V., Strekalova, T., Zacher, C , Layer, L.E., Kvello, A., Reschke, M., Spanagel, R., Sprengel, R. Wagner, E.F. and Gass, P. (2003) Impaired long-term memory and NR2A-type NMDA receptor-dependent synaptic plasticity in mice lacking c-Fos in the CNS. J. Neurosci., 23: 9116-9122. Flexner, L.B. and Flexner, J.B. (1966) Effect of acetoxycycloheximide and of an acetoxycycloheximide-puromycin mixture on cerebral protein synthesis and memory in mice. Proc. Natl. Acad. Sci. USA, 55: 369-374. Frankland, P W , O'Brien, C , Ohno, M., Kirkwood, A. and Silva, A.J. (2001) Alpha-CaMKII-dependent plasticity in the cortex is required for permanent memory. Nature, 411: 309-313. Frankland, P W , Ohno, M., Takahashi, E., Chen, A.R., Costa, R.M., Kushner, S. A. and Silva, A.J. (2003) Pharmacologically regulated induction of silent mutations (PRISM): combined pharmacological and genetic approaches for learning and memory. Neuroscientist, 9: 104-109. Frey, U., Huang, YY and Kandel, E.R. (1993) Effects of cAMP simulate a late stage of LTP in hippocampal CAl neurons. Science, 260:1661-1664. Frey, U. and Morris, R.G. (1997) S5maptic tagging and long-term potentiation. Nature, 385: 533-536.

177

Gandolfo, G., Gottesmann, C , Bidard, J.N. and Lazdunski, M. (1989) Subtypes of K+ channels differentiated by the effect of K+ channel openers upon K+ channel blocker-induced seizures. Brain Res., 495:189-192. Genoux, D., Haditsch, U., Knobloch, M., Michalon, A., Storm, D. and Mansuy, I.M. (2002) Protein phosphatase 1 is a molecular constraint on learning and memory. Nature, 418: 970-975. Giese, K.P, Fedorov, N.B., Filipkowski, R.K. and Silva, A.J. (1998) Autophosphorylation at Thr286 of the alpha calcium-calmodulin kinase II in LTP and learning. Science, 279: 870-873. Giese, K.R, Friedman, E., Telliez, J.B., Fedorov, N.B., Wines, M., Feig, L.A. and Silva, A.J. (2001) Hippocampus-dependent learning and memory is impaired in mice lacking the Ras-guanine-nucleotide releasing factor 1 (Ras-GRFl). Neuropharmacology, 41: 791-800. Giese, K.P, Storm, J.F., Reuter, D., Fedorov, N.B., Shao, L.R., Leicher, T., Pongs, O. and Silva, A.J. (1998) Reduced K+ channel inactivation, spike broadening, and after-hyperpolarization in Kvbetal.l-deficient mice with impaired learning. Learn. Mem., 5: 257-273. Gonzalez, G.A. and Montminy, M.R. (1989) Cyclic AMP stimulates somatostatin gene transcription by phosphorylation of CREB at serine 133. Cell, 59: 675-680. Gossen, M. and Bujard, H. (1992) Tight control of gene expression in mammalian cells by tetracycline-responsive promoters. Proc. Natl. Acad. Sci. USA, 89: 5547-5551. Gossen, M., Freundlieb, S., Bender, G., MuUer, G., Hillen, W and Bujard, H. (1995) Transcriptional activation by tetracyclines in mammalian cells. Science, 268:1766-1769. Hayashi, M.L., Choi, S.Y, Rao, B.S., Jung, H.Y, Lee, H.K., Zhang, D., Chattarji, S., Kirkwood, A. and Tonegawa, S. (2004) Altered cortical synaptic morphology and impaired memory consolidation in forebrain-specific dominant-negative PAK transgenic mice. Neuron, 42: 773-787. Hebb, D.O. (1949) The Organisation of Behaviour, Wiley, New York. Ho, N., Liauw, J.A., Blaeser, R, Wei, R, Hanissian, S., Muglia, L.M., Wozniak, D.F., Nardi, A., Arvin, K.L., Holtzman, D.M., Linden, D.J., Zhuo, M., Muglia, L.J. and Chatila, T.A. (2000) Impaired synaptic plasticity and cAMP response element-binding protein activation in Ca2+/Calmodulin-dependent protein kinase type IV/Gr-deficient mice. J. Neurosci, 20: 6459-6472. Huang, YY and Kandel, E.R. (1994) Recruitment of long-lasting and protein kinase A-dependent long-term potentiation in the CAl region of hippocampus requires repeated tetanization. Learning & Memory, 1:74-82. Huerta, PT., Sun, L.D., Wilson, M.A. and Tonegawa, S. (2000) Formation of temporal memory requires NMDA receptors within CAl pyramidal neurons. Neuron, 25: 473-480. Jones, M.W, Errington, M.L., French, P.J., Fine, A., Bliss, TV., Garel, S., Chamay, P., Bozon, B., Laroche, S. and Davis, S. (2001) A requirement for the immediate early gene Zif268 in the expression of late LTP and long-term memories. Nat. Neurosci., 4: 289-296. Josselyn, S.A., Kida, S. and Silva, A.J. (2004) Inducible repression of CREB function disrupts amygdala-dependent memory. Neurobiol. Learn Mem., 82:159-163. Josselyn, S.A., Shi, C , Carlezon Jr., W.A., Neve, R.L., Nestler, E.J. and Davis, M. (2001) Long-term memory is facilitated by cAMP response element-binding protein overexpression in the amygdala. J. Neurosci., 21: 2404-2412. Kang, H., Sun, L.D., Atkins, CM., Soderling, T.R., Wilson, M.A. and Tonegawa, S. (2001) An important role of neural activitydependent CaMKIV signaling in the consolidation of long-term memory Cell, 106: 771-783.

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

178

13. GENETIC MANIPULATION OF LEARNING AND MEMORY

Kelleher, R.J. Ill, Govindarajan, A., Jung, H.Y., Kang, H. and Tonegawa, S. (2004) Translational control by MAPK signaling in long-term synaptic plasticity and memory. Cell, 116: 467-479. Kempermann, G., Kuhn, H.G. and Gage, F.H. (1997) More hippocampal neurons in adult mice living in an enriched environment. Nature, 386: 493-i95. Kennedy, M.B. (1993) The postsynaptic density. Curr. Opin. NeurobioL, 3: 732-737. Kida, S., Josselyn, S.A., de Ortiz, S.P., Kogan, J.H., Chevere, I., Masushige, S. and Silva, A.J. (2002) CREB required for the stability of new and reactivated fear memories. Nat. Neurosci., 5: 348-355. Kim, S., Mizoguchi, A., Kikuchi, A. and Takai, Y. (1990) Tissue and subcellular distributions of the smg-21/rapl/Krev-l proteins which are partly distinct from those of c-ras p21s. Mol. Cell Biol, 10, 2645-2652. BClugmann, M., Wymond Symes, C , Leichtlein, C.B., Klaussner, B.K., Duniung, J., Fong, D., Young, D. and During, M.J. (2005) AAVmediated hippocampal expression of short and long Homer 1 proteins differentially affect cognition and seizure activity in adult rats. Mol. Cell Neurosci., 28: 347-360. Koponen, E., Voikar, V., Riekki, R., Saarelainen, T., Rauramaa, T., Rauvala, H., Taira, T. and Castren, E. (2004) Transgenic mice overexpressing the full-length neurotrophin receptor trkB exhibit increased activation of the trkB-PLCgarmna pathway, reduced anxiety, and facilitated learning. Mol. Cell Neurosci., 26: 166-181. Korzus, E., Rosenfeld, M.G. and Mayford, M. (2004) CBP histone acetyltransferase activity is a critical component of memory consolidation. Neuron, 42: 961-972. Landgraf, R., Frank, E., Aldag, J.M., Neumann, I.D., Sharer, C.A., Ren, X., Terwilliger, E.F., Niwa, M., Wigger, A. and Young, L.J. (2003) Viral vector-mediated gene transfer of the vole Via vasopressin receptor in the rat septum: improved social discrimination and active social behaviour. Eur. J. Neurosci., 18, 403-411. Levenson, J.M. and Sweatt, J.D. (2005) Epigenetic mechanisms in memory formation. Nat. Rev. Neurosci., 6: 108-118. Lisman, J. (1989) A mechanism for the Hebb and the anti-Hebb processes underlying learning and memory. Proc. Natl. Acad. Sci. USA, 86: 9574-9578. Lisman, J.E. and Goldring, M.A. (1988) Feasibility of long-term storage of graded information by the Ca2+/calmodulin-dependent protein kinase molecules of the postsynaptic density. Proc. Natl. Acad. Sci. USA, 85: 5320-5324. Malleret, G., Haditsch, U., Genoux, D., Jones, M.W., Bliss, T.V, Vanhoose, A.M., Weitlauf, C , Kandel, E.R., Winder, D.G. and Mansuy, I.M. (2001) Inducible and reversible enhancement of learning, memory, and long-term potentiation by genetic inhibition of calcineurin. Cell, 104: 675-686. Mansuy, I.M., Mayford, M., Jacob, B., Kandel, E.R. and Bach, M.E. (1998) Restricted and regulated overexpression reveals calcineurin as a key component in the transition from short-term to longterm memory. Cell, 92: 39-^9. Martin, S.J., Grimwood, PD. and Morris, R.G. (2000) Synaptic plasticity and memory: an evaluation of the hypothesis. Annu. Rev. Neurosci., 23: 649-711. Mayford, M., Bach, M.E., Huang, YY, Wang, L., Hawkins, R.D. and Kandel, E.R. (1996) Control of memory formation through regulated expression of a CaMKII transgene. Science, 274: 1678-1683. Mazzucchelli, C , Vantaggiato, C , Ciamei, A., Fasano, S., Pakhotin, P, Krezel, W, Welzl, H., Wolfer, D.P, Pages, G., Valverde, O., Marowsky, A., Porrazzo, A., Orban, P C , Maldonado, R.,

Ehrengreuber, M.U., Cestari, V, Lipp, H.P, Chapman, PR, Pouyssegur, J. and Brambilla, R. (2002) Knockout of ERKl MAP kinase enhances synaptic plasticity in the striatum and facilitates striatalmediated learning and memory. Neuron, 34: 807-820. Migaud, M., Charlesworth, P., Dempster, M. and Webster, L.C. (1998) Enhanced long-term potentiation and impaired learning in mice with mutant posts)aiaptic density-95 protein. Nature, 396: 433-439. Miller, S., Yasuda, M., Coats, J.K., Jones, Y, Martone, M.E. and Mayford, M. (2002) Disruption of dendritic translation of CaMKIIalpha impairs stabilization of synaptic plasticity and memory consolidation. Neuron, 36: 507-519. Morozov, A., Muzzio, I.A., Bourtchouladze, R., Van-Strien, N., Lapidus, K., Yin, D., Winder, D.G., Adams, J.P, Sweatt, J.D. and Kandel, E.R. (2003) Rapl couples cAMP signaling to a distinct pool of p42/44MAPK regulating excitability, synaptic plasticity, learning, and memory. Neuron, 39: 309-325. Murphy, G.G., Fedorov, N.B., Giese, K.P, Ohno, M., Friedman, E., Chen, R. and Silva, A.J. (2004) Increased neuronal excitability, synaptic plasticity, and learning in aged Kvbetal.l knockout mice. Curr. Biol., 14: 1907-1915. Nakanishi, S. (1992) Molecular diversity of glutamate receptors and implications for brain function. Science, 258: 597-603. Nakazawa, K., Quirk, M.C., Chitwood, R.A., Watanabe, M., Yeckel, M.E, Sun, L.D., Kato, A., Carr, C.A., Johnston, D., Wilson, M.A. and Tonegawa, S. (2002) Requirement for hippocampal CA3 NMDA receptors in associative memory recall. Science, 297: 211-218. Nakazawa, K., Sun, L.D., Quirk, M.C., Rondi-Reig, L., Wilson, M.A. and Tonegawa, S. (2003) Hippocampal CA3 NMDA receptors are crucial for memory acquisition of one-time experience. Neuron, 38: 305-315. Nolan, M.E, Malleret, G., Dudman, J.T., Buhl, D.L., Santoro, B., Gibbs, E., Vronskaya, S., Buzsaki, G., Siegelbaum, S.A., Kandel, E.R. and Morozov, A. (2004) A behavioral role for dendritic integration: HCNl channels constrain spatial memory and plasticity at inputs to distal dendrites of CAl pyramidal neurons. Cell, 119: 719-732. Nolan, M.E, Malleret, G., Lee, K.H., Gibbs, E., Dudman, J.T., Santoro, B., Yin, D., Thompson, R.R, Siegelbaum, S.A., Kandel, E.R. and Morozov, A. (2003) The hyperpolarization-activated H C N l channel is important for motor learning and neuronal integration by cerebellar Purkinje cells. Cell, 115: 551-564. Ogryzko, VV, Schiltz, R.L., Russanova, V, Howard, B.H. and Nakatani, Y (1996) The transcriptional coactivators p300 and CBP are histone acetyltransferases. Cell, 87: 953-959. Oike, Y, Hata, A., Mamiya, T., Kaname, T., Noda, Y, Suzuki, M., Yasue, H., Nabeshima, T., Araki, K. and Yamamura, K. (1999) Truncated CBP protein leads to classical Rubinstein-Taybi syndrome phenotypes in mice: implications for a dominant-negative mechanisms. Human Molecular Genetics, 8: 387-396. Pittenger, C , Huang, YY, Paletzki, R.R, Bourtchouladze, R., Scanlin, H., Vronskaya, S. and Kandel, E.R. (2002) Reversible inhibition of CREB/ATF transcription factors in region CAl of the dorsal hippocampus disrupts hippocampus-dependent spatial memory. Neuron, 34: 447-462. Rattiner, L.M., Davis, M., French, C.T. and Ressler, K.J. (2004) Brainderived neurotrophic factor and tyrosine kinase receptor B involvement in amygdala-dependent fear conditioning. J. Neurosci., 24: 4796-4806. Rich, R.C. and Schulman, H. (1998) Substrate-directed function of calmodulin in autophosphorylation of Ca2+/calmodulin-dependent protein kinase II. J. Biol. Chem., 273: 28424-28429.

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

CONCLUSION Rondi-Reig, L., Libbey, M., Eichenbaum, H. and Tonegawa, S. (2001) CAl-specific N-methyl-D-aspartate receptor knockout mice are deficient in solving a nonspatial transverse patterning task. Proc. Natl. Acad. Sci. USA, 98: 3543-3548. Saura, C.A., Choi, S.Y., Beglopoulos, V., Malkani, S., Zhang, D., Shankaranarayana Rao, B.S., Chattarji, S., Kelleher, R.J. Ill, Kandel, E.R., Duff, K., Kirkwood, A. and Shen, J. (2004) Loss of presenilin function causes impairments of memory and synaptic plasticity followed by age-dependent neurodegeneration. Neuron, 42: 23-36. Scammell, T.E., Arrigoni, E., Thompson, M.A., Ronan, P.J., Saper, C.B. and Greene, R. W. (2003) Focal deletion of the adenosine Al receptor in adult mice using an adeno-associated viral vector. J. Neurosci., 23: 5762-5770. Selcher, J.C, Nekrasova, T., Paylor, R., Landreth, G.E. and Sweatt, J.D. (2001) Mice lacking the ERKl isoform of MAP kinase are unimpaired in emotional learning. Learn Mem., 8:11-19. Shimizu, E., Tang, Y.P, Rampon, C. and Tsien, J.Z. (2000) NMDA receptor-dependent synaptic reinforcement as a crucial process for memory consolidation. Science, 290:1170-1174. Silva, A.J., Paylor, R., Wehner, J.M. and Tonegawa, S. (1992) Impaired spatial learning in alpha-calcium-calmodulin kinase II mutant mice. Science, 257: 206-211. Soderling, T.R. (1996) Structure and regulation of calcium/calmodulin-dependent protein kinases II and IV. Biochim. Biophys. Acta, 1297: 131-138. Tang, Y.R, Shimizu, E., Dube, G.R., Rampon, C , Kerchner, G.A., Zhuo, M., Liu, G. and Tsien, J.Z. (1999) Genetic enhancement of learning and memory in mice. Nature, 401: 63-69. Taubenfeld, S.M., Milekic, M.H., Monti, B. and Alberini, C M . (2001) The consolidation of new but not reactivated memory requires hippocampal C/EBPbeta. Nat. Neurosci. 4: 813-818. Tsien, J.Z., Chen, D.F., Gerber, D., Tom, C , Mercer, E.H., Anderson, D.J., Mayford, M., Kandel, E.R. and Tonegawa, S. (1996) Subregion- and cell type-restricted gene knockout in mouse brain. Cell, 87:1317-1326. Tsien, J.Z., Huerta, PT. and Tonegawa, S. (1996) The essential role of hippocampal CAl NMDA receptor-dependent synaptic plasticity in spatial memory. Cell, 87:1327-1338.

179

Wang, H., Ferguson, G.D., Pineda, V.V., Cundiff, RE. and Storm, D.R. (2004) Overexpression of t5rpe-l adenylyl cyclase in mouse forebrain enhances recognition memory and LTP. Nat. Neurosci., 7: 635-642. Wang, H., Shimizu, E., Tang, Y.R, Cho, M., Kyin, M., Zuo, W , Robinson, D.A., Alaimo, P.J., Zhang, C , Morimoto, H., Zhuo, M., Feng, R., Shokat, K.M. and Tsein, J.Z. (2003). Inducible protein knockout reveals temporal requirement of CaMKII reactivation for memory consolidation in the brain. Proc. Natl. Acad. Sci. USA, 100: 4287-4292. Watkins, J.C, Krogsgaard-Larsen, P. and Honore, T. (1990) Structureactivity relationships in the development of excitatory amino acid receptor agonists and competitive antagonists. Trends Pharmacol. Sci., 11: 25-33. Wei, R, Qui, C.S., Liauw, J., Robinson, D.A., Ho, N., Chatila, T. and Zhuo, M. (2000) Calcium calmodulin-dependent protein kinase IV is required for fear memory. Nat. Neurosci., 5: 573-579. Wittenberg, G. and Tsien, J. (2002). An emerging molecular and cellular framework for memory processing by the hippocampus. Trends Neurosci., 25: 501. Wong, S.T., Athos, J., Figueroa, X.A., Pineda, VV, Schaefer, M.L., Chavkin, C C , Muglia, L.J. and Storm, D.R. (1999) Calciumstimulated adenylyl cyclase activity is critical for hippocampusdependent long-term memory and late phase LTP Neuron, 23: 787-798. Woo, N.H., Abel, T. and Nguyen, RV (2002) Genetic and pharmacological demonstration of a role for cyclic AMP-dependent protein kinase-mediated suppression of protein phosphatases in gating the expression of late LTP Eur. J. Neurosci., 16: 1871-1876. Wu, G.Y, Deisseroth, K. and Tsien, R.W. (2001) Activity-dependent CREB phosphorylation: convergence of a fast, sensitive calmodulin kinase pathway and a slow, less sensitive mitogenactivated protein kinase pathway. Proc. Natl. Acad. Sci. USA, 98: 2808-2813. Zeng, H., Chattarji, S., Barbarosie, M., Rondi-Reig, L., Philpot, B.D., Miyakawa, T., Bear, M.F. and Tonegawa, S. (2001) Forebrainspecific calcineurin knockout selectively impairs bidirectional synaptic plasticity and working/episodic-like memory. Cell, 107: 617-629.

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

C H A P T E R

14 Psychiatric Applications of Viral Vectors Thomas A. Green, Eric], Nestler

Abstract: There is no doubt that viral vectors hold the promise of transforming the field of psychiatry. However, safe and efficacious gene therapy for psychiatric disorders is several years from becoming reality. Regardless, viral vectors are already providing useful information to basic science in the investigation into the molecular determinants of behavior. Viral vectors expressing the transcription factor, cAMP-response-element binding protein (CREB) or a dominant-negative mutant form of CREB (mCREB) affect behavior in rodent models ranging from depression to anxiety to drug dependence. This chapter compares and contrasts four viral vector systems (herpes simplex 1 virus, adeno-associated virus, sindbis virus, and lentivirus) best suited for gene transfer into brain and describes basic strategies where viral-mediated gene transfer has proven useful. The final section outlines novel uses and improvements to vector technology and describes current obstacles preventing viral vectors from becoming clinically viable. Keywords: AAV; HSV; lentivirus; sindbis; CREB; ICER; drug abuse; addiction; depression; anxiety; reward; behaviour

In the past century, some of the most exciting advances in psychiatry, and in medicine, have come from pharmacological therapeutic interventions. Even without knowing the exact mechanisms or pathologies mediating depression, anxiety, and schizophrenia, the use of antidepressants, anxiolytics, and antipsychotic drugs has revolutionized the field of psychiatry. Despite these tremendous advancements, the vast majority of potentially valuable therapeutic targets cannot be manipulated with known pharmaceuticals. For example, out of perhaps 30,000 genes in the human genome, far fewer than 100 have been targeted for pharmaceutical development. Further, factors such as the high cost of pharmaceuticals, their side effects, and the risk of noncompliance are extremely disadvantageous given the protracted nature of most psychiatric conditions. Targeted gene transfer using viral vectors has the potential for further revolutionizing the field of psychiatry. Although it will likely take

Gene Therapy of the Central Nervous System: From Bench to

decades to develop safe and therapeutic gene transfer therapy for human psychiatric conditions, viral-mediated gene transfer is already providing fundamentally novel information for basic science research into the molecular control of behavior in animal models of these illnesses. L

RECOMBINANT VECTORS

In a very basic sense, viruses perform two fundamental tasks: infect a suitable host cell with genetic information and hijack the machinery of the host cell to replicate itself. Because one major hurdle of gene therapy (or equivalent gene transfer studies in animal models) is delivery of the target gene to appropriate cells, viral vectors are attractive vehicles for gene delivery. However, in the course of infection and replication, most wild-type viruses have some degree of

181

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

182

14. PSYCHIATRIC APPLICATIONS OF VIRAL VECTORS

the most promising candidates for future gene therapy in human brain (Table 1).

pathogenicity, and such uncontrolled replication and infection is unacceptable for obvious reasons. Thus, a common strategy for designing viral vectors suitable for gene transfer in animals, and ultimately for gene therapy in humans, is to separate the infection and replication components to reduce toxicity and ensure adequate control. Recombinant vectors are designed such that the target gene(s) are packaged into the viral capsid in lieu of the wild-type virus's replication genes, thus exploiting the virus's natural ability to infect cells yet rendering the engineered pseudo-virion deficient for further replication after infection. A.

1.

The HSV is a very attractive vehicle for gene transfer because of its ability to package and transfer a large segment of DNA. The wild-type HSV genome is over 150 kb and has over 80 genes. An amplicon (plasmid) vector system has been developed such that the HSV origin of replication and packaging/cleavage site are inserted into a plasmid, which targets the expression cassette of the target transgene to be packaged as an HSV pseudo-virion. Interestingly, the plasmid, usually less than 10 kb, is replicated multiple times and packaged as a large series of repeats called a concatemer, thus packaging several copies of the expression cassette in each virion (Spaete and Frenkel, 1982; Stow and McMonagle, 1982; Neve and Geller, 1995; Coopersmith and Neve, 1999; Carlezon et al., 2000). The second striking feature of HSV is the pattern of temporal expression. When infecting trigeminal sensory neurons, wild-type HSVl expresses several immediate/early genes at very high levels for several days but then enters a latent phase where only a few "latent" genes are expressed. The HSV DNA remains episomal indefinitely and can be reactivated periodically to cause orolabial sores. As a viral vector, engineered HSV vectors (including amplicon systems) utilizing strong viral promoters exhibit a similar pattern of high expression for less than a week before being silenced, despite the continued presence of the transgene DNA,

Various Recombinant Viral Vectors

Recombinant viral vector systems have been developed from many different v^ild-type viruses. The ideal viral vector for psychiatric research in animals, and eventually for treatments of humans, must have good neurotropism, cause little or no toxicity, and express the target protein(s) for extended periods of time. Further, adequate procedures must be available to produce highly concentrated and extremely well-purified stocks for injection into discrete brain nuclei. Contamination with debris from cell culture or the presence of large amounts of empty capsids could produce an inflammatory response and damage the very neurons targeted for gene transfer. Thus far, recombinant herpes simplex virus (HSV) and adeno-associated virus (AAV) have been used most extensively in animal studies of the nervous system and would appear to be

TABLE 1

Herpes Simplex^ 1 Virus (HSV)

Viral Vectors used for Neuronal Gene Transfer

HSV

Lentivirus

Sindbis

rAAV

Virus type

Herpes dsDNA enveloped

Parvovirus ssDNA encapsulated

Alphavirus ssRNA enveloped

Retrovirus ssRNA encapsulated

Relationship to host genome Capacity (kb)

Episomal 130

Some random integration 4.7

Episomal 7.5

Integrative (random) 8

Toxicity

Moderate

Minimal

Severe (except for

Minimal

Can express two transgenes?

Yes

Only for very small genes

Yes

Yes

Temporal expression pattern

Usually < 1 week

> 15 months

< 1 week

> 1 year

Used in human trials?

Yes

Yes

No

Yes

Major strength

Large capacity

Long expression

High expression level

Long expression

Major weakness

Short expression

Small capacity

Short expression

Random integration

newest vectors)

Possibly permanent

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

RECOMBINANT VECTORS

and no sores or other evidence of infection ever occur. However, longer neuronal-specific promoters such as the tyrosine hydroxylase and preproenkephalin promoters have lengthened expression up to 2 months or longer (Kaplitt et al., 1994; Jin et al., 1996; Song et al., 1997, 1998), although this is often associated v^ith a loss of titer and infectivity. The lack of long-term stable expression can be very useful when investigating gene mechanisms in animal models, but is one major hurdle preventing HSV from becoming the vector of choice for gene delivery in humans. There are two general approaches to packaging amplicon-based HSV vectors: helper virus systems and helper-free systems. Helper virus systems generally use replication-deficient HSV virus to construct the components and to control replication and packaging of the pseudo-virions (Neve and Cellar, 1995; Coopersmith and Neve, 1999; Carlezon et al., 2000). Using this system, the transgene-containing amplicon plasmid is transfected into permissive cells infected with replication-deficient helper virus. The amplicon, because it contains an HSV origin of replication and a packaging signal, is replicated, concatemerized, and packaged into an HSV capsid constructed by the helper virus. After 3 days, the cells are harvested and a fresh plate of cells is infected with the mixture of amplicon and helper virus (this is termed "passaging"). The amplicon, but not the replication-deficient helper virus, is further amplified by the fresh cells and the virus is passaged once more. After the third round of amplification the virus is purified on a sucrose gradient and the ratio of amplicon to helper is usually about 100:1. The drawback of this technique is that the final product contains small amounts of helper virus that produces a slight inflammatory response in vivo. However, when properly purified, this inflammatory response does not add much toxicity beyond that of the stereotaxic injection itself (see Carlezon et al., 1997). Systems are being developed that do not necessitate active helper virus, but these systems often suffer from decreased titer (Sun et al., 2000; Bowers et al., 2001). The high titer of the helper virus system outweighs the slight increase in toxicity for basic science applications in laboratory animals, but scalable helper-free systems will be necessary for therapeutic use in humans. The most important factor in designing viral vectors for use in the laboratory and the clinic is that of safety. Despite wild-type HSV being extremely pathogenic, engineered HSV vectors are vastly less cytotoxic as a result of removing endogenous HSV genes. However, spontaneous homologous recombination of an engineered virus to produce a replication-competent

183

HSV expressing the transgene is of primary concern when designing viral vectors for human use. Fortunately, because amplicon-based systems are missing the vast majority of the viral sequence of the wild-type HSV, these pseudo-virions are unlikely to revert to replication-competent viruses via recombination. 1.

AAznO'kssociatedi Virus 2 (AAV)

If the defining characteristic of HSV is its high-cargo capacity, the defining characteristic of the AAV is its low capacity. Typically, AAV can package only about 4.7 kb of DNA cargo. For experimental and therapeutic purposes, the entire expression cassette including promoter, insulating sequences, transgene, polyA, and any reporter elements must be contained within this 4.7 kb limit. There are at least eight known serotypes of AAV with each serotype having slightly different preferences for specific cell types; AAV2 is best suited for infecting neurons in the forebrain. The gracile nature of AAV stems from its existence as a "dependovirus" — a virus that can infect a cell, but does not possess all of the necessary genes to propagate itself. As the name suggests, AAV is dependent upon coinfection with adenovirus, and was originally identified as a contaminant of adenovirus stocks. Despite its size disadvantage, AAV is advantageous because it is naturally replication deficient, produces no inflammation response, readily infects neurons, and produces stable transgene expression for long periods of time. Expression is detectable in vivo around 1 week postinfection, with expression peaking around 2-4 weeks and gradually decreasing with time. Significant transgene expression can be detected well beyond 1 year (Xu et al., 2001). One controversial subject regarding AAV is its ability to integrate into the genome of the host cell. Insertional mutagenesis could potentially disrupt genes of infected cells and could possibly be oncogenic. Wildtype AAV preferentially integrates into a specific site on human chromosome 19 called AAVSl (Kotin et al., 1992); however, recombinant AAV does not show preferential integration at AAVSl and may even integrate into active genes (Rivadeneira et al., 1998; Nakai et al., 2003). Although integration has been demonstrated for both wild-type and recombinant AAV, it is likely that the frequency of integration is very low (Nakai et al., 2001). It seems that most rAAV DNA persists episomally as large circular concatemers. Despite the small risk of insertional mutagenesis, rAAVs are generally considered safe (see Tenenbaum et al., 2003 for review of safety concerns). Commercial helper-free systems have been developed to eliminate

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

184

14. PSYCHIATRIC APPLICATIONS OF VIRAL VECTORS

any helper adenovirus contamination (Stratagene Cat# 240071). Most recombinant systems have removed about 95% of the w^ild-type viral sequence, aU but the inverted terminal repeats (ITRs) necessary for packaging of rAAV. Thus, there is very little risk of recombination events producing wild-type or replication-competent rAAV. 3.

Sindbis Virus

Sindbis virus is a member of the alphavirus family that has been studied extensively for use as a treatment for many different kinds of cancer. The virus infects a cell and shuts dov^^n production of endogenous protein for the host cell, thereby increasing expression of viral proteins. Consequently, the expression profile of sindbis virus is very rapid onset, extremely high expression, and then rapid shutdow^n due to cytotoxic insult. Although cytotoxicity may be advantageous for cancer gene therapy, cytotoxicity of this sort w^ould be unacceptable for in vivo use in brain, both in animal models and in humans. As a result, most current studies w^ith sindbis virus apply it to brain slices or cell culture to achieve gene transfer (Dong et al., 2004). A recently modified sindbis vector is missing the gene responsible for shutting dow^n host-cell protein production, thereby increasing transgene expression length (>6 days) and eliminating most if not all toxicity (Jeromin et al., 2003). The vector is still in early stages of development, but now^ has the potential to become a useful tool for in vivo studies in psychiatry. 4.

Lentivirus

Lentivirus stands out among other retroviruses for its unique ability to infect nondividing cells. Lentiviral vectors are being developed from human, simian, and feline immunodeficiency viruses. Although more development of lentiviral vectors is necessary, lentivirus is attractive for in vivo use because it readily infects neurons, produces sustained (if not permanent) expression in brain, and elicits little toxicity (Blomer et al., 1997). Lentiviral vectors have a substantial capacity of about 8 kb that integrates into the host genome. The virulent nature of wrild-type immunodeficiency viruses has caused considerable fear of vectors reverting to replication-competent viruses, but these fears seem to be unfounded w^ith the new^er generation of vectors. Insertional mutagenesis, hovsrever, is a safety issue that has not been adequately studied vsrith these vectors (recall that the insertional profile of rAAV differs from that of wild-type viruses). Despite these concerns, a new modular "kit" style system has recently been developed that now makes lentiviral vector technology accessible to the broader scientific community (Mitta et al., 2002).

Although development lags behind HSV and AAV vectors, lentiviral vectors have already proven useful in altering behavior. Lentiviral vectors can influence the degree of locomotor stimulation subsequent to repeated cocaine administration either by overexpressing the protein CD81 or by expressing small interfering RNAs to block CD81 protein synthesis (Bahi et al., 2004, 2005). A separate group has used a lentiviral vector to show that expression of a dominant negative mutant of the TrkB tyrosine kinase receptor in the amygdala specifically blocked acquisition of fear conditioning without affecting expression of fear or baseline startle (Rattiner et al., 2004).

IL APPLICATION TO THE STUDY OF BEHAVIOR By far the most important (and difficult) factor in using viral gene transfer to study behavior and other in vivo endpoints is the generation of extremely clean preparations of high titer (~1 X 10^/ml). Virus is delivered to discrete brain regions using common stereotactic techniques.

A. HSVs Our laboratory utilizes an amplicon-based system using replication-deficient helper virus to package rHSV in a permissive 2-2 cell line (Neve and Cellar, 1995; Neve et al., 1997). We use several variants of the HSV PrpUC amplicon depending on the needs of the experiment. Transgene expression is high from 1 to 3 days postsurgery, drops by about 50% by day 4, and is nondetectable by day 7 (Carlezon et al., 1997; Barrot et al., 2002). 1. pHSVPrpUC. The basic amplicon is 4.9 kb and contains an HSV origin of replication and packaging site necessary for amplification and packaging into HSV, and an ampicillin-resistance gene to facilitate DNA amplification in bacterial cells for cloning purposes. The transgene expression cassette is driven by an IE4/5 promoter, contains a pUC 19 multiple cloning site, and terminates with an SV-40 polyA (see Fig. 1). This amplicon is best used to express proteins that are easily detected by immunohistochemistry, or those that are epitopetagged or CFP fusion proteins. 2. myc-PrpUC. One technical problem in overexpressing a protein that is also endogenously expressed is that it is difficult to determine by immunohistochemistry, which cells are infected and expressing protein driven by the transgene

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

185

APPLICATION TO THE STUDY OF BEHAVIOR

MCS

SV-40 PolyA

beta-Globin

R-rm

hGH PolyA

FIGURE 1 Viral expression plasmids for HSV and AAV vectors. The target gene of interest is cloned into the multiple-cloning site (MCS) of HSV amplicons (A-C). (A) The basic HSV amplicon contains an HSV origin of replication and packaging site in addition to the expression cassette driven by the HSV immediate/early promoter IE4/5. (B) A myc epitope tag is incorporated into the expression cassette. When cloned "in frame/' this plasmid expresses a protein with an N-terminal myc tag fusion. (C) A bicistronic HSV amplicon incorporates a second expression cassette to express eGFP driven by a CMV promoter. This bicistronic plasmid expresses two proteins, each from a different transcript. For AAV plasmids (D and E), the gene of interest is driven by a CMV promoter. Only the expression cassette (the sequence between the inverted terminal repeats) is packaged into viral capsid. (D) Stratagene's basic AAV expression plasmid can accommodate genes up to 3 kb (Cat# 240071). (E) A bicistronic AAV expression plasmid expresses the gene of interest and a GFP reporter protein from a single transcript using an internal ribosomal entry site (IRES) for translation of GFP (Cat# 240075). All HSV and AAV plasmids contain ampicillin resistance genes to facilitate cloning in bacteria.

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

186

14. PSYCHIATRIC APPLICATIONS OF VIRAL VECTORS

and which cells are merely expressing endogenous protein. Often one can identify infected cells by the very high levels of the expressed product, but often one cannot differentiate conclusively which cells are infected. Further, many (if not most) proteins cannot be visualized via immunohistochemistry due to a lack of effective and selective antibodies. Therefore, a modification of the basic FrpUC amplicon incorporates an N-terminal myc epitope tag that when cloned "in frame" fuses a myc epitope to the expressed protein so that the protein can be readily visualized using antibodies directed at the myc epitope rather than the expressed protein. 3. Bicistronic GFP amplicon. Many electrophysiological and cell culture experiments are designed to investigate properties of a single infected neuron while the cell is still alive. Thus, for viral gene transfer manipulations, one would need to identify infected cells without the aid of immunohistochemistry (e.g., without killing the cell). Traditionally, scientists would express the protein of interest fused to GFP, a protein that can be detected by simple (and noninvasive) fluorescent microscopy. A major drawback to the GFP fusion approach is that the resulting chimeric protein may not function identically to the endogenous protein of interest. Therefore, one must first spend considerable effort to verify the functionality of the GFP-fused protein before the intended experiment can begin. Fiowever, expressing GFP as a separate protein along with the protein of interest would still identify the infected cell without interfering with the function of the protein of interest, and thereby increase throughput by eliminating the time spent verifying function of a chimeric protein. Accordingly, a bicistronic HSV amplicon has been developed that expresses GFP from a second expression cassette driven by a CMV promoter (Clark et al., 2002). One concern of using a bicistronic vector is that expression of the reporter (GFP) does not necessarily mean that the protein of interest is being expressed. For example, the expression cassettes might compete for transcriptional machinery, meaning that one protein might be expressed at the expense of the other. However, in our hands this bicistronic vector coexpresses both proteins at high levels about 95% of the time. Another potential application of the bicistronic vector is that the GFP can be used to examine morphological differences resulting from expression of the protein of interest.

B,

AAV

Our laboratory also utilizes a commercially available AAV helper-free system (Stratagene Cat# 240071). The basic expression cassette is driven by a CMV promoter with a j8-globin intron before the inserted transgene and a human growth hormone polyA. This vector can accommodate transgenes approaching 3 kb in size. For smaller transgenes (<1.7kb) we utilize a plasmid with an expression cassette that includes an internal ribosomal entry site (IRES) to promote translation of a GFP reporter as a separate protein from the same transcript (Stratagene Cat# 240075).

III. EXAMPLE OF THE UTILITY OF VIRAL VECTORS IN PSYCHIATRY: MODULATION OF CRE^DEPENDENT TRANSCRIPTION ALTERS BEHAVIOR We and many other groups have been interested for some time in the role of the transcription factor, cAMP-response-element binding protein (CREB), in the regulation of complex behavior. Over the past 7 years, utilizing viral-mediated overexpression of CREB itself or a dominant negative mutant of CREB (termed mCREB), we and others have demonstrated a dramatic influence of CREB in animal models of several psychiatric conditions, including drug abuse, anxiety, depression, and fear conditioning. CREB is a constitutively expressed basic-region leucine zipper (BZip) transcription factor that is activated upon phosphorylation of Serl33 by protein kinase A, Ca^^/calmodulin-dependent protein kinase IV, or growth factor activated ribosomal S6 kinases (RSKs) (see Mayr and Montminy, 2001, for review). CREB homodimerizes and binds to DNA at CRE sites in promoter/enhancer regions of many genes to increase transcription (Impey et al., 2004). Several environmental factors such as drugs of abuse, stress, and antidepressants increase CRE-mediated transcription in a regionspecific manner in brain (Thome et al., 2000; Barrot et al., 2002; Shaw-Lutchman et al., 2003). mCREB is a single amino acid substitution mutant of CREB where Serl33 is mutated to Ala. mCREB can dimerize with CREB and bind to DNA, but the dimer is incapable of activation. As a result, mCREB blocks transcription at CRE sites in a dominant-negative fashion. The nucleus accumbens is one terminal region of the mesolimbic dopamine pathway and is important in motivation and reward (Robinson and Berridge, 1993). Most drugs of abuse evoke a robust release of dopamine

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

EXAMPLE OF THE UTILITY OF VIRAL VECTORS IN PSYCHIATRY: MODULATION OF ORE-DEPENDENT TRANSCRIPTION ALTERS BEHAVIOR

in the nucleus accumbens and other terminal fields of ascending dopaminergic projections (Wise and Bozarth, 1987). As mentioned above, drugs of abuse also increase CRE-mediated transcription in the nucleus accumbens (Barrot et al., 2002; Shaw-Lutchman et al., 2002), likely as a result of dopaminergic neurotransmission. Viralmediated gene transfer experiments using HSV vectors expressing w^ild-type CREB or mCREB are providing evidence as to the function of CREB in response to drugs of abuse. Overexpressing CREB in the nucleus accumbens of rats decreases the rev^^arding effects of cocaine or morphine in the conditioned place preference paradigm, suggesting that CREB acts as an inhibitory feedback mechanism to oppose the drug effect (Carlezon et al., 1998; Barrot et al., 2002). Conversely decreasing CRE-mediated transcription by overexpressing mCREB increases the rewarding effect of these drugs. When not being hijacked by drugs of abuse, the mesolimbic dopamine system functions to regulate natural revsrard. Accordingly, a similar viral gene transfer experiment show^ed that overexpression of CREB in the nucleus accumbens also decreases, and mCREB increases, natural reward as measured by preference for a sucrose solution over water (Barrot et al., 2002). Interestingly, CREB and mCREB overexpression in the nucleus accumbens also affect behavioral responses to aversive stimuli. CREB overexpression decreases and mCREB increases naloxone-induced place aversion (Barrot et al., 2002). In tests of anxiety, mCREB decreases time spent on open arms of an elevated plus maze and time in the center of an open field, both evidence of increased anxiety (Barrot et al., 2002). Conversely, overexpression of CREB decreases anxiety in the open field (unpublished data). The fact that CREB decreases behavioral responses to negative as well as positive stimuli suggests that CREB functions not to decrease reward specifically, but to blunt responses to all emotional stimuli regardless of their valence (Barrot et al., 2002). Depression is a complex psychiatric disorder with a multitude of symptoms; one of the most common symptoms is anhedonia. While the neural substrate for anhedonia is not known, it makes sense that the mesolimbic dopamine system is involved (Nestler et al., 2002). As described above, CREB overexpression in the nucleus accumbens decreases sucrose preference and mCREB increases preference (Barrot et a l , 2002). Therefore, mCREB overexpression in the accumbens produces an antidepressant-like effect with regard to anhedonic symptoms. Other experiments not clearly related to hedonia, also suggest mCREB acts as an antidepressant; overexpression of mCREB in the nucleus accumbens

187

produces an antidepressant-like behavioral phenotype in the forced swim test and the learned helplessness paradigm (Pliakas et al., 2001; Newton et al., 2002). Viral gene transfer technology is also now being used to provide mechanistic information as to how CREB manipulation in the nucleus accumbens alters behavior. Recent studies show that CREB increases and mCREB decreases the excitability of nucleus accumbens neurons (Dong et al., 2004). These studies utilized sindbis vectors expressing a GFP-fused constitutively active mutant CREB (VP16-CREB) or mCREB to infect brain slices taken from rats. In electrophysiology experiments, GFP-positive neurons expressing VP16-CREB exhibit increased firing at a broad range of injected currents; those expressing mCREB show the opposite effect. These electrophysiology experiments provide insight into how CREB may gate emotional responses. It is well known that drugs of abuse, natural rewards, and stressful stimuli all release dopamine in the mesolimbic dopamine pathway and that dopamine, among other actions, affects a subpopulation of nucleus accumbens neurons expressing inhibitory D2 receptors (Wise and Bozarth, 1987; Robinson et al., 2001; Stevenson et al., 2003). These neurons show increased CRE-mediated transcription in response to drugs of abuse or stress (Barrot et al., 2002; Shaw-Lutchman et al., 2002). We also know that the increase in CRE-mediated transcription causes these neurons to be more excitable (Dong et al., 2004) and decreases responses to subsequent dopamine-releasing stimuli. This negative feedback of dopamine signaling is one possible explanation for why drug addicts, who repeatedly bombard their dopamine system with drug, become less responsive to hedonic stimuli (Uslaner et a l , 1999). The studies above expressed a mutant form of CREB to disrupt CRE-dependent transcription, but more recent studies are investigating the effects of an endogenous protein named inducible cAMP early repressor (ICER) on behavior. ICER is an alternate protein product of the cAMP-response-element modulator (CREM) gene and is driven by an intronic promoter with four CRE sites in tandem, making ICER highly inducible by increases in CRE-mediated transcription. ICER has a basic region and a leucine zipper domain similar to CREB, but lacks activation domains. ICER therefore dimerizes and binds to CRE sites and blocks transcription. An HSV vector expressing ICER causes behavioral effects very similar to those observed with mCREB. HSV-mediated overexpression of ICER in the nucleus accumbens increases natural reward (sucrose preference and social grooming), anxiety (elevated plus maze and sucrose neophobia), and has antidepressant-like effects in the forced

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

188

14. PSYCHIATRIC APPLICATIONS OF VIRAL VECTORS

swim test (Green et al., 2004). Thus, ICER is a natural antagonist of CREB function in the nucleus accumbens and serves to increase behavioral responses to emotional stimuli. It is important to note that, as transcription factors, CREB and ICER presumably affect neuronal firing by increasing and decreasing, respectively, the transcription of a protein (or proteins) that affects neuronal excitability. One possible candidate is brain-derived neurotrophic factor (BDNF), a known target of CREB that has been implicated in psychiatric conditions such as depression and addiction (Mayr and Montminy, 2001; Sen et al., 2003; Bolanos and Nestler, 2004; Matsushita et al., 2004). A viral gene transfer approach with AAV was used to express the high-affinity BDNF receptor TrkB or its dominant negative in the nucleus accumbens to investigate the role of BDNF signaling in the forced swim test (Eisch et al., 2003). The results of the experiment showed that blocking BDNF signaling with the dominant negative TrkB receptor produces an antidepressant phenotype similar to that seen with HSV overexpression of mCREB or ICER, consistent with the notion that blocking CRE-mediated transcription produces antidepressant-like effects in animal models. Together, these studies demonstrate the unique utility of viral vectors to gain new knowledge of the pathophysiology of diverse types of psychiatric conditions in animals. It is interesting to speculate on potential clinical applications of this technology. For example, new insight into disease pathophysiology can be used to target novel proteins (CREB or ICER themselves or any of the many genes they regulate) for traditional drug development or to even consider potential gene therapy for these conditions. Clearly, however, gene therapy for psychiatry disorders is still a long way off in the future. A.

On the Horizon

As our knowledge of neurobiology and virology increase, we will have much more flexibility in how we can study and treat psychiatric conditions. As discussed above, there are several different viral vector systems being developed based on a number of different kinds of wild-type viruses. Each vector design has inherent strengths and weaknesses that can be exploited for gene therapy.

IV.

NOVEL USES FOR VIRAL VECTORS

To date, most studies aimed at reducing the function of a protein within a given brain region have

overexpressed a dominant-negative mutant. While this approach has yielded important findings, blocking expression of the endogenous protein would be preferable and less prone to artifact inherent in overexpression systems. New research into RNA interference technology is being used widely in cell culture to selectively decrease expression of a particular protein. This technology is rapidly being adapted for use with viral vectors and may one day become an important experimental or even a therapeutic tool to disrupt the expression of a protein in vivo (Hommel et al., 2003; Abrams et al., 2004; Ventura et al., 2004; Xia et al., 2004; Bahi et al., 2005). One advantage of RNAi over other methods is that specific splice variants can be targeted for knockdown rather than suppressing all products of the gene. Traditional methods for studying the knockout of loxP-flanked genes use transgenic driver lines to express Cre recombinase. However, uneven or chimeric Cre expression in driver lines, lack of specific focal expression, and the time and expense of developing new driver lines are limitations that have hampered the widespread utility of this knockout technology for behavioral research. Viral vectors can be used to express Cre recombinase with great focal precision in the fully differentiated adult brain (Scammell et al., 2003; Ahmed et a l , 2004). Even more exciting is the fact that a single viral vector expressing Cre recombinase can be used in any brain region without development, breeding, and validation of new driver lines. A. Obstacles to a Clinically Viable Gene Therapy for Psychiatric Disorders 1.

Finding Qood Targets

Psychiatric conditions are complex to diagnose due to the fact that behavior is so variable, and categorizing behavioral syndromes is often subjective. The Diagnostic and Statistical Manual for psychiatrists has only general criteria and loose descriptions of psychiatric conditions. For example, the criteria for clinical depression include a list of symptoms, with only five of nine symptoms necessary for a diagnosis of major depressive disorder. Further, the symptoms listed are broadly defined, such as either an increase or a decrease in appetite. To complicate matters further, it is almost certain that there are multiple subtypes of depression, each likely with its own etiology and best course of treatment (see Nestler et al., 2002). Even with a proper diagnosis, basic science has not yet provided knowledge of the mechanisms responsible for behavior as complex as psychiatric disorders.

HI. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

NOVEL USES FOR VIRAL VECTORS

One must keep in mind that altering one behavior often affects many others. For example, the experiments described above suggest that decreasing CREB function in the nucleus accumbens, e.g., by mCREB or ICER gene therapy, might be a useful treatment for depression. Viral-mediated overexpression of these proteins not only had an antidepressant-like effect in the forced swim and other tests, but also increased natural reward, an action that would be therapeutic for the anhedonia often associated with depression. However, these viruses also increase anxiety and sensitivity to aversive stimuli at the same time. It is hoped that studies like these will provide specific information about how CREB controls behavior, leading eventually to more specific targets. It is possible, for instance, that inhibition of a specific target gene of CREB might exert most of the beneficial consequences of CREB inhibition without the negative. 2.

Designing Better Vectors

The advancements in viral vector technology over the past decade are remarkable, and is providing scientists with an increasing number of vector choices. The bulk of future advancement will be to incorporate some degree of finesse into viral vectors. Currently, vectors infect whatever neurons are within reach, express at extremely high levels for a period of time, and are usually silenced to some degree. While these features make the vectors useful for many basic science studies, the goal for psychiatric gene therapy applications is sustained, controlled expression at appreciable levels only in specific neuronal subtypes at specified times. More development is necessary, but there are several strategies that are being employed in designing expression cassettes more appropriate for gene therapy (see Papadakis et al., 2004, for review). 3.

Controlling Expression

Even well-defined nuclei of the brain consist of a heterogeneous population of neuronal subtypes. As a hypothetical example, it is possible that a therapy would need to specifically target D2-receptor/enkephalin-expressing neurons in nucleus accumbens with an HSV vector, but not infect the Dl-receptor/dynorphinexpressing neurons interspersed among them. Current vectors do not discriminate between the two. Eliminating ectopic expression of this sort can be accomplished either by targeting infection only to the specific neuron or by infecting all neurons, but confining expression to the specified neuron via a subtype-specific promoter. Targeting infection to specific cell types can be achieved by altering envelope or capsid proteins of the viral

189

vector. Wild-type viruses target cells by binding to specific proteins at the cell surface. An engineered HSV vector targeting specific neurons must first be relieved of its natural tropism by removal of these protein-binding domains from envelope proteins and replacing them with protein binding domains specific for the cell being targeted. In the example above, the specific targeting of D2-enkephalin neurons might be achieved by incorporation of a domain into the envelope proteins that can bind to the extracellular N-terminal tail of the D2 receptor. Although this is a hypothetical example, the technology for targeting of this sort has been demonstrated (Grandi et al., 2004). An advantage of this approach is that the specificity of the virus is engineered into the helper system rather than the amplicon itself and, once a vector has been validated, any suitable transgene-containing amplicon can be targeted to that cell type without the need for additional cloning or design. Further, once separate vectors have been engineered for several different cell subtypes, a single transgene amplicon can be used to target any cell subtype simply by choosing the appropriate helper. While the technology of targeted infection is still in early stages of development, considerable progress has been made in utilizing neural promoters to restrict the expression of a viral-encoded transgene to a particular neuronal cell type. The success of this approach has been demonstrated using large endogenous promoters in HSV vectors (Kaplitt et al., 1994; Jin et al., 1996). However, concerns remain about the impact of such promoters on viral titer and infectivity. In addition to controlling which cells express a transgene, temporal control of transgene expression is a much needed advance in vector technology. One approach is use of a tetracycline (tet)-inducible system, where transgene expression can be turned on and off by administering very low concentrations of doxycycline (below bactericidal concentrations). A tet-inducible lentiviral vector has already proven useful in altering behavioral responses to cocaine (Bahi et al., 2004,2005). Inducible AAVs of this sort are also being developed, but efforts are severely hampered by the size limitation of AAV (McGee et al., 2001; Chtarto et al., 2003). Temporal regulation of this sort could one day be important for cyclic psychiatric conditions such as bipolar disorder where expression of the therapeutic transgene may be advantageous during one phase of the cycle and disadvantageous at another. Even greater temporal control could conceivably be achieved in the future by designing expression cassettes that are induced by derangements in intracellular stimuli related to psychiatric illnesses rather than pharmaceutical induction.

IIL PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

190

14. PSYCHIATRIC APPLICATIONS OF VIRAL VECTORS

This would alleviate issues regarding the difficulty of detecting neuronal imbalances and, more importantly, would eliminate issues of compliance so important to episodic conditions such as bipolar dis-order or schizophrenia. Thus far, most viral vectors used for the study of behavior have been monocistronic or bicistronic (expressing GFP for easy identification of infected cells). However, it may well be, given the complex nature of psychiatric conditions, that an effective gene therapy would need to express two or more therapeutic transgenes. The large capacity of the HSV vector is ideal for such polycistronic expression (the wildtype HSV has over 80 genes). Many strategies have been developed for expressing multiple proteins in viral vectors, including multiple transcript approaches, internal translation initiation sequences, and self-cleaving fusion proteins (see de Felipe, 2002, for review). Other approaches are being developed to allow for greater control of expression in viral vectors. One such approach is to employ hybrid vectors. For example, flanking the expression cassette of an HSV amplicon with inverted terminal repeats of an AAV (e.g., the elements flanking the expression cassette of AAVs) allows the resultant amplicon to replicate and package into an HSV capsid but exhibits the extended temporal expression pattern of the AAV vector (Wang et al., 2002; Heister et al., 2002). 4.

Safety Issues

The most important obstacle for a clinically viable gene-based treatment is that of safety. Although viral vectors appear to be quite safe, at least in some settings, the field of virology is still striving to understand the fundamentals of wild-type and engineered vectors. One fear, mentioned earlier, is that of homologous recombination of the vector with wild-type or helper elements to produce replication-competent vectors expressing the engineered transgene. This problem, however, has been greatly reduced by removing as much wild-type viral sequence from the vector as possible. The best viral vectors would be those that only contain the transgene cassette(s) and the viral elements necessary for packaging (and integration if necessary). For example, the current generation of rAAV vectors contains only the expression cassette and the inverted terminal repeats of the AAV. The inverted terminal repeats of the vector comprise only 5% of the wildtype AAV genome. Further, the helper elements do not contain the inverted terminal repeats and, because the helper and expression elements contain no common sequence, homologous recombination is impossible.

One safety concern not often mentioned is that of vector mobilization. Wild-type viruses contain all of the genes necessary for replication and packaging, so if a viral vector is used on a patient already infected with wild-type virus, the wild-type virus can act in effect as a "helper" virus to propagate the replication deficient vector in the patient. This may seem trivial, but this is the exact same phenomenon that propagates wild-type AAVs, which are naturally replication deficient unless infecting coincidentally with adenovirus. In effect, our replication-deficient viral vectors are nothing more than engineered "dependoviruses" that cannot replicate but can be amplified and packaged with wild-type viruses. Recombinant AAV vectors, however, are double-dependent on coinfection with both wild-type AAV and adenovirus. Viral vector mobilization is a potentially serious issue due to the fact that herpes and AAV are common in the general population. Another important current safety issue is that of insertional mutagenesis from random insertion of the transgene cassette into the host genome. Although HSV vectors are thought to be relatively immune from insertion, AAV, lentivirus and HSV/AAV hybrids likely have some risk associated with insertion. One way to limit this liability would be to engineer these vectors to remain episomal. This might be a viable option for AAV given that a very small percentage integrates into the genome (Nakai et al., 2001). However, not enough is known about lentivirus vectors to know if sustained transgene expression can be achieved without insertion. Another strategy for improving the safety of vectors that insert into the genome is to target insertion to areas of the genome where insertion will not disrupt endogenous genes. As discussed above, wild-type AAV integrates somewhat preferentially to a specific region (AAVSl) of human chromosome 19, but rAAV vectors integrate randomly into the host genome (Kotin et al., 1992; Rivadeneira et al., 1998). Recently, hybrid HSV/AAV vectors have been designed to coexpress the r ^ genes of the wild-type AAV thought to mediate targeted insertion of wild-type AAV, thereby causing the hybrid vector to insert the transgene into the AAVSl region (Heister et al., 2002; Wang et al., 2002). To eliminate unnecessary integration of the rep genes into the host genome, the rep genes were placed outside the ITRs and did not integrate into the genome. Although these hybrid vectors did promote site-specific integration, some random integration still persisted, suggesting that the hybrid vectors still retain some danger of insertional mutagenesis. Thus far, all of the technology described has been based on strategies currently in development and

HI. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

ACKNOWLEDGMENT

proven, at least in concept, to be viable solutions for improving current vectors. However, one current issue regarding safety of viral vectors for use as gene therapies for psychiatric conditions remains in the realm of science fiction — that of noninvasive targeting. Current strategies require precise intracranial stereotaxic delivery of vectors to specific brain regions to produce changes in behavior. Any such delivery to human patients would currently require drilling through the skull and threading a cannula to the specific brain region. An invasive technique of this sort would only be acceptable for the most serious of conditions, and only for those conditions that cannot be treated pharmacologically. This issue will likely be the limiting factor for the widespread use of viral vectors in psychiatric gene therapy In fact, it is commonly thought that the blood-brain barrier evolved specifically to block access of microorganisms and other blood-borne pathogens to the most irreplaceable cells of the body — neurons of the central nervous system.

V.

CONCLUSIONS

Although gene therapy for psychiatric conditions is still many years from being a reality, current experiments with animals have proven the concept of gene therapy to influence behavior and to even correct abnormalities in behavior caused by genetic mutations or environmental exposures (e.g., drugs of abuse, stress, etc.). In addition to being an invaluable tool in basic neuroscience research in psychiatry, continued development of vector technology may someday provide a powerful alternative to pharmacotherapy for serious mental disorders.

ACKNOWLEDGMENT This work was supported by grants from the National Institute of Mental Health and National Institute on Drug Abuse. Financial disclosure: EJN serves on the Scientific Advisory Board of Neurologix, Inc., which is working toward the development of viral gene therapy in humans. References Abrams, M.T., Robertson, N.M., Yoon, K. and Wickstrom, E. (2004) Inhibition of glucocorticoid-induced apoptosis by targeting the

191

major splice variants of BIM mRNA with small interfering RNA and short hairpin RNA. J. Biol. Chem., 279: 55809-55817. Ahmed, B.Y., Chakravarthy, S., Eggers, R., Hermens, W.T., Zhang, J.Y., Niclou, S.R, Levelt, C , Sablitzky, R, Anderson, P.N., Lieberman, A.R. and Verhaagen, J. (2004) Efficient delivery of Cre-recombinase to neurons in vivo and stable transduction of neurons using adeno-associated and lentiviral vectors. BMC Neurosci., 5:4. Bahi, A., Boyer, R, Kafri, T. and Dreyer, D. L. (2004) CD81-induced behavioural changes during chronic cocaine administration: in vivo gene delivery with regulatable lentivirus. Eur. J. Neurosci., 19(6): 1621-1633. Bahi, A., Boyer, R, Kolira, M. and Dreyer, D.L. (2005) In vivo gene silencing of CD81 by lentiviral expression of small interference RNAs suppresses cocaine-induced behavior. J. Neurochem., 92(5): 1243-1255. Barrot, M., Olivier, J.D., Perrotti, L.L, DiLeone, R.J., Berton, O., Eisch, A.J., Impey, S., Storm, D.R., Neve, R.L., Yin, J.C, Zachariou, V. and Nestler, E.J. (2002) CREB activity in the nucleus accumbens shell controls gating of behavioral responses to emotional stimuli. Proc. Natl. Acad. Sci. USA, 99:11435-11440. Blomer, U., Naldini, L., Kafri, T., Trono, D., Verma, I.M. and Gage, RH. (1997) Highly efficient and sustained gene transfer in adult neurons with a lentivirus vector. J. Virol., 71: 6641-6649. Bolanos, C.A. and Nestler, E.J. (2004) Neurotrophic mechanisms in drug addiction. J. Neuromol. Med., 5: 69-83. Bowers, W.J., Howard, D.R, Brooks, A.I., Halterman, M.W. and Federoff, H.J. (2001) Expression of vhs and VP16 during HSV-1 helper virus-free amplicon packaging enhances titers. Gene Ther., 8:111-120. Carlezon, W.A. Jr., Boundy, V.A., Haile, C.N., Lane, S.B., Kalb, R.G., Neve, R.L. and Nestler, E.J. (1997) Sensitization to morphine induced by viral-mediated gene transfer. Science, 277: 812-814. Carlezon, W.A. Jr., Nestler, E.J. and Neve, R.L. (2000) Herpes simplex virus-mediated gene transfer as a tool for neuropsychiatric research. Grit. Rev. Neurobiol., 14: 47-67. Carlezon, W.A. Jr., Thome, J., Olson, V.G., Lane-Ladd, S.B., Brodkin, E.S., Hiroi, N., Duman, R.S., Neve, R.L. and Nestler, E.J. (1998) Regulation of cocaine reward by CREB. Science, 282: 2272-2275. Chtarto, A., Bender, H.U., Hanemann, C O . , Kemp, T., Lehtonen, E., Levivier, M., Brotchi, J., Velu, T. and Tenenbaum, L. (2003) Tetracycline-inducible transgene expression mediated by a single AAV vector. Gene Ther., 10: 84-94. Clark, M.S., Sexton, T.J., McClain, M., Root, D., Kohen, R. and Neumaier, J.F. (2002) Overexpression of 5-HTlB receptor in dorsal raphe nucleus using Herpes Simplex Virus gene transfer increases anxiety behavior after inescapable stress. J. Neurosci., 22: 4550^562. Coopersmith, R. and Neve, R.L. (1999) Expression of multiple proteins within single primary cortical neurons using a replication deficient HSV vector. Biotechniques, 27:1156-1160. de Felipe, P. (2002) Polycistronic viral vectors. Curr. Gene Ther., 2: 355-378. Dong, Y, Saal, D., Marie, H., Xu, W., Nestler, E. J. and Malenka, R.C. (2004) CREB-mediated regulation of excitability and synaptic transmission in nucleus accumbens neurons. Program No. 577.2. 2004 Abstract Viewer/Itinerary plaimer. Society for Neuroscience, Washington, DC. Eisch, A.J., Bolanos, C.A., de Wit, J., Simonak, R.D., Pudiak, CM., Barrot, M., Verhaagen, J. and Nestler, E.J. (2003) Brain-derived neurotrophic factor in the ventral midbrain-nucleus accumbens pathway: a role in depression. Biol. Psychiatry, 54: 994-1005.

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

192

14. PSYCHIATRIC APPLICATIONS OF VIRAL VECTORS

Grandi, P., Spear, M., Breakefield, X.O. and Wang, S. (2004) Targeting HSV amplicon vectors. Methods, 33:179-186. Green, T.A., Hommel, J.D., DiLeone, R.J., Alibhai, I.N., Neve, R.L. and Nestler, E.J. (2004) Viral-mediated overexpression of ICER in the nucleus accumbens increases behavioral responses to emotional stimuli. Program No. 574.14. 2004 Abstract Viewer/Itinerary planner. Society for Neuroscience, Washington, DC. Heister, T., Heid, I., Ackermann, M. and Fraefel, C. (2002) Herpes simplex virus type 1/adeno-associated virus hybrid vectors mediate site-specific integration at the adeno-associated virus preintegration site, AAVSl, on human chromosome 19. J. Virol., 76: 7163-7173. Hommel, J.D., Sears, R.M., Georgescu, D., Simmons, D.L. and DiLeone, R.J. (2003) Local gene knockdown in the brain using viral-mediated RNA interference. Nat. Med., 9: 1539-1544. Impey S., McCorkle, S.R., Cha-Molstad, H., Dwyer, J.M., Yochum, G.S., Boss, J.M., McWeeney, S., Dimn, J.J., Mandel, G. and Goodman, R.H. (2004) Defirung the CREB regulon: a genome-wide analysis of transcription factor regulatory regions. Cell, 119: 1041-1054. Jeromin, A., Yuan, L.L., Frick, A., Pfaffinger, P. and Johnston, D. (2003) A modified Sindbis vector for prolonged gene expression in neurons. J. Neurophysiol., 90: 2741-2745. Jin, B.K., Belloni, M., Conti, B., Federoff, H.J., Starr, R., Son, J.H., Baker, H. and Joh, T.H. (1996) Prolonged in vivo gene expression driven by a tyrosine hydroxylase promoter in a defective herpes simplex virus amplicon vector. Hum. Gene Ther., 7: 2015-2024. Kaplitt, M.G., Kwong, A.D., Kleopoulos, S.P, Mobbs, C.V., Rabkin, S.D. and Pfaff, D.W (1994) 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., 91: 8979-8983. Kotin, R.M., Linden, R.M. and Bems, K.I. (1992) Characterization of a preferred site on human chromosome 19q for integration of adeno-associated virus DNA by non-homologous recombination. EMBO J., 11: 5071-5078. Matsushita, S., Kimura, M., Miyakawa, T., Yoshino, A., Murayama, M., Masaki, T. and Higuchi, S. (2004) Association study of brainderived neurotrophic factor gene polymorphism and alcoholism. Alcohol Clin. Exp. Res., 28: 1609-1612. Mayr, B. and Montminy, M. (2001) Transcriptional regulation by the phosphorylation-dependent factor CREB. Nat. Rev. Mol. Cell Biol., 2: 599-609. McGee, S.L., Rendahl, K.G., Quiroz, D., Coyne, M., Ladner, M., Manning, WC. and Flannery J.G. (2001) Recombinant AAV-mediated delivery of a tet-inducible reporter gene to the rat retina. Mol. Ther., 3: 688-696. Mitta, B., Rimann, M., Ehrengruber, M.U., Ehrbar, M., Djonov, V., Kelm, J. and Fussenegger, M. (2002) Advanced modular selfinactivating lentiviral expression vectors for multigene interventions in mammalian cells and in vivo transduction. Nucleic Acids Res., 30(21): 1-18. Nakai, H., Montiiu, E., Fuess, S., Storm, T.A., Grompe, M. and Kay, M.A. (2003) AAV serotype 2 vectors preferentially integrate into active genes in mice. Nat. Genet., 34: 297-302. Nakai, H., Yant, S.R., Storm, T.A., Fuess, S., Meuse, L. and Kay M.A. (2001) Extrachromosomal recombinant adeno-associated virus vector genomes are primarily responsible for stable liver transduction in vivo. J. Virol., 75: 6969-6976. Nestler, E.J., Barrot, M., DiLeone, R.J., Eisch, A.J., Gold, S.J. and Monteggia, L.M. (2002) Neurobiology of depression. Neuron, 34: 13-25.

Neve, R.L. and Geller, A.I. (1995) A defective herpes simplex virus vector system for gene delivery into the brain: comparison with alternative gene delivery systems and usefulness for gene therapy Clin. Neurosci., 3: 262-267. Neve, R.L., Howe, J.R., Hong, S. and Kalb, R.G. (1997) Introduction of the glutamate receptor subunit 1 into motor neurons in vitro and in vivo using a recombinant herpes simplex virus. Neuroscience, 79: 435-447. Newton, S.S., Thome, J., Wallace, T.L., Shirayama, Y, Schlesinger, L., Sakai, N., Chen, J., Neve, R., Nestler, E.J. and Duman, R.S. (2002) Inhibition of cAMP response element-binding protein or dynorphin in the nucleus accumbens produces an antidepressant-like effect. J. Neurosci., 22: 10883-10890. Papadakis, E.D., Nicklin, S.A., Baker, A.H. and White, S.J. (2004) Promoters and control elements: designing expression cassettes for gene therapy. Curr. Gene Ther., 4: 89-113. Pliakas, A.M., Carlson, R.R., Neve, R.L., Konradi, C , Nestler, E.J. and Carlezon, WA. Jr. (2001) Altered responsiveness to cocaine and increased immobility in the forced swim test associated with elevated cAMP response element-binding protein expression in nucleus accumbens. J. Neurosci., 21: 7397-7403. Rattiner, L.M., Davis, M., French, C.T. and Ressler, K.J. (2004) Brainderived neurotrophic factor and tyrosine kinase receptor B involvement in amygdala-dependent fear conditioning. J. Neurosci., 24(20): 4796-4806. Rivadeneira, E.D., Popescu, N.C., Zimonjic, D.B., Cheng, G.S., Nelson, RJ., Ross, M.D., DiPaolo, J.A. and Klotman, M.E. (1998) Sites of recombinant adeno-associated virus integration. Int. J. Oncol., 12: 805-810. Robinson, D.L., Phillips, PE.M., Budygin, E.A., Trafton, B.J., Garris, PA. and Wightman, R.M. (2001) Sub-second changes in accumbal dopamine during sexual behavior in male rats. Neuroreport, 12(118): 2549-2552 Robinson, T.E. and Berridge, K.C. (1993) The neural basis of drug craving: an incentive-sensitization theory of addiction. Brain Res. Brain Res. Rev, 18: 247-291. Scammell, T.E., Arrigoni, E., Thompson, M.A., Ronan, P.J., Saper, C.B. and Greene, R.W (2003) Focal deletion of the adenosine A l receptor in adult mice using an adeno-associated viral vector. J. Neurosci., 23: 5762-5770. Sen, S., Nesse, R.M., Stoltenberg, S.R, Li, S., Gleiberman, L., Chakravarti. A., Weder, A.B. and Burmeister, M. (2003) A BDNF coding variant is associated with the NEO personality inventory domain neuroticism, a risk factor for depression. Neuropsychopharmacology 28: 397-401. Shaw-Lutchman, S.Z., Impey S., Storm, D., and Nestler, E.J. (2003) Regulation of CRE-mediated transcription in mouse brain by amphetamine. Synapse, 48:10-17. Song, S., Wang, Y, Bak, S.Y, During, M.J., Bryan, J., Ashe, O., Ullrey D.B., Trask, L.E., Grant, RD., O'MaUey K.L., Riedel, H., Goldstein, D.S., Neve, K. A., LaHoste, G.J., MarshaU, J.R, Haycock, J. W, Neve, R.L. and Geller, A.I. (1998) Modulation of rat rotational behavior by direct gene transfer of constitutively active protein kinase C into nigrostriatal neurons. J. Neurosci., 18: 4119^132. Song, S., Wang, Y, Bak, S.Y, Lang, P , Ullrey D., Neve, R.L., O'Malley K.L. and Geller, A.I. (1997) An HSV-1 vector containing the rat tyrosine hydroxylase promoter enhances both long-term and cell type-specific expression in the midbrain. J. Neurochem., 68: 1792-1803. Spaete, R.R. and Frenkel, N. (1982) The herpes simplex virus amplicon: a new eucaryotic defective-virus cloning-amplifying vector. Cell, 30: 295-304.

[I. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

ACKNOWLEDGMENT Stevenson, C.W., Sullivan, R.M. and Gratton, A. (2003) Effects of basolateral amygdala dopamine depletion on the nucleus accumbens and medial prefrontal cortical dopamine responses to stress. Neuroscience, 116: 285-293. Stow, N. and McMonagle, E. (1982) Propagation of foreign DNA sequences linked to herpes simplex virus origin of replication. In: Gluzman Y. (Ed.), Eucaryotic Viral Vectors. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, pp. 199-204. Sun, M., Lee, J., Yu, L. and Geller, A.I. (2000) Modest increases in the titers of helper virus-free herpes simplex virus 1 (HSV-1) vectors by packaging in a cell line with inducible expression of HSV-1 VP16 or by treatment with N,N'-hexamethylene-bis-acetamide. Acta Virol., 44: 365-369. Tenenbaum, L., Lehtonen, E. and Monahan, P.E. (2003) Evaluation of risks related to the use of adeno-associated virus-based vectors. Curr. Gene Ther., 3: 545-565. Thome, J., Sakai, N., Shin, K., Steffen, C , Zhang, Y.J., Impey, S., Storm, D. and Duman, R.S. (2000) cAMP response element-mediated gene transcription is upregulated by chronic antidepressant treatment. J. Neurosci., 20: 4030-4036. Uslaner, J., Kalechstein, A., Richter, T., Ling, W. and Newton, T. (1999) Association of depressive symptoms during abstinence

193

with the subjective high produced by cocaine. Am. J. Psychiatry, 156(9): 1444-1446. Ventura, A., Meissner, A., Dillon, C.P., McManus, M., Sharp, P.A., Van Parijs, L., Jaenisch, R. and Jacks, T. (2004) Cre-lox-regulated conditional RNA interference from transgenes. Proc. Natl. Acad. Sci. USA, 101:10380-10385. Wang, Y, Camp, S.M., Niwano, M., Shen, X., Bakowska, J.C, Breakefield, X.O. and Allen, P.D. (2002) Herpes simplex virus type 1 / adeno-associated virus rep(+) hybrid amplicon vector improves the stability of transgene expression in human cells by site-specific integration. J. Virol, 76: 7150-7162. Wise, R.A. and Bozarth, M.A. (1987) A psychomotor stimulant theory of addiction. Psychol. Rev, 94: 469^92. Xia, H., Mao, Q., Eliason, S.L., Harper, S.Q., Martins, I.H., Orr, H.T., Paulson, H.L., Yang, L., Kotin, R.M. and Davidson, B.L. (2004) RNAi suppresses polyglutamine-induced neurodegeneration in a model of spinocerebellar ataxia. Nat. Med., 10: 816-820. Xu, R., Janson, C.G., Mastakov, M., Lawlor, P., Young, D., Mouravlev, A., Fitzsimons, H., Choi, K.L., Ma, H., Dragunow, M., Leone, P., Chen, Q., Dicker, B. and During, M.J. (2001) Quantitative comparison of expression with adeno-associated virus (AAV-2) brainspecific gene cassettes. Gene Ther., 8(17): 1323-1332.

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

C H A P T E R

15 Use of Viral Vectors to Influence Behavior Sergei Musatov, Michael G. Kaplitt, Donald W, Pfaff, Sonoko Ogawa

Abstract: Viral-mediated gene transfer is an important approach to delineate the role of individual genes in regulating a specific behavior. Current vector systems allow stable expression of various proteins and peptides in a desired brain region without eliciting substantial toxicity. Furthermore, viralmediated RNA interference is a revolutionary new approach to silence a gene of interest in a discrete area of the brain. These technologies can be applied to virtually any animal species at different stages of development thus obviating limitations of the current transgenic and knockout models in behavioral neuroscience. Here, we discuss how recombinant adeno-associated virus vectors can be explored to study complex behaviors as well as regulation of gene expression in the brain. Keywords: adeno-associated virus; RNA interference; estrogen receptor a; progesterone receptor; sexual behavior; lordosis; body weight regulation

brain regions in vivo. It was long thought that this would be difficult for several reasons. First, the efficiency of viral vectors, initially, was poor and most behaviors are sufficiently complex that it is likely that most cells in a region of interest would need to be altered to reliably demonstrate significant behavioral changes. As is clear from the abundant literature and other contributions to this volume, this is no longer a significant concern. Similarly, vector-mediated toxicity could potentially confound data. Although this can be readily controlled for, the use of viral vectors currently in several human trials for brain disorders reflects the consensus that current systems do not cause substantial toxicity. Two niajor factors contributed to the approach that is described in this chapter. First, while viral vectors have been used for more than 15 years to express genes of interest (Kaplitt et al., 1991) the ability to silence gene expression with viral vectors has only recently been developed. The use of vectors to express Cre recombinase in transgenic animals to sitespecifically delete genes with flanking loxP sites has been available for several years (Le and Sauer, 2000).

The most complex functions of the brain in higher mammals control behaviors in response to a variety of environmental and hormonal cues. Most central nervous system disorders, even those believed to affect specific systems such as motor diseases, ultimately have some effect on behavior. New technologies such as functional imaging are now providing previously unimagined data regarding the potential functions of specific brain regions. Nonetheless, the temporal and spatial resolutions of such technologies are necessarily limited. Furthermore, the correlation of such data with specific subcellular functions or behavioral outputs would be impossible. Therefore, behavioral studies in relevant animal models remain a key tool particularly for deciphering the role of individual genes on particular behaviors. The generation of data in human patients has only increased the importance of good behavioral studies in animal systems, in order to provide correlates for findings of particular interest. Viral vectors represent an increasingly important method for determining the effect on particular behaviors of altering expression of specific genes in specific

Gene Therapy of the Central Nervous System: From Bench to

195

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

196

15. USE OF VIRAL VECTORS TO INFLUENCE BEHAVIOR

However, this requires expression of a foreign protein product, the Cre recombinase, in the setting of an animal which has been manipulated to harbor a transgene through development. Both of these are potential confounds for sensitive behavioral studies, and the transgenic issue is more fully addressed below. However, the recent introduction of somatic gene silencing by expression of a small-interfering RNA from a viral vector now permits site-specific silencing of individual genes in otherwise normally developed animals (Xia et a l , 2002, 2004; Hommel et al., 2003). Another key factor in behavioral studies is the choice of system to work with. Many complex behaviors involve interactions of several brain regions and redundant or overlapping signal transduction pathways, which could require an effort to decipher which would be prohibitive with current technologies. Nonetheless, viral vectors have been used quite effectively to influence several different types of behavior, such as learning and memory or addictive behaviors, both of which are detailed elsewhere in this volume. We have long studied female sexual behaviors, and in particular the lordosis response. These are among the most well-defined behaviors currently available. In addition, traditional drug or hormonal infusion, lesioning and physiological studies have suggested that a limited number of brain regions are responsible for these behaviors. Furthermore, these are dependent upon the presence of estrogens in ovariectomized female rats. Since estrogens are believed to function via specific receptors, this provides an opportunity to study a well-characterized behavior by altering expression of a defined set of genes in one or more focal brain regions. We have therefore chosen to use our substantial experience with adeno-associated virus (AAV) vectors to both silence and overexpress estrogen receptors in ovariectomized mice to both demonstrate the power of viral vector technology to alter sexual behavior and to address remaining unanswered questions regarding the role of specific estrogen receptors and individual brain regions on the response of females to estrogens.

I.

VIRAL^MEDIATED R N A INTERFERENCE FOR BEHAVIORAL NEUROSCIENCE

Gene knockout in transgenic mice has been traditionally used to study behaviors. While powerful, this approach has several limitations. First and foremost, it cannot distinguish among discrete brain regions or developmental epochs involved in facilitating a specific effect. Therefore, a regional site-specific

knockdown of gene expression in a normal animal could provide new important opportunities to analyze the influence of individual genes in particular brain regions on complex behaviors. In addition, gene deletion in a transgenic aniraal can cause developmental changes, which might result in a behavioral abnormality reflecting these abnormalities rather than the primary effect of the gene in a normally developed adult animal. These issues and the effort required to generate transgenic animals also makes it difficult to easily knock out multiple genes of interest simultaneously. Finally, gene deletion in transgenic animals is largely limited to mice, but many behavioral studies have been performed in rats or even larger animals. Therefore, it would be desirable to develop a method to manipulate behavior via gene deletion in animals of any species, including humans if therapeutic interventions were warranted. RNA interference (RNAi) has emerged recently as a powerful tool to generate such region-specific knockouts. Traditional RNAi involved application of a double-stranded synthetic small-interfering RNA molecule to cells, resulting in specific degradation of target gene messages through an RISC-mediated mechanism (Elbashir et a l , 2001; Martinez et al., 2002). A major advance for gene therapy, however, was the development of plasmid vectors capable of expressing a small-interfering RNA. This utilizes a PolIII promoter to permit high-level expression of molecules with very specific start and stop sites (Brummelkamp et al., 2002; Hommel et al., 2003). A self-complementary RNA is then expressed with a small hinge region, which is ultimately cleaved and trimmed by cellular enzymes which resolve single-stranded RNA. This leaves a fairly defined small, double-stranded RNA which can act very efficiently as an siRNA molecule. Subsequently, this was applied to viral vectors, and has now been shown to function efficiently in animal brains in vivo. This approach has been successfully used to study motor deficits in models of neurodegenerative diseases, and is increasingly being applied to other CNS models (Xia et al., 2002, 2004; Hommel et al., 2003). 11. SILENCING ESTROGEN^RECEPTOR EXPRESSION IN MOUSE HYPOTHALAMUS U S I N G AAV VECTORS Of particular interest is the application of this technique to study the influence of genes in discrete brain nuclei on complex behaviors. To illustrate this point, we have selected estrogen receptor a (ERa) as a target

III PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

SILENCING ESTROGEN'RECEPTOR EXPRESSION IN MOUSE HYPOTHALAMUS USING AAV VECTORS

for siRNA-mediated gene silencing. Estrogens elicits many important biochemical responses in the brain, and the respective roles as well as the mechanisms of action of the two estrogen receptors, ERa and ERjS, in mediating these effects are not yet completely understood (Ogawa et al., 2000, 2002). One of the most significant sites of estrogen action is a group of neurons in the ventromedial nucleus (VMN) of hypothalamus, which is critical for several behaviors including female reproductive behavior. VMN is a key player in a neural circuit controlling the execution of lordosis, a hormone-dependent reflexive posture exhibited by sexually receptive female rodents in response to male mounting (Pfaff and Sakuma, 1979). Analysis of homozygous ERa knockout (ERKO) females revealed that they completely lack lordosis behavior (Rissman et al., 1997). Furthermore, these mice are also deficient in sexual behavioral interactions that precede the lordosis response, as they are extremely aggressive and rejective toward attempted mounts by male mice, thus preventing copulation (Ogawa et al., 1996, 1998; Rissman et al., 1997). In addition, estrogens appears to negatively regulate energy expenditure and homozygous ERKO mice have a significantly higher body weight, mostly due to increased fat deposits compared to their wild-type littermates (Heine et al., 2000). In dramatic contrast, reproductive behaviors of ERjS knockout females are undistinguishable from those of wild-type mice (Ogawa et a l , 1999). While these studies identified ERa as a gene essential for several mammalian behaviors, it remained imclear as to which of these effects are due to a mere absence of this receptor in the mature neurons and which deficits result from the lack of ERa expression during gestation, since estrogens can influence brain development. Moreover, recent findings indicate a potential confound in the design of ERKO mice, as these animals express an abnormal splicing variant of ERa as a result of genetic manipulation (Moffatt et al., 1998). This truncated form of the receptor retains both the DNA- and ligand-binding domains and is able to mediate, albeit far less efficiently, at least some estrogen effects. For example, while sexual behaviors

are diminished, some induction of progesterone receptor expression by estrogen is retained. To address these questions and to develop a tool for manipulating behaviors in normal animals of any species, we believed that somatic gene knockdown in individual nuclei through viral vector-mediated RNAi could provide a valuable alternative to conventional transgenic techniques. Figure 1 depicts a basic rAAV vector (AAV.Hl) containing the human PolIII H I promoter for expression of small hairpin RNA (shRNA) as well as an enhanced green fluorescent protein (EGFP) expression cassette as an independent marker to detect infected cells. Finding a potent siRNA still remains one of the most challenging and time-consuming steps in the process of developing an efficient vector for gene silencing. Several web-based tools are currently available that help select an efficient sequence. However, choosing a potent siRNA is still mostly an empirical process. Shown in Fig. 2 is one of the mouse ERa-specific siRNAs that have been tested for efficiency of silencing in cultured cells. As a control, we have chosen a vector expressing luciferase siRNAs as well as EGFP to control for any possible nonspecific adverse effects due to viral-mediated RNAi or EGFP toxicity. Although some utilize scrambled or mutated RNA sequences as a negative control, by design these do not have gene targets, therefore they cannot be tested for functionality as a true siRNA. With our luciferase siRNA control, we have confirmed functionality through a significant knockdown of luciferase expression in cell culture. There have been a number of reports that small dsRNAs operate via both miRNA and siRNA pathways (Doench et al., 2003). While the former is characterized by a reduction of expression at a protein but not mRNA level, the latter pathway leads to a cleavage of mRNA. As shown in Fig. 2, the degree of silencing at mRNA and protein levels are similar indicating that our AAV.Hl.ERl indeed directed RNAi. To generate conditional ERa knockdown mice, adult females were microinjected with AAV.Hl.ERl or AAV.Hl.Luc bilaterally into the VMN. Eight weeks after surgery animals were sacrificed and ERa expression was analyzed by immunocytochemistry to test

CMV/CBA promoter

Stuffer FIGURE 1

197

Intron

Poly(A)

Map of the rAAV vectors for siRNA delivery.

HI. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

ShRNA

198

15. USE OF VIRAL VECTORS TO INFLUENCE BEHAVIOR

ERa \

Histone

(a) 140 0

^20-

1 100o Z o -£ o> Ji a> 0(b)

I

80-

60 402u0 -n

n-

1

/ \AV.H1.Lu( :

1

AAV.H1.ER1

FIGURE 2 AAV-mediated RNAi reduces ERa expression in vitro, (a) Western blot analysis of ERa protein levels following transduction of cultured cells with different AAV vectors, (b) Q-PCR analysis of ERa mRNA level in the same cells.

AAV.H1.Luc

efficacy of siRNA-mediated ERa silencing in vivo. As can be seen in Fig. 3, injections into the hypothalamus resulted in an efficient transduction of VMN neurons. Double labeling for ERa (purple nuclear staining) and EGFP (brown cytoplasmic staining) revealed that virtually, all EGFP-positive cells in the area were also ERa-negative in mice injected with AAV.Hl.ERl. Concomitantly, no change in ERa staining was observed in control animals treated with AAV.Hl.Luc. Of importance is the fact that AAV.Hl.ERl-infected neurons produced EGFP, retained normal morphology and were as abundant as AAV.Hl.Luc-infected cells, suggesting that lack of ERa immunoreactivity was not due to cell loss caused by ERa siRNA toxicity but was rather a result of specific inhibition of ERa expression. To ensure that ERa knockdown is specific, we examined the expression of a homologous gene, ERjS, which is also present in the ventral part of the VMN and has some overlapping functions with ERa. In fact, both genes share high sequence similarity in the DNA- as well as the ligand-binding domains (Mosselman et al., 1996). Furthermore, our ERa-specific siRNA sequence ERl (GGCATGGAGCATCTCTACA) was also similar to a corresponding sequence of ERjS with only four mismatches (GGCATGGAACATCTGCICA, mismatched nucleotides are underlined). Although it

AAV.H1.ER1

O LU

+ LU

y^ ^^':ij^

«^^^;%:^ =

^ ** '•^*|

^.:'

' ^. -\..^^:i4^^ss

FIGURE 3 AAV-mediated knockdown of ERa expression in the brain. (Top panel) Double-label immunostaining for EGFP (brown) and ERa (purple) in the VMN of female mice 8 weeks after stereotaxic surgery (top two panels). (Bottom panel) expression of ERj5 in the VMN of the same animals. III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

SILENCING ESTROGEN-RECEPTOR EXPRESSION IN MOUSE HYPOTHALAMUS USING AAV VECTORS

AAV,H1.Luc

199

AAV.H1.ER1

FIGURE 4 Knockdown of ERa in the VMN prevents estrogen-induced upregulation of PR immunoreactivity. Note silencing of ERa by AAV.Hl.ERl (top panel) and concomitant inhibition of PR induction in the VMN but not the ARC (bottom panel).

has been demonstrated that a single base pair substitution in the antisense strand of siRNA duplex would prevent RNAi in vitro (Brummelkamp et al., 2002), the fidelity of this process in vivo is not well characterized. Results presented in Fig. 3 demonstrate that transduction of VMN neurons with AAV.Hl.ERl did not diminish ERjS immunoreactivity, as evident from a similar number of ERjS-positive cells and staining intensity To characterize the effects of ERa silencing in the brain, the mice were injected bilaterally into the VMN with either AAVHl.Luc or AAV.Hl.ERl and then treated with jS-estradiol 3-benzoate (EB). Consistent with the previous experiment, infection with AAV. Hl.ERl resulted in a complete loss of ERa immunoreactivity in the VMN compared to AAV.Hl.Lucinjected mice (Fig. 4, top panel). In addition, we did not observe any decrease in the ERa immunoreactivity in other ERa-positive brain regions, such as the juxtaposed arcuate nucleus (ARC, Fig. 4, top panel) as well as amygdala (Fig. 5, top panel), where projections of EGFP-positive fibers from the hypothalamus can be readily detected (Fig. 5, bottom panel). These results demonstrate the power of this technology as a vehicle for highly focal, highly specific silencing of gene expression in the brain of a normally developed adult animal.

One of the major targets of ERa signaling pathway in the brain following estrogens exposure is upregulation of progesterone receptor (PR) transcription. We and others have previously shown that upregulation of PR following estrogens surge is critical for female reproductive behavior, as inhibition of PR translation by antisense oligonucleotides significantly reduced female proceptive and receptive responses (PoUio et al., 1993; Mani et al., 1994; Ogawa et a l , 1994). We therefore set out to determine if ERa silencing would suppress activation of PR expression after estrogen administration. As anticipated, in mice injected with AAVHl.Luc, estrogen treatment resulted in a robust PR immunoreactivity in the VMN and ARC (Fig. 4, bottom panel), as well as other brain regions such as medical preoptic area MPOA (data not shown). In contrast, in AAV.Hl.ERl-treated mice detectable PR expression in the VMN was completely eliminated, yet it was unaffected in the ARC (Fig. 4, bottom panel). As indicated earlier, this is in distinction to the ERKO transgenic knockout mice, which retained PR induction by estrogens in the VMN presumably due to the presence of the aberrant transcript (see below). These findings demonstrate that rAAV-mediated siRNA delivery can be used to achieve a precise, region-specific silencing of ERa to a level, sufficient to suppress the normal physiological signaling cascade of this nuclear receptor

HI. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

200

15. USE OF VIRAL VECTORS TO INFLUENCE BEHAVIOR

AAV. H1.Luc

AAV.H1.ER1

FIGURE 5 Downregulation of ERa is region-specific as expression of ERa in the amygdala is not affected by AAV.Hl.ERl (top panel). Note EGFP-positive projections from the hypothalamus in both animals (bottom panel).

in neurons, and unequivocally establishes ERa as the mediator of estrogen induction of PR expression in the VMN in vivo. Having characterized the effects of ERa knockdown at a histological level, we examined these effects on complex behaviors. As expected, after priming with estrogen, mice in the control group injected with AAV. HI.Luc became sexually receptive, displaying proceptive still posture (Fig. 6a) and the lordosis response (Fig. 6b). In addition, they demonstrated very few rejections toward male mounting (Fig. 6c). This was equivalent to the response seen in naive, untreated females, confirming that neither the AAV vector nor expression of EGFP or RNAi inherently causes a reduction in female sexual responses to estrogens. In female mice treated with AAV.Hl.ERl, however, sexual receptivity toward males was completely abolished (Fig. 6a, b). Instead, these female mice showed vigorous rejection such as kicking and defensive fight back toward male approach and attempted mounts (Fig. 6c). Since the female rejections were very strong, stud males could hardly show normal mounts or intromissions. It thus appears that silencing of ERa restricted to the VMN of normal adult mice confers a behavioral response very similar to that of transgenic ERKO mice. It was also found that ERa silencing in the VMN had a profound

effect on body weight. Female mice injected with AAV. Hl.ERl in the VMN gained significantly more weight over a period of several weeks after surgery compared to those treated with AAV.Hl.Luc (Fig. 7). The degree of weight gain following ERa knockdown in the VMN was similar to that reported for ERKO mice (Heine et al., 2000). Several conclusions can be drawn from these data regarding estrogen signaling in the VMN specifically, as well as regarding viral vector-mediated siRNA in general. As indicated above, this can clearly be a highly focused and efficient method, in this case completely silencing ERa expression only in the VMN and not in surrounding regions, without influencing local expression of the highly homologous ERjS gene. Furthermore, this silencing was physiologically significant since this completely blocked estrogen induction of PR expression. This is in opposition to the ERKO transgenic knockout which did not demonstrate this phenotype, thereby raising some questions regarding the true cause of behavioral changes seen in those animals. The similarly profound behavioral change, which we observed now conclusively confirms without such ambiguity prior suggestions that ERa in the VMN is the major mediator of estrogen action on female sexual behavior. Furthermore, the change in

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

RESTORATION OF ESTROGEN RECEPTOR EXPRESSION IN TRANSGENIC KNOCKOUTS USING AAV VECTORS

1 40-.

g AAV.H1.Luc • AAV.H1 .ER1

3 O

-^ 40-| 3 ^ 30-

2 30CO

4> (A

o 10Q. (0 Q

^

Q J

AAV.H1.Luc AAV.H1.ER1

(0

1 ,

1 B .

,

Experiment 1

• •

201

Experiment 2

1:i ^

(a)

Experiment 1

(b) |100-1

• •

3

2o ^^ 80 7B jo 6 0 -

^ Experiment 2

"^ AAV.H1.Luc AAV.H1.ER1

««. 8 40o Q. 20 (A

^

^

oJ

Experiment 1

Experiment 2

FIGURE 6 Silencing of ERa in the VMN suppresses estrogen-mediated female sexual behavior. In two independent experiments, mice treated with AAV.Hl.ERl did not show any proceptive postures (a) or lordosis (b) compared to the animals injected with AAV.Hl.Luc. Instead, they demonstrated vigorous rejection toward males (c).

2 5 0 -1 200

body weight in female animals. Finally, this represents one of the few examples, whereby viral vector-mediated RNAi resulted in site-, gene-specific silencing of expression, leading to both significant physiological and profound behavioral changes, all consistent with the mechanism of action of the single silenced gene in the targeted brain region.

AAV.H1.Luc AAV.H1.ER1

c (0

u 150

<•-•

C

o 100 0)

I

50 A Experiment 1

Experiment 2

FIGURE 7 VMN-specific ERa knockdown leads to an increase in body weight. Mice injected with AAV.Hl.ERl have gained significantly more weight compared to control animals 8 weeks (Experiment 1) or 13 weeks (Experiment 2) after surgery.

body weight in our animals is of particular interest, since these animals retained normal ERa expression everywhere in the body and everywhere in the brain other than the VMN, unlike the ERKO animals in which this gene was deleted in every cell in the body. Therefore, this further implicates ERa expression in the VMN as a major mediator of estrogen effects on

IIL RESTORATION OF ESTROGEN RECEPTOR EXPRESSION IN TRANSGENIC KNOCKOUTS U S I N G AAV VECTORS An alternative to viral-mediated silencing of individual genes in normal animals is a knock-in approach, i.e., delivery of a gene of interest into the brain of transgenic knockout animals. We have designed a recombinant AAV vector that harbors ERa and injected it into the VMN of ERKO mice. As evident from Fig. 8, infusion of this vector resulted in a robust expression of ERa in the VMN. Furthermore, treatment of these mice with estrogen led to a significant enhancement of PR immunoreactivity in this region (Fig. 9). Thus, these experiments demonstrate that expression of ERa in the

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

202

15. USE OF VIRAL VECTORS TO INFLUENCE BEHAVIOR

Recombinant ERa in ERKO mouse

FIGURE 8

Endogenous ERa in wild-type mouse

AAV-mediated delivery of ERa into the brain of ERKO mice.

brain of ERKO mice returns to wild-type levels, one of the major downstream effects of estrogen action, the upregulation of PR, which is present but diminished in these animals. This observation suggests that at least one ERa-dependent signaling pathway is preserved in the brain of ERKO mice and is not compromised by the lack of this receptor during development. According to this line of reasoning, once a functional ERa is delivered, its interaction with estrogens triggers a chain of events leading to the initiation of PR gene transcription. However, despite the profound effect on PR expression, ERa delivery to ERKO mice did not lead to any behavioral changes discussed above, i.e., rescue of sexual behavior, decreased aggression or body weight. While our RNAi data demonstrate that elimination of ERa expression in the VMN is sufficient to abolish estrogen-induced female sexual behaviors, these data suggest that induction of female sexual behavior may be more complex. It is possible, for example, that the simultaneous presence of ERa in other brain regions may be necessary for full function, and elimination of ERa in any one of these regions could abolish behavior. This could readily be addressed with both systems, either by restoring ERa expression in multiple regions in ERKO mice or by silencing expression in multiple areas of normal animals using our AAV-mediated siRNA. It is also possible, as suggested earlier, that expression of this receptor during development to govern maturation of neural circuits may be necessary for a specific behavior, but that in the adult the

primary mediator remains the VMN without any significant need for ERa in other regions. Further studies such as those outlined above should readily clarify these questions and are most efficiently addressed through the use of viral vector technology. Nonetheless, site-specific silencing of ERa expression in a single nucleus (VMN) of normally developed adult mice blocked estrogen-induced sexual behavior. These findings extend previous reports that document the importance of this brain region in female reproductive behavior. Local infusions of estrogen and lesion studies have identified the VMN as a critical region to mediate female sexual behavior in rodents (Pfaff, 1999). The neural network that governs lordosis, an estrogen-dependent female sexual posture induced by stimulation by a male, involves projections from the VMN neurons to aqueductal gray, followed by the medullary reticular formation, lumbar ventral horn and finally the deep-back muscles directly involved in the execution of the reflex (Pfaff, 1999). Thus, the VMN appears to function as a gate, allowing transmission of a sensory stimulatory signaling to motorneurons only in the presence of steroid hormones. Furthermore, this process is contingent upon estrogen-mediated activation of transcription, since the VMN neurons cannot be immediately stimulated by estrogens, but rather require priming with the hormone and this effect can be blocked by transcription or translation inhibitors (Pfaff, 1999). We have previously shown that female mice lacking ERa (ERKO) but not ERjS demonstrate

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

RESTORATION OF ESTROGEN RECEPTOR EXPRESSION IN TRANSGENIC KNOCKOUTS USING AAV VECTORS

203

500-1

-480-360-240-120 0

120 240 360 480 600

-480-360-240-120 0

120 240 360 480 600

FIGURE 9 Distribution of ERa- and PR-positive cells in the brain of ERKO mice injected with AAV.ERa. ERa-positive cells (top) and PR-positive cells (bottom) were counted in alternate sections of animals injected bilaterally with AAV. ERa and treated with estrogen. Position "0'' is the injection site.

profoundly reduced receptive and proceptive (lordosis) behaviors and are very aggressive toward males (Ogawa et a l , 1996, 1998, 1999). Extending the above findings writh novel methodology, disruption of neural circuitry in the VMN by silencing of ERa in our experiments abolished estrogen-induced female sexual behaviors and mimicked ERKO phenotype. One intriguing question that has raised debate in the literature is the role of ERa in mediating activation of signal transduction and gene expression by estrogens in the brain and other organs. While ovariectomized v^ild-type mice respond to estrogen treatment by a robust increase in PR immunoreactivity in VMH, MPOA and other brain regions, in ERKO mice this effect is also observed, albeit to a considerably lesser degree and in a smaller area compared to wild-type

animals (Shugrue et al., 1997). Furthermore, estrogens have been shown to trigger the mitogen-activated protein kinase (MAPK) cascade in the cerebral cortex of ERKO mice (Singh et a l , 2000). Several hypotheses have been proposed to explain this paradox. First, this effect could be mediated by the other estrogen receptor, ERj8. However, analysis of double knockouts (ERa/j8KO) generated by crossing ERKO mice with ERjS knockout mice has revealed that estrogen-binding sites in the brain as well as PR induction by estrogen are still preserved, suggesting that ERj8 cannot fully account for this effect (Shugrue et a l , 2002). Shughrue et al. (2002) have discovered in ERKO mice the expression of a truncated form of ERa. It appeared to result from a splicing event between a cryptic donor site in the Neo cassette and the acceptor site of exon 3. Thus,

III. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

204

15. USE OF VIRAL VECTORS TO INFLUENCE BEHAVIOR

this splice variant consisted of exon 1, the 5' half of exon 2, the first 20 bp of the Neo cassette, foUov^^ed by exons 3-9. Ironically, the ERa open-reading frame was preserved and the protein contained both the DNAand ligand-binding domains (Shugrue et al., 2002). While the presence of this truncated receptor could certainly explain the continued estrogen response in ERKO mice, the presence of another, yet unknown, estrogen receptor (ERy) cannot be ruled out. In fact, an existence of this putative receptor has been suggested by several research groups, and all of them have arrived at this conclusion using studies that involved ERKO mice (Shugrue et al., 1997, 2002; Ghosh et al., 1999; Singh et al., 2000). If such a receptor existed, it could confound our data only if its mRNA contained a sequence identical to our ERl siRNA. In order to minimize this possibility, v^e preformed a BLAST search and did not find any known mouse mRNA containing ERl sequence other than ERa. Clearly, our results that ERl did not silence ERj8 in vivo and other reports on the fidelity of RNAi in vitro (Brummelkamp et al., 2002) suggest that a very close homology between siRNA and a target messenger is required to initiate cleavage. While possible, it seems unlikely that a gene of such importance would have a high similarity to ERa and yet remain unidentified. In addition, the primary rational for postulating existence of ERa was conflicting data generated from ERKO mice (Shugrue et a l , 1997, 2002; Ghosh et al., 1999; Singh et al., 2000). The identification of a functional mutant receptor in these animals combined with our data demonstrating concordance between silencing of ERa, loss of PR expression and subsequent inhibition of sexual behavior suggests that the presence of an unidentified estrogen receptor (ERy) is no longer necessary to explain these effects. IV.

CONCLUSION

Viral vector-mediated RNAi technology provides several advantages over existing transgenic techniques. First, it allows generation of conditional knockdown animals without a significant investment of time and resources within only a few weeks after a potent siRNA sequence is identified in vitro. Equally important, gene silencing can be restricted to a single brain nucleus (as in the present study) and can be performed in a normally developed animal. This is particularly valuable when developmental consequences of gene silencing might confound analysis of data generated in the resulting adult animals. There are, however, some technical challenges one should consider. One

obstacle in generating conditional knockdown animals using RNAi technology is an infusion of a viral vector precisely into a particular brain region. This is especially difficult in the case of small nuclei in mice. In our experience, approximately 50% of animals are usually eliminated at the end of the experiment following immunohistochemical staining of the tissue. This further emphasizes the importance of using an independent reporter (e.g., EGFP) for precise evaluation of transduced regions, particularly if there is no clean, efficient antibody available for immunohistochemical detection of the target gene or if gene expression is so low that immunohistochemistry would be unreliable. The choice of the vector can also significantly influence the efficiency and fidelity of gene silencing. We have selected rAAV serotype 2 for the above study because of its limited spread in the brain compared to more efficient vectors of serotypes 1 and 5. At issue in these studies was the proximity of the ARC, another brain region reach in ERa-positive neurons. Thus, using a virus with a minimal diffusion rate from a needle opening allowed us to achieve a better control of the VMN infection. In contrast, other serotypes may be preferable when a bigger area needs to be transduced especially in larger animals. Unlike traditional knockout techniques, which are established primarily for mice, somatic gene silencing using appropriate viral vectors can be equally easily applied to several mammal species, including primates and humans, provided that the sequence of the gene of interest is known. Furthermore, in contrast to gene silencing induced by antisense or siRNA oligonucleotides, which is usually limited to only a few days, AAV-mediated siRNA delivery is expected to provide long-term effects. Consistent with other recent reports, in the present study we did not observe any decline in the efficiency of ERa knockdown from 3 weeks to 3 months after surgery. It is well established that in several animal species rAAV vectors can mediate stable transgene expression up to a period of several years (Kaplitt et al., 1994; Tenenbaum et al., 2004). While supporting data regarding long-term RNAi effects in vivo is still lacking, data from our 3-month study supports the idea that AAV-mediated gene silencing will be analogous to stable expression achieved with these vectors in other paradigms. This would greatly facilitate studies aimed at identifying genes and neural networks involved in complex brain functions, such as social and maternal behaviors, anxiety, drug addiction, long-term memory, sleep, etc., and may provide novel therapeutic options for diseases where gene silencing would be desirable.

HI. PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

CONCLUSION

References Brummelkamp, T.R., Bernards, R. and Agami, R. (2002) A system for stable expression of short interfering RNAs in mammalian cells. Science, 296: 550-553. Doench, J.G., Petersen, C.R and Sharp, RA. (2003) siRNAs can function as miRNAs. Genes Dev., 17: 438-442. Elbashir, S.M., Harboth, J., Lendeckel, W., Yalchin, A., Weber, K. and Tuschi, T. (2001) Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature, 411: 494-498. Ghosh, D., Taylor, J.A., Green, J.A. and Lubahn, D.B. (1999) Methoxychlor stimulates estrogen-responsive messenger ribonucleic acids in mouse uterus through a non-estrogen receptor (non-ER) alpha and non-ER beta mechanism. Endocrinology, 140: 3526-3533. Heine, RA., Taylor, J.A., Iwamoto, G.A., Lubahn, D.B. and Cooke, P.S. (2000) Increased adipose tissue in male and female estrogen receptor-alpha knockout mice. Proc. Natl. Acad. Sci. LJSA., 97: 12729-12734. Hommel, J.D., Sears, R.M., Georgescu, D., Simmons, D.L. and DiLeone, R.J. (2003) Local gene knockdown in the brain using viral-mediated RNA interference. Nat. Med., 9: 1539-1544. Kaplitt, M.G., Pfaus, J.G., Kleopoulos, S.P, Hanlon, B.A., Rabkin, S.D., Pfaff, D.W. (1991) Expression of a functional foreign gene in adult mammalian brain following in vivo transfer via a herpes simplex virus type 1 defective viral vector. Mol. Cell Neurosci., 2: 320-330. Kaplitt, M.G., Leone, R, Samulski, R.J., Xiao, X., Pfaff, D.W., O'Malley, K.L. and During, M.J. (1994) Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat. Genet., 8:148-154. Le, Y. and Sauer, B. (2000) Conditional gene knockout using ere recombinase. Methods Mol. Biol, 136: 477-485. Mani, S.K., Blaustein, J.D., Allen, J.M.C., Law, S.W., O'Malley, B.W., and Clark, J.H. (1994) Inhibition of rat sexual behavior by antisense oligonucleotides to the progesterone receptor. Endocrinology, 135:1409-1414. Martinez, J., Patkaniowska, A., Urlaub, H., Luhrmann, R. and Tuschi, T. (2002) Single-stranded antisense siRNAs guide target RNA cleavage in RNAi. Cell, 110: 563-574. Moffatt, C.A., Rissman, E.R, Shupnik, M.A. and Blaustein, J.D. (1998) Induction of progestin receptors by estradiol in the forebrain of estrogen receptor-alpha gene-disrupted mice. J. Neurosci., 18: 9556-9563. Mosselman, S., Polman, J. and Dijkema, R. (1996) ER beta: identification and characterization of a novel human estrogen receptor. FEBS Lett., 392: 49-53. Ogawa, S., Olazabal, U.E., Parhar, I.S. and Pfaff, D.W. (1994) Effects of intrahypothalamic administration of antisense DNA for progesterone receptor mRNA on reproductive behavior and progesterone receptor immunoreactivity in female rat. J. Neurosci., 14: 1766-1774.

205

Ogawa, S., Taylor, J.A., Lubahn, D.B., Korach, K.S. and Pfaff, D.W. (1996) Reversal of sex roles in genetic female mice by disruption of estrogen receptor gene. Neuroendocrinology, 64: 467-470. Ogawa, S., Eng. V., Taylor, J.A., Lubahn, D.B., Korach, K.S. and Pfaff, D.W. (1998) Roles of estrogen receptor-a gene expression in reproduction-related behaviors in female mice. Endocrinology, 139: 5070-5081. Ogawa, S., Chan, J., Chester, A.E., Gustafsson, J.A., Korach, K.S., and Pfaff, D.W. (1999) Survival of reproductive behaviors in estrogen receptor beta gene-deficient (betaERKO) male and female mice. Proc. Natl. Acad. Sci. USA, 96:12887-12892. Ogawa, S., Chester, A.E., Hewitt, S.C, Walker, V.R., Gustafsson, J.A., Smithies, O., Korach, K.S. and Pfaff, D.W. (2000) Abolition of male sexual behaviors in mice lacking estrogen receptors alpha and beta (alpha beta ERKO). Proc. Natl. Acad. Sci. USA, 97: 14737-14741. Ogawa, S., Korach, K.S. and Pfaff, D.W (2002) Differential roles of two types of estrogen receptors in reproductive behavior. Cuur. Opin. Endocrinol. Diabetes, 9: 224-229. Pfaff, D.W. (1999) Drive. MIT Press, Cambridge. Pfaff, D.W. and Sakuma, Y. (1979) Facilitation of the lordosis reflex of female rats from the ventromedial nucleus of the hypothalamus. J. Physiol., 288:189-202. PoUio, G., Xue, R, Zanisi, M., Nicolin, A. and Maggi, A. (1993) Antisense oligonucleotide blocks progesterone-induced lordosis behavior in ovariectomized rats. Mol. Brain Res., 19:135-139. Rissman, E.R, Early, A.H., Taylor, J.A., Korach, K.S. and Lubahn, D.B. (1997) Estrogen receptors are essential for female sexual receptivity. Endocrinology, 138: 507-510. Shughrue, P.J., Askew, G.R., Dellovade, T.L. and Merchenthaler, I. (2002) Estrogen-binding sites and their functional capacity in estrogen receptor double knockout mouse brain. Endocrinology, 143:1643-1650. Shughrue, P.J., Lubahn, D.B., Negro-Vilar, A., Korach, K.S. and Merchenthaler, I. (1997) Responses in the brain of estrogen receptor alpha-disrupted mice. Proc. Natl. Acad. Sci. USA., 94: 11008-11012. Singh, M., Setalo, G. Jr., Guan, X., Frail, D.E. and Toran-Allerand, C D . (2000) Estrogen-induced activation of the mitogen-activated protein kinase cascade in the cerebral cortex of estrogen receptor-alpha knock-out mice. J. Neurosci., 20:1694-1700. Tenenbaum, L., Chtarto, A., Lehtonen, E., Velu, T, Brotchi, J.E., Levivier, M. (2004) Recombinant AAV-mediated gene delivery to the central nervous system. J. Gene Med., 6 (Suppl. 1): S212-S222. Xia, H., Mao, Q., Paulson, H.L. and Davidson, B.L. (2002) siRNAmediated gene silencing in vitro and in vivo. Nat. Biotechnol., 20: 1006-1010. Xia, H., Mao, Q., Eliason, S.L., Harper, S.Q., Martens, I.H., Orr, H.T., Paulson, H.L., Yang, L., Kotin, R.M. and Davidson, B.L. (2004) RNAi suppresses polyglutamine-induced neurodegeneration in a model of spinocerebellar ataxia. Nat. Med., 10: 816-820.

IIL PSYCHIATRIC AND BEHAVIORAL GENE THERAPY

C H A P T E R

16 Herpes Simplex Vector-Mediated Gene Transfer to the Peripheral Nervous System for the Treatment of Polyneuropathy and Chronic Pain Marina Mata, ]oseph C. Glorioso, David], Fink

Abstract: The distributed anatomy of the peripheral sensory system combined with the relative inaccessibility of the dorsal root ganglion (DRG) poses a challenge to gene delivery, but gene transfer to achieve local release of peptides from DRG neurons provides unique opportunities for the development of novel therapies. Nonreplicating neurotrophic vectors created from recombinant herpes simplex virus efficiently transduce DRG neurons from peripheral application. Production and release of neurotrophic factors from transduced DRG neurons can be used to prevent the progression of peripheral polyneuropathy in rodent models of neuropathy caused by neurotoxic drugs or resulting from diabetes. Production and release of inhibitory neurotransmitters from DRG neurons achieved by HSV-mediated gene transfer reduces pain-related behavior in rodent models of chronic regional pain caused by inflammation, nerve damage, or cancer. In this chapter, recent animal studies and prospects for moving towards a human trial are reviewed. Keywords: Gene therapy; pain; polyneuropathy; herpes simplex virus; neurotrophins; opioid peptides; gamma amino butyric acid

the brain, brainstem, and spinal cord can be effectively achieved with any one of a number of vectors including those based on adenovirus (Friedmann and Roblin, 1972), adeno-associated virus, lentivirus, and herpes simplex virus (HSV) (Davidson and Breakefield, 2003). Effective gene transfer to substantia nigra or striatum, for example, has been demonstrated using each of these vectors in models of Parkinson disease (Choi-Lundberg et al., 1997; Mandel et al., 1997; Bensadoun et al., 2000; Puskovic et al., 2004), and the choice between these vectors for a particular application may depend upon considerations of safety, immunogenicity, or ease of preparation of clinical grade vector. But gene transfer to the peripheral nervous system, specifically sensory neurons located in the dorsal root ganglia (DRG) whose bipolar axons project peripherally toward the skin and other organs and centrally into spinal cord faces special

Gene therapy was initially proposed as a novel approach for the treatment of inherited metabolic diseases by correction of the responsible genetic defect (Friedmann and Roblin, 1972); but in experimental animal models of multifactorial or nongenetic conditions it has become clear that gene transfer may also be useful in the treatment of diseases that are not genetic in etiology. In the nervous system, gene transfer can be used to circumvent barriers to the penetration of therapeutic peptides into central nervous system, or to achieve the local release of peptides to activate receptors at discrete sites in the nervous system. As in other therapeutic situations, the choice of vector for nervous system applications depends on the details of the application envisioned, and must reflect a balance of the effectiveness and safety. Gene transfer to neurons in the parenchyma of the central nervous system, including structures of

Gene Therapy of the Central Nervous System: From Bench to Bedside

209

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

210

16. HERPES SIMPLEX VECTOR-MEDIATED GENE TRANSFER TO THE PERIPHERAL NERVOUS SYSTEM

constraints for which HSV is uniquely suited as a gene transfer vector. There are two major disease conditions for which gene transfer to the DRG provides a therapeutic option. The first is the degeneration of peripheral sensory neurons characteristic of sensory polyneuropathy, a clinical condition that results in progressive sensory loss. The most common cause of polyneuropathy in the United States is diabetes, but sensory neuropathy may also occur as a dose-limiting complication of cancer chemotherapy, or as an inherited condition. There are no effective treatments commonly occurring for any of the forms of neuropathy. Second, chronic pain, defined as a pain lasting for more than 3 months, is a common problem estimated to effect more than 60 million people annually. While there are many drugs and other treatments for chronic pain, more than 50% of patients with chronic pain receiving conventional therapy report that the therapy is ineffective. Neurons of the DRG represent an excellent target for gene transfer to prevent progression of neuropathy or to treat chronic regional pain. Gene transfer to the DRG can be achieved by surgical exposure of the DRG to allow direct inoculation of a gene-transfer vector into the ganglion (Xu et al., 2003). Gene transfer in this manner can be achieved with any one of the standard vectors used for gene transfer to neural parenchyma but this approach is cumbersome, invasive, and unlikely to be widely applicable to the treatment of human disease. In contrast HSV which naturally establishes a lifelong latent state in DRG following a primary infection of skin or mucous membranes can be used effectively to transduce neurons of the DRG from peripheral inoculation. In this chapter, we will review the biology of wild-type HSV, the construction of replication defective HSV recombinants from the wild-type virus, and the use of these vectors in the treatment of animal models of polyneuropathy and chronic pain. Special consideration will be given to issues of vector safety, and the propagation of quantities of recombinant-free vector appropriate for the treatment of human disease. L

BIOLOGY O F H S V

Type 1 HSV is a natural human pathogen, the cause of the "cold sore" skin lesion among other manifestations. In nature, wild-type HSV is spread by contact with infected skin or mucous membranes, but a defining characteristic of HSV infection is the tendency for lesions to recur in the same site over a period of years (Roizman and Sears, 1996). This phe-

nomenon of recurrent epithelial infection results from the establishment by wild-type virus of a lifelong latent infection of sensory neurons of the trigeminal ganglion (TG) or DRG that send projections to the site of the primary infection. Targeting of virions from the skin to sensory neurons of the TG or DRG is unique to herpes viruses (type 2 HSV and varicella zoster share this characteristic) and makes vectors derived from HSV ideally suited for gene transfer to the DRG. HSV is an enveloped double-stranded DNA virus (Hao et al., 2003) whose core of DNA is surrounded by an icosadeltahedral capsid and an amorphous protein layer (tegimient) (Newcomb et al., 1999) between the capsid and the lipid envelope (Fig. lA). Virus entry into the cell is mediated by glycoproteins embedded in the lipid envelope (Spear et al., 2000); nonspecific interactions between heparan sulfate glycoprotein moieties on the cell surface and glycoproteins gB and gC in the HSV envelope are followed by specific binding of cell surface receptors including HveA, a member of the TNF receptor family and HveC, a homotypic cell adhesion molecule, with HSV gD. Fusion, a complex process requiring gB, gD, gH, and gL, is followed by entry of the capsid and tegument into the cytoplasm (Soreq et a l , 1990; Yura et al., 1992; Montgomery et al., 1996; Geraghty et al., 1998). Sensory nerve terminals contain high amounts of HveC, and entry of the virion into the sensory nerve terminal is followed by retrograde axonal transport along axonal microtubules to the neuronal cell body, where the viral genome is injected into the nucleus. Uptake requires specific interactions of viral glycoproteins with high-affinity receptors; retrograde axonal transport requires specific interactions between viral proteins and the retrograde motor dynein (Smith and Enquist, 2002). The HSV genome contains more than 85 open reading frames within its 152 kb of DNA arranged as largely contiguous genetic elements in a unique long (UL), and unique short (Ug) segment (McGeoch et al., 1988), each flanked by inverted repeat sequences (Fig. IB). In lytic infection of epithelial cells by wild-type HSV, viral genes are expressed in a rigidly ordered temporal cascade (Honess and Roizman, 1974) of immediate early (IE), early (E), followed by late (L) gene expression (Fig. IC) that results in DNA replication, production of structural and functional proteins, and assembly into daughter virions that bud through the membrane to acquire the lipid envelope (Roizman and Sears, 1996). In sensory neurons of the DRG, the virus may naturally establish a lifelong latent state in which expression genes characteristic of lytic replication is suppressed (Woolf and Costigan, 1999). During latency only a single

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

211

HSV VECTOR CONSTRUCTION AND PROPAGATION

ICPO ICP27

IE—^E ICP4 ICP22

B Accessory D. £ t yyzi

32 3 z)DD

3D

3

D

ii 31

Us—D

UL

Essential FIGURE 1 (A) HSV is a dsDNA Virus With a Structured Protein Capsid, an Amorphous Protein Tegument, and a Surrounding Lipid Envelope Containing Glycoproteins. (B) Schematic Representation of the HSV Genome, Consisting of a Unique Long (UL) and a Unique Short (Ug) Sequence Each Surrounded by an Inverted Repeat. The Location of Accessory (Above) and Essential (Below) Viral Genes are Shown. (C) In Lytic Replication HSV Genes are Expressed in a Temporal Cascade. Four IE Genes (ICPO, ICP4, ICP22, and ICP27) are Involved in Regulating Transcription of E and L Viral Genes.

region of the viral genome remains transcriptionally active to produce a family of nonpolyadenylated latency associated transcripts (LATs), RNAs that can be detected in the nucleus of latency infected neurons years after the primary infection (Stevens, 1989). Latently infected neurons survive and appear to function normally, even though minor alterations in expression of cellular genes can be identified by microarray analysis. The combination of high-affinity uptake into sensory neurons of the DRG from peripheral inoculation and the natural establishment of a lifelong latent state makes HSV a prime candidate for gene transfer to the DRG.

II. HSV VECTOR CONSTRUCTION AND PROPAGATION HSV replication in neurons, but not in other cell types, requires expression of the HSV thymidine kinase (tk), and HSV recombinants defective in expression of tk are therefore incapable of reactivating from latent infection in neurons (Tenser et al., 1989). The simplest HSV-based vectors used for gene transfer to the DRG

are defective only in HSV tk. These recombinants can be propagated in the African green monkey kidney cells (VERO cells) employed routinely for propagation of wild-type virus, and following inoculation into animals are capable of replicating in epithelial cells of the skin, but not in neurons where they are forced into a pseudo latent state, from which they are incapable of reactivating. Vectors based on tk-recombinants have been used to transfer genes to DRG in several models of chronic pain (reviewed below), but because tkrecombinants do replicate in the skin, these recombinants may not be appropriate for the creation of vectors for use in human trials. Nonreplicating HSV vectors have been created by interruption of genes required for the replication of HSV in vitro. Of more than 85 genes that are expressed during the lytic replication cycle, only five are expressed with IE kinetics (expression does not require synthesis of viral proteins by the cell) and only four of these (coding for infected cell polypeptides ICPO, ICP4, ICP22, and ICP27) are involved in promoting viral gene transcription (Roizman and Sears, 1996) (Fig. IC). Two of the IE genes, those coding for

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

212

16. HERPES SIMPLEX VECTOR-MEDIATED GENE TRANSFER TO THE PERIPHERAL NERVOUS SYSTEM

ICP4 and ICP27, are essential viral genes, meaning that recombinant virions defective in the expression of these gene products are incapable of replicating even in the permissive environment of VERO cells in vitro. HSV recombinants defective for ICP4, for example, are incapable of replication but can be propagated to high titers as pure preparations in cell lines engineered to express the missing HSV gene product from the cellular genome (DeLuca et al., 1985). Although nonreplicating HSV vectors deleted only for both copies of ICP4 have been used successfully in animal models of neurologic disease without evidence of significant toxicity (see below^), there is reason to believe from tissue culture studies that some of the remaining IE genes that are expressed from this vector may be toxic to neurons, at least in vitro. Vectors designed for human applications in our laboratories have therefore been deleted for at least four of the five IE genes (Krisky et al., 1998). These multiply deleted vectors are propagated in 7B cells that provide functional complementation for the essential IE genes, ICP4 and ICP27. ICP22 is an accessory IE gene that is not absolutely required for viral propagation, but deletion of ICP22 reduces the yield of a recombinant vector (Krisky et al., 1998). Because it proved difficult to generate stable cell lines expressing ICP4, ICP22, and ICP27 v^e have exploited a mutated ICP22 promoter that converts ICP22 expression kinetics from that of an IE gene to an E gene (Hobbs and DeLuca, 1999), thus preventing expression of ICP22 in noncomplementing cells but allowing for therapeutic expression in 73 cells once the essential IE genes are expressed. A similar strategy has been employed for inhibiting expression of ICP47, to create a vector that expresses only the IE gene ICPO in noncomplementing cells. Our laboratories have used a range of nonreplicating vector constructs in rodent models of neuropathy and chronic pain that will be reviewed below. A substantial practical advantage to the development of HSV vectors for human gene therapy applications is that the safety of the parental virus has been established in ongoing and completed phase I/II human trials for treatment of glioblastoma. For those trials, replication compromised recombinant HSVbased vectors were constructed to infect and destroy recurrent malignant glioblastoma by lytic replication. The use of these conditional replication competent "oncolytic viruses" is based on the assumption that the vector would replicate and spread in tumor tissue without damaging normal brain. In contrast to the nonreplicating vectors that are described below, the conditional replication competent viruses are deleted

for nonessential viral genes such as the viral thymidine kinase (Kosz-Vnenchak et al., 1990; Markert et al., 1993; Boviatsis et al., 1994), the ribonucleotide reductase (Yamada et al., 1991; Mineta et al., 1994), a protein kinase (Tenser et al., 1989), or a gene (734.5) required for growth specifically in neurons (MacLean et al., 1991; Whitley et al., 1993; Chambers et al., 1995). In the two phase I studies involving intracranial inoculation in patients that have now been reported using these conditional replicating vectors, no evidence of viral encephalitis was observed, there were no episodes consistent with reactivation of wild-type virus, and no adverse events that could be unequivocally attributed to the vectors were reported. One study, which used a virus deleted for the y34.5 neurovirulence gene and UL39 ribonucleotide reductase gene, showed safety up to a dose of 3 X 10^ pfu (Markert et al., 2000). A second study using a ^34.5 mutant with a complementing mutation in the USll gene, which restores the growth characteristics of wild-type virus without reversing the attenuated neurovirulence phenotype of the 734.5 mutant, showed safety to a final dose of 10^ pfu (Rampling et al., 2000). A total of 30 patients were treated in the two trials. The evidence that even replication-competent HSV-1 may be safely inoculated into brain tumors in humans suggests that the multiply deleted replication-incompetent vectors should be safe for human use. TIL HSV^MEDIATED GENE TRANSFER IN THE TREATMENT OF SENSORY POLYNEUROPATHY Dysfunction of the sensory neurons in the peripheral nervous system resulting from axonal degeneration, demyelination, or progressive loss of cell bodies in the DRG results in a syndrome of loss of sensation in the extremities that is commonly subacute to chronic and relentlessly progressive (Thomas and Ochoa, 1993). Neuropathy may occur as a result of an inherited genetic defect, as a complication of systemic illness, or as a side effect of chemotherapeutic drugs or unintentional exposures to other toxic substances (Thomas and Ochoa, 1993). In the United States, the most common cause of neuropathy is diabetes mellitus. With the exception of the immune-mediated inflammatory neuropathies, there are no currently approved effective treatments for most of the commonly occurring polyneuropathies. In diabetes, for example, patients present with numbness affecting the distal lower extremities. Rigorous treatment of

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

HSV'MEDIATED GENE TRANSFER IN THE TREATMENT OF SENSORY POLYNEUROPATHY

the underlying diabetes and management of cardiovascular risk factors may slow the progression of the disease (Thomas, 1999), but over the course of years these patients uniformly develop progressive severe sensory loss often complicated by spontaneous neuropathic pain. Experimental studies in rodent models over the past 20 years have convincingly demonstrated that the progression of polyneuropathy in models of diabetes and toxic neuropathies can be prevented by the administration of peptide factors of the neurotrophin family including nerve growth factor (NGF), neurotrophin 3 (NT-3) (Schmidt et al., 2001), or unconventional neurotrophic factors such as erythropoietin (EPO) (Bianchi et al., 2004), but the widespread effectiveness of these peptide factors in rodent models of polyneuropathy has not translated into successful treatments of human disease. There are likely many causes of this result, but one major factor must relate to dose and delivery. For example, NGF is effective in preventing the progression of diabetic neuropathy in rodent models when administered intraperitoneally at a dose of 3-5 mg/kg, but human subjects given NGF by injection complained of irritating side effects beginning at doses as low as 1 |Lig/kg, and in the subsequent phase 3 trial of NGF for the treatment of diabetic polyneuropathy in which the peptide was administered by subcutaneous inoculation at a dose of 0.1 |Lig/kg, NGF was ineffective (Apfel, 2002). One approach to exploit the neuroprotective effects of these pleiotropic highly potent but short-lived peptides would be to use gene transfer to provide continuous production of the peptide factor. Adenoviral-mediated gene transfer of neurotrophic factor genes to muscle, resulting in continuous systemic release of the peptide from the transduced muscle, can be used to prevent progression of neuropathy (Pradat et al., 2001, 2002), but it is difficult to know whether patients would notice side effects from systemic release of the peptide from muscle, and it is certainly the case that the amount of vector required to scale up from rodent models to human treatment would preclude the use of this approach in patients. For the reasons outlined above, we have used HSV-mediated gene transfer to DRG by subcutaneous inoculation of the vector to achieve local release of neurotrophic factors and have used this approach to examine HSVmediated gene transfer of NT-3 and NGF in models of polyneuropathy caused by (1) overdose of pyridoxine (PDX), (2) treatment with cisplatin, and (3) diabetes induced by injection of streptozotocin (STZ). The empirical data are reviewed below.

213

Overdose of PDX causes selective degeneration predominantly of the large myelinated fibers in the peripheral nerve, resulting in numbness and loss of proprioception to produce a sensory ataxia, without weakness (Schaumburg et al., 1983; Windebank et al., 1985). The rodent model of PDX overdose provides the opportunity to test treatments in a model of peripheral neuropathy in which the animal is otherwise healthy, unlike the neuropathies induced by diabetes or chemotherapeutic agents. Intraperitoneal inoculation of NT-3 (2-20 mg/kg) has been demonstrated to effectively prevent PDX toxicity to nerve (Helgren et al., 1997). Vector QL2HNT3 was constructed to contain the NT-3 coding sequence under control of a fusion LAP2 human cytomegalovirus immediate early promoter (HCVM lEp) in the UL41 locus of a nonreplicating HSV vector defective in ICP4 and ICP27 and with ICP22 and ICP47 converted to P genes as described above (Mochizuki et al., 2001). Preliminary studies in vitro demonstrated that DRG neurons in culture transduced with the vector at a multiplicity of infection (MOI) of 1 produced and released NT-3 into the medium. Subcutaneous inoculation of the vector under the skin of the foot resulted in transduction of DRG neurons and production of NT-3 RNA detected by RT-PCR. Expression of NT-3 peptide in the DRG was confirmed by ELISA (unpublished results). Subcutaneous inoculation of the NT-3 expressing vector QL2HNT3 but not control vector QOZHG 3 days prior to the start of an 8-day course of PDX protected against PDX-induced nerve degeneration measured by the sensory evoked response in the sciatic nerve, preservation of the H-reflex, and behavioral measures of proprioceptive function (Chattopadhyay et al., 2003). Morphometric evaluation of nerve cross sections confirmed protection of the larger myelinated fibers that characteristically degenerate in response to PDX intoxication. Similar results were obtained using a nonreplicating HSV vector (SHN) that expresses NGF under the control of the HCMV lEp (Chattopadhyay et al., 2002). Subcutaneous inoculation of SHN under the skin of the foot 3 days prior to the start of a 8-day course of PDX preserved sensory evoked response in the sciatic nerve, H-reflex, and behavioral measures of proprioceptive function in a fashion similar to that observed with the NT-3 expressing vector QL2HNT3 (Chattopadhyay et al., 2002) (Fig. 3). This result was not initially anticipated because the large DRG neurons that give rise to the large myelinated fibers in the sciatic nerve express the high-affinity neurotrophin receptor trkC, while NGF acts predominantly through the trkA receptor.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

214

16. HERPES SIMPLEX VECTOR-MEDIATED GENE TRANSFER TO THE PERIPHERAL NERVOUS SYSTEM

In fact, the large DRG neurons proved to express low levels of the trkA receptor and the results of this study demonstrate that these trkA receptors on the large neurons are functional. The effect of HSV-mediated neurotrophin gene transfer in models of neuropathy has been extended to the clinically relevant model of cisplatin-induced neuropathy in the rodent. Cisplatin is a commonly used chemotherapeutic agent, but cumulative doses greater than 300mg/m^ result in a severe sensory neuropathy that is often a dose limiting side effect of the drug. In addition to the NT-3 expressing vector QL2HNT3 and the NGF-expressing vector SHN described above we tested the NGF-expressing vector SLN that contains the NGF coding sequence under the control of the latency associated promoter element (LAP2). Subcutaneous inoculation of each of those vectors (but not subcutaneous inoculation of the control HSV vector QOZHG) 3 days prior to the start of a 2-month course of cisplatin administration (3 m g / k g twice a week) protected against the development of neuropathy measured by the electrophysiologic measures of evoked sensory response and H-wave, behavioral measures of proprioception (beam walking and rotarod testing), and histologically by measures of innervation of the skin using histochemistry with PGP9.5 (Chattopadhyay et a l , 2004). One measure of small fiber survival, the expression of the neuropeptides substance P (SP) and calcitonin gene related peptide (CGRP) in sensory nerve terminals in the dorsal horn of spinal cord showed that both the NGFexpressing and NT-3-expressing vectors protected small fibers against the toxic effects of the chemotherapeutic agent. Again, this result would not have been entirely anticipated because the small fibers expressed predominantly the trkA high-affinity neurotrophin receptor, and there is little known crosstalk between NT-3 and the trkA receptor. Similar results were observed in the mouse model of diabetic polyneuropathy. Diabetes was induced in Swiss Webster mice by the intraperitoneal injection of STZ, a relatively selective pancreatic islet cell toxin. Two weeks after the onset of diabetes animals were inoculated subcutaneously in both hind feet with the NGF-expressing nonreplicating HSV vectors, SHN or SLN. We chose to use the mouse model of diabetes because these animals do not become very sick but develop a mild neuropathy characterized by reduced sensory nerve amplitude that has some characteristics that are similar to the human neuropathy. Animals inoculated with the NGF-expressing vectors SHN or SLN but not animals inoculated with the control

vector SHZ showed preservation of foot sensory amplitude as well as preservation of expression of the neuropeptides SP and CGRP in the DRG at 6 weeks after the onset of diabetes (Goss et al., 2002a). Taken together, the results of the studies summarized above demonstrate that HSV-mediated gene transfer to the DRG by subcutaneous inoculation is effective in preventing the progression of neuropathy in rodent models. The studies performed to date have not evaluated whether neurotrophin gene transfer might be effective in reversing an established neuropathy, but because all of the clinically relevant neuropathies are subacute to chronic and progressive, an effective treatment that could be used to prevent the progression of neuropathy would represent a major advance. In this regard, chemotherapy-induced neuropathy represents an ideal target for the first phase I or phase I/II of human trials because (1) the time of onset of the disease is known; (2) the course, which is largely predictable based on the dose of chemotherapeutic agent will be given; (3) the progression of the disease is relatively rapid so that outcomes could be conveniently measured; and (4) the patients have an underlying serious disease that is often fatal, so that safety may be assessed ethically. Although the first studies in this model could be done using the HCMV lEp to achieve relatively short-term expression of the neurotrophin, extension of this approach to chronic neuropathies such as that resulting from diabetes will require the development of a vector with regulatable transgene expression, so that were the patient to experience some adverse effect of neurotrophin expression, expression could be terminated by omission of the activator drug. IV. HSV-MEDIATED GENE TRANSFER I N THE TREATMENT OF CHRONIC PAIN A second group of conditions in which HSV-mediated gene transfer to the DRG may prove to be useful is in the treatment of chronic pain. In this approach, gene transfer is used to effect the local release of neurotransmitters or neurotransmitter related substances to modulate the physiology of the nervous system without affecting the structure of the neural components (Fig. 2). An understanding of this approach requires some consideration of the neuroanatomy and neuropharmacology of pain perception. Acute pain, defined as an unpleasant experience with sensory and affective components, serves an essential role in a looking the individual to potential

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

215

HSV-MEDIATED GENE TRANSFER IN THE TREATMENT OF CHRONIC PAIN

J 1 \J A

FIGURE 2 Gene transfer to the DRG using HSV. Vector inoculated subcutaneously (Arrowhead) is carried by retrograde axonal transport to the DRG (arrow in Figure and Inset). Nonreplicating vectors establish a persistent latent infection of DRG and produce the transgene product that is transported centrally (Inset) and peripherally (Figure) in the bipolar DRG axon.

0.06

1000

0.09

800 0.06

600

0.03

0.02

200

1000

0.08

800 600

0.04

400

0.02

aSHZ-treated SHN-treated

FIGURE 3 HSV-mediated transfer and expression of the NGF gene 3 days prior to an 8-day course of PDX intoxication preserves H-Wave (A) proprioceptive function (B) and large myelinated fibers in sciatic nerve (C) measured 1 week after the conclusion of PDX treatment. From Chattopadhyay (2003), with permission.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

216

16. HERPES SIMPLEX VECTOR-MEDIATED GENE TRANSFER TO THE PERIPHERAL NERVOUS SYSTEM

or existing tissue damage. Specialized nociceptive neurons in the peripheral nervous system contain receptors that allow them to convert mechanical, thermal (hot or cold), chemical, or biochemical signals into electrophysiologic activity (Raja et al., 2002). These neurons synapse largely in the dorsal horn of spinal cord on second order neurons that project rostrally in the nervous system, terminating principally in the sensory thalamus (Doubell et a l , 2002). Third order neurons in the thalamus project mainly to the sensory cortex to mediate the discriminative aspects of pain, and to limbic structures to mediate the affective components of the experience. Neurotransmission of nociceptive information may be modulated by factors intrinsic to primary nociceptive neurons, by modulators of neurotransmission at the first synapse in the dorsal horn of spinal cord, by the state of the second order neuron in the dorsal horn, and at higher levels in the brain stem and cortex (Doubell et al., 2002; Raja et al., 2002). Chronic pain is defined as pain that persists beyond the termination of the inciting event, or pain that accompanies a peripheral lesion that cannot be treated. For the purpose of approval of pharmaceuticals and drugs the United States Food and Drug Administration defines chronic pain as a painful condition that persists for more than 3 months. The dorsal horn of spinal cord represents an excellent anatomic target for the treatment of chronic pain, because in the transition from acute pain to chronic pain continued activation of nociceptors results in alterations in the neurophysiologic response of second order neurons in the spinal cord, changes in neurotransmitters and receptors or expression of cytokines in the dorsal horn. While acute pain is important for the survival of the organism and does not warrant treatment, chronic pain represents a cormnon and difficult to treat clinical problem that severely impacts quality of life and whose cost is estimated in billions of dollars annually The first use of HSV vectors in the treatment of pain was reported by Wilson and coworkers who constructed a tk-deleted replication competent HSV vector containing the human proenkephalin gene, coding for the precursor protein that is processed to produce the endogenous pentapeptide opiates met- and leuenkephalin (Wilson et al., 1996). Subcutaneous inoculation of the vector in the dorsal surface of the hind paw resulted in transduction of DRG neurons in vivo to express enkephalin, and anti-hyperalgesic effects of the vector were demonstrated using chemical hypersensitization followed by thermal (laser) stimulation. We subsequently demonstrated that a nonreplicating HSV vector expressing the same construct reduced

spontaneous pain behavior in the delayed phase of the formalin test in rats, a widely used standard test of inflammatory pain (Goss et a l , 2001) (Fig. 4A). Similar results have been observed in the complete Freund's adjuvant model of chronic inflammatory pain (unpublished observation). The underlying rationale for the use of an enkephalin producing vector is the observation that opiate receptors are found presynaptically on primary nociceptive afferents and postsynaptic on second order neurons in the dorsal horn. Activation of these inhibitory receptors in the spinal cord represents one site at which opiate drugs such as morphine and to provide an analgesic effect; the vector-mediated effect is blocked by opiate receptor antagonists such as naloxone and naltrexone demonstrating that the proenkephalin expressing vector exerts its affect through the same receptor. Opiate receptors are widely distributed throughout the nervous system and in nonneural tissues including gut and bladder, and one of the limitations of the use of opiate drugs for the treatment of chronic pain occurs because systemically administered opiates activate receptors at these other sites. The local release of enkephalin in dorsal horn from the transduced DRG neurons allows the selective activation of opiate receptors at the spinal level, which represents a major advantage for the treatment of chronic regional pain. Chronic pain may also result from damage to neural structures in the absence of any extra neural tissue damage. This so-called "neuropathic" pain is a common complication of diabetic neuropathy and is often difficult to treat effectively. We examined the effect of a proenkephalin-expressing nonreplicating HSV vector in the standard spinal nerve ligation model of peripheral neuropathic pain (Hao et al., 2003). Subcutaneous inoculation of the vector into the plantar surface of the hind foot 1 week after spinal nerve ligation produced a substantial reduction in mechanical allodynia that was continuous over time (animals tested throughout the day showed an identical response), lasted for several weeks, and could be reversed by naloxone (Hao et al., 2003). The analgesic effect of vector SHPE that expresses proenkephalin under the control of the HCMV lEp diminished over the course of several weeks. It is unlikely that this diminution reflected the emergence of tolerance to the opiate effect, because reinoculation of the vector 6 weeks after the initial inoculation reestablished the antiallodynic effect (Fig. 4B). Interestingly the effect of enkephalin expression from the vector appeared to be independent of ^-opioid receptor, as animals inoculated with the vector continue to demonstrate the antiallodynic effect despite

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

HSV-MEDIATED GENE TRANSFER IN THE TREATMENT OF CHRONIC PAIN

217

re-lnoculatlon

Reinoculation •SHPE

5 10 weeks after spinal nerve ligation

r Inoculation

re-inoculation o QHGAD67 A QOZHG X Vehicle

FIGURE 4 Summary figures demonstrating the antinociceptive effect of HSV-mediated gene transfer for proenkephalin in the formalin model of inflammatory pain (A) (Goss, 2001) and neuropathic pain (B) (Hao et al., 2003), and HSV-mediated transfer of GAD in central pain following spinal cord injury (C) (Liu et al., 2004). In each model the effect of vector-mediated transgene expression (driven by the HCMV lEp) waned over a course of weeks, but was effectively reestablished by reinoculation. (A) From Goss et al. (2001) (B) from Hao et al. (2003) (C) from Liu et al. (2004).

the development of tolerance to morphine induced by twice a day injections of high doses of morphine. In addition, in the absence of tolerance the vector-mediated effect was additive with that of morphine, reducing the ED5Q of morphine from 1.8 to 0.15 m g / k g in animals with spinal nerve ligation (Hao et al., 2003). Enkephalin produced by the proenkephalin vector in the DRG is transported in both directions in the bipolar axon of the DRG, centrally toward the spinal cord as well as peripherally toward the end organ (Antunes bras et al., 2001). The effect of enkephalin released from central terminals in the dorsal horn of spinal cord can be assessed by its block by intrathecal inoculation of naltrexone, but peripheral effects are also observed. In the polyarthritis model of chronic inflammatory pain, Pohl and coworkers used a tk-deleted replicating HSV vector to transduced DRG neurons and showed that enkephalin released from the peripheral terminals of the DRG neurons acting directly on immune cells reduced inflammatory response to prevent bone loss in this model of inflammatory pain (Braz et al., 2001).

Identifying a patient population appropriate for phase I safety and dose-finding studies of FiSV-mediated gene transfer for the treatment of pain involves many of the same considerations discussed in the treatment of neuropathy. We believe that one appropriate patient population for this type of study is made up of patients with chronic pain resulting from cancer. This is a condition that not infrequently results in severe pain refractory to maximal medical and surgical management. We therefore tested the effectiveness of the proenkephalin expressing vector SFiPE in animal model of cancer pain created by injection of osteosarcoma cells into the distal femur (Schwei et a l , 1999). Growth for the tumor results in pain that can be measured by a spontaneous ambulatory pain score, and inoculation of the vector under the skin of the foot 1 week after tumor implantation resulted in a substantial and statistically significant reduction in spontaneous pain (Goss et al., 2002b). Based on the data presented above we proceeded to propose a phase I trial of a nonreplicating FiSV

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

218

16. HERPES SIMPLEX VECTOR-MEDIATED GENE TRANSFER TO THE PERIPHERAL NERVOUS SYSTEM

vector expressing proenkephalin in the treatment of refractory cancer pain. The proposed study would have enrolled 18 patients with cancer metastatic to a vertebral body causing pain unresponsive to maximal medical therapy. Inoculation of the vector into the dermatome corresponding to the radicular distribution of the pain was propose to begin and 10^ pfu an increase in half-life intervals with the patients at each dose. The study was reviewed by the Recombinant DNA Advisory Committee (RAC) of the NIH in June 2002. Enkephalin is only one of a number of inhibitory neurotransmitters that are present in the dorsal horn of spinal cord that may be used to modulate nociceptive neurotransmission in chronic pain. Gamma amino butyric acid (GABA) is a major inhibitory neurotransmitter in the central nervous system, and tonic GABAergic inhibition in the dorsal horn of spinal cord plays an important role in sensory processing. Inhibition of GABAergic neurotransmission produces pain, and reduction in GABAergic tone in the spinal cord is one of the changes that occur in the spinal cord in models of peripheral neuropathic pain. GABA-mimetic drugs are used to provide an analgesic effect in some types of neuropathic pain but the dose is severely limited by central (CNS) side effects. We constructed a nonreplicating HSV-based vector containing coding sequence for glutamic acid decarboxylase (GAD) under the control of the HCMV lEp (vector QHGAD67) (Liu et al., 2004). DRG neurons in tissue culture transduced with QHGAD67 at an MOI of 1 express GAD67 and release GABA into the medium, by a mechanism that is independent of depolarization that appears to involve a reversal of the GABA transporter-1 (unpublished data). Subcutaneous inoculation of QHGAD67 in the skin transduces DRG neurons to produce GAD67 which is transported to the nerve terminals in the dorsal horn of spinal cord (Liu et al., 2004). Release of the GABA in vivo from transduced DRG neurons can be demonstrated by microdialysis of dorsal horn and superfusion of isolated spinal cord and root preparations (Liu et al., 2004). The release in vivo is also constitutive of the unaffected by electrostimulation of the roots at parameters chosen to stimulate small fibers predominantly or all fibers in the nerve (unpublished results). In the spinal cord hemisection model of belowlevel central neuropathic pain resulting from spinal cord injury, inoculation with QHGAD67 reduces mechanical allodynia and thermal hyperalgesia (Fig. 4C), an effect that is partially reversed by block of GABA A and GABA B receptors by bicuculline and phaclofen, respectively (Liu et al., 2004). Like the effect

of enkephalin expression from the proenkephalin expressing vector SHPE, the effect of QHGAD67mediated GABA release lasted about 6 weeks, and reinoculation reestablished the analgesic effect. Similar results have been observed using QHGAD67 in the spinal nerve ligation model of peripheral neuropathic pain (unpublished observations). The duration of the analgesic effect of inoculation with either SHPE expressing enkephalin or QHGAD67 expressing GAD67 was 4-6 weeks. The fact that in each of the models studied with both of these vectors the analgesic effect could be reestablished by reinoculation (Fig. 4) suggests strongly that the loss of effect is not due to the development of tolerance but rather results from reduced expression of the transgene over time. Both of these vectors use the HCMV lEp to drive transgene expression. We have not measured transgene expression from these vectors in these models, but from other experiments we have established that expression from this promoter reaches a peak in the days immediately following inoculation and then falls exponentially over time. Transient expression would be beneficial phase I trials of dose-finding and safety, and reinoculation could certainly be used in clinical situations. However long-term expression would be desirable for some applications. In that case, the HSV LAP2 promoter element could be used to drive transgene expression. LAP2 is the genetic element responsible for lifelong expression of latency associated transcripts from latent HSV genomes in vivo (Goins et al., 1994), and in studies using expression of the glial cell-derived neurotrophic factor in models of Parkinson disease we have found that LAP2 is capable of driving biologically relevant levels of transgene expression for a minimum of 6 months after vector inoculation (Puskovic et al., 2004). We demonstrated a similar duration of effect in the peripheral nervous system in a delayed toxicity model of peripheral neuropathy (unpublished results). V-

SUMMARY

Conditions affecting the peripheral nervous system are often treated as orphan diseases. The application of trophic factors to protect peripheral nerves from degeneration is limited by the widespread distribution of receptors for these factors throughout the nervous system and in nonneural tissues. Similarly, chronic pain can be effectively treated by modulation of neurotransmission in the dorsal horn of spinal cord, but there are no neurotransmitters that are selective for the

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

ACKNOWLEDGMENTS

nociceptive pathways. HSV-mediated gene transfer to DRG from skin inoculation in vivo provides a novel means to deliver genes to the DRG in order to achieve local and/or focal release of neurotrophic factors were neurotransmitter substances to prevent the progression of neuropathy or to provide an analgesic effect in models of chronic pain. Phase I/Il trials of conditionally replicating HSV vectors injected into brain have been carried out without complications, suggesting that the vector is safe. Nonreplicating vectors deleted for multiple immediate early genes have been constructed and can be propagated into preparations to high titers. We await the completion of the first human trials to determine if these vectors can be used effectively in the treatment of the peripheral nervous system in patients.

ACKNOWLEDGMENTS We acknowledge the work of Munmun Chattopadhyay, James Goss, Shuanglin Hao, and Jun Liu in the animal models, and William Goins, Darren Woolf, David Krisky, and Shaohua Huang in vector construction. This work was supported by grants from the NIH, The Department of Veterans Affairs, and the Juvenile Diabetes Research Foundation.

References Antunes bras, J., Becker, C , Bourgoin, S., Lombard, M., Cesselin, R, Hamon, M. and Pohl, M. (2001) Met-enkephalin is preferentially transported into the peripheral processes of primary afferent fibres in both control and HSVl-driven proenkephalin A overexpressing rats. Neuroscience, 103: 1073-1083. Apfel, S.C. (2002) Nerve growth factor for the treatment of diabetic neuropathy: what went wrong, what went right, and what does the future hold? Int. Rev. Neurobiol., 50: 393-413. Bensadoun, J.C., Deglon, N., Tseng, J.L., Ridet, J.L., Zurn, A.D. and Aebischer, R (2000) 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 GDNR Exp. Neurol., 164: 15-24. Bianchi, R., Buyukakilli, B., Brines, M., Savino, C , Cavaletti, G., Oggioni, N., Lauria, G., Borgna, M., Lombardi, R., Cimen, B., Comelekoglu, U., Kanik, A., Tataroglu, C., Cerami, A. and Ghezzi, P. (2004) Erythropoietin both protects from and reverses experimental diabetic neuropathy. Proc. Natl. Acad. Sci. USA, 101: 823-828. Boviatsis, E.J., Chase, M., Wei, M.X., Tamiya, T., Hurford, R.K., Jr., Kowall, N.W., Tepper, R.I., Breakefield, X.O. and Chiocca, E.A. (1994) Gene transfer into experimental brain tumors mediated by adenovirus, herpes simplex virus, and retrovirus vectors. Hum. Gene Then, 5: 183-191. Braz, J., Beaufour, C., Coutaux, A., Epstein, A.L., Cesselin, R, Hamon, M. and Pohl, M. (2001) Therapeutic efficacy in experimental

219

polyarthritis of viral-driven enkephalin overproduction in sensory neurons. J. Neurosci., 21: 7881-7888. Chambers, R., Gillespie, G.Y., Soroceanu, L., Andreansky, S., Chatterjee, S., Chou, J., Roizman, B. and Whitley, R.J. (1995) Comparison of genetically engineered herpes simplex viruses for the treatment of brain tumors in a scid mouse model of human malignant glioma. Proc. Natl. Acad. Sci. USA, 92: 1411-1415. Chattopadhyay, M., Wolfe, D., Huang, S., Goss, J., Glorioso, J., Mata, M. and Fink, D. (2002) In vivo gene therapy of pyridoxineinduced neuropathy by HSV-mediated gene transfer of neurotrophin-3. Ann. Neurol., 51: 19-27. Chattopadhyay, M., Goss, J., Lacomis, D., Goins, W , Glorioso, J.C., Mata, M. and Rink D.J. (2003) Protective effect of HSV-mediated gene transfer of nerve growth factor in pyridoxine neuropathy demonstrates functional activity of trkA receptors in large sensory neurons of adult animals. Eur. J. Neurosci., 17: 732-740. Chattopadhyay, M., Goss, J., Wolfe, D., Goins, W , Huang, S., Glorioso, J., Mata, M. and Rink, D.J. (2004) Protective effect of herpes simplex virus-mediated neurotrophin gene transfer in cisplatin neuropathy Brain, 127: 929-939. Choi-Lundberg, D.L., Lin, Q., Chang, Y.-N., Chiang, Y.L., H a y CM., Mohajeri, H., Davidson, B.L. and Bohn, M.C. (1997) Dopaminergic neurons protected from degeneration by GDNR gene therapy. Science, 275: 838-841. Davidson, B.L. and Breakefield, X.O. (2003) Viral vectors for gene delivery to the nervous system. Nat. Rev. Neurosci., 4: 353-364. DeLuca, N.A., McCarthy, A.M. and Schaffer, RA. (1985) Isolation and characterization of deletion mutants of herpes simplex virus type 1 in the gene encoding immediate-early regulatory protein ICP4. J. Virol., 56: 558-570. Doubell, T.P, Mannion, R.J. and Woolf, C.J. (2002) The dorsal horn: state-dependent sensory processing, plasticity and the generation of pain. In: Melzack R. and Wall PD. (Eds.),Textbook of Pain, Rourth Edition. Churchill Livingstone, Edinburgh, pp. 165-182. Rriedmann, T. and Roblin, R. (1972) Gene therapy for human genetic disease? Science, 175: 949-955. Geraghty, R.J., Krummenacher, C , Cohen, G.H., Eisenberg, R.J. and Spear, PG. (1998) Entry of alphaherpesviruses mediated by poliovirus receptor-related protein 1 and poliovirus receptor. Science, 280: 1618-1620. Goins, W.R, Sternberg, L.R., Croen, K.D., Krause, PR., Hendricks, R.L., Fink, D.J., Straus, S.E., Levine, M. and Glorioso, J.C. (1994) A novel latency-active promoter is contained within the herpes simplex virus t3Ape 1 UL flanking repeats. J. Virol., 68: 2239-2252. Goss, J.R., Mata, M., Goins, W.R, Wu, H.H., Glorioso, J.C. and Fink, D.J. (2001) Antinociceptive effect of a genomic herpes simplex virus-based vector expressing human proenkephalin in rat dorsal root ganglion. Gene Then, 8: 551-556. Goss, J.R., Goins, W.R, Lacomis, D., Mata, M., Glorioso, J.C. and Fink, D.J. (2002a) Herpes simplex-mediated gene transfer of nerve growth factor protects against peripheral neruropathy in streptozotocin-induced diabetes in the mouse. Diabetes, 51: 2227-2232. Goss, J.R., Harley, C.R, Mata, M., O'Malley, M.E., Goins, W.R, Hu, X-P, Glorioso, J.C. and Fink, D.J. (2002b) Herpes vector-mediated expression of proenkephalin reduces pain-related behavior in a model of bone cancer pain. Ann. Neurol., 52: 662-665. Hao, S., Mata, M., Goins, W., Glorioso, J.C. and Fink, D.J. (2003) Transgene-mediated enkephalin release enhances the effect of morphine and evades tolerance to produce a sustained antiallodynic effect. Pain, 102: 135-142.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

220

16. HERPES SIMPLEX VECTOR-MEDIATED GENE TRANSFER TO THE PERIPHERAL NERVOUS SYSTEM

Helgren, M.E., differ, K.D., Torrento, K., Cavnor, C , Curtis, R., DiStefano, P.S., Wiegand, S.J. and Lindsay, R.M. (1997) Neurotrophin-3 administration attenuates deficits of pyridoxine-induced largefiber sensory neuropathy. J. Neurosci., 17: 372-382. Hobbs, W.E., II. and DeLuca, N.A. (1999) Perturbation of cell cycle progression and cellular gene expression as a function of herpes simplex virus ICPO. J. Virol., 73: 8245-8255. Honess, R.W. and Roizman, B. (1974) Regulation of herpesvirus macromolecular synthesis I. Cascade regulation of the synthesis of three groups of viral proteins. J. Virol., 14: 8-19. Kosz-Vnenchak, M., Coen, D.M. and Knipe, D.M. (1990) Restricted expression of herpes simplex virus lytic genes during establishment of latent infection by thymidine kinase-negative mutant viruses. J. Virol., 64: 5396-5402. Krisky, D.M., Wolfe, D., Coins, W.F., Marconi, P C , Ramakrishnan, R., Mata, M., Rouse, R.J., Fink, D.J. and Glorioso, J.C. (1998) Deletion of multiple immediate-early genes from herpes simplex virus reduces cytotoxicity and permits long-term gene expression in neurons. Gene Ther., 5: 1593-1603. Liu, J., Wolfe, D., Hao, S., Huang, S., Glorioso, J.C, Mata, M. and Fink, D.J. (2004) Peripherally delivered glutamic acid decarboxylase gene therapy for spinal cord injury pain. Mol. Ther., 10: 57-66. MacLean, A.R., Ul-Fareed, M., Robertson, L., Harland, J. and Brown, S.M. (1991) Herpes simplex virus type 1 deletion variants 1714 and 1716 pinpoint neurovirulence-related sequences in Glasgow strain 17+ between immediate early gene 1 and the 'a' sequence. J. Gen. Virol., 72: 631-639. Mandel, R.J., Spratt, S.K., Snyder, R.O. and Leff, S.E. (1997) 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, 94: 14083-14088. Markert, J.M., Malick, A., Coen, D.M. and Martuza, R.L. (1993) Reduction and elimination of encephalitis in an experimental glioma therapy model with attenuated herpes simplex mutants that retain susceptibility to acyclovir. Neurosurgery, 32: 597-603. Markert, J.M., Medlock, M.D., Rabkin, S.D., Gillespie, G.Y., Todo, T., Hunter, W.D., Palmer, CA., Feigenbaum, F., Tornatore, C , Tufaro, F. and Martuza, R.L. (2000) Conditionally replicating herpes simplex virus mutant, G207 for the treatment of malignant glioma: results of a phase I trial. Gene Ther., 7: 867-874. McGeoch, D.J., Dalrymple, M.A., Davison, A.J., Dolan, A., Frame, M.C, McNab, D., Perry, L.J., Scott, J.E. and Taylor, R (1988) The complete DNA sequence of the long unique region in the genome of herpes simplex virus type 1. J. Gen. Virol., 69: 1531-1574. Mineta, T., Rabkin, S. and Martuza, R. (1994) Treatment of malignant gliomas using ganciclovir-hypersensitive, ribonucleotide reductase-deficient HSV mutant. Cancer Res., 54: 3963-3966. Mochizuki, H., Hayakawa, H., Migita, M., Shibata, M., Tanaka, R., Suzuki, A., Shimo-Nakanishi, Y, Urabe, T., Yamada, M., Tamayose, K., Shimada, T., Miura, M. and Mizuno, Y (2001) An AAV-derived Apaf-1 dominant negative inhibitor prevents MPTP toxicity as antiapoptotic gene therapy for Parkinson's disease. Proc. Natl. Acad. Sci. USA, 98:10918-10923. Montgomery, R.L, Warner, M.S., Lum, B.J. and Spear, P.G. (1996) Herpes simplex virus-1 entry into cells mediated by a novel member of the TNF/NGF receptor family. Cell, 87: 427-436. Newcomb, W W , Homa, F.L., Thomsen, D.R., Trus, B.L., Cheng, N., Steven, A., Booy, F. and Brown, J.C. (1999) Assembly of the herpes simplex virus procapsid from purified components and

identification of small complexes containing the major capsid and scaffolding proteins. J. Virol., 73: 4239-4250. Pradat, P.F., Kennel, P., Naimi-Sadaoui, S., Finiels, R, Orsini, C , Revah, R, Delaere, P. and Mallet, J. (2001) Continuous delivery of neurotrophin 3 by gene therapy has a neuroprotective effect in experimental models of diabetic and acrylamide neuropathies. Hum. Gene Ther., 12: 2237-2249. Pradat, PR, Kennel, P., Naimi-Sadaoui, S., Finiels, R, Scherman, D., Orsini, C , Delaere, P, Mallet, J. and Revah, R (2002) Viral and non-viral gene therapy partially prevents experimental cisplatininduced neuropathy. Gene Ther., 9: 1333-1337. Puskovic, v., Wolfe, D., Goss, J., Huang, S., Mata, M., Glorioso, J.C. and Fink, D.J. (2004) Prolonged biologically active transgene expression driven by HSV LAP2 in brain in vivo. Mol. Then, 10: 67-75. Raja, S.N., Meyer, R.A., Ringkamp, M. and Campbell, J.N. (2002) Peripheral neural mechanisms of nociception. In: Wall P.D. and Melzack R. (Eds.), Textbook of Pain (4th ed.). Churchill Livingstone, Edinburgh, pp. 11-58. Rampling, R., Cruickshank, G., Papanastassiou, V, Nicoll, J., Hadley, D., Brennan, D., Petty, R., MacLean, A., Harland, J., McKie, E., Mabbs, R. and Brown, M. (2000) Toxicity evaluation of replication-competent herpes simplex virus (ICP 34.5 null mutant 1716) in patients with recurrent malignant glioma. Gene Ther., 7: 859-866. Roizman, B. and Sears, A. (1996) Herpes simplex viruses and their replication. In: Fields B.N., Knipe D.M., Howley P.M., Chanock R.M., Hirsch M.S., Melnick J.L., Monath T.P and Roizman B. (Eds.), Fields Virology (3rd ed.) Lippincott-Raven, Philadelphia, PA, pp. 2231-2295. Schaumburg, H., Kaplan, J., Windebank, A., Vick, N., Rasmus, S., Pleasure, D. and Brown, M.J. (1983) Sensory neuropathy from pyridoxine abuse. A new megavitamin syndrome. N. Engl. J. Med., 309: 445-448. Schmidt, R.E., Dorsey, D.A., Beaudet, L.N., Parvin, C A . and Escandon, E. (2001) Effect of NGF and neurotrophin-3 treatment on experimental diabetic autonomic neuropathy. J. Neuropathol. Exp. Neurol, 60: 263-273. Schwei, M.J., Honore, P., Rogers, S.D., Salak-Johnson, J.L., Finke, M.P, Ramnaraine M.L., Clohisy, D.R. and Mantyh, P W (1999) Neurochemical and cellular reorganization of the spinal cord in a murine model of bone cancer pain. J. Neurosci., 19:10886-10897. Smith, G.A. and Enquist, L.W. (2002) BREAK INS AND BREAK OUTS: Viral Interactions with the Cytoskeleton of Mammalian Cells. Annu. Rev Cell Dev Biol., 18: 135-161. Soreq, H., Ben-Aziz, R., Prody, CA., Seidman, S., Gnatt, A., Neville, L., Lieman-Hurwitz, J., Lev-Lehman, E., Ginzberg, D. Lipidot-Lifson, Y Zakut, H. (1990) Molecular cloning and construction of the coding region for human acetylcholinesterase reveals a G + C-rich attenuating structure. Proc. Natl. Acad. Sci. USA, 87: 9688-9692. Spear, P C , Eisenberg, R.J. and Cohen, G.H. (2000) Three classes of cell surface receptors for alphaherpesvirus entry. Virology, 275:1-8. Stevens, J.G. (1989) Human herpesvisruses: a consideration of the latent state. Microbiol. Rev, 53: 318-332. Tenser, R.B., Hay, K.A. and Edris, WA. (1989) Latency-associated transcript but not reactivatable virus is present in sensory ganglion neurons after inoculation of thymidine kinase-negative mutants of Herpes Simplex Virus Type 1. J. Virol., 63: 2861-2865. Thomas, P.K. (1999) Diabetic peripheral neuropathies: their cost to patient and society and the value of knowledge of risk factors for development of interventions. Eur. Neurol., 41: 35-43.

IV GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

ACKNOWLEDGMENTS Thomas, RK. and Ochoa, J. (1993) Symptomatology and differential diagnosis of peripheral neuropathy: clinical features and differential diagnosis. In: Dyck RJ., Thomas RK., Griffin J.W., Low RA. and Roduslo J.R (Eds.), Peripheral Neuropathy (3rd ed.). W.B. Saunders, Philadelphia, pp. 749-774. Whitley, R.J., Kern, E.R., Chatterjee, S., Chou, J. and Roizman, B. (1993) Replication, establishment of latency, and induced reactivation of herpes simplex virus gamma 1 34.5 deletion mutants in rodent models. J. Clin. Invest., 91: 2837-2843. Wilson, S.R, Yeomans, D.C., Bender, M.A. and Glorioso, J.C. (1996) Delivery of enkephalins to mouse sensory neurons by a herpes virus encoding proenkephalin. Soc. Neurosci. Abs., 22: 1362. Windebank, A.J., Low, R.A., Blexrud, M.D., Schmelzer, J.D. and Schaumburg, H.H. (1985) Ryridoxine neuropathy in rats: specific degeneration of sensory axons. Neurology, 35: 1617-1622.

221

Woolf, C.J. and Costigan, M. (1999) Transcriptional and posttranslational plasticity and the generation of inflammatory pain. Rroc. Natl. Acad. Sci. USA, 96: 7723-7730. Xu, Y, Gu, Y, Xu, G.Y, Wu, R, Li, G.W and Huang, L.Y (2003) Adeno-associated viral transfer of opioid receptor gene to primary sensory neurons: a strategy to increase opioid antinociception. Rroc. Natl. Acad. Sci. USA, 100: 6204-6209. Yamada, Y, Kimura, H., Morishima, T., Daikoku, T., Maeno, K. and Nishiyama, K. (1991) The pathogenicity of ribonucleotide reductase-null mutants of herpes simplex virus type 1 in mice. J. Infect. Dis., 164: 1091-1097. Yura, Y, Iga, H., Kondo, Y, Harada, K., Tsujimoto, H., Yanagawa, T., Yoshida, H. and Sato, M. (1992) Heparan sulfate as a mediator of herpes simplex virus binding to basement membrane. J. Invest. Dermatol., 98: 494-498.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

C H A P T E R

17 Characterization of Pain Receptors in the Spinal Cord Using a Viral Vector for Spatial-Temporal Gene Targeting Charles E. Inturrisi

Abstract: The purpose of this chapter is to describe the use of viral vectors to facilitate a spatial and temporal loss of function mutation. The goal of these studies was to spatially localize the deletion so that the locus and contribution of the gene of interest to pain transmission and amplification could be unequivocally established. The gene target is the NRl subunit of the N-methyl-D-aspartate (NMDA) receptor. To determine the importance of the NMDA receptor in pain hypersensitivity following injury, the NRl subunit was selectively deleted in the lumbar spinal cord of adult mice by the localized injection of an adeno-associated virus expressing the Cre recombinase into floxed NRl mice. NRl subunit mRNA and dendritic protein are reduced by 80% in the area of the virus injection and NMDA, but not AMPA, currents are abolished in lamina II neurons. The spatial NRl knockout does not alter heat or cold paw withdrawal latencies, mechanical threshold, or motor function. However, injury-induced pain produced by intraplantar formalin is reduced by 70%. Our results demonstrate conclusively that the postsynaptic NRl receptor subunit in the lumbar dorsal horn of the spinal cord is required for central sensitization, the central facilitation of pain transmission produced by peripheral injury. Keywords: NMDA receptor; conditional knockout; Cre-loxP; injury-induced pain; fromalin; spinal cord dorsal horn; adeno-associated virus; synaptic transmission; central sensitization

L

Intense or sustained noxious stimuli that are associated with tissue injury result in a temporal summation of postsynaptic depolarizations, which remove the voltage-dependent magnesium block of the NMDA receptor allowing calcium influx into the postsynaptic cell (Woolf and Salter, 2000). These events also generate an intracellular cascade involving a number of pain-activated proteins that results in a prolonged increase in the excitability of spinal cord pain transmission neurons, a process described as central sensitization (Woolf and Costigan, 1999; Woolf and Salter, 2000). Central sensitization produces changes in gene expression and phenotype changes in spinal cord dorsal horn (SCDH) and dorsal root ganglia (DRG) neurons that results in pain hypersensitivity characterized by spontaneous pain, hyperalgesia (an increase in the intensity of a painful stimulus) and allodynia

INTRODUCTION

With the recognition of the ability of viral vectors to deliver and express genes in the CNS (Kaplitt et al., 1994a, b), a number of recent studies have utilized viral vectors to express proteins with analgesic activity (Fink et al., 2003; Glorioso et a l , 2003) (see also Chapter 16 in this volume). The purpose of this chapter is to describe the use of viral vectors to facilitate a spatial and temporal loss of function mutation. The goal of these studies was to spatially localize the deletion so that the locus and contribution of the gene of interest to pain transmission and amplification could be unequivocally established. The gene target is the NRl subunit of the N-Methyl-D-Aspartate (NMDA) receptor.

Gene Therapy of the Central Nervous System: From Bench to Bedside

223

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

224

17. CHARACTERIZATION OF PAIN RECEPTORS IN THE SPINAL CORD USING A VIRAL VECTOR FOR SPATIAL-TEMPORAL GENE TARGETING

(an innocuous stimulus is now painful). Thus, central plasticity changes that are believed to be triggered by activation of NMDA receptors in the SCDH results in facilitated pain transmission. Many pharmacological studies have identified the role of NMDA receptors in injury-induced hypersensitivity states. How^ever, neither a systemic nor a centrally applied NMDA receptor antagonist can provide assurance that all or even most of the available receptors have been blocked or those that are blocked, are limited to the dorsal horn or even that, at the higher doses sometimes used, that the antagonism is limited to NMDA receptors. Furthermore, the narrow in vivo therapeutic index has not provided convincing evidence of the efficacy of targeting only spinal cord dorsal NMDA receptors or a strategy for targeting only these receptors (Kristensen et al., 1992; Boyce and Rupniak, 2002; South et a l , 2003). What was required for the essential behavioral studies was a knockout of the NMDA receptor that was confined to the L4-L6 region of the lumbar SCDH, which comprises the sensory input to the SCDH. To allow the contralateral side to serve as a control for these studies, the knockout was limited to one side of the SCDH. Native NMDA receptors are heterooligomers of subunits that in the mouse include NRl (CI), el through 4 (NR2A through NR2D) and NR3 (Dingledine et al., 1999). Studies of expressed NMDA receptors (Dingledine et al., 1999) and with neurons of a CAl hippocampal NRl KO mouse (Tsien et al., 1996) show that the NRl subunit is required to generate characteristic NMDA currents. Therefore, our target was deletion of the NRl subunit of the NMDA receptor. However, the NMDA receptor is essential for normal development and its constitutive deletion results in early postnatal lethality (Li et al., 1994). A conditional knockout can delete a gene both spatially and temporally (Tsien et al., 1996), avoiding the complications resulting from deletion of the gene in systems other than that which affects the response, gene redundancy during development or embryonic or postnatal lethality (Tsien et al., 1996). The requirements of a temporal, as well as, a very localized spatial knockout of the NRl gene favored the Cre/loxP gene deletion system (Wilson and Kola, 2001). Fortunately, a mouse engineered with loxP sites in the NRl gene had been prepared in the Tonegawa laboratory (Tsien et al., 1996) and was available to us. The floxed NRl (fNRl) gene locus has loxP sites between exons 10 and 11 and 3' downstream, after the exon 22, the last exon. The two loxP sequences flank the coding region for the four membrane domains (3TMs and the M2 recurrent

loop) and the entire C-terminal sequence of the NRl polypeptide chain (Tsien et al., 1996). Thus, the Cremediated recombination could be expected to result in deletion of essential coding regions of the NRl gene. The floxed mice are in a C57/B6 background and all comparisons were made among mice with this background (South et al., 2003). This floxed NRl mouse had been bred to transgenic mice that express Cre recombinase to demonstrate the ability of loxP sites to serve as substrates for Cre, and to produce an NRl knockout that was conditionally restricted to the CA-1 region of the hippocampus (Tsien et al., 1996). For the localized delivery of Cre recombinase to the SCDH, we selected a recombinant adeno-associated virus that uses a cytomegalovirus promoter to drive the expression of a fusion protein of the reporter gene green fluorescent protein (GFP) and Cre recombinase (rAAV-GFP-Cre) (South et al., 2003). The expression of GFP allowed us to identify in postmortem tissue sections those regions of the CNS that had been transduced by the virus. rAAV vectors can transduce postmitotic neurons in the SCDH in a highly spatially localized manner (Kaplitt et al., 1994b; Peel et al., 1997; Kaspar et a l , 2002; Washbourne and McAllister, 2002; South et al., 2003). The cytotoxicity and immunogenicity of rAAV vectors is very low (Kaplitt et al., 1994b; Kaspar et al., 2002) and at our institution rAAV-based vectors are BioSafety Level-1 and approved for use in vivo in rodents. This safety feature makes rAAV vectors much more convenient than herpes or lentivirus vectors for use in rodents that are also undergoing a variety of repeated behavioral tests. Attempts to transduce the SCDH neurons following intrathecal administration of viral vectors into CSF were unsuccessful since the virus is avidly taken up by ependimal cells and does not penetrate into the SCDH (Finegold et al., 2001). Therefore, the virus is microinjected directly into the SCDH (intraparenchymal injection, IPX). The mouse (South et al., 2003) is anesthetized and placed in a specially designed spinal frame that supports the abdomen and pelvis, and a laminectomy is used to remove spinous process VL2 and part of VL3. We found that three unilateral injections of 1 |Lil (1 X 10^ genomic particles/|Lil) provided the requisite transduction of the ipsilateral SCDH (see below). The injections were made 0.5 mm apart, at a depth of 0.3 mm into SCDH, using a glass pipette with a 40 \im diameter tip attached to a 5 ]jil Hamilton syringe. The syringe is mounted on a microinjector (KOPF Model 5000) attached to a stereotaxic unit (KOPF Model 960). In the target area 95% of neurons are transduced (South et al., 2003).

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

225

INTRODUCTION

GFP 5 0 0 jim

SOOjiim

FIGURE 1 The intraparenchymal injection of rAAV-GFP-Cre into the spinal cord dorsal horn of a floxed NRl mouse results in viral transduction, Cre-mediated recombination and a spatial-temporal knockout of the NRl gene. (A, B) On the side ipsilateral to the injection of rAAV-GFP-Cre, viral transduction results in the expression of GFP immunolabeling and the disappearance of NRl mRNA as measured by in situ hybridization. Modified from South et al. (2003). Copyright 2003 by the Society for Neuroscience.

Two weeks after IPI injection of rAAV-GFP-Cre into the lumbar SCDH of an adult fNRl mouse, a highly localized and reproducible pattern of expression of GFP is observed. GFP expression is restricted to the ipsilateral dorsal and part of the ipsilateral ventral horn (Fig. lA). This expression persists for as long as 10 months after administration of the viral vector. Region-specific Cre-mediated recombination is evident where reduction in NRl mRNA labeling was measured in an adjacent section (Fig. IB) by in situ hybridization. An antisense riboprobe was used whose sequence will be deleted at the DNA level by the Cre/loxP recombination. Three injections of rAAV-GFP-Cre at 0.5 mm intervals result in a segment-specific expression of GFP and a deletion of NRl mRNA that extends for 3.75 ± 0.25 (SEM) mm rostro-caudally to incorporate the L4-L6 spinal segments. These are the target areas that comprise the sciatic sensory distribution from the paw to the dorsal horn. The extent of the GFP label correlated almost perfectly with the area of reduced NRl mRNA (Figs. lA and B). Using this injection protocol effectively, the entire ipsilateral dorsal horn is depleted of NRl mRNA leaving the contralateral dorsal horn and non-lumbar spinal cord completely intact. Residual mRNA labeling represents NRl mRNA in presynaptic terminals, which are not affected by the virus (South et al., 2003). The ratio of ipsilateral to contralateral labeling of the neuronal-specific nuclear marker NeuN is not significantly different indicating no loss of neurons see (South et al., 2003). SCDH sections from animals that received the control vector (rAAV-GFP) show no difference in the ratio percent of ipsilateral to contralateral labeling of NRl mRNA or of NeuN (South et al., 2003).

NRl protein labeling of individual neurons in lamina II was evident on the contralateral side and absent in lamina II of the rAAV-GFP-Cre-injected side (South et al., 2003). At the ultrastructural level, the mean numbers of NRl-labeled dendritic profiles were reduced by 80% (Figs. 2A and B). This represents a conservative estimate as the profiles in lamina II indicate nearly complete depletion. No partial labeling was observed, which is consistent with a complete deletion of NRl (a NRl KO) in each cell that was transduced. Profiles on the rAAV-GFP-Cre-injected side that lacked NRl labeling were not degenerating, indicating that the absence of NRl-labeling was not due to death or damage to NRl-expressing cells (South et al., 2003). In the target area 95% of neurons, but not glia, are transduced (Kaspar et a l , 2002; South et a l , 2003). To determine whether IPI administration of rAAVGFP-Cre was associated with retrograde transport of the vectors, we examined the ipsilateral and contralateral DRG for GFP immunolabeling, and NRl mRNA (in situ) in sections that correspond to the lumbar spinal cord segments where recombination had occurred following rAAV-GFP-Cre or after rAAVGFP. No labeling for GFP was observed in the DRG nor was any evidence obtained of a decrease in NRl mRNA in the ipsilateral DRG compared with the contralateral DRG. We have also examined sections from the thoracic and cervical spinal cord and the brain and find no evidence for the expression of GFP. The deletion of the NRl subunit in fNRl mice appears to be confined to SCDH cells, primarily in postsynaptic elements (dendrites and somata) and does not extend to the corresponding DRG cells.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

226

17. CHARACTERIZATION OF PAIN RECEPTORS IN THE SPINAL CORD USING A VIRAL VECTOR FOR SPATIAL-TEMPORAL GENE TARGETING

A

Contralateral

rAAV-GFP-Cre

.*^'-.'

1 |im

B

I] contralateral side 300 (A

•= 250 o ^ cc z o

200-1

c

100

•

rAAV-GFP-Cre side

X

150

XJ

0)

50

FIGURE 2 The intraparenchymal injection of rAAV-GFP-Cre into the spinal cord dorsal horn of a floxed NRl mouse results in a decrease in NRl protein that is confined to the ipsilateral (rAAV-GFP-Cre-injected side) of the SCDH (A) Ultrastructural analysis (electron microscopy) of the spinal cord dorsal horn of an rAAV-GFP-Cre-injected mouse demonstrates a loss of NRl protein labeling in dendrites (u-d) and somata (u-s) compared to the contralateral side (NRl-d and NRl-s) n = nucleus. (B) Measurement of the number of mean dendritic NRl profiles indicates an 80% reduction in NRl protein labeling on the rAAV-GFP-Cre-injected side. Modified from South et al. (2003). Copyright 2003 by the Society for Neuroscience.

v^hich constitute the major presynaptic source of NRl (South et al., 2003). To investigate w^hether the viral vectors produced changes in NMDA-receptor function electrophysiological studies w^ere conducted. We found that Spontaneous Excitatory Postsynaptic Current (EPSC) activity, membrane resting potential and AMPA receptor-mediated current responses were unaffected in vector control (rAAV-GFP) or in rAAV-GFP-Cre mice (South et a l , 2003). In contrast the current amplitudevoltage relationship mediated by NMDA receptors w^as abolished in rAAV-GFP-Cre mice (Figs. 3A and B). In addition, synaptic responses mediated by the NMDA receptor, but not the AMPA receptor, w^ere

almost completely abolished in rAAV-GFP-Cre mice (Fig. 3C). We focused our initial behavioral evaluation on any consequences that might result from the IPI procedure. Since the blockade of NMDA receptors by NRl antagonists is often associated w^ith significant effects of motor coordination a n d / o r on threshold responses to stimuli (Boyce and Rupniak, 2002), we also compared groups of fNRl mice before and at 7.-\ weeks after IPI of either the control vector (rAAV-GFP) or the Cre-expressing vector (rAAV-GFP-Cre). Motor coordination was not affected by the IPI of either viral vector (Fig. 4A). Next, we determied whether IPI affected non-injury inducing noxious stimuli. Compared to the responses prior to IPI, the administration of either viral vector did not alter the latency for thermal tail withdrawal from a water bath maintained at 48, 52.5 or 55°C (Fig. 4B), the latency for thermal paw withdrawal at low (Fig. 4C) or high intensity (Fig. 4D) the mechanical stimulus threshold to von Frey filaments (Fig. 4E), the cold stimulus threshold to evaporating acetone (Fig. 4F). When the paw ipsilateral to the IPI was compared with the contralateral paw, no difference was observed in the responses to acute mechanical or thermal stimuli (Figs. 4C-F). Neither the IPI procedure for delivering the virus nor the spatial-temporal KO of NRl affected the response to non-injury-inducing stimuli. Thus, deletion of the NMDA receptor from the SCDH does not alter the transmission of "protective'' pain. Intraplantar formalin evokes two phases of spontaneous injury-induced pain behaviors, an immediate, short-lasting phase due to the direct irritant effects of the chemical, and a slower onset, longer-lasting phase that reflects a combination of ongoing sensory input and central sensitization. Figure 5 shows that formalininduced pain (licking and biting the affected paw) that occurred during phase 2 (10-60 min after the formalin injection) was decreased by 70-75% in the rAAV-GFPCre (NRl KO) group but not in the vehicle and rAAVGFP treated. Our results demonstrate conclusively that the postsynaptic NMDARl receptor subunit in the lumbar dorsal horn of the spinal cord is required for central sensitization, the central facilitation of pain transmission produced by peripheral injury. Importantly, we could attenuate this response to formalin without any of the adverse effects that often accompany the systemic or IT administration of NMDA receptor antagonists. We are currently extending these studies to other models of persistent injury-induced inflammatory or nerve injury-induced pain. In addition, by comparing gene and protein expression in pain models in the presence and absence of the NRl knockout, we have

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

227

INTRODUCTION A. Exogenously applied NMDA Untreated NMDA

V

rAAV-GFP

rAAV-GFP-Cre

_

<

-60

L

Q.

^

* Y j | ^

|¥l'C|if|pi"H"lll'

30 s

J25pA

nv^

. 4 o . ! ! ^ j ^

NMDA Current (-70mV)

-40

L

•D

I -20 E <

0

^

<

-70

^

-50

NMDA Current (-40mV)

1,

•o

i

-30

Q.

I -10 ^-^
+40 mV

0 •D 13 Q.

F

<

1?0

NMDA Current (+40mV)

L

90 60 30 0

Untreated

rAAVGFP

rAAVGFP-Cre

B. Current Amplitude-Voltage relationship -•- Untreated - ^ rAAV-GFP -^ rAAV-GFP-Cre

120 80

0. NMDA EPSCs evoked by dorsal root stimulation at +40 mV Voltage (mV) -70 -40

Untreated 0 =5. £

<

-40

Evoked NMDA Current (-H40mV) 1001

rAAV-GFP-Cre

f-

j

50 pA

I I I

-80

60 40 20 0

Untreated

rAAVGFP-Cre

FIGURE 3 Current and synaptic responses mediated by NMDA receptors are diminished in rAAVGFP-Cre mice. (A-C) Whole cell patch clamp recordings were made in spinal cord slices from C57BL/6 (untreated controls), or floxed NRl mice that received intraparenchymal injections of rAAV-GFP or rAAV-GFP-Cre. (A) NMDA (50 viM) was bath-applied to spinal cord slices from untreated, rAAV-GFP and rAAV-GFP-Cre mice with the voltage of the lamina II neurons clamped at four different holding potentials, - 7 0 , - 4 0 , 0 and +40 mV in the presence of tetrodotoxin (TTX, 1 ]iM) to inhibit synaptic effects resulting from the application of NMDA. Exogenously applied NMDA elicits an inward current at - 7 0 or - 4 0 mV and induces outward current at +40 mV. Most neurons from untreated and rAAV-GFP mice had NMDA receptor components at - 7 0 , - 4 0 or +40 mV. In contrast, the neurons from rAAV-GFPCre-injected mice exhibited minimal or absent NMDA receptor responses at each of these potentials (*;?<0.01). (B) The current-voltage {I/V) relationship that is characteristic of the NMDA ion channel is present in untreated and rAAV-GFP mice but absent in rAAV-GFP-Cre mice. (C) NMDA receptor-mediated primary afferent-evoked EPSC amplitudes (at +40 mV) are significantly decreased in rAAV-GFP-Cre mice compared with untreated mice. The responses were measured in the presence of a glycine receptor antagonist (strychnine, 1 |LiM), a GABA^ receptor antagonist (bicuculline, 20 |LiM) and an AMPA/kainate receptor antagonist (CNQX, 20 ]iM) {y < 0.01). From South et al. (2003). Copyright 2003 by the Society for Neuroscience.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

228

17. CHARACTERIZATION OF PAIN RECEPTORS IN THE SPINAL CORD USING A VIRAL VECTOR FOR SPATIAL-TEMPORAL GENE TARGETING

. Motor c o o r d i n a t i o n ^ Before virus • After rAAV-GFP • i After rAAV-GFP-Cre

D. T h e r m a l p a w w i t h d r a w a l t h r e s h o l d (high s t i m u l u s ) I ^ 8 1

I injected side Non-injected side

6

m m

A. Formalin-induced injury behavior (Phases) I I fNR1 (Vehicle) ^ ^ fNR1 (rAAV-GFP) ^ — ^

CD After rAAV-GFP •

After rAAV-GFP-Cre

E. M e c h a n i c a l t h r e s h o l d ( v o n Prey) I I Injected side ! • Non-injected side :D

1.00 ^

o

I

0.75 ] 0.50 0.25

48

52.5

55

0.00

T e m p e r a t u r e (°C) C. T h e r m a l p a w w i t h d r a w a l t h r e s h o l d (low stimulus) I I Injected side I Non-injected side 12 .10

B

4

m

Before After After rAAVvirus rAAV-GFP GFP-Cre

Before After After rAAVvlrus rAAV-GFP GFP-Cre

F. Cold sensitivity t h r e s h o l d I I Injected side • • Non-injected side

0.8 § 0.6

o CO 0.4 0.2 0.0

(iiiii

Before After After rAAVvirus rAAV-GFP GFP-Cre

FIGURE 4 A spatial-temporal knockout of NRl in the spinal cord dorsal horn of floxed NRl mice does not affect motor coordination and non-injury-inducing stimuli. (A) Motor coordination was measured using the rotarod test. (B) A thermal stimulus was applied to the tail (thermal tail withdrawal threshold) at 48, 52.5 and 55°C. (C, D) A brief thermal stimulus of low (C) and higher (D) intensity was applied to the paw. (E) Mechanical (tactile) stimuli were applied using von Frey hairs. (F) Cold sensitivity was measured by applying a drop of acetone to the paw. From South et al. (2003). Copyright 2003 by the Society for Neuroscience.

begun to identify the genes and proteins that signal pain downstream of the NMDA receptor in SCDH. The rAAV vector can be used to produce a localized spatial knockout virtually anywhere in the CNS. In a preliminary study we have produced an NRl knockout that is localized to the ventral tegmental area, which is part of the supraspinal opioid and psychostimulant reward circuitry (Loonam and Inturrisi, unpublished observations). We have demonstrated that rAAV vectors provide a means for the precise and reproducible spatial deliv-

300

? 250

_LT

I 200-1 c 150 o tCD 100

Before After After rAAVvlrus rAAV-GFP GFP-Cre B. T h e r m a l tail w i t h d r a w a l threshold &B Before virus

fNR1 (rAAV-GFP-Cre)

Q. CO CD

E

50 -{ Phase 1

Phase 2

FIGURE 5 A spatial-temporal knockout of NRl in the spinal cord dorsal horn of floxed NRl (fNRl) mice significantly decreases the painful response to an injury-inducing stimulus measured during phase 2 of the formalin test. The intraplantar injection of formalin (5%) results in licking activity that can be divided into two phases, 0-10 min (phase 1) and a second phase, which begins at 10 and continues to 60 min (phase 2). During phase 1 the area under the response to formalin over time curve is the same in fNRl mice that received vehicle, rAAV-GFP or rAAV-GFP-Cre treatment. During phase 2 the response of rAAV-GFP-Cre mice is reduced 70% compared to vehicle and rAAV-GFP mice {*p ^ 0.05). Modified from South et al. (2003). Copyright 2003 by the Society for Neuroscience.

ery of a gene designed to delete a target gene in the CNS. A recent report has demonstrated that rAAV can efficiently deliver siRNA sequences into the CNS (Hommel et al,, 2003). In the near future this vector should prove very useful in delivering siRNAs to silence the expression of genes involved in pain transmission and amplification.

ACKNOWLEDGMENTS This research was supported in part by NIDA grants DA001457, DA005130 and DA000198. References Boyce, S. and Rupniak, N.M. (2002) Behavioral studies on the potential of NMDA receptor antagonists as analgesics. In: Sirinathsinghji D.J.S. and Hill R.G (Eds.), NMDA antagonists as potential analgesic drugs. Birkhauser Verlag, Basel, pp. 147-164. Dingledine, R., Borges, K., Bowie, D, and Traynelis, S.F. (1999) The glutamate receptor ion channels. Pharmacol. Rev., 51(1): 7-61. Finegold, A.A., Perez, KM. and ladarola, M.J. (2001) In vivo control of NMDA receptor transcript level in motoneurons by viral transduction of a short antisense gene. Brain Res. Mol. Brain Res., 90(1): 17-25. Fink, D., Mata, M. and Glorioso, J.C. (2003) Cell and gene therapy in the treatment of pain. Adv. Drug Deliv Rev, 55(8): 1055-1064.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

ACKNOWLEDGMENTS Glorioso, J.C., Mata, M. and Fink, D.J. (2003) Therapeutic gene transfer to the nervous system using viral vectors. J. NeuroviroL, 9(2): 165-172. Hommel, J.D., Sears, R.M., Georgescu, D., Simmons, D.L. and DiLeone, R.J. (2003) Local gene knockdown in the brain using viral-mediated RNA interference. Nat. Med., 9(12): 1539-1544. Kaplitt, M.G., Kwong, A.D., Kleopoulos, S.R, Mobbs, C.V., Rabkin, S.D. and Pfaff, D.W. (1994a) 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, 91(19): 8979-8983. Kaplitt, M.G., Leone, R, Samulski, R.J., Xiao, X., Pfaff, D.W., O'Malley, K.L. and During, M.J. (1994b) Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat. Genet., 8(2): 148-154. Kaspar, B.K., Vissel, B., Bengoechea, T., Crone, S., Randolph-Moore, L., Muller, R., Brandon, E.R, Schaffer, D., Verma, I.M., Lee, K.R, Heinemann, S.R and Gage, F.H. (2002) Adeno-associated virus effectively mediates conditional gene modification in the brain. Proc. Natl. Acad. Sci. USA, 99(4): 2320-2325. Kristensen, J.D., Svensson, B. and Gordh, T, Jr. (1992) The NMDAreceptor antagonist CPP abolishes neurogenic 'wind-up pain' after intrathecal administration in humans. Pain, 51(2): 249-253. Li, Y, Erzurumlu, R.S., Chen, C , Jhaveri, S. and Tonegawa, S. (1994) Whisker-related neuronal patterns fail to develop in the tri-

229

geminal brainstem nuclei of NMDARl knockout mice. Cell, 76(3): 427-437. Peel, A.L., Zolotukhin, S., Schrimsher, G.W., Muzyczka, N. and Reier, P.J. (1997) Efficient transduction of green fluorescent protein in spinal cord neurons using adeno-associated virus vectors containing cell type-specific promoters. Gene Then, 4(1): 16-24. South, S.M., Kohno, T., Kaspar, B.K., Hegarty D., Vissel, B., Drake, C.T., Ohata, M., Jenab, S., Sailer, A.W., Malkmus, S., Masuyama, T., Horner, R, Bogulavsky J., Gage, RH., Yaksh, T.L., Woolf, C.J., Heinemann, S.R and Inturrisi, C.E. (2003) A conditional deletion of the NRl subunit of the NMDA receptor in adult spinal cord dorsal horn reduces NMDA currents and injury-induced pain. J. Neurosci., 23(12): 5031-5040. Tsien, J.Z., Huerta, RT. and Tonegawa, S. (1996) The essential role of hippocampal CAl NMDA receptor-dependent synaptic plasticity in spatial memory Cell, 87(7): 1327-1338. Washbourne, P. and McAllister, A.K. (2002) Techniques for gene transfer into neurons. Curr. Opin. NeurobioL, 12(5): 566-573. Wilson, T.J. and Kola, L (2001) The LoxP/CRE system and genome modification. Methods Mol. Biol., 158: 83-94. Woolf, C.J. and Costigan, M. (1999) Transcriptional and posttranslational plasticity and the generation of inflammatory pain. Proc. Natl. Acad. Sci. USA, 96(14): 7723-7730. Woolf, C.J. and Salter, M.W. (2000) Neuronal plasticity: increasing the gain in pain. Science, 288(5472): 1765-1769.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

C H A P T E R

18 Viral Vector Mediated Gene Therapy of Pain Luc]asmin, Peter T. Ohara

Abstract: Gene therapy using viral vectors targeting primary sensory neurons or the central nervous system is fast becoming a promising therapeutic option. In vivo pre-clinical studies show that various vectors reliably infect pain related neurons and gene expression is observed for many weeks. Several groups have already demonstrated the feasibility of viral vector mediated delivery of enkephalin, mu-opioid receptor, glutamic acid decarboxylase (GAD), pro-opiomelanocortin (POMC) precursor, ^-endorphin, glial-derived neurotrophic factor (GDNF) and semaphorin 3A genes, and clinical trials are planned. Direct delivery of these genes to restricted regions of the nervous system clearly reduces pain behavior and is likely to avoid many of the problems of systemic acting analgesic drugs. Keywords: Chronic pain; GAD; GABA; opioids; enkephalin; viral vector; spinal cord; rostral agranular insular cortex; amygdala

L

occurs or whether it is the same for all neuropathic pain is not known. Most of our understanding and current studies are concentrated on events in the peripheral nerves or spinal cord dorsal horn and to a lesser extent in parts like the brainstem, thalamus and cerebral cortex. The changes found in the spinal cord in experimental models of neuropathic pain range from loss or growth of primary afferents, loss of neurons and up- and/or down-regulation of a variety of factors including neurotransmitter receptors, nerve growth factors (NGFs) and inflammatory mediators. It would be convenient for developing therapies if there was a single causative factor on which one could concentrate, but it is unlikely that for a disease as complex as neuropathic pain, targeting a single factor will ever be sufficient. Instead, one has a bewildering variety of choices for intervention and, although initially daunting, it encourages the idea that among the many possibilities a few will provide good targets for therapy. In view of the challenge that treating chronic pain presents, gene therapy using viral vectors has clear

INTRODUCTION

Pain can be broadly divided into two categories. The first is acute pain, and is usually associated with some form of tissue injury, has a protective role and subsides with the remission of the injury. This type of pain is relatively easy to ameliorate with short acting drugs such as antiinflanimatory agents, opiates or local anesthetics. The second major category is the pain that persists long after the causative injury has healed, and serves no practical function. Such chronic pain often comes under the heading neuropathic pain indicating that the nervous system, rather than the tissue damage, has become the pain generator. Treatment for chronic, neuropathic pain is usually complex and multidimensional and has driven the generation of pain clinics and pain specialties over the last two decades. Neuropathic pain might be easier to treat if we fully understand the underlying cause(s). While it is clear that the underlying cause is a change in the nervous system what exactly that change is, where it

Gene Therapy of the Central Nervous System: From Bench to Bedside

231

Copyright © 2006, Elsevier, Iru:. All rights of reproduction in any form reserved.

232

18. VIRAL VECTOR MEDIATED GENE THERAPY OF PAIN

advantages over other presently available methods. Chronic pain, by its nature, requires ongoing treatment and thus w^ith drug therapy the issue of side effects, addiction and resistance are important considerations. Other treatments such as implanting stimulators within, or over the brain or spinal cord involve the usual risks of surgery, are not universally effective, and sometimes stop working after a few months. IL

OPIATE^BASED STRATEGY

The initial attempts at gene therapy aimed at reducing pain have been directed at the same targets as conventional treatments. One of the most powerful and consistently successful classes of pain relieving drugs are those that act on opioid receptors, the most well known and naturally occurring being morphine. There are several subtypes of opioid receptors and they occur at all levels of the nervous system including nociceptive receptors and their axons, the spinal cord, the brainstem and cerebral cortex. Opioid receptors are also present in non-neural tissue such as the gut and cells of the immune system. Thus, the opioid receptors are a clear target for any antinociceptive therapy but unfortunately opioid acting drugs also have well-known side effects such as respiratory depression, somnolence, nausea, itching, addiction and tolerance, many of which are due to the occurrence of the opioid receptors outside the pain pathway. The hope is that gene therapy aimed at opioid receptors will avoid some of the problems encountered with systemic administration of opiates. The majority of viral vector-mediated gene delivery has been aimed at the primary sensory neurons (Fig. 1) and in some ways follow the path laid by previous pain research and methodology. Early experiments showed that it was possible by introducing herpes simplex virus (HSV) into peripheral tissue and through axonal transport of the virus it was possible to infect dorsal root ganglion (DRG) cells and induce the expression of marker genes (Dobson et al., 1990; Ecob-Prince et al., 1995). The advantages of HSV for this application is that it naturally establishes a latent infection in sensory neurons that does not cause any pathology (Fig. 2). The relative simplicity of this approach led to experiments injecting replication-deficient HSV into the foot of mice to deliver the proenkephalin cDNA to the DRG, where the ceU bodies of primary sensory neurons are located (Wilson et al., 1999). Expression of this gene results in the production of the opioid receptor ligands metenkephalin and leu-enkephalin, which are then released in the dorsal horn of the spinal cord. Enkephalins act on

the delta-opioid receptor. Testing showed the treated mice became analgesic to the painful effects of capsaicin or dimethylsulfoxide (DMSO) applied to the skin and that this analgesia was mediated through opiate receptors (Wilson et al., 1999; Wilson and Yeomans, 2002). A number of subsequent studies using the proenkephalin gene have produced similar reduction in pain behavior in mice resulting from bone cancer (Goss et al., 2002b; Mata et al., 2002), thermal hyperalgesia (increased response to nociceptive heat) from pertussis toxin injection (Yeomans et a l , 2004) and in a neuropathic pain model (Hao et al., 2003a). Recently the same strategy was tested in a model of trigeminal neuropathic pain produced by a constriction of the infra-orbital nerve. Inoculation of the vibrissal pad with the HSV vector carrying the proenkephalin A gene produced an opioid receptor-mediated, antiallodynic effect that lasted at least 3 weeks (Meunier et al., 2005). The broad conclusion that emerges from these studies is that the proenkephalin gene introduced into sensory ganglion neurons by peripheral injection is analgesic and antiallodynic. Interestingly, there seems to be no change in normal sensation. The effects of the opioid gene expression are somatotopic meaning that analgesia occurs only in the body area where the vector is injected. The effect is transient in the sense that the effect decreases after several months (in part depending on the promoter used), but during that period the transgene is continually expressed and, of clinical importance, tolerance to the opioid peptide does not develop. Although not specifically examined, the side effects associated with systemic administration of opiates (cited above) are also not expected to occur with the local administration of the opioid gene. In addition to the enkephalins, therapy using other endogenous opioid genes have been investigated, ladarola and co-workers (Finegold et al., 1999) used ^-endorphin delivered by a replication deficient adenovirus (Ad5) to meningeal cells by intracerebroventricular or intrathecal injection. The viral transfer resulted in an opiate receptor-mediated analgesia (i.e. naloxone reversible) in a carageenin induced inflammation model. In a different approach Xu and co-workers (Xu et al., 2003) injected the L4 and L5 sensory ganglia with a recombinant adeno-associated viral (rAAV) vector encoding the /i-opioid receptor itself rather than the ligand. In this case, when tested in a complete Freund's adjuvant-induced inflammation model, there was a reduction of pain responses. It was also found that in addition to reducing the level of spontaneous pain, systemically administered morphine was effective at much lower doses than in control animals. Thus,

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

GROWTH FACTOR-BASED STRATEGY

233

FIGURE 1 Pain sensation is carried by nerve fibers that have receptors in the periphery (A) and cell bodies in DRG (B) or in the trigeminal ganglia for sensation from the head. The path continues through tract neurons in the spinal dorsal horn (C) and travels in axons to the thalamus and then to the cerebral cortex (D). The most common strategy for introducing viral vectors is to inoculate the skin (A) in the target area. This results in the virus entering the free nerve endings and being transported to the cell body in the dorsal root (or trigeminal) ganglion where expression of the delivered gene takes place. The product of the delivered gene is released within the ganglion or in the spinal dorsal horn via the central axon (C). Another strategy is to deliver the viral vector directly into the ganglion by injection with micropipettes. In a few cases, a laminectomy is used to expose the spinal cord and vectors injected directly into the spinal dorsal horn (C). Viral vectors have also been delivered into several different locations in the forebrain (D) by direct injection or nasal inoculation.

increasing the expression of an opioid receptor presents the possibility of avoiding some of the side effects of conventional opioid receptor-mediated therapy by making it possible to obtain sufficient analgesia with lower doses of morphine or similar opiates (the side effects opiates are proportional to their dose). Another class of opioid receptor acting ligands are the dynorphins that are derived from the pro-opiomelanocortin (POMC) precursor gene. POMC has been delivered via gene-gun and electroporation in a number of studies and found to be effective in relieving pain from bladder stimulation (Chuang et al., 2003),

the formalin test (Lu et al., 2002) and chronic nerve constriction injury (Wu et al., 2004). The successes of the administration of the POMC gene using nonviral means of application suggest that this precursor would also be a good candidate for gene therapy. III.

GROWTH FACTOR-BASED STRATEGY

A second broad strategy has been aimed at modifying the events that occur in the spinal cord following injury. Even though initially the same receptors.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

234

18. VIRAL VECTOR MEDIATED GENE THERAPY OF PAIN

Jasmin and Ohara FIGURE 2 Low-power (A) and high-power (B) micrographs of green fluorescent protein (GFP)-expressing HSV vectors injected into the trigeminal ganglion. In (A), a small patch of neurons (e.g. arrow) and axons (arrowhead) are visible at the injection site. Some of the smaller GFP labeled structures are probably glial cells, which take up the virus. (B) Typical smooth outlined appearance of a GFP-labeled DRG neuron (arrow). (C , D) Low- and highpower micrographs of the rostral agranular cortex (RAIC). The rhinal fissure (RF) marks the ventral boundary. A small injection of a viral vector containing both GFP and the GAD67 gene results in a patch at GFP expressing cells at the injection site. The region outlined in (C) is shown at greater magnification in (D). Typical pyramidal neuron cell bodies are shown (GFP labeled in (D), arrow) and the apical dendrite is indicated with the arrowhead.

pathways and neurotransmitters mediate both acute and chronic pain, ultimately chronic pain involves changes in gene expression that results in a spinal cord that is in some w^ay different from normal. Some of the changes that have been documented are the result of increases or decreases in the expression of growth factors and these observations have provided the rationale for therapy targeting such factors. Brainderived neurotrophic factor (BDNF), for instance is a potent survival and differentiation factor for many developing and injured neuronal systems, including neurons expressing y-amino-butyric acid (GABA) and serotonin. Eaton and coworkers (Eaton et al., 2002) injected rAAV expressing BDNF into the dorsal horn of rats exhibiting neuropathic pain resulting from a nerve constriction. Both allodynia (pain caused by normally innocuous stimuli) and hyperalgesia (increased response to nociceptive stimuli) were reduced for as long as 8 weeks following injection. Another important

growth factor is glial cell-derived neurotrophic factor (GDNF) that is known to increase the survival of many neuronal populations including sensory and motor neurons. GDNF delivered via an FiSV-based vector inoculated into the skin was found to be expressed in the DRG (where sensory neurons are located) and reduce the allodynia resulting from spinal nerve ligation (Fiao et al., 2003b). In contrast to the above strategies designed to increase growth factors, an attempt to obtain the opposite effect was tried using semaphorin 3A (SemaSA) (Tang et al., 2004). Semaphorins are a family of secreted and membrane anchored glycoproteins that act to selectively repel axons. Sema3A in particular repels axons from a subset of embryonic DRG neurons that are of small diameter, NGF responsive, and involved in thermoreception and nociception (Tang et al., 2004). Production of NGF is known to increase in some injury states (Nakamura and Bregman, 2001) and, of importance in the

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

NON-PAIN DIRECTED STRATEGIES

present context, can increase the sprouting of nociceptive axons and release of the neuropeptide they use as neurotransmitters (Christensen and Hulsebosch, 1997; Gwak et al., 2003). To examine the effects of Sema3A, two adeno-viral vectors were injected into the rat dorsal horn, one vector expressing NGF and the other Sema3A. The Sema3A reduced the NGF-mediated growth and sprouting of axons in a dose dependant manner (Tang et al., 2004). The reduction in axonal sprouting was accompanied by a reduction in hyperalgesia. IV.

GABA^BASED STRATEGIES

Possibly second in importance to opioids in terms of analgesia are the GAB A acting drugs (Jasmin et al., 2004). GABA is the most abundant inhibitory neurotransmitter in the central nervous system (CNS) and is found at all levels of the neuraxis. Intrathecal administration of GABA acting drugs is a currently used treatment for pain and spasticity. GABAergic intemeurons are present in the spinal dorsal horn and changes in GABA levels have been reported in the dorsal horn following injury (Demediuk et al., 1989; Eaton et al., 1998; Moore et al., 2002). One strategy to increase GABA in the dorsal horn following injury has been to use an HSV-based vector to deliver the glutamic acid decarboxylase (GAD67) gene to DRGs by subcutaneous inoculation of the virus (Liu et al., 2004). GAD67 is one of the key enzymes responsible for the production of GABA. When the vector was inoculated into both hind paws of rat with neuropathic pain following a T13 hemisection, both the mechanical allodynia and thermal hyperalgesia were ameliorated. Intrathecal application of GABA receptor antagonists in these same animals partially blocked the viral vector-mediated effects indicating that the antinociceptive effect was partly mediated by GABA.

V.

FOREBRAIN TARGETS

Very few studies have been carried out on the analgesic effects of viral vector-mediated gene transfer in regions of the CNS other than the spinal cord or peripheral nerve. HSV has been used to deliver preproenkephalin to the amygdala in rats (Kang et al., 1998). Subsequent nociceptive testing using the formalin test showed the treatment had no effect on the first phase (generally considered to be more like acute pain) of the test but reduced flinching in the second phase (considered to be more like chronic pain) of this test. The amelioration of flinching in the second phase was reversed

235

by application of naloxone. Recently, we (Jasmin et al., 2003) used a defective HSV vector to deliver the GAD67 gene to an area of the insular cortex called the rostral agranular insular cortex (RAIC) (Fig. 2) in rats. This treatment resulted in an increase in withdrawal threshold (i.e. analgesia) for both heat and mechanical tests. The analgesia was reversed by application of an adrenergic receptor antagonist to the spinal cord suggesting the pain reduction was mediated by the descending pain inhibitory system, which uses noradrenaline as one of its key neurotransmitters. One key consideration when delivering genes to the peripheral nerve, sensory ganglia or spinal cord dorsal horn is that this approach will probably have topographically limited effects. This is of great advantage when one is dealing with pain that originates in a well-defined part of the body for in this case the gene therapy can be aimed at that specific location. Even in cases where pain might originate from internal, visceral structures it should be possible to inoculate the skin in the region of the body to which the pain is referred as a way to target internal structures. Decreasing the intensity of the discriminative component of pain will clearly have repercussions for the affective component, although one might not be directly altering the parts of the forebrain involved in the affective component. In contrast, the effects of genes delivered to forebrain structures are likely to be more global. Parts such as the amygdala and cerebral cortex have extensive and often bilateral connection with other brain areas and are thus able to have whole body effects. In some cases it might be that cortical manipulation turns on descending inhibitory circuits (Jasmin et al., 2003) and therefore act by interrupting the transmission of information from the periphery (i.e. discriminative aspects) but even in this case the effect on pain sensation will involve the entire body rather than a specific region. When the pain is generalized and not centered on any particular location, a widespread effect is more desirable and thus forebrain targets might be the best choice for some neuropathic pain conditions. VL

NON^PAIN DIRECTED STRATEGIES

Finally, there are a number of studies that have used gene therapy for conditions in which there occurs a reduction of pain as a consequence of reduction of the causative factor. For example, vectors have been use to deliver vascular endothelial growth factor (VEGF) to promote angiogenesis in myocardium of patients with angina pectoralis (Merkle and Montgomery, 2003).

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

236

18. VIRAL VECTOR MEDIATED GENE THERAPY OF PAIN

Similarly neutrophin 3 (Pradat et al., 2001, 2002) and NGF (Goss et al., 2002a) have been used to prevent axonal degeneration in a rat model of diabetic neuropathy. Although nociception w^as not specifically tested, pain is often associated w^ith this condition and could be expected to be ameliorated by this therapy.

VII.

OTHER CONSIDERATIONS

The use of virally mediated gene therapy in pain control is in its infancy. The complex nature of pain makes it difficult to pinpoint any one best target but at the same time offers a number of productive routes one may consider for successful treatment. Presently, opioid-based targets have provided the largest base of positive results and are encouraging for development of this therapy. Hov^ever a number of issues require further investigation. Introducing the viral vector by skin inoculation has advantages in being non-invasive and amenable to several different methods of application. It should be noted that many of the experimental studies have used abrasion of the skin to introduce the virus and this is not necessarily the best method in the clinical setting. A point already raised is that antinociceptive effects of peripheral application w^ill be limited to the sites of inoculation so this method of application might not be useful for diffuse pain. The other methods used in the experimental model such as direct injection of virus into ganglia, spinal cord or forebrain all involve surgery that is in some cases quite extensive. There are also basic science questions that remain unanswered. Using current techniques the viral vectors do not target specific cell types. Thus following peripheral application, fibers associated with non-nociceptive sensation in addition to those associated with nociceptive sensation are transfected. Similarly, when directly injected into ganglia or neuropil, both neurons and glia cells become transfected. However, practical experience is that delivery genes to ganglion cells or to the brain without regard to which specific cell population is involved, is effective. Possibly treatment will be more effective if we could restrict the gene transfer to specific neuronal or glial populations. We also do not know the long-term consequences of gene therapies and if there might be any long-term neuronal loss as the result of the viral load. Because the nervous system has limited natural ability to regenerate itself, neuronal survival is a significant issue. Ultimately, understanding the mechanism(s) through which these treatments

work will also provide important insight into general mechanisms of nociception as well as being necessary before widespread human use can be contemplated. Finally it has to be realized that the models of pain we use to test the effectiveness of analgesic therapies are limited. Simple threshold measurement, the formalin test and even constriction injury models are at best poor n\odels of neuropathic pain in humans. The simple animal models of pain do not always reliably model the complex cultural and emotional aspects of pain as experienced by humans.

ACKNOWLEDGMENT This work was supported by NIH and the Koret Foundation. References Christensen, M.D. and Hulsebosch, C.E. (1997) Spinal cord injury and anti-NGF treatment results in changes in CGRP density and distribution in the dorsal horn in the rat. Exp. Neurol., 147: 463-475. Chuang, Y.C., Chou, A.K., Wu, R C , Chiang, RH., Yu, T.J., Yang, L.C., Yoshimura, N. and Chancellor, M.B. (2003) Gene therapy for bladder pain with gene gun particle encoding pro-opiomelanocortin cDNA. J. Urol., 170: 2044-2048. Demediuk, R, Daly, M.R and Faden, A.I. (1989) Effect of impact trauma on neurotransmitter and nonneurotransmitter amino acids in rat spinal cord. J. Neurochem., 52: 1529-1536. Dobson, A.T., Margolis, T.R, Sedarati, R, Stevens, J.G. and Feldman, L.T. (1990) A latent, nonpathogenic HSV-1-derived vector stably expresses beta-galactosidase in mouse neurons. Neuron, 5: 353-360. Eaton, M.J., Blits, B., Ruitenberg, M.J., Verhaagen, J. and Oudega, M. (2002) Amelioration of chronic neuropathic pain after partial nerve injury by adeno-associated viral (AAV) vector-mediated over-expression of BDNF in the rat spinal cord. Gene Ther., 9: 1387-1395. Eaton, M.J., Plunkett, J.A., Karmally, S., Martinez, M.A. and Montanez, K. (1998) Changes in GAD- and GABA-immunoreactivity in the spinal dorsal horn after peripheral nerve injury and promotion of recovery by lumbar transplant of immortalized serotonergic precursors. J. Chem. Neuroanat., 16: 57-72. Ecob-Prince, M.S., Hassan, K., Denheen, M.T. and Rreston, C M . (1995) Expression of beta-galactosidase in neurons of dorsal root ganglia which are latently infected with herpes simplex virus type 1. J. Gen. Virol, 76( Pt 6): 1527-1532. Finegold, A.A., Mannes, A.J. and ladarola, M.J. (1999) A paracrine paradigm for in vivo gene therapy in the central nervous system: treatment of chronic pain. Hum. Gene Ther., 10: 1251-1257. Goss, J.R., Coins, W.R, Lacomis, D., Mata, M., Glorioso, J.C. and Fink, D.J. (2002a) Herpes simplex-mediated gene transfer of nerve growth factor protects against peripheral neuropathy in streptozotocin-induced diabetes in the mouse. Diabetes, 51: 2227-2232.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

ACKNOWLEDGMENT Goss, J.R., Harley, C.E, Mata, M., O'Malley, M.E., Coins, W.R, Hu, X., Glorioso, J.C. and Fink, DJ. (2002b) Herpes vector-mediated expression of proenkephalin reduces bone cancer pain. Ann. Neurol., 52: 662-665. Gwak, Y.S., Nam, T.S., Paik, K.S., Hulsebosch, C.E. and Leem, J.W. (2003) Attenuation of mechanical hyperalgesia following spinal cord injury by administration of antibodies to nerve growth factor in the rat. Neurosci. Lett., 336: 117-120. Hao, S., Mata, M., Coins, W., Glorioso, J.C. and Fink, D.J. (2003a) Transgene-mediated enkephalin release enhances the effect of morphine and evades tolerance to produce a sustained antiallodynic effect in neuropathic pain. Pain, 102: 135-142. Hao, S., Mata, M., Wolfe, D., Huang, S., Glorioso, J.C. and Fink, D.J. (2003b) HSV-mediated gene transfer of the glial cell-derived neurotrophic factor provides an antiallodynic effect on neuropathic pain. Mol. Ther., 8: 367-375. Jasmin, L., Rabkin, S.D., Granato, A., Boudah, A. and Ohara, P.T. (2003) Analgesia and hyperalgesia from GABA-mediated modulation of the cerebral cortex. Nature, 424: 316-320. Jasmin, L., Wu, M.V. and Ohara, PT. (2004) GABA puts a stop to pain. Curr. Drug Targets CNS Neurol. Disord., 3: 487-505. Kang, W., Wilson, M.A., Bender, M.A., Glorioso, J.C. and Wilson, S. P. (1998) Herpes virus-mediated preproenkephalin gene transfer to the amygdala is antinociceptive. Brain Res., 792: 133-135. Liu, J., Wolfe, D., Hao, S., Huang, S., Glorioso, J.C, Mata, M. and Fink, D.J. (2004) Peripherally delivered glutamic acid decarboxylase gene therapy for spinal cord injury pain. Mol. Ther., 10: 57-66. Lu, C.Y., Chou, A.K., Wu, C.L., Yang, C.H., Chen, J.T., Wu, P C , Lin, S.H., Muhammad, R. and Yang, L.C (2002) Gene-gun particle with pro-opiomelanocortin cDNA produces analgesia against formalin-induced pain in rats. Gene Ther., 9: 1008-1014. Mata, M., Glorioso, J.C and Fink, D.J. (2002) Targeted gene delivery to the nervous system using herpes simplex virus vectors. Physiol. Behav, 17\ 483-488. Merkle, C.J. and Montgomery, D.W (2003) Gene therapy with vascular endothelial growth factor reduces angina. J. Cardiovasc. Nurs., 18: 3 8 ^ 3 . Meunier, A., Latremoliere, A., Mauborgne, A., Bourgoin, S., Kayser, v., Cesselin, E, Hamon, M. and Pohl, M. (2005) Attenuation of pain-related behavior in a rat model of trigeminal neuropathic

237

pain by viral-driven enkephalin overproduction in trigeminal ganglion neurons. Mol. Ther., 11: 608-616. Moore, K.A., Kohno, T., Karchewski, L.A., Scholz, J., Baba, H. and Woolf, C.J. (2002) Partial peripheral nerve injury promotes a selective loss of GABAergic inhibition in the superficial dorsal horn of the spinal cord. J. Neurosci., 22: 6724-6731. Nakamura, M. and Bregman, B.S. (2001) Differences in neurotrophic factor gene expression profiles between neonate and adult rat spinal cord after injury. Exp. Neurol, 169: 407-415. Pradat, RE, Kennel, P., Naimi-Sadaoui, S., Finiels, E, Orsini, C , Revah, E, Delaere, P. and Mallet, J. (2001) Continuous delivery of neurotrophin 3 by gene therapy has a neuroprotective effect in experimental models of diabetic and acrylamide neuropathies. Hum. Gene Ther., 12: 2237-2249. Pradat, RE, Kennel, P., Naimi-Sadaoui, S., Finiels, E, Scherman, D., Orsini, C , Delaere, P., Mallet, J. and Revah, E (2002) Viral and non-viral gene therapy partially prevents experimental cisplatininduced neuropathy. Gene Ther., 9: 1333-1337. Tang, X.Q., Tanelian, D.L. and Smith, C M . (2004) Semaphorin3A inhibits nerve growth factor-induced sprouting of nociceptive afferents in adult rat spinal cord. J. Neurosci., 24: 819-827. Wilson, S.P and Yeomans, D.C (2002) Virally mediated delivery of enkephalin and other neuropeptide transgenes in experimental pain models. Ann. NY Acad. Sci., 971: 515-521. Wilson, S.P, Yeomans, D.C, Bender, M.A., Lu, Y, Coins, W E and Glorioso, J.C. (1999) Antihyperalgesic effe. cts of infection with a preproenkephalin-encoding herpes virus. Proc. Natl. Acad. Sci. USA, 96: 3211-3216. Wu, CM., Lin, M.W, Cheng, J.T., Wang, Y.M., Huang, Y W , Sun, W.Z. and Lin, CR. (2004) Regulated, electroporation-mediated delivery of pro-opiomelanocortin gene suppresses chronic constriction injury-induced neuropathic pain in rats. Gene Ther., 11: 933-940. Xu, Y, Gu, Y, Xu, G.Y, Wu, P , Li, G.W and Huang, L.Y (2003) Adeno-associated viral transfer of opioid receptor gene to primary sensory neurons: a strategy to increase opioid antinociception. Proc. Natl. Acad. Sci. USA, 100: 6204-6629. Yeomans, D.C, Jones, T., Laurito, C.E., Lu, Y and Wilson, S.P (2004) Reversal of ongoing thermal hyperalgesia in mice by a recombinant herpesvirus that encodes human preproenkephalin. Mol. Ther., 9: 24-29.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

C H A P T E R

19 VEGF, an Angiogenic Factor with Neurotrophic Activity, Useful for Treatment of ALS ? Diether Lambrechts, Peter Carmeliet

Abstract: Vascular endothelial growth factor (VEGF) is a key regulator of angiogenesis. Recent evidence indicated, however, that VEGF also plays a significant role in the development and maintenance of the neural system. In addition to its ability to promote the survival of various populations of neurons and glial cells, transgenic mice expressing low levels of VEGF develop late-onset motoneuron degeneration, reminiscent of amyotrophic lateral sclerosis (ALS) in humans. Recent data further revealed that intracerebroventricular (ICV) delivery of recombinant VEGF protein delays disease onset, improves motor performance and prolongs survival of ALS rats. Notably, intramuscular administration of a VEGFexpressing rabies G-pseudotyped lentivirus, which is retrogradely transported to the neuronal cell body, increased the life expectancy of ALS mice by as much as 30%. Together with the positive effects achieved with viral gene delivery for insulin-like growth factor 1 (IGFl), these efforts have primed widespread interest in applying these vectors for therapeutic use in ALS patients. Keywords: VEGF; ALS; retrograde transport; AAV; EIAV; ICV delivery

developing VEGF-based therapies for neurodegenerative disorders.

Two millennia ago, opposing schools of thought in ancient Greece debated about the relative role of the brain and blood vessels as the central source of life. One school, headed by Plato, proposed that the brain was harbored the soul, while Aristotle considered the heart and blood vessels of vital importance. In Aristotle's words: 'the system of blood vessels can be compared with those of watercourses in gardens: they start from one source, and branch off into numerous channels, so as to carry a supply to every part of the garden'. Now, centuries later, blood vessels and nerve cells are no longer considered to be antagonists, as the Greeks once thought, but they have been recognized to share much more in common than originally anticipated. Perhaps, the best illustration of such cross-talk is provided by the activities of vascular endothelial growth factor (VEGF) — originally discovered as a blood vessel-specific growth factor, but recently recognized to have important neuronal activities. These findings have primed interest in

Gene Therapy of the Central Nervous System: From Bench to Bedside

L

VEGF — A KEY ANGIOGENIC PLAYER

Originally purified as a growth factor, capable of increasing vascular permeability and endothelial cell proliferation, VEGF distinguishes itself from other angiogenic factors by a unique combination of properties. Indeed, VEGF is produced and released by cells in close vicinity to endothelial cells (Breier et al., 1995; Dumont et al., 1995) and, after binding VEGF receptor-1 (VEGFRl, Fltl), VEGFR2 (Flkl), and neuropilins (Nrpl; Nrp2), VEGF induces endothelial cells to proliferate, migrate, survive, and assemble into an interconnected vessel network (Ferrara et al., 2003). VEGF is transcribed from a single gene into various isoforms (Stalmans et al., 2002). Because of their different affinities for extracellular matrix components and receptor

239

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

240

19. VEGF, AN ANGIOGENIC FACTOR WITH NEUROTROPHIC ACTIVITY, USEFUL FOR TREATMENT OF ALS?

subtypes, these isoforms provide a spatial gradient, critical for vessel patterning (Stalmans et al., 2003). The expression of VEGF is upregulated by hypoxia, thereby providing a feedback mechanism to accommodate insufficient tissue oxygenation. Loss of even a single VEGF allele caused fatal vascular defects in the mouse embryo, indicating that VEGF is a key regulator of vascular development (Carmeliet et al., 1996; Ferrara et al., 1996). A large body of evidence has now^ implicated VEGF as the key determinant of the angiogenic sw^itch in numerous pathological disorders, such as tissue ischemia, inflammation, malignancy, blindness, and many other diseases (Carmeliet, 2003a, b; Ferrara, 2005). Last year, an anti-VEGF antibody w^as approved as the first successful angiogenesis inhibitor for the treatment of colorectal cancer (Ferrara et al., 2004; Hurwitz et al., 2004).

n . ANGIOGENIC ROLE OF VEGF IN N E U R A L DEVELOPMENT A N D DISEASE Several lines of evidence indicate that VEGF regulates neural development because of its angiogenic activities. In the developing brain, VEGF is expressed by neurons of the subventricular zone and induces VEGFR2-expressing endothelial cells of the perineural vascular plexus to invade the neural tissue, thereby supplying the growing CNS with sufficient oxygen (Breier et al., 1992; Millauer et a l , 1993; Dumont et al., 1995). The importance of neuronal VEGF production is illustrated by findings that conditional inactivation of VEGF in the brain results in decreased vessel density, ischemia, neuronal degeneration, and cell death (Haigh et al., 2003). Apart from stimulating vessel formation, VEGF, expressed by sensory neurons or Schwann cells, also instructs arteries to branch alongside nerves (Mueller et al., 2000). Neurodegenerative diseases are generally considered to result from neuronal dysfunction, but recent insights indicate that a vascular component may also be involved in the pathogenesis of these disorders. We illustrate this concept for three diseases. Alzheimer's disease (AD) is characterized by amyloid plaques, neurofibrillary tangles, cerebral hypoperfusion, and amyloid angiopathy. The observation that cerebral hypoperfusion precedes the onset of clinical symptoms in AD suggests that chronic brain hypoperfusion increases the risk for AD (Kalaria, 2002). Amyloid-jS has direct effects on the cerebral vasculature: it causes vasoconstriction, reduces the blood flow and even induces endothelial cell damage. Furthermore, VEGF

co-localizes with and binds amyloid-j5 plaques in the brains of AD patients (Yang et al., 2004). Such a deposition of VEGF in plaques may reduce the availability of VEGF, and thereby contribute to cerebral hypoperfusion and further aggravate memory decline in AD. A second VEGF-dependent disease is amyotrophic lateral sclerosis (ALS). Insufficient VEGF levels cause adult-onset motoneuron degeneration and paralysis in VEGF^/^ knock-in mice with a subtle VEGF promoter mutation (see below) (Oosthuyse et al., 2001). At least one mechanism, responsible for the development of ALS-like motoneuron degeneration in these mice, is the hypoperfusion of the brain and spinal cord, which leads to chronic ischemia of the metabolically active motoneurons. VEGF^^^ mice are also more susceptible to transient spinal cord ischemia and remain paralyzed for much longer periods (Lambrechts et al., 2003). A third example of a VEGF-dependent disorder is diabetic neuropathy, characterized by basementmembrane thickening, endothelial cell hyperplasia and other vascular abnormalities, which all together impair endoneural perfusion and induce chronic neuronal ischemia. VEGF has a favorable effect on such diabetic neuropathies by restoring perfusion and, in addition, improving neural cell function (Schratzberger et al., 2000, 2001). III. A N E U R A L ROLE FOR VEGF IN N E U R A L DEVELOPMENT A N D DISEASE VEGF is also capable of affecting neural cells directly. In vitro studies revealed that VEGF has direct effects on a variety of neuronal cell types. For instance, VEGF increases the survival of cultured hippocampal, cortical, cerebellar, granule, dopaminergic, autonomic, and sensory neurons under various conditions of stress (reviewed by Greenberg and Jin, 2004; Storkebaum et al., 2004). In general, the survival activity of VEGF relies on the VEGFR2-mediated activation of PI3K, Akt, and NFk^ pathways and suppression of the proapoptotic caspase-3 (Greenberg and Jin, 2004). When applied to primary cortical neurons, VEGF increases the expression of the neuronal microtubule markers TUJl and MAP-2, suggesting that VEGF also regulates the growth, development and structural stability of neurons (Rosenstein et al., 2003). VEGF also protects mouse NSC34 motoneuron-like cells against hypoxia and oxidative-stress-induced apoptosis — an effect, blocked by a combination of anti-Nrpl and -VEGFR2 antibodies (Oosthuyse et al., 2001). In vivo studies indicate that VEGF regulates neuronal cell behavior. By binding its

IV GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

A NEURAL ROLE FOR VEGF IN NEURAL DEVELOPMENT AND DISEASE

receptor N r p l , VEGF determines the correct migration of motoneuron somata of the developing facial nerve (Schwarz et al., 2004). Guidance of the axons of these facial nerves is also determined by N r p l , but requires binding of Semaphorin 3A (Sema3A) (Schwarz et al., 2004). Thus, by acting through a shared receptor, VEGF and SemaSA coordinately pattern different cellular compartments of the same nerve. Notably, knockdown of VEGFR2 expression by antisense oligonucleotides induces motoneuron death in the spinal cord of rats exposed to hypoxia (Shiote et al., 2005). VEGF also affects other neural cell types. For instance, VEGF has mitogenic effects on astrocytes, in mesencephalic explant cultures (Silverman et al., 1999) and following intracerebral VEGF delivery in vivo (Krum et al., 2002). VEGF also prolongs the survival and stimulates the proliferation of Schwann cells in explant cultures of superior cervical and dorsal root

241

ganglia (Sondell et al., 1999). Schwann cells express VEGFRl, VEGFR2, and Nrpl, but the effect of VEGF on Schwann cell migration appears to be mediated by VEGFR2, as addition of a neutralizing antibody against VEGFR2 completely blocks VEGF-induced migration (Schratzberger et al., 2000). Activation of VEGFRl by VEGF has also been reported to induce migration and proliferation of microglial cells (Forstreuter et al., 2002). Thus, the direct effects of VEGF on neurons and Schwann cells are mediated predominantly by VEGFR2 (and Nrpl), whereas the effects of VEGF on astrocytes and microglia are directed by VEGFRl (Fig. 1). Besides its effects on differentiated neurons, VEGF also stimulates neurogenesis in vitro and in vivo. As a competitive antagonist of Sema3A for Nrpl binding, VEGF also promotes the survival, migration, and proliferation of neural stem cells (Bagnard et al., 2001). After infusion of VEGF into the lateral ventricle of adult rats, the number

FIGURE 1 VEGF in the nervous system. Increasing experimental evidence supports the concept that VEGF (Red) has direct effects on different neuronal cell types including neurons, Schwann cells, oligodendrocytes, astrocytes and microglia. The effects of VEGF on neurons and Schwann cells are mediated predominantly by VEGF-receptor 2 or Flkl (purple), whereas the effects of VEGF on astrocytes and microglia are directed by VEGF-receptor 1 or Fltl (yellow). After intramuscular injection, VEGF-expressing viral vectors (EIAV-VEGF; yellow) are retrogradely transported to infect the motoneuron soma. Intracerebroventricularly delivered VEGF (Red) diffuses from the cerebrospinal fluid through the ependymal layer into the parenchyma, where it directly affects the motoneurons. In addition, the VEGF protein itself can be both anterogradely and retrogradely transported in the axons of neurons.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

242

19. VEGF, AN ANGIOGENIC FACTOR WITH NEUROTROPHIC ACTIVITY, USEFUL FOR TREATMENT OF ALS?

of BrdU-positive neurons increases, not by stimulating the proliferation of neural stem cell progenitors, but by preventing them to undergo apoptosis (Schanzer et al., 2004). Enriched environments, exercise and hippocampus-dependent learning tasks stimulate neurogenesis. VEGF, by acting through VEGFR2, seems to mediate some of the enriched environment effect on neurogenesis and cognition (Cao et al., 2004). The fact that VEGF and its receptors have both neural and vascular effects raises the intriguing question which activity developed first during evolution. The genome of the nematode Caemorhabditis elegans, an invertebrate devoid of a vascular system, contains a family of four receptor tyrosine kinases, structurally related to the VEGF receptors. As these receptors are expressed in neural cells, VEGF may have evolved originally as a neural regulator (Popovici et al., 2002). Vessels, w^hich developed later in evolution than nerves, seem to have co-opted VEGF as a critical regulator of their formation.

IV.

VEGF A N D ALS

ALS, also called Lou Gehrig or Charcot disease, is a devastating incurable disorder with a yearly incidence of 1-2 per 100,000 individuals (Cleveland, 1999; Brown and Robberecht, 2001; Rowland and Shneider, 2001; Shaw et al., 2001; Turner et al., 2001). It is characterized by progressive wasting and weakness of limb, bulbar, and respiratory muscles due to degeneration of motoneurons in the spinal cord, brain stem, and motor cortex. Disease onset is usually in the fifth or sixth decade of life and, in most affected individuals, it progresses to death due to respiratory failure within 3-5 years after onset. There are, however, unexplained large individual differences in clinical presentation (initial symptoms may be spinal or bulbar in nature), age of onset (ranging from juvenile to the very elderly) and survival time (from a few months to more than 20 years). In approximately 5-10% of cases, ALS is familial in origin (Andersen, 2001; Siddique and Lalani, 2002). Several types of familial ALS (FALS) have been assigned to loci of the human genome (Andersen, 2001; Hand and Rouleau, 2002; Siddique and Lalani, 2002). ALSl, an autosomal dominant form of adult ALS affects 14-23% of individuals with familial ALS and is associated with more than 100 different mutations in the gene encoding C u / Z n superoxide dismutase (SODl) on chromosome 21 (Rosen et al., 1993). Mutant SOD is generally believed to cause motoneuron degeneration by gaining a toxic property (Bruijn et al., 2004). Mice and rats, expressing a mutant

SODl transgene, develop ALS with clinical and anatomopathological features, that are highly reminiscent of those found in human ALS and are by far the most frequently studied small animal model for ALS. However, the mechanisms by which this toxic function causes ALS are still debated and remain elusive. Only very recently, additional genes associated with familial ALS were discovered: (i) identical missense mutations in the vesicle-trafficking protein VAPB gene have been identified in families with atypical ALS (ALS8) (Nishimura et al., 2004); (ii) homozygous mutations in the gene ALSIN, a putative guanine nucleotide factor for GTPase, have been found in a family with juvenileonset ALS (ALS2) (Yang et al., 2001; Hand et al., 2003) and (iii) missense mutations in senataxin, a gene that encodes a novel DNA/RNA helicase, have been found in autosomal dominant forms of juvenile ALS (ALS4) (Chen et al., 2004). The genetic causes for the other forms of ALS are not yet known. Recently, a novel role for VEGF in ALS was documented. Reduced levels of VEGF in the brain and spinal cord in knock-in mice with a subtle deletion of the hypoxia-response element in the VEGF gene promoter (VEGF^'/^'mice) caused adult-onset motoneuron degeneration, reminiscent of ALS (Oosthuyse et al., 2001). The neuropathological features in these mice, such as the specific loss of choline acetyltransferase (CHAT)positive motoneurons, the ultrastructural signs of motoneuron degeneration, axonal spheroids, aberrant neurofilament inclusions, Wallerian degeneration in peripheral nerves, and selective loss of large myelinated motor axons were strikingly similar to those found in mutant SODl mice, a mouse model of ALS, and in ALS patients. Because neural perfusion in these mice was reduced and VEGF confers neuroprotective effects on motoneurons, both chronic vascular insufficiency and insufficient VEGF-dependent neuroprotection may lead to the degeneration of motoneurons in mice. In a follow-up study, we showed that VEGF was also implicated in ALS in humans (Lambrechts et al., 2003, 2004). Spontaneous mutations of the hypoxia response element were not detected in ALS individuals (Gros-Louis et al., 2003; Lambrechts et al., 2003). However, human subjects homozygous for the - 2 5 7 8 A / - 1 1 5 4 A / - 6 3 4 G or - 2 5 7 8 A / - 1 1 5 4 G / —634G haplotypes were more common in the population of ALS patients than healthy individuals. These 'at-risk' haplotypes reduced VEGF gene transcription, impaired IRES-dependent translation, and interfered with translational initiation of a novel long VEGF isoform ( L - V E G F ) , resulting in reduced circulating

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

GENE THERAPY APPROACHES FOR ALS

VEGF levels in vivo. To further characterize the role of VEGF in ALS, VEGF^/^ mice were intercrossed with SOD1G93A mice. Normally, S O D I ^ ^ S A ^lice develop clinical symptoms at the age of 90 days, and die at the age of 130 days. In contrast, VEGF^/^ mice only develop symptoms beyond the age of 6 months and survive for up to 2 years. Notably, double-transgenic VEGF^/7SODlG93A mice died sooner than singletransgenic SODl^^^^ mice, indicating that VEGF also modified motoneuron degeneration in the standard mouse model of ALS (Lambrechts et al., 2003). By now, our observations have been confirmed by additional studies. Indeed, in SODl^^^^ rats, VEGF levels are reduced before disease onset and become progressively lower with disease progression (Xie et al., 2004). In humans, low levels of VEGF in the cerebrospinal fluid (CSF) are also a hallmark of early ALS (Devos et al., 2004). In a much smaller association study by Terry (2004), individuals homozygous for the 'at-risk' haplotypes exhibited a threefold increased risk for ALS (Terry et al., 2004). A role for VEGF in motoneuron degeneration is further suggested by findings that the VEGF levels in the spinal cord were reduced before onset of spinal and bulbar muscular atrophy in a transgenic mouse model of Kennedy's disease — a lower motoneuron disease caused by GAG repeat expansion in the androgen receptor gene (Sorenson, 2004). Interestingly, the -2578A VEGF allele, which correlates with reduced VEGF expression levels in plasma and serum, also confers an increased risk for patients with AD (Del Bo et al., 2005). To provide a better understanding of the molecular basis of the VEGF system in motoneuron degeneration, we generated transgenic mice overexpressing VEGFR2 in neurons of the adult nervous system by using the Thyl.2 promoter (Storkebaum et al., 2005). The Thy-Flkl transgenic mice were fertile, appeared healthy and their motor performance under basal conditions was normal. Intercrossing of the Thy-Flkl transgenic mice with SODl^^^^ mice revealed that the motor performance of the double-transgenic ThyFlkl/SODl^^^^ transgenic mice deteriorated at a later age as compared to the single-transgenic SODl^^^^ mice. Thy-Flkl/SODl^^^^ transgenic mice also lived longer than SODl^^^^ single-transgenic mice, revealing that overexpression of Flkl in neurons delays the degeneration of spinal motoneurons in SODl^^^^ mice by transmitting survival signals of endogenous VEGF (Storkebaum et al., 2005). All together, these genetic studies raised the intriguing question whether VEGF could have therapeutic effects in animal models for ALS.

¥•

243

FINDING A THERAPY FOR ALS

Despite intensive efforts, ALS remains an incurable disorder. Riluzole, a glutamate antagonist, is currently the only drug that has been approved for the treatment of ALS, but it provides only a marginal therapeutic benefit by retarding the decline in muscle strength and prolonging survival by only a few months (Bensimon et al., 1994). Neurotrophic factors have been considered for therapy of motoneuron degeneration, but previous clinical trials have all failed. Indeed, intrathecal infusion of brain-derived growth factor (BDNF) and ciliary neurotrophic factor (CNTF), ICV delivery of glial cell line-derived neurotrophic factor (GDNF) or systemic administration of BDNF or CNTF did not result in clinical improvement in ALS patients (reviewed by Aebischer and Ridet, 2001; Thoenen and Sendtner, 2002; Mitsumoto et al., 2004). When delivered systemically, insulin-like growth factor (IGFl) marginally reduced disease progression in one but not in another trial (Lai et al., 1997; Borasio et al., 1998). In S O D I ^ ^ S A mice, leukemia inhibitory factor (LIF) improved motoneuron survival without prolonging lifespan in one study, but had no effect in another study (Azari et al., 2003; Feeney et al., 2003), whereas delivery of BDNF, GDNF, CNTF, or IGFl has never been evaluated in SODl^^^^ mice. Some of these clinical trials with neurotrophic agents may have failed because of inappropriate delivery of these neurotrophic factors. Indeed, when delivered systemically, neurotrophic factors often fail to cross the blood-brain barrier and thus never reach the motoneurons. In addition, these polypeptides are usually rapidly cleared from the circulation and may evoke a neutralizing immune reaction after repetitive injections (Thorne and Frey, 2001). When delivered into the CSF, neurotrophins may also be rapidly cleared into the venous or lymphatic system, or be unable to penetrate into the surrounding parenchyma because they are sequestered by cellular receptors on the ependymal lining — all preventing these neurotrophic factors to reach the degenerating motoneuron. Optimizing the appropriate delivery route is thus critical for therapeutic success.

VL

A.

GENE THERAPY APPROACHES FOR ALS

General Requirements: Targeting Motoneurons

Treatment of motoneuron degeneration with neurotrophins requires chronic delivery. This can be achieved

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

244

19. VEGF, AN ANGIOGENIC FACTOR WITH NEUROTROPHIC ACTIVITY, USEFUL FOR TREATMENT OF ALS?

by gene therapy or repetitive protein delivery. An advantage of viral gene transfer is that the transduced cells produce a 'natural' neurotrophic protein with a mammalian-type glycosylation pattern, w^hich is not recognized as a foreign antigen. In the past, only viral vectors were available which were capable of infecting the motoneuron cell bodies when the viral vector was injected directly into the anterior horn of the spinal cord. Such invasive surgery is obviously not a therapeutic option in humans. More recently, improved vectors have been developed, which facilitate the delivery of neurotrophic factors to the motoneuron soma. These vectors take advantage of the intrinsic property of the motoneuron to retrogradely transport cargo from the nerve terminal via its long axon to the cell body (Fig. 1). The polio, herpes, and rabies viruses are capable of infecting the nerve terminals at the neuromuscular junction. Thereafter, these neurotropic viruses highjack the axonal transport system and are thereby capable of reaching the motoneuron cell body (Charlton, 1994; Mazarakis et al., 2001; Kaspar et al., 2003). Adeno-associated adenoviral vectors (AAV) and lentiviral vectors pseudotyped with rabies-G envelopes have been constructed, which are retrogradely transported via the axons upon intramuscular injection (see below). These vectors are particularly relevant for the treatment of motoneuron disorders, as they offer the opportunity to administer these vectors via peripheral intramuscular injection. Therapeutic effects following gene therapy in SODl^^^^ mice have so far only been described for adeno-associated adenoviruses and rabies-derived lentiviral viruses. B.

Adeno'Associated Viruses for ALS

AAV-based vectors are promising vectors for gene therapy of ALS (Wang et al., 2002; Kaspar et al., 2003; Lu et al., 2003). AAV is a naturally occurring nonpathogenic virus that is inherently replicationdefective since adenovirus coinfection is required to initiate a productive infection (VandenDriessche et al., 2003). AAV vectors are capable of achieving long-term transgene expression in the absence of viral gene expression and transduce nondividing cells in vivo (Monahan and Samulski, 2000). The prototype AAV vector, AAV serotype 2, has been used safely and successfully in gene therapy trials (Manno et al., 2003) and animal models (Monahan and Samulski, 2000). When injected in the brain or spinal cord, AAV2-based vectors are neurotropic and preferentially transduce neurons, including motoneurons, with a greater efficiency than glial cells (Bartlett et a l , 1998; Wang

et al., 2002; Kaspar et al., 2003; Lu et al., 2003). However, when injected in the muscle, AAV is also retrogradely transported from presynaptic nerve terminals via the axon to the cell soma, where it enters the cell nucleus to provide sustained gene delivery. Indeed, injection of AAV vectors carrying green fluorescent protein (AAVGFP) injected into the hindlimb (HL) quadriceps and intercostal muscles of SODl^^^^ mice resulted in GFP expression in CHAT-positive motoneurons within the ventral horn of the lumbar and thoracic spinal cord, respectively, with no evidence of glial cell transduction (Kaspar et al., 2003). Up to 1.1% of the virus injected at a dose of 1 X 10^^ viral particles was transported to the lumbar region of the spinal cord (Kaspar et al., 2003). Several groups have successfully employed AAV2based vectors to deliver, either by intramuscular delivery or intraspinal injection, therapeutic transgenes such as IGFl, anti-apoptotic bcl-2, cardiotrophin-1, and GDNF to spinal motoneurons in ALS mouse models (Table 1) (Azzouz et al., 2000; Bordet et al., 2001; Wang et al., 2002; Lu et al., 2003). Intraspinal injection of AAV encoding bcl-2 in SODl^^^^ mice resulted in sustained bcl-2 expression in motoneurons and increased the number of surviving motoneurons around the injection site at the end-stage of disease (Azzouz et al., 2000). Although bcl-2 expression in spinal motoneurons delayed the onset of motor deficiency, it failed to prolong the survival of SODl^^^^ mice, likely due to the restricted transduction pattern, which was limited to the injection site (Azzouz et al., 2000). The most robust effects obtained with AAV vectors were those reported by Kaspar et al., who injected AAV2-IGF1 into the HL quadriceps and intercostal muscles of SODl*^^^^ mice, resulting in a 30% increase in lifespan (Kaspar et al., 2003). In the same study, delivery of AAV2-GDNF had only very modest effects, prolonging survival by only 9%. Importantly, injection of IGFl-expressing viral particles at the onset of disease still prolonged the lifespan of these mice by approximately 18%. Some of the observed therapeutic effect might be attributable to retrograde transport of the IGFl protein itself. Indeed, a lentiviral vector, that produced IGFl in the muscle but was not retrogradely transported to the spinal cord, increased the survival of SODl^^^"^ mice by only 9 days. Thus, viral vectors that are retrogradely transported to the motoneuron cell body improve the overall therapeutic efficiency The observed therapeutic benefit with the AAV2-IGF1 vector is one of the greatest effects in the field. Why is AAV2-IGF1 gene transfer so effective? Possibly, because of the pleiotropic and multitasking activity of this factor. In the brain, IGFl promotes growth, differentiation

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

245

GENE THERAPY APPROACHES FOR ALS

TABLE 1

Therapeutic Effects of Gene or Protein Therapy in ALS Rodent Models

Delay in disease onset (days)

Delay in survival (days)

Delay in survival after treatment at disease onset (days)

Reference

InSODlG93Aniice AAV-GDNF

13

17

n.d.

Wang et al. (2002)

AAV-CTl

27

11-13

n.d.

Bordet et al. (2001)

AAV2-bcl-2

10

n.s.

n.d.

Azzouz et al. (2000)

AAV2-GDNF

16

11

6 days

Kaspar et al. (2003)

AAV2-IGF1

31

35

22 days

Kaspar et al. (2003)

EIAV-VEGF

28

38

19 days

Azzouz (2004b)

EIAV-hSODi

108

100

n.d.

Ralph et al. (2005)

Lenti-hSODi

20

n.d.

n.d.

Raoul et al. (2005)

17

22

10

Masu et al. (1993)

InSODlG93Arats VEGF ICV

Notes. AAV, adeno-associated virus; EIAV, equine infectious anaemia virus; GDNF, glial cell line-derived neurotrophic factor; CTl, cardiotrophin-1; VEGF, vascular endothelial growth factor; IGFl, insulin-like growth factor 1; hSODi, human SOD interfering RNA; ICV, intracerebroventicular; n.d., not determined; and n.s., not significant.

and survival of various neural and glial cell types, i.e. neurons, Schwann cells, oligodendrocytes, astrocytes, and neural stem cells (Dore et al., 1996; Aberg et al., 2000; Lichtenwalner et al., 2001). Interestingly, IGFlnuU mice exhibit hypomyelination in the CNS and have less white matter in the spinal cord, suggesting a prominent role for IGFl in axonal growth and maturation and myelination (Beck et al., 1995; Gao et a l , 1999). IGFl also has effects on the motoneuron system. Motoneurons express the IGFl receptor (Dore et al., 1997) and ^^^I-IGFl binding sites in the ventral spinal cord are increased in ALS patients (Dore et al., 1996). IGFl is transported in an antero- and a retrograde direction in the sciatic nerve (Ishii et al., 1994) and increases neuromuscular function and muscle strength (Rabinovsky et al., 2003). In addition to its survival effect on motoneurons following axotomy IGFl increases nerve terminal sprouting, accelerates the functional recovery of injured nerves, overcomes cycloheximide-mediated inhibition of outgrowth of sciatic nerves after a crush lesion, and attenuates peripheral motor neuropathies induced by chemotherapeutic agents such as vincristine, cisplatin, and paclitaxel (reviewed in Dore et al., 1996). All these pleiotropic effects of IGFl may have contributed to its overall therapeutic effect. C.

Lentiviral Vectors for ALS

The prototypical lentiviral vectors are suitable for gene therapy approaches because of the following reasons: (i) lentiviral vectors have a large insert

capacity (8-10 kb); (ii) they efficiently transduce not only dividing but also nondividing cells, such as neurons (Mazarakis et al., 2001; Azzouz and Mazarakis, 2004); (iii) Antiviruses elicit a minimal (if any at all) inflammatory response, which can compromise the viability of the transduced cells (Bensadoun et al., 2000, 2003; Kordower et al., 2000; Mazarakis et al., 2001); and (iv) they integrate genes into the chromosome of the target cells, leading to stable long-term expression (Mitrophanous et al., 1999; Bensadoun et al., 2000, 2003; Kordower et al., 2000; Mazarakis et al., 2001). Lentiviral vectors can be derived from primates, e.g. the human immunodeficiency virus (HIV), or from non-primates, e.g. the equine infectious anaemia virus (EIAV). Lentiviral vectors have been pseudotyped with a wide variety of envelope glycoproteins, for example, those from the vesicular stomatitis virus (VSV-G) or from the rabies virus (rabies-G) (Mazarakis et al., 2001). Pseudotyping of vectors with such envelopes confers a specific tissue tropism in vivo. After injection into the brain, lentiviral vectors pseudotyped with VSV-G envelope preferably transduce neurons. The expressed protein is anterogradely transported throughout the cell bodies and axons; retrograde transport of the VSV-G vectors has however never been demonstrated (Mazarakis et al., 2001). Rabies-G-pseudotyped EIAV vectors, in contrast, enter the central nervous system by retrograde transport and transduce spinal motoneurons after intramuscular injection in mice (Mazarakis et al., 2001; Wong et al., 2004). Indeed, 4 weeks after

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

246

19. VEGF, AN ANGIOGENIC FACTOR WITH NEUROTROPHIC ACTIVITY, USEFUL FOR TREATMENT OF ALS?

the injection of a high titer EIAV vector, carrying the LacZ reporter gene (EIAV-LacZ), unilaterally into the gastrocnemius and facial muscles of 21-day-old SODl^^^^ mice, extensive reporter gene expression w^as observed in both the spinal cord and brainstem (Azzouz et al., 2004b). By staining for the calcitonin gene-related peptide (CGRP) to directly identify the spinal motoneurons, more than 60% of CGRP-positive motoneurons could be transduced, illustrating that EIAV-based lentiviral vectors are effective in delivering transgenes to motoneurons after injection into the muscle. This methodology has already been successfully applied for spinal muscular atrophy (SMA), which is characterized by muscle weakness due to degeneration of the lower motoneurons in the spinal cord and brain stem early in childhood (Gendron and MacKenzie, 1999; Sendtner, 2001). SMA is a recessive autosomal inherited disorder caused by mutations or deletions in the telomeric survival motoneuron gene 1 (SMNl), but not the centromeric SMN2. In mice, a homozygous knockout of the SMN gene — mice have only a single SMN gene that is equivalent to human SMNl — results in early embryonic lethality due to massive cell death (Schrank et al., 1997). Transgenic mice that express one or two copies of the human SMN2 gene and that were mated onto the null SMN"/~ background have a normal number of motoneurons at birth, and subsequently develop massive motoneuron loss by postnatal day 5 leading to death around postnatal day 13 (Monani et al., 2000). This phenotype closely resembles the severe SMA phenotype in humans and serves as an animal model of the disease. Intercrossing mice that overexpress eight copies of the human transgene rescues the phenotype in these SMA mice, indicating that increasing SMN2 levels may be useful for treatment of SMA (Monani et al., 2000). By injecting rabies-G pseudotyped EIAV lentivectors expressing human SMN2 in various muscles of SMA mice, SMN expression in motoneurons was substantially restored, motoneuron death reduced and survival expectancy increased by 5 days (Azzouz et al., 2004a). Although this may seem a modest effect, it is relevant considering that the time window in SMA mice is relatively short to obtain full expression of the SMN transgene, i.e. lentiviral-SMN administration was performed at postnatal day 2, whereas control-injected mice already die at postnatal day 13. Gene therapy approaches for neurodegenerative diseases can also be used to knockdown the expression of dominant disease genes. In a recent study from Xia et al., short hairpin RNAs directed against ataxin

1 were expressed by AAV viruses delivered into the brain of spinocerebellar ataxia type 1 (SCAl) mice. This RNAi-based approach improved the neuropathological hallmarks and protected the transduced areas against neuronal loss. Most familial forms of ALS are dominant in nature and the gene silencing approach may therefore also be promising for FALS. However, whereas Xia et al. targeted both the mutant and normal ataxin 1 transcripts, the challenge for ALS in particular will be to silence the mutant but not the normal SODl allele. Silencing of the wild-type allele would be detrimental since SODl-knockout mice develop various defects, including age-associated loss of neuromuscular synapses and fertility problems (Reaume et al., 1996; Matzuk et al., 1998; Shefner et al., 1999). In SODl^^^^ transgenic mice, the mutated SODl protein can be more selectively targeted than in humans. Indeed, based on sequence differences, Ralph et al., designed an EIAV vector expressing a human SOD interfering RNA, EIAV-hSODi, specifically targeting the human but not the endogeneous murine SODl gene. When delivered intramuscularly to SODl^^^^ mice, the therapeutic effect of the EIAV-hSODi vector was impressive: the onset of ALS symptoms was delayed from 94 to 202 days and the lifespan was increased by almost 80% (128 days in control-treated versus 228 days in EIAV-hSODi treated mice; Table 1) (Ralph et al., 2005). In humans, the mutant SODl protein would need to be targeted more selectively. Several studies already described the in vitro generation of silencing RNAs specific for mutant SODl-alleles, but targeting individual point mutations in the SODl protein does not always represent a realistic therapeutic approach in a rare disease such as FALS, for which more than 100 different point mutations in the same gene are responsible for only 2% of all ALS patients. Therefore, Raoul et al. (2005) designed a single vector allowing for the knockdown of all SODl forms, while expressing a wild-type protein, which is refractory to RNAi-based silencing. In case such a vector turns out to be as effective as the EIAV-hSODi, gene therapybased RNA interference might offer novel promising treatment opportunities for FALS and other dominantly inherited neurodegenerative disorders. D,

Herpes Simplex Virus 1-Derived Vectors

Herpes simplex virus (HSV) vectors have many properties that make them attractive for gene delivery to motoneurons following intramuscular application. HSV naturally infects postmitotic cells such as neurons. Up to 20% of the 152 kb of the endogenous

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

THE THERAPEUTIC POTENTIAL OF VEGF FOR ALS

viral genome can be replaced by exogenous transgenes without significantly altering the properties of the virus (Palmer et al., 2000). This implicates that large or multiple transgenes can be expressed by a single viral construct. Similar to AAV2 and EIAV, HSV is retrogradely transported to the cell body where it enters a latent state for the lifetime of the infected cell (Perez et al., 2004). As this vector does not integrate into the host cell genome, the risk of insertional (in)activation of host cell genes is minimal. The lifetime presence of the HSV genome within the host cell as a nuclear extrachromosomal entity allows longterm transgene expression. During HSV latency, a unique region in the HSV genome remains transcriptionally active and drives the expression of transgenes in neighboring regions for extended periods (Coins et al., 1994). Therefore, these vectors are potentially useful tools to investigate the functions of genes involved in motoneuron survival and regeneration (Perez et a l , 2004). HSV-VEGF gene transfer to dorsal root ganglia has been shown to prevent diabetic neuropathy, but the potential of HSV gene transfer in ALS remains to be determined (Chattopadhyay et al., 2005).

VIL

THE THERAPEUTIC POTENTIAL OF VEGF FOR ALS

A. VEGF Delivery with the Retrogradely Transported EIAV Lentiviral Vector To evaluate the therapeutic potential of VEGF for ALS, a rabies-G pseudotyped EIAV lentiviral vector encoding the human VEGF gene, EIAV-VEGF, was constructed. After confirming that this vector was able to produce active VEGF, EIAV vectors were injected bilaterally into the HL gastrocnemius, diaphragm, intercostal, facial, and tongue muscles of SODl^^^^ mice (Azzouz et al., 2004b). VEGF treatment significantly extended the lifespan of these mice by an average of 38 days. Thereby, VEGF treatment achieved one of the most effective therapies reported in the field so far, prolonging survival by as much as 30% (Table 1). The size of the observed effect was comparable to that reported for IGFl when delivered with AAV2 vectors. Notably, administration of VEGF therapy at the time of disease onset was also effective, prolonging survival by as much as 19 days. When injecting a similar amount of a rabies-G pseudotyped EIAV vector expressing GDNF, a negligible increase in survival of 6 days was observed. EIAV-VEGF gene

247

transfer was well tolerated and did not cause vascular side effects. B.

Recombinant VEGF Protein Therapy

Though clinical trials with viral vectors are being considered, the clinical applicability, feasibility, efficacy, and safety of gene therapy for ALS remains to be established. Delivery of recombinant neurotrophic growth factors, on the other hand, is clinically very attractive, as it offers flexible control of the dose and duration of the administered protein. Previous clinical trials with delivery of a neurotrophic factor failed, however, to show a therapeutic effect, presumably because the factors were insufficiently active, immunogenic, rapidly cleared or inappropriately delivered. By using osmotic mini-pumps implanted subcutaneously on the back of SODl^^^^ rats, another rodent model of ALS (Nagai et al., 2001; Howland et al., 2002), we were capable of delivering the recombinant protein continuously via a catheter stereotactically implanted in the left lateral ventricles of rats for more than 100 days (Storkebaum et al., 2005). Distribution studies revealed that ^^^I-VEGF, after bolus injection into the left lateral ventricle, diffused from the CSF into the parenchyme of the brain and spinal cord and reached the motoneurons, where it remained intact for several hours (Storkebaum et al., 2005) (Fig. 1). Compared to rats treated with artificial cerebrospinal fluid (aCSF), SODl^^^^ rats receiving 0.2 ]ig VEGF/kg/day exhibited a delayed disease onset, an improved motor performance, a longer spontaneous activity and prolonged survival. When scoring spontaneous activity, for instance, VEGF-treated animals consistently performed better than controls for as long as 35-42 days, whereas survival of SODl^^^"^ rats was prolonged by 22 days after VEGF treatment (Storkebaum et al., 2005). SOD1G93A rats tolerated the VEGF treatment well without adverse effects. To evaluate whether ICV delivery of VEGF would be capable of prolonging survival when initiated at the time of disease onset, 0.2 ]ig VEGF/kg/day was administered from 85 days of age, when the first signs of motor impairment emerged. Despite the very rapidly progressing disease course of this model, treatment of SOD1G93A rats with ICV delivered VEGF significantly improved survival by 10 days (Storkebaum et al., 2005). VEGF treatment at 60 days also changed the disease subtype from a severe to a much milder form. Indeed, significantly fewer SODl^^^^ rats suffered from the severe forelimb (FL) onset type of disease after VEGF than aCSF delivery. Compared to aCSF-treated

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

248

19. VEGF, AN ANGIOGENIC FACTOR WITH NEUROTROPHIC ACTIVITY, USEFUL FOR TREATMENT OF ALS?

rats with HL- or FL-onset, VEGF-treated rats survived 17 or 27 days longer, respectively (Storkebaum et al., 2005). This is a relevant finding, as involvement of brain stem and cervical disease results in a worse prognosis in both rats and humans. The more pronounced therapeutic effect of VEGF on FL than hindlimb muscles is likely attributable to the higher VEGF levels in the bulbar/cervical than in the lumbar spinal cord upon ICV delivery. Recently, systemic delivery of recombinant VEGF was also shown to prolong the survival of SODl^^^^ mice (Zheng et al., 2004). As VEGF is unlikely to cross the blood-brain barrier, these data may possibly suggest that VEGF affects motoneuron survival indirectly via effects on the vasculature. C.

Therapeutic Mechanisms of VEGF

How can we explain the therapeutic potential of VEGF? Apart from the importance of the chosen route of delivery (retrograde axonal transport for EIAV-VEGF; ICV delivery of VEGF protein), VEGF also has distinct biological activities, which differ from those of other neurotrophic factors. Similar to IGFl, VEGF has pleiotropic effects on various different types of neuronal and glial cells, stimulates neurogenesis, is axonally transported in an antero- and retrograde direction and stimulates neural perfusion (Cao et al., 2004; Greenberg and Jin, 2004; Storkebaum et al., 2005; Vande Velde and Cleveland, 2005) — all these effects might have contributed to promote motoneuron survival. In addition, while VEGF and other neurotrophins are capable of affecting motoneurons, their importance in the pathogenesis of ALS may differ substantially. Indeed, even though over- or under-expression of BDNF, GDNF, CNTF, or LIE affected motoneurons during development and injured motoneurons after birth (Sendtner et al., 1992; Masu et al., 1993; Holtmann et al., 2005), and although adenoviral gene transfer of some of these factors stimulated motoneuron survival in ALS mouse models (Wang et a l , 2002; Kaspar et al., 2003; Lu et al., 2003; Azzouz et al., 2004b; ), loss of these neurotrophins did not cause adult-onset ALS-like motoneuron degeneration and paralysis in mice. In addition, little or no human genetic evidence supports a role for these factors in human ALS. All these findings therefore raise the question as to whether these factors significantly participate in the pathogenesis of chronic motoneuron degeneration in the adult. In contrast, the findings that reduced levels of VEGF in mice and humans increase the risk of ALS (Lambrechts et al., 2003, 2004) indicate that VEGF is an important endogenous

regulator of motoneuron survival and is implicated in the pathogenesis of ALS. Thus, VEGF may have a more important role in the pathogenesis of ALS and therefore also be a better therapeutic target for ALS. D.

Perspectives of VEGF Therapy

The challenge for the future will be to translate the findings, obtained in preclinical animal models, to the clinic. VEGF therapy showed significant promise in SODl^^^^ rodent models of ALS (Azzouz et al., 2004b; Storkebaum et al., 2005). However, several hurdles need to be overcome. For instance, ALS is caused by SODl mutations in only a small fraction of human subjects and it thus remains to be determined whether VEGF will be useful for sporadic cases of ALS with unknown etiology. Our genetic findings that VEGF gene variations, lowering VEGF levels, also increased the risk of sporadic ALS (Lambrechts et al., 2003) might raise hope that VEGF could be effective in these patients as well. Will VEGF therapy be safe? VEGF is a potent angiogenic factor, which can also increase vascular permeability. Administration of excessive doses of VEGF may lead to neovascularization, edema formation and ventriculomegaly (Harrigan et al., 2003). Flexible control over the dose and duration of VEGF gene or protein delivery is thus crucial. This is currently more readily achieved with ICV delivery of recombinant VEGF, but development of viral vector systems with regulatable expression may perhaps also offer such control in the future. Also, a single agent such as VEGF or IGFl is unlikely to be sufficient to cure ALS, raising the question which combination or cocktail of agents need to be delivered. In this respect, it would be interesting to assess whether VEGF and IGFl would confer additive or synergistic effects. Since IGFl activates the expression of VEGF in colon cancer cells (Akagi et al., 1998), the therapeutic effect of IGFl may at least partly be mediated by VEGF. Clinical trials in the near future will help to address these outstanding issues. References Aberg, M.A., Aberg, N.D., Hedbacker, H., Oscarsson, J. and Eriksson, RS. (2000) Peripheral infusion of IGF-I selectively induces neurogenesis in the adult rat hippocampus. J. Neurosci., 20: 2896-2903. Aebischer, P. and Ridet, J. (2001) Recombinant proteins for neurodegenerative diseases: the delivery issue. Trends Neurosci., 24: 533-540. Akagi, Y, Liu, W., Zebrowski, B., Xie, K. and Ellis, L.M. (1998) Regulation of vascular endothelial growth factor expression in

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

THE THERAPEUTIC POTENTIAL OF VEGF FOR ALS human colon cancer by insulin-like growth factor-L Cancer Res., 58: 4008-4014. Andersen, P.M. (2001) Genetics of sporadic ALS. Amyotroph. Lateral Scler Other Motor Neuron. Disord., 2(Suppl 1): S37-S41. Azari, M.R, Lopes, E.G., Stubna, G., Turner, B.J., Zang, D., Nicola, N.A., Kurek, J.B. and Gheema, S.S. (2003) Behavioural and anatomical effects of systemically administered leukemia inhibitory factor in the SODl(G93A GIH) mouse model of familial amyotrophic lateral sclerosis. Brain Res., 982: 92-97. Azzouz, M., Hottinger, A., Patema, J.G., Zum, A.D., Aebischer, P. and Bueler, H. (2000) Increased motoneuron survival and improved neuromuscular function in transgenic ALS mice after intraspinal injection of an adeno-associated virus encoding Bcl-2. Hum. Mol. Genet., 9: 803-811. Azzouz, M., Le, T., Ralph, G.S., Walmsley L., Monani, U.R., Lee, D.G., Wilkes, P., Mitrophanous, K.A., Kingsman, S.M., Burghes, A.H. and Mazarakis, N.D. (2004a) Lentivector-mediated SMN replacement in a mouse model of spinal muscular atrophy. J. Glin. Invest., 114:1726-1731. Azzouz, M. and Mazarakis, N. (2004) Non-primate EIAV-based lentiviral vectors as gene delivery system for motor neuron diseases. Gurr. Gene Ther., 4: 277-286. Azzouz, M., Ralph, G.S., Storkebaum, E., Walmsley, L.E., Mitrophanous, K.A., BCingsman, S.M., Garmeliet, P. and Mazarakis, N.D. (2004b) VEGF delivery with retrogradely transported lentivector prolongs survival in a mouse ALS model. Nature, 429: 413^17. Bagnard, D., Vaillant, G., Khuth, S.T., Dufay N., Lohrum, M., Puschel, A.W., Belin, M.R, Bolz, J. and Thomasset, N. (2001) Semaphorin 3A-vascular endothelial growth factor-165 balance mediates migration and apoptosis of neural progenitor cells by the recruitment of shared receptor. J. Neurosci., 21: 3332-3341. Bartlett, J.S., Samulski, R.J. and McGown, T.J. (1998) Selective and rapid uptake of adeno-associated virus type 2 in brain. Hum. Gene Ther., 9: 1181-1186. Beck, K.D., Powell-Braxton, L., Widmer, H.R., Valverde, J. and Hefti, F. (1995) Igfl gene disruption results in reduced brain size, GNS hypomyelination, and loss of hippocampal granule and striatal parvalbumin-containing neurons. Neuron, 14: 717-730. Bensadoim, J.G., Deglon, N., Tseng, J.L., Ridet, J.L., Zurn, A.D. and Aebischer, P. (2000) 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, 164: 15-24. Bensadoun, J.G., Pereira de Almeida, L., Fine, E.G., Tseng, J.L., Deglon, N. and Aebischer, P. (2003) Comparative study of GDNF delivery systems for the GNS: polymer rods, encapsulated cells, and lentiviral vectors. J. Control Release, 87:107-115. Bensimon, G., Lacomblez, L. and Meininger, V. (1994) A controlled trial of riluzole in amyotrophic lateral sclerosis. ALS/Riluzole Shidy Group. N. Engl. J. Med., 330: 585-591. Borasio, G.D., Robberecht, W , Leigh, PN., Emile, J., Guiloff, R.J., Jerusalem, R, Silani, V., Vos, RE., Wokke, J.H. and Dobbins, T. (1998) A placebo-controlled trial of insulin-like growth factor-I in amyotrophic lateral sclerosis. European ALS/IGF-I Study Group. Neurology, 51: 583-586. Bordet, T, Lesbordes, J.G., Rouharu, S., Castelnau-Ptakhine, L., Schmalbruch, H., Haase, G. and Kahn, A. (2001) Protective effects of cardiotrophin-1 adenoviral gene transfer on neuromuscular degeneration in transgenic ALS mice. Hum. Mol. Genet., 10:1925-1933. Breier, G., Albrecht, U., Sterrer, S. and Risau, W. (1992) Expression of vascular endothelial growth factor during embryonic angio-

249

genesis and endothelial cell differentiation. Development, 114: 521-532. Breier, G., Glauss, M. and Risau, W. (1995) Coordinate expression of vascular endothelial growth factor receptor-1 (flt-1) and its ligand suggests a paracrine regulation of murine vascular development. Dev Dyn., 204: 228-239. Brown, R.H., Jr. and Robberecht, W. (2001) Amyotrophic lateral sclerosis: pathogenesis. Semin. Neurol., 21:131-139. Bruijn, L.I., Miller, T.M. and Cleveland, D.W. (2004) Unraveling the mechanisms involved in motor neuron degeneration in ALS. Aimu. Rev Neurosci., 27: 723-749. Cao, L., Jiao, X., Zuzga, D.S., Liu, Y, Fong, D.M., Young, D. and During, M.J. (2004) VEGF links hippocampal activity with neurogenesis, learning and memory. Nat. Genet., 36: 827-835. Carmeliet, P. (2003a) Angiogenesis in health and disease. Nat. Med., 9: 653-660. Carmeliet, P. (2003b) Blood vessels and nerves: common signals, pathways and diseases. Nat. Rev. Genet., 4: 710-720. Carmeliet, P., Perreira, V, Breier, G., PoUefeyt, S., Kieckens, L., Gertsenstein, M., Fahrig, M., Vandenhoeck, A., Harpal, K., Eberhardt, C., Declercq, C , Pawling, J., Moons, L., CoUen, D., Risau, W and Nagy, A. (1996) Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature, 380: 435-439. Charlton, K.M. (1994) The pathogenesis of rabies and other lyssaviral infections: recent studies. Gurr. Top Microbiol. Immunol., 187: 95-119. Chattopadhyay, M., Krisky, D., Wolfe, D., Glorioso, J.C., Mata, M. and Fink, D.J. (2005) HSV-mediated gene transfer of vascular endothelial growth factor to dorsal root ganglia prevents diabetic neuropathy. Gene Therapy advance online publication, 21 April 2005. Chen, Y.Z., Bennett, C.L., Huynh, H.M., Blair, LP, Puis, I., Irobi, J., Dierick, I., Abel, A., Kennerson, M.L., Rabin, B.A., Nicholson, G.A., Auer-Grumbach, M., Wagner, K., De Jonghe, P., Griffin, J.W., Pischbeck, K.H., Timmerman, V, Comblath, D.R. and Chance, PR (2004) DNA/RNAhelicase gene mutations in a form of juvenile amyotrophic lateral sclerosis (ALS4). Am. J. Hum. Genet., 74:1128-1135. Cleveland, D.W. (1999) Prom Charcot to SODl: mechanisms of selective motor neuron death in ALS. Neuron, 24: 515-520. Del Bo, R., Scarlato, M., Ghezzi, S., Martinelli Boneschi, R, Fenoglio, C., Galbiati, S., Virgilio, R., Galimberti, D., Galimberti, G., Crimi, M., Ferrarese, C , Scarpini, E., Bresolin, N. and Comi, G.P (2005) Vascular endothelial growth factor gene variability is associated with increased risk for AD. Ann. Neurol, b7\ 373-380. Devos, D., Moreau, C., Lassalle, P., Perez, T, De Seze, J., BrunaudDanel, V, Destee, A., Tonnel, A.B. and Just, N. (2004) Low levels of the vascular endothelial growth factor in CSF from early ALS patients. Neurology, 62: 2127-2129. Dore, S., Kar, S. and Quirion, R. (1997) Rediscovering an old friend, IGF-I: potential use in the treatment of neurodegenerative diseases. Trends Neurosci., 20: 326-331. Dore, S., Krieger, C., Kar, S. and Quirion, R. (1996) Distribution and levels of insulin-like growth factor (IGF-I and IGF-II) and insulin receptor binding sites in the spinal cords of amyotrophic lateral sclerosis (ALS) patients. Brain Res. Mol. Brain Res., 41: 128-133. Dumont, D.J., Fong, G.H., Puri, M.C., Gradwohl, G., Alitalo, K. and Breitman, M.L. (1995) Vascularization of the mouse embryo: a study of flk-1, tek, tie, and vascular endothelial growth factor expression during development. Dev. Dyn., 203: 80-92.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

250

19. VEGF, AN ANGIOGENIC FACTOR WITH NEUROTROPHIC ACTIVITY, USEFUL FOR TREATMENT OF ALS?

Feeney, S.J., Austin, L., Bennett, T.M., Kurek, J.B., Jean-Francois, MJ., Muldoon, C. and Byrne, E. (2003) The effect of leukaemia inhibitory factor on SODl G93A murine amyotrophic lateral sclerosis. Cytokine, 23: 108-118. Ferrara, N. (2005) The role of VEGF in the regulation of physiological and pathological angiogenesis. EXS 94: 209-231. Ferrara, N., Carver-Moore, K., Chen, H., Dowd, M., Lu, L., O'Shea, K.S., Powell-Braxton, L., Hillan, K.J. and Moore, M.W. (1996) Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature, 380: 439-i42. Ferrara, N., Gerber, H.P. and LeCouter, J. (2003) The biology of VEGF and its receptors. Nat. Med., 9: 669-676. Ferrara, N., Hillan, K.J., Gerber, H.P. and Novotny, W. (2004) Discovery and development of bevacizumab, an anti-VEGF antibody for treating cancer. Nat. Rev. Drug Discov., 3: 391^00. Forstreuter, F., Lucius, R. and Mentlein, R. (2002) Vascular endothelial growth factor induces chemotaxis and proliferation of microglial cells. J. NeuroimmunoL, 132: 93-98. Gao, W.Q., Shinsky, N., Ingle, G., Beck, K., Elias, K.A. and PowellBraxton, L. (1999) IGF-I deficient mice show reduced peripheral nerve conduction velocities and decreased axonal diameters and respond to exogenous IGF-I treatment. J. NeurobioL, 39: 142-152. Gendron, N.H. and MacKenzie, A.E. (1999) Spinal muscular atrophy: molecular pathophysiology. Curr. Opin. Neurol., 12: 137-142. Coins, W.F., Sternberg, L.R., Croen, K.D., Krause, PR., Hendricks, R.L., Fink, D.J., Straus, S.E., Levine, M. and Glorioso, J.C. (1994) A novel latency-active promoter is contained within the herpes simplex virus type 1 UL flanking repeats. J. Virol., 68: 2239-2252. Greenberg, D.A. and Jin, K. (2004) VEGF and ALS: the luckiest growth factor? Trends Mol. Med., 10: 1-3. Gros-Louis, R, Laurent, S., Lopes, A.A., Khoris, J., Meininger, V., Camu, W. and Rouleau, G.A. (2003) Absence of mutations in the hypoxia response element of VEGF in ALS. Muscle Nerve, 28: 774r-775. Haigh, J.J., Morelli, PL, Gerhardt, H., Haigh, K., Tsien, J., Damert, A., Miquerol, L., Muhlner, U., Klein, R., Ferrara, N., Wagner, E.F., Betsholtz, C. and Nagy, A. (2003) Cortical and retinal defects caused by dosage-dependent reductions in VEGF-A paracrine signaling. Dev Biol., 262: 225-241. Hand, C.K., Devon, R.S., Gros-Louis, R, Rochefort, D., Khoris, J., Meininger, V., Bouchard, J.P, Camu, W., Hayden, M.R. and Rouleau, G.A. (2003) Mutation screening of the ALS2 gene in sporadic and familial amyotrophic lateral sclerosis. Arch. Neurol., 60: 1768-1771. Hand, C.K. and Rouleau, G.A. (2002) Familial amyotrophic lateral sclerosis. Muscle Nerve, 25: 135-159. Harrigan, M.R., Ennis, S.R., Sullivan, S.E. and Keep, R.R (2003) Effects of intraventricular infusion of vascular endothelial growth factor on cerebral blood flow, edema, and infarct volume. Acta Neurochir. (Wien) 145: 49-53. Holtmann, B., Wiese, S., Samsam, M., Grohmann, K., Pennica, D., Martini, R. and Sendtaier, M. (2005) Triple knock-out of CNTF, LIE, and CT-1 defines cooperative and distinct roles of these neurotrophic factors for motoneuron maintenance and function. J. Neurosci., 25: 1778-1787. Howland, D.S., Liu, J., She, Y., Goad, B., Maragakis, N.J., Kim, B., Erickson, J., Kulik, J., DeVito, L., Psaltis, G., DeGeimaro, L.J., Cleveland, D.W. and Rothstein, J.D. (2002) Focal loss of the glutamate transporter EAAT2 in a transgenic rat model of SODl mutant-mediated amyotrophic lateral sclerosis (ALS). Proc. Natl. Acad. Sci. USA, 99: 1604-1609.

Hurwitz, H., Fehrenbacher, L., Novotny, W., Cartwright, T., Hainsworth, J., Heim, W., Berlin, J., Baron, A., Griffing, S., Holmgren, E., Ferrara, N., Fyfe, G., Rogers, B., Ross, R. and Kabbinavar, R (2004) Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N. Engl. J. Med., 350: 2335-2342. Ishii, D.N., Glazner, G.W. and Pu, S.R (1994) Role of insulin-like growth factors in peripheral nerve regeneration. Pharmacol. Then, 62:125-144. Kalaria, R.N. (2002) Small vessel disease and Alzheimer's dementia: pathological considerations. Cerebrovasc. Dis. 13(Suppl 2): 48-52. Kaspar, B.K., Llado, J., Sherkat, N., Rothstein, J.D. and Gage, RH. (2003) Retrograde viral delivery of IGP-1 prolongs survival in a mouse ALS model. Science, 301: 839-842. Kordower, J.H., Emborg, M.E., Bloch, J., Ma, S.Y., Chu, Y, Leventhal, L., McBride, J., Chen, E.Y, Palfi, S., Roitberg, B.Z., Brown, W.D., Holden, J.E., Pyzalski, R., Taylor, M.D., Carvey, P, Ling, Z., Trono, D., Hantraye, P., Deglon, N. and Aebischer, P. (2000) Neurodegeneration prevented by lentiviral vector delivery of GDNF in primate models of Parkinson's disease. Science, 290: 767-773. Krum, J.M., Mani, N. and Rosenstein, J.M. (2002) Angiogenic and astroglial responses to vascular endothelial growth factor administration in adult rat brain. Neuroscience, 110: 589-604. Lai, E.G., Felice, K.J., Festoff, B.W., Gawel, M.J., Gelinas, D.R, Kratz, R., Murphy M.R, Natter, H.M., Norris, RH. and Rudnicki, S.A. (1997) Effect of recombinant human insulin-like growth factor-I on progression of ALS. A placebo-controlled study. The North America ALS/IGP-I Study Group. Neurology, 49: 1621-1630. Lambrechts, D., Storkebaum, E. and Carmeliet, P (2004) VEGF: necessary to prevent motoneuron degeneration, sufficient to treat ALS? Trends Mol. Med., 10: 275-282. Lambrechts, D., Storkebaum, E., Morimoto, M., Del-Favero, J., Desmet, R, Marklund, S.L., Wyns, S., et al. (2003) VEGR is a modifier of amyotrophic lateral sclerosis in mice and humans and protects motoneurons against ischemic death. Nat. Genet., 34: 383-394. Lichtenwalner, R.J., Porbes, M.E., Bennett, S.A., L)mch, C D . , Sonntag, W.E. and Riddle, D.R. (2001) Intracerebroventricular infusion of insulin-like growth factor-I ameliorates the age-related decline in hippocampal neurogenesis. Neuroscience, 107: 603-613. Lu, YY, Wang, L.J., Muramatsu, S., Ikeguchi, K., Rujimoto, K., Okada, T., Mizukami, H., Matsushita, T., Hanazono, Y, Kume, A., Nagatsu, T., Ozawa, K. and Nakano, I. (2003) Intramuscular injection of AAV-GDNF results in sustained expression of transgenic GDNF, and its delivery to spinal motoneurons by retrograde transport. Neurosci. Res., 45: 3 3 ^ 0 . Manno, C.S., Chew, A.J., Hutchison, S., Larson, P.J., Herzog, R.W., Arruda, V.R., Tai, S.J., Ragni, M.V., Thompson, A., Ozelo, M., Couto, L.B., Leonard, D.G., Johnson, F.A., McClelland, A., Scallan, C , Skarsgard, E., Flake, A.W., Kay, M.A., High, K.A. and Glader, B. (2003) AAV-mediated factor IX gene transfer to skeletal muscle in patients with severe hemophilia B. Blood, 101: 2963-2972. Masu, Y, Wolf, E., Holtmann, B., Sendtner, M., Brem, G. and Thoenen, H. (1993) Disruption of the CNTF gene results in motor neuron degeneration. Nature, 365: 27-32. Matzuk, M.M., Dionne, L., Guo, Q., Kumar, T.R. and Lebovitz, R.M. (1998) Ovarian function in superoxide dismutase 1 and 2 knockout mice. Endocrinology, 139: 4008^011. Mazarakis, N.D., Azzouz, M., Rohll, J.B., EUard, RM., Wilkes, R.J., Olsen, A.L., Carter, E.E., Barber, R.D., Baban, D.R, Kingsman, S.M., Kingsman, A.J., O'Malley, K. and Mitrophanous, K.A. (2001) Rabies virus glycoprotein pseudotyping of lentiviral vectors

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

THE THERAPEUTIC POTENTIAL OF VEGF FOR ALS enables retrograde axonal transport and access to the nervous system after peripheral delivery. Hum. Mol. Genet., 10: 2109-2121. Millauer, B., Wizigmann-Voos, S., Schnurch, H., Martinez, R., Moller, N.P., Risau, W. and Ullrich, A. (1993) High affinity VEGF binding and developmental expression suggest Flk-1 as a major regulator of vasculogenesis and angiogenesis. Cell, 72: 835-846. Mitrophanous, K., Yoon, S., Rohll, J., Patil, D., Wilkes, P., Kim, V., Kingsman, S., Kingsman, A. and Mazarakis, N. (1999) Stable gene transfer to the nervous system using a non-primate lentiviral vector. Gene Ther., 6: 1808-1818. Mitsumoto, H., Gordon, P., Kaufmann, P., Gooch, C.L., Przedborski, S. and Rowland, L.P. (2004) Randomized control trials in ALS: lessons learned. Amyotroph. Lateral Scler. Other Motor Neuron. Disord., 5 (Suppl 1): 8-13. Monahan, RE. and Samulski, R.J. (2000) Adeno-associated virus vectors for gene therapy: more pros than cons? Mol. Med. Today, 6: 433^40. Monani, U.R., Sendtner, M., Coovert, D.D., Parsons, D.W., Andreassi, C., Le, T.T., Jablonka, S., Schrank, B., Rossol, W., Prior, T.W., Morris, G.E. and Burghes, A.H. (2000) The human centromeric survival motor neuron gene (SMN2) rescues embryonic lethality in Smn(-/-) mice and results in a mouse with spinal muscular atrophy. Hum. Mol. Genet., 9: 333-339. Mueller, M.D., Vigne, J.L., Minchenko, A., Lebovic, D.I., Leitman, D.C. and Taylor, R.N. (2000) Regulation of vascular endothelial growth factor (VEGF) gene transcription by estrogen receptors alpha and beta. Proc. Natl. Acad. Sci. USA, 97:10972-10977. Nagai, M., Aoki, M., Miyoshi, I., Kato, M., Pasinelli, P., Kasai, N., Brown, R.H., Jr. and Itoyama, Y. (2001) Rats expressing human cytosolic copper-zinc superoxide dismutase transgenes with amyotrophic lateral sclerosis: associated mutations develop motor neuron disease. J. Neurosci., 21: 9246-9254. Nishimura, A.L., Mitne-Neto, M., Silva, H.C., Richieri-Costa, A., Middleton, S., Cascio, D., Kok, R, Oliveira, J.R., Gillingwater, T, Webb, J., Skehel, P. and Zatz, M. (2004) A mutation in the vesicle-trafficking protein VAPB causes late-onset spinal muscular atrophy and amyotrophic lateral sclerosis. Am. J. Hum. Genet., 1^\ 822-831. Oosthuyse, B., Moons, L., Storkebaum, E., Beck, H., Nuyens, D., Brusselmans, K., Van Dorpe, J., et al. (2001) Deletion of the hypoxia-response element in the vascular endothelial growth factor promoter causes motor neuron degeneration. Nat. Genet., 28: 131-138. Palmer, J.A., Branston, R.H., Lilley, C.E., Robinson, M.J., Groutsi, R, Smith, J., Latchman, D.S. and Coffin, R.S. (2000) Development and optimization of herpes simplex virus vectors for multiple long-term gene delivery to the peripheral nervous system. J. Virol, 74: 5604-5618. Perez, M.C., Hunt, S.P, Coffin, R.S. and Palmer, J.A. (2004) Comparative analysis of genomic HSV vectors for gene delivery to motor neurons following peripheral inoculation in vivo. Gene Ther., 11: 1023-1032. Popovici, C , Isnardon, D., Birnbaum, D. and Roubin, R. (2002) Caenorhabditis elegans receptors related to mammalian vascular endothelial growth factor receptors are expressed in neural cells. Neurosci. Lett., 329:116-120. Rabinovsky, E.D., Gelir, E., Gelir, S., Lui, H., Kattash, M., DeMayo, F.J., Shenaq, S.M. and Schwartz, R.J. (2003) Targeted expression of IGF-1 transgene to skeletal muscle accelerates muscle and motor neuron regeneration. FASEB J., 17: 53-55. Ralph, G.S., Radcliffe, PA., Day D.M., Carthy J.M., Leroux, M.A., Lee, D.C, Wong, L.F., Bilsland, L.G., Greensmith, L., Kingsman, S.M., Mitrophanous, K.A., Mazarakis, N.D. and Azzouz, M. (2005)

251

Silencing mutant SODl using RNAi protects against neurodegeneration and extends survival in an ALS model. Nat. Med., 11: 429-433. Raoul, C , Abbas-Terki, T., Bensadoun, J.C, Guillot, S., Haase, G., Szulc, J., Henderson, C.E. and Aebischer, P. (2005) Lentiviralmediated silencing of SODl through RNA interference retards disease onset and progression in a mouse model of ALS. Nat. Med., 11(4): 423-428. Reaume, A.G., Elliott, J.L., Hoffman, E.K., Kowall, N.W., Ferrante, R.J., Siwek, D.R, Wilcox, H.M., Flood, D.G., Beal, M.R, Brown, R.H., Jr., Scott, R.W. and Snider, W D . (1996) Motor neurons in C u / Z n superoxide dismutase-deficient mice develop normally but exhibit enhanced cell death after axonal injury. Nat. Genet., 13: 4 3 ^ 7 . Rosen, D.R., Siddique, T., Patterson, D., Figlewicz, D.A., Sapp, P., Hentati, A., Donaldson, D., Goto, J., O'Regan, J.P, Deng, H.X., Rouleau, G.A., Gusella, G.S., Horvitz, R. and Brown, R.H. (1993) Mutations in C u / Z n superoxide dismutase gene are associated with familial amyotrophic lateral sclerosis. Nature, 362: 59-62. Rosenstein, J.M., Mani, N., Khaibullina, A. and Krum, J.M. (2003) Neurotrophic effects of vascular endothelial growth factor on organotypic cortical explants and primary cortical neurons. J. Neurosci., 23: 11036-11044. Rowland, L.P. and Shneider, N.A. (2001) Amyotrophic lateral sclerosis. N. Engl. J. Med., 344: 1688-1700. Schanzer, A., Wachs, F.P., Wilhelm, D., Acker, T., Cooper-Kuhn, C , Beck, H., Winkler, J., Aigner, L., Plate, K.H. and Kuhn, H.G. (2004) Direct stimulation of adult neural stem cells in vitro and neurogenesis in vivo by vascular endothelial growth factor. Brain Pathol., 14: 237-248. Schrank, B., Gotz, R., Gunnersen, J.M., Ure, J.M., Toyka, K.V, Smith, A.G. and Sendtner, M. (1997) Inactivation of the survival motor neuron gene, a candidate gene for human spinal muscular atrophy, leads to massive cell death in early mouse embryos. Proc. Natl. Acad. Sci. USA, 94: 9920-9925. Schratzberger, P., Schratzberger, G., Silver, M., Curry, C , Kearney, M., Magner, M., Alroy, J., Adelman, L.S., Weinberg, D.H., Ropper, A.H. and Isner, J.M. (2000) Favorable effect of VEGF gene transfer on ischemic peripheral neuropathy. Nat. Med., 6: 405-413. Schratzberger, P., Walter, D.H., Rittig, K., Bahlmann, F.H., Pola, R., Curry, C , Silver, M., Krainin, J.G., Weinberg, D.H., Ropper, A.H. and Isner, J.M. (2001) Reversal of experimental diabetic neuropathy by VEGF gene transfer. J. Clin. Invest., 107: 1083-1092. Schwarz, Q., Gu, C , Fujisawa, H., Sabelko, K., Gertsenstein, M., Nagy, A., Taniguchi, M., Kolodkin, A.L., Ginty, D.D., Shima, D.T. and Ruhrberg, C. (2004) Vascular endothelial growth factor controls neuronal migration and cooperates with Sema3A to pattern distinct compartments of the facial nerve. Genes Dev., 18: 2822-2834. Sendtner, M. (2001) Molecular mechanisms in spinal muscular atrophy: models and perspectives. Curr. Opin. Neurol., 14: 629-634. Sendtner, M., Holtmann, B., Kolbeck, R., Thoenen, H. and Barde, Y.A. (1992) Brain-derived neurotrophic factor prevents the death of motoneurons in newborn rats after nerve section. Nature, 360: 757-759. Shaw, C.E., al-Chalabi, A. and Leigh, N. (2001) Progress in the pathogenesis of amyotrophic lateral sclerosis. Curr. Neurol. Neurosci. Rep., 1: 69-76. Shefner, J.M., Reaume, A.G., Flood, D.G., Scott, R.W., Kowall, N.W., Ferrante, R.J., Siwek, D.R, Upton-Rice, M. and Brown, R.H., Jr. (1999) Mice lacking cytosolic copper/zinc superoxide dismutase display a distinctive motor axonopathy Neurology, 53: 1239-1246.

IV GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

252

19. VEGF, AN ANGIOGENIC FACTOR WITH NEUROTROPHIC ACTIVITY, USEFUL FOR TREATMENT OF ALS?

Shiote, M., Nagano, L, Ilieva, H., Murakami, T., Narai, H., Ohta, Y, Nagata, T., Shoji, M. and Abe, K. (2005) Reduction of a vascular endothelial growth factor receptor, fetal liver kinase-1, by antisense oligonucleotides induces motor neuron death in rat spinal cord exposed to hypoxia. Neuroscience, 132: 175-182. Siddique, T. and Lalani, I. (2002) Genetic aspects of amyotrophic lateral sclerosis. Adv. Neurol, 88: 21-32. Silverman, W.F., Krum, J.M., Mani, N. and Rosenstein, J.M. (1999) Vascular, glial and neuronal effects of vascular endothelial growth factor in mesencephalic explant cultures. Neuroscience, 90: 1529-1541. Sondell, M., Lundborg, G. and Kanje, M. (1999) Vascular endothelial growth factor has neurotrophic activity and stimulates axonal outgrowth, enhancing cell survival and Schwann cell proliferation in the peripheral nervous system. J. Neurosci., 19: 5731-5740. Sorenson, E. (2004) Great Lakes ALS Study Group. Amyotroph. Lateral Scler Other Motor Neuron. Disord., 5(Suppl 1): 128. Stalmans, I., Lambrechts, D., De Smet, R, Jansen, S., Wang, J., Maity, S., Kneer, R, et al. (2003) VEGF: a modifier of the del22qll (DiGeorge) syndrome? Nat. Med., 9:173-182. Stalmans, I., Ng, Y.S., Rohan, R., Fruttiger, M., Bouche, A., Yuce, A., Fujisawa, H., Hermans, B., Shard, M., Jansen, S., Hicklin, D., Anderson, D.J., Gardiner, T., Hammes, H.R, Moons, L., Dewerchin, M., Collen, D., Carmeliet, R and D'Amore, RA. (2002) Arteriolar and venular patterning in retinas of mice selectively expressing VEGF isoforms. J. Clin. Invest., 109: 327-336. Storkebaum, E., Lambrechts, D. and Carmeliet, R (2004) VEGF: once regarded as a specific angiogenic factor, now implicated in neuroprotection. Bioessays, 26: 943-954. Storkebaum, E., Lambrechts, D., Dewerchin, M., Moreno-Murciano, M.R, Appelmans, S., Oh, H., Van Damme, P., Rutten, B., Man, W.Y, De Mol, M., Wyns, S., Manka, D., Vermeulen, K., Van Den Bosch, L., Mertens, N., Schmitz, C , Robberecht, W., Conway, E.M., Collen, D., Moons, L. and Carmeliet, R (2005) Treatment of motoneuron degeneration by intracerebroventricular delivery of VEGF in a rat model of ALS. Nat. Neurosci., 8: 85-92. Terry, RD., Kamel, R, Umbach, D.M., Lehman, T.A., Hu, H., Sandler, D.R and Taylor, J. A. (2004) VEGF promoter haplotype and amyotrophic lateral sclerosis (ALS). J. Neurogenet., 18: 429-434.

Thoenen, H. and Sendtner, M. (2002) Neurotrophins: from enthusiastic expectations through sobering experiences to rational therapeutic approaches. Nat. Neurosci., 5 (Suppl): 1046-1050. Thorne, R.G. and Frey, W.H., 2nd (2001) Delivery of neurotrophic factors to the central nervous system: pharmacokinetic considerations. Clin. Rharmacokinet., 40: 907-946. Turner, M.R., Rarton, M.J. and Leigh, RN. (2001) Clinical trials in ALS: an overview. Semin. Neurol. 21: 167-175. Vande Velde, C. and Cleveland, D.W. (2005) VEGF: multitasking in ALS. Nat. Neurosci., 8: "5-7. Vanden Driessche, T., Collen, D. and Chuah, M.K. (2003) Gene therapy for the hemophilias. J. Thromb. Haemost., 1: 1550-1558. Wang, L.J., Lu, YY, Muramatsu, S., Ikeguchi, K., Fujimoto, K., Okada, T., Mizukami, H., Matsushita, T., Hanazono, Y, Kume, A., Nagatsu, T., Ozawa, K. and Nakano, I. (2002) Neuroprotective effects of glial cell line-derived neurotrophic factor mediated by an adeno-associated virus vector in a transgenic animal model of amyotrophic lateral sclerosis. J. Neurosci., 22: 6920-6928. Wong, L.R, Azzouz, M., Walmsley L.E., Askham, Z., Wilkes, F.J., Mitrophanous, K.A., Kingsman, S.M. and Mazarakis, N.D. (2004) Transduction patterns of pseudotyped lentiviral vectors in the nervous system. Mol. Ther., 9:101-111. Xie, Y, Weydt, R, Rowland, D.S., Kliot, M. and Moller, T. (2004) Inflammatory mediators and growth factors in the spinal cord of G93A SODl rats. Neuroreport, 15: 2513-2516. Yang, S.R, Bae, D.G., Kang, H.J., Gwag, B.J., Gho, YS. and Chae, C.B. (2004) Co-accumulation of vascular endothelial growth factor with beta-amyloid in the brain of patients with Alzheimer's disease. Neurobiol. Aging, 25: 283-290. Yang, Y, Hentati, A., Deng, H.X., Dabbagh, O., Sasaki, T., Hirano, M., Hung, W.Y, Ouahchi, K., Yan, J., Azim, A.C., Cole, N., Gascon, G., Yagmour, A., Ben-Hamida, M., Pericak-Vance, M., Hentati, F. and Siddique, T. (2001) The gene encoding alsin, a protein with three guanine-nucleotide exchange factor domains, is mutated in a form of recessive amyotrophic lateral sclerosis. Nat. Genet., 29: 160-165. Zheng, C , Nennesmo, I., Fadeel, B. and Henter, J.I. (2004) Vascular endothelial growth factor prolongs survival in a transgenic mouse model of ALS. Ann. Neurol., 56: 564-567.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

C H A P T E R

20 Viral Vector Axonal Uptake and Retrograde Transport: Mechanisms and Applications Qingshan Teng, Thais Federici, Nicholas M. Boulis

Abstract: In this chapter, we discuss the forms and mechanisms of neuronal axonal transport, with emphasis on the retrograde axonal transport of viruses and viral vectors as vehicles for therapeutic gene delivery into the central nervous system (CNS). We first describe the axonal uptake as a physiological pathway for neurotropic viruses and other agents, such as the tetanus toxin, to invade the CNS. We then concentrate on the advantages of viral vectors that naturally undergo retrograde axonal transport, as well as the strategies to enhance or modify their tropism for CNS gene delivery Finally, we discuss the applications of retrograde axonal transport of viral vectors for the treatment of neurological disorders, including neurodegenerative diseases, neuropathy, pain, and spasticity. Keywords: axonal transport; nervous system; gene therapy; neurodegenerative diseases; pain; viral vectors; neurotropic virus

in the cell body and delivered through the long axon to the axon terminals (Hirokawa and Takemura, 2004). The axons making up the long tracts of the spinal cord within the blue whale are often invoked to illustrate the distance that these processes must traverse relative to the size of the soma. In turn, the delivery of material synthesized in the cell body across this distance poses a unique challenge for the cell biology of neurons. In the nervous system, neurons are not only connected to each other forming complex circuits, they also form extensive synaptic connections with their target tissues. These target cells and tissues secret neurotrophic factors, such as nerve growth factor (NGF), which are instrumental to neuronal functionality (Reynolds et a l , 2000; Campenot and Maclnnis, 2004). It is therefore equally important for neurons to receive these factors from their terminals through axons. Neurons also need to sense changes in their surroundings and transmit information using surface receptors on the soma or nerve terminals. This bidirectional flow of materials or information between

L NEURONAL AXONAL TRANSPORT Neurons and neuroglial cells are the major cells that form the nervous system. Three distinct parts compose most neurons: a cell body, multiple dendrites, and a single axon. The cell body of a neuron, also known as the perikaryon or soma, contains the nucleus and organelles responsible for the metabolism and gene expression that support cell survival. The cell processes, i.e. axons and dendrites, are a unique feature of these neurons that underlie their cell-to-cell signaling properties. Dendrites project from the cell body. They are specialized in receiving stimuli from other cells or axons. Neurons usually possess multiple dendrites whose structure plays a role in the integration of synaptic information. In contrast, most neurons have a single long axon. The axon conducts impulses away from the cell body to other neurons or target organs. Axons communicate with other cells through synapses. They lack protein synthesis machinery. Therefore, all materials within axons or at synapses are synthesized

Gene Therapy of the Central Nervous System: From Bench to Bedside

253

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

254

20. VIRAL VECTOR AXONAL UPTAKE AND RETROGRADE TRANSPORT MECHANISMS AND APPLICATIONS

the soma and nerve terminal occurs through a process called axonal transport (Hirokav^a, 1993). There are two forms of axonal transport between neuronal cell body and terminal axons (Hirokawa and Takemura, 2004). Anterograde axonal transport involves the movement of proteins and subcellular structures from the cell body through the axon to the synapses. Retrograde transport is regarded as an exaggerated manifestation of the normal endocytosis process seen in cells in general. It moves lysosomes and multivesicular bodies from the nerve terminals back to the cell body (Oztas, 2003). Depending upon the material transport rate, axonal transport can be further divided into fast axonal transport and slow axonal transport (Oztas, 2003; Vallee and Bloom, 1991). Fast axonal transport is composed of both anterograde and retrograde components and moves peptides between soma and nerve terminals (Ochs, 1972). Slow axonal transport involves the cytoskeletal protein distribution within the axon. Slow transport has been traditionally considered to be an anterograde transport process. Now it is clear that slow axonal transport is also a retrograde transport component (Glass and Griffin, 1994). The force of fast axonal transport derives from molecular motors, mainly members of the kinesin and dynein superfamilies (Schiavo and Chao, 2004). Like myosin proteins, which are actin (microfilaments)dependent molecules, kinesin and dynein proteins are microtubule-dependent motor molecules. Microtubules in axons and distal dendrites are unipolar. The axon terminus end points toward the terminal, and the minus end points toward the soma. Kinesin family proteins move toward the plus end of microtubules and may carry vesicles. Thus the movement of kinesin is responsible for anterograde transport (Hirokawa and Takemura, 2004). In contrast, dynein proteins move toward the minus end of the microtubules and therefore provide the driving force for retrograde axonal transport. This process transfers materials from the periphery back to the cell body (Fig. 1 and Table 1) (Oztas, 2003; Vallee et al., 2004). II.

MECHANISMS OF UPTAKE A N D RETROGRADE TRANSPORT OF NEUROTROPIC VIRUSES

As alluded to above, retrograde transport plays a critical role in the survival and functionality of neurons by delivering target-derived neurotrophic signals and material recycling from the axon terminals back

AXOM Belrograde

Vesfcte ,

O CO

c

Dynein

Microtubule Anterograde

2 C X

< c

FIGURE 1. Molecular mechanisms of axonal transport. From Oztas (2003, Fig. 1, p. 3), with permission.

TABLE 1 A Summary of Components of Axonal Transport. Axonal transport

Velocity (mm/day)

Transporting substances

Fast transport

200^00

Synaptic vesicle, enzymes, neurotransmitters

Mitochondrial transport

50-100

Mitochondria

0.1-1.0

Tubulin, neurofilament protein Actin, clathrine, calmodulins, spectrin, cytoplasmic enzymes Prelysosomal vesicles, recycled proteins, HRP, WGA, neurotrophic viruses

Anterograde transport

Slow transport Slow components a (Sea) Slow components b (Scb) Retrograde transport

2-6

100-200

Note: HRP, horseradish peroxidase; WGA, wheat germ agglutinin.

From Oztas (2003, Table 1, p. 4).

to the soma. This physiological process also serves as an important pathway whereby some neurotropic viruses and toxins invade the central nervous system (CNS) from peripheral sites (Bisby, 1982; Schwab and Thoenen, 1977). As early as in 1884, Pasteur postulated that rabies virus could be transported along the peripheral nerve (Pasteur et al., 1884). This hypothesis was later extended to other viruses such as herpes virus. Rabies virus is an enveloped, negative-sense RNA virus. Its RNA genome is about 11.9 kb. The virion contains a nucleoprotein, a phosphoprotein, and an RNA-dependent RNA polymerase. Its lipoprotein envelope is a host-derived lipid bilayer embedded with rabies glycoproteins. The virus is transmitted through bites by infected animals such as dogs, foxes, and raccoons, among others. During the incubation period, the virus replicates in the muscles. Rabies virus binds to nicotinic acetylcholine

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

MECHANISMS OF UPTAKE AND RETROGRADE TRANSPORT OF NEUROTROPIC VIRUSES

receptors at the neuromuscular junction and migrates retrogradely to the neuronal cell body (Lentz et al., 1982). There is also recent evidence showing that the neural cell adhesion molecule (Thoulouze et al., 1998), the N-methyl-D-aspartate NRl receptor (Gosztonyi and Ludw^ig, 2001), and the neurotrophin receptor p75NTR (Langevin et al., 2002; Tuffereau et al., 2001) are also rabies virus receptors. It was reported that rabies virus migrates through the fast axonal transport component at a rate of 12-100 mm per day (Tsiang, 1993). Reports also indicated that molecular motors such as cytoplasmic dynein and myosin are involved in this axonal transport process. Rabies virus phosphoprotein interacts with cytoplasmic dynein light chain (LC8). The latter is a component in both cytoplasmic dynein, which is involved in microtubule-directed organelle transport, and myosin V, which is involved in actin-based vesicle transport in axons (Jacob et al., 2000; Poisson et al., 2001; Raux et al., 2000). The exact form in which the rabies virus is transported into the CNS from the periphery remains an area of research. It is postulated that rabies virus could be transported either in the form of naked nucleocapsid (de-enveloped viral particle) or as an intact virus (Warrell and Warrell, 2004). Rabies virus belongs to the genus Lyssavirus. It is therefore not surprising that other viruses in this group, such as Lagos bat virus, Duvenhage virus, and European bat Lyssavirus (EBL), employ a similar pathway to gain access to CNS. Herpes virus is also considered neurotropic. Herpes simplex virus (HSV) is an enveloped double-stranded (ds) DNA virus (Roizman and Sears, 1993). The HSV genome consists of 152 kb of dsDNA arranged as unique long and short segments (UL and US) and flanked by inverted repeated sequences (McGeoch et al., 1988). The mature wild-type virion consists of a core of dsDNA surrounded by an icosahedral capsid. The outer viral coat consists of a lipid envelope. The viral envelope contains viral glycoproteins, which principally mediate viral entry into the host cells (Spear et al., 2000). An amorphous layer known as the tegument exists between the viral capsid and the envelope. The tegument contains a number of proteins that are responsible for induction of viral gene expression and shutoff of host cell protein synthesis (Kwong and Frenkel, 1989). Viral infection is initiated by the contact of host skin or mucous with HSV Viruses released at the site of the primary infection may infect neighboring epithelial cells or are taken up by sensory nerve terminals innervating the region of infection. The latter process could be enhanced by the high level of expression of herpes virus entry mediator

255

C (HveC) on sensory nerve terminal membranes in the skin (Geraghty et al., 1998; Mata et al., 2001). It was reported that the viral envelope is not a necessary component for retrograde transport of the virus. The de-enveloped viral particle (capsid particle) could be efficiently transported along the axon after peripheral injection into the squid axon (Bearer et al., 2000). Viral nucleocapsids and surrounding tegument are carried by retrograde axonal transport along axonal microtubules to the neuronal cell body in the peripheral sensory ganglion (Cook and Stevens, 1973; Sodeik et al., 1997). It is very interesting that both herpes virus and rabies virus also undergo anterograde axonal transport after amplification in cell soma. Miranda-Saksena et al. (2000) reported that herpes virus envelope glycoproteins and the capsid were transported by different pathways from the nucleus to the axon. Interestingly, the direction of herpes virus transneuronal transport is strain-dependent. Zemanick et al. (1991) reported the isolation of one herpes virus strain that spreads preferentially in the anterograde direction and another strain that spreads in the retrograde direction. Both herpes virus and pseudorabies virus belong to the Alphaherpesvirinae subfamily A swine herpes virus known as porcine pseudorabies virus (PrV) was also reported to undergo retrograde transport after infection. This virus is not a human virus, but is the causative agent of Aujeszky's disease in livestock. While pseudorabies virus is not currently being engineered for human gene transfer, it is widely used as a tracer in neuroanatomy to uncover neuronal connections and to map neural circuits in the CNS (Dolivo et al., 1978,1979; Martin and Dolivo, 1983). Other neurotropic viruses such as poliovirus and GD VII strain of Theiler's virus also spread through axonal transport pathway (Martinat et al., 1999; Ohka et al., 1998). Viruses are not the only hazard agents that use axonal transport machinery to invade the CNS (von Bartheld, 2004). During evolution, clostridial tetanus toxin has gained the ability to reach CNS through retrograde transport mechanism. Tetanus toxin is composed of a heavy chain (HC) and a light chain (LC) linked through a disulphide bond. The LC is a protease, responsible for its intracellular activity. The HC mediates binding and internalization into the neurons (Sinha et al., 2000). After binding to the nerve terminals, the toxin is transported through fast retrograde axonal transport back to the cell body. This retrograde transport was reported to rely on microtubule-dependent motors (Lalli et al., 2003). The neurotrophin receptor p75NTR has been reported

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

256

20. VIRAL VECTOR AXONAL UPTAKE AND RETROGRADE TRANSPORT MECHANISMS AND APPLICATIONS

to act as an axon terminal receptor for the HC. This protein is also a receptor for NGR Like neurotropic viruses, the HC component of tetanus toxin has been used as a tracer to map neuronal network in the CNS (Kissa et al., 2002). More importantly the HC component of tetanus toxin has been employed as a gene transfer vehicle to ferry foreign genes specifically into the motor neurons (Francis et al., 2004). Interestingly, the prion proteins were also reported to traffic in the nervous system in both the retrograde and anterograde fashion (Borchelt et al., 1994; Moya et al., 2004). Ill- C U R R E N T DATA O N UPTAKE A N D RETROGRADE TRANSPORT OF VIRAL VECTORS Besides being used in neuroanatomy as a tracer (Oztas, 2003), neurotropic viruses are logical tools for gene transfer into the CNS. Before a detailed discussion of the application of neurotropic viruses in gene transfer, we will briefly review literature on nonneurotropic viral vectors used for gene transfer in the context of retrograde transport. Adenovirus and adeno-associated virus (AAV) are not traditionally considered to be neurotropic viruses. However, they have been used widely as retrograde axonal transporters in CNS gene transfer. Ridoux et al. (1994) reported that an adenoviral vector containing the E. coli ^-galactosidase gene (AdRSVj^gal) was transported to nigral neurons from the striatum through the retrograde transport pathway They observed that the distribution of ^-galactosidase labeled neurons in the nigral region was similar to that observed following injection of horseradish peroxidase (HRP) or fluorescence dye into the rat striatum. Furthermore, they also demonstrated that dopaminergic neurons were the only cell population that was targeted by the retrogradely transported virus. By using a similar approach, other researchers demonstrated that spinal motor neurons or sensory neurons in the CNS could be transduced through the retrograde axonal flow of virus after intramuscular (IM) injection of recombinant adenovirus (Finiels et al., 1995; Ghadge et al., 1995). Early studies by our laboratory (Boulis et al., 1999) compared retrograde axonal transport of adenoviral vectors following intraneural, IM and subcutaneous (SC) viral injection. Sciatic nerve viral injection resulted in a significant higher CNS uptake than that obtained through SC or IM injection (Fig. 2). Intraneural injection into both crushed and uncrushed

sciatic resulted in retrograde transport. However, the mechanism whereby viral uptake could occur into intact axons is unclear and counterintuitive. In an attempt to enhance adenoviral vector-mediated retrograde transport, Millecamps et al. (2002) used botulinum neurotoxin to induce nerve sprouting at the neuromuscular junction. They demonstrated that treatment with botulinum neurotoxin significantly enhanced gene transfer to projection motor neurons after IM injection. Although botulinum neurotoxin is one of the most toxic poisons reported, it is widely used in spasticity treatment. Theoretically, this toxin could be used to enhance other viral vectors to achieve more robust retrograde transport. The mechanism underlying the non-neurotropic viral vector-mediated retrograde gene transfer remained unclear. Ridoux et al. (1994) demonstrated that adenoviral vectors expressing ^-galactosidase injected into the striatum were retrogradely transported to the substantia nigra. On the other hand, injection of jS-galactosidase protein into the striatum failed to produce any j6-galactosidase signal in the substantia nigra, indicating that the retrogradely transported agent was adenoviral vectors but not the transgene product. Colchicine is an antigout agent that acts by suppressing microtubule polarization. Our laboratory (Boulis et al., 2003b) employed colchicine to implicate a microtubule-dependent retrograde transport mechanism in the trafficking of non-neurotropic viral vectors. Sciatic colchicine injection concurrent with or before viral vector intraneural delivery inhibited CNS uptake of peripherally injected vectors. Colchicine injection inhibited retrograde transport of both adenoviral vectors and adeno-associated viral vectors. This observation is consistent with studies performed on rabies virus, which showed that colchicine or vinblastin prevented rabies virus transport from the peripheral site of inoculation to the CNS (Tsiang, 1979). It is accepted that rabies virus is transported through retrograde transport via dynein-microtubule interaction. This study therefore provided direct evidence that recombinant adenoviral or AAV vectors could be transported through microtubule-mediated retrograde transport system, even though the nerve terminal receptors for these vectors are unclear at the present time. Canine adenovirus serotype 2 (CAV-2)-based vectors are a recently generated vector system showing features that support CNS application. When compared to the human adenoviruses type 2- and 5-based vectors, these vectors possess a larger cloning capacity and lower induced immunogenicity. Further, these

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

CURRENT DATA ON UPTAKE AND RETROGRADE TRANSPORT OF VIRAL VECTORS

257

FIGURE 2 Spinal cord lacZ gene expression after injection of an attenuated adenoviral vector, AdSRSVntLacZ, into the sciatic nerve. (A) Cellular staining in both gray and white matter 12 days after injection of the sciatic nerve. Red Arrows: small white matter cellular jS-gal expression; Green Arrows: dorsal horn neuronal staining; Blue Arrows: staining in ventral horn motor neurons. (B) DRG neuronal staining 6 days after injection. (C) Gene expression is detected in the perineurium but not in axons at the site of injection. From Boulis et al. (1999, Fig. 1, p. 133), with permission.

vectors show preferential transduction of neurons and remarkable retrograde transport when injected into different regions of the brain and skeletal muscles (Kremer et a l , 2000; Soudais et a l , 2001; Peltekian et al., 2002). Soudais et al. (2001) reported that a CAV-2 vector injected into the rat striatum was successfully transferred to the ipsilateral cortex, substantia nigra and centromedian nucleus. Furthermore, IM injection of the CAV-2 vector resulted in a significant number of ventral horn motor neuron transduction. These authors reported that CAV-2 vectors induced a more robust retrograde axonal transport than a serotype 5 adenoviral vector. The same CAV-2 viral vector was also reported to have specifically transduced dopami-

nergic neurons of the substantia nigra following virus injection into the striatum (Peltekian et al., 2002). Recombinant AAV viral vectors have also been applied to therapeutic models through retrograde transport. A recombinant AAV2 vector harboring the Bcl-xL gene was retrogradely transported into the projection areas of entorhinal cortex and substantia nigra after hippocampus and striatum injection (Kaspar et al., 2002). Interestingly, Kaspar also characterized the retrograde transport of other AAV serotypes to motor neurons in the spinal cord after IM injection, with serotypes 1, 2, and 6 showing the best performances among all the serotypes tested (Kaspar et al., 2003, 2004). The concern that robust AAV binding in

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

258

20. VIRAL VECTOR AXONAL UPTAKE AND RETROGRADE TRANSPORT: MECHANISMS AND APPLICATIONS

FIGURE 3 Fluorescence microscopy following sciatic nerve injection of rAAVCAG-GFP. (A) Magnified view of GPP expressing spinal cord motor neurons 21 days after remote rAAVCAG-GFP delivery (40X). (B) DRG primary sensory neurons expressing GFP 21 days after sciatic injection (40X). (C) Anti-GFP immunofluorescence reveals neuronal GFP expression in the spinal cord ventral horn (40X). (D) Anti-GFP immunofluorescence reveals abundant GFP in the DRG (40X). The superimposed green bar measures 50 |jim in Size. From Boulis et al. (2003, Fig. 2, p. 538), with permission.

muscle would undermine peripheral axonal binding and uptake motivated our group to investigate intraneural injection as an alternative approach. In our hands, sciatic AAV injection resulted in motor neuron gene delivery. In comparison with Ad5 vectors, this expression was markedly more robust 3 weeks after injection (Boulis et al., 2003a) (Fig. 3). It is important to point out that retrograde axonal transport of AAV2 vectors has been disputed. Martinov et al. (2002) reported that IM injection of an AAV type 2 virus containing green fluorescent protein (GFP) did not induce any spinal motor neuron transduction. Direct injection into CNS parenchyma takes advantage of the privileged immunological environment to reduce the vector directed inflammatory response (Thomas et al., 2001). In contrast, peripheral injection necessarily presents vector and transgene epitopes to antigen presenting cells, potentially triggering immunemediated shutdown of gene expression. A variety of evidence supports a non-cytolytic immune-mediated mechanism for the deterioration in vector gene expression (Boulis et al., 2002; Teng et al., 2005). Adenovirus and AAV are not the only non-neurotropic viruses that undergo retrograde axonal transport. Li et al. (2004) reported that a baculoviral vector injected into the rat striatum resulted in gene expression in the dopaminergic neurons in the cerebral cortex, and substantia nigra to a lesser extent. The most

striking feature of the baculovirus in this study was its transsynaptic spread in the CNS. Following baculovirus vector injection into the vitreous body of the rat eye, gene expression was detected in the retina, optic nerve, the lateral geniculate body, the superior coUiculus and the primary visual cortex. It is probably too early to speculate about the disadvantages or advantages of this transsynaptic property of baculovirus vector. But it is possible that a controlled and targeted gene expression in a specific neurological pathway might be desired for CNS gene therapy. IV.

NEUROTROPIC VIRUSES AS GENE TRANSFER VECTORS

The application of neurotropic viruses, mainly herpes and rabies virus, as gene transfer vectors comes in two forms. First, genetically modified herpes virus itself can be used as a gene transfer vector. Second, considering the pathogenic characteristics of rabies virus, rabies virus G glycoprotein (Rabies-G) can be used to pseudotype commonly used gene transfer vectors to lend these vectors enhanced CNS penetration. As previously mentioned, HSV gains access to the CNS of the host via the epithelial surface. HSV initially replicates within the epithelial cells. The progeny HSV particles are released and bind to the peripheral nerve terminals, which innervate the site of primary

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

TETANUS TOXIN HC AS A PROTEIN CARRIER TARGETING CNS

%i^^^/^5|#|.

m•

259

, ^ ^ >:

m.

]^y\r^

J ^

o

^

FIGURE 4 In vivo expression of jS-gal activity in the Wistar rat spinal cord mediated by a live attenuated HSV-1 vector. Lumbar (LI) spinal cord, 182 days after inoculation of a live attenuated HSV-1 vector in the right gastrocnemius muscle. Arrows show p-ga\ positive cells in motor neurons. The scale bars are 1.0 mm in lower magnification and 50 ]xn\ in higher magnification. Adapted from Yamamura et al. (2000, Fig. 1, p. 936).

viral infection. Following fusion of the virus envelope with the nerve cell surface membrane, the viral nucleocapsid is transported back to the cell body via retrograde transport process (Sodeik et al., 1997; Smith et al., 2001). As detailed in other chapters of this text, attenuated FiSV-based vectors are made by deletion of essential genes to make its replication defective, and deletion of the accessory genes to make room for insertion of therapeutic genes. FiSV-mediated gene transfer can be achieved by direct inoculation into neural parenchyma to achieve local transduction or by SC inoculation to transduce sensory neurons that innervate the inoculation site (Glorioso and Fink, 2004). Fiere, we focus our discussion on its application in the retrograde transport gene transfer. HSV-based vectors have been used extensively to target dorsal root ganglia (DRGs) and trigeminal ganglion to control pain in animal models (Glorioso and Fink, 2004; Liu et al., 2004b; Meunier et al., 2005), but they can also be used to deliver transgenes to the spinal cord motor neurons. Yamamura et al. used a live attenuated FFSV virus expressing j^-galactosidase (j^FFl) to target spinal cord motor neurons following IM virus injection. Injection of 1 X 10^ pfu jSFil virus into the rat gastrocnemius muscle induced jS-galactosidase gene expression in the spinal cord motoneurons of the innervating motor unit. Gene expression was observed in the spinal cord until 14 days after gene transfer. They reported that up to 90% of the anterior horn motor neurons were transduced and gene expression lasted for over 182 days (Yamamura et al., 2000) (Fig. 4). Because Rabies-G glycoprotein mediates axon terminal uptake via the neuromuscular junction (Etessami et al., 2000), the use of the Rabies-G envelope for vector

pseudotyping may be especially useful to access motor neurons in the spinal cord, the target of motor neuron diseases (MNDs). Lentiviral vectors for CNS gene transfer are usually based on the human immunodeficiency virus type 1 (HIV-1) and the majority of these vectors are pseudotyped with the vesicular stomatitis virus G glycoprotein (VSV-G) envelope protein. VSV-G is a surface glycoprotein responsible for viral attachment to host cells. More recently, rabies glycoproteins, as well as glycoproteins from other members of the Rhabdoviridae family have been utilized to pseudotype FilV vectors. Pseudotyping of lentiviral vectors with rabies glycoproteins will be discussed in detail in the vector pseudotyping section. V-

T E T A N U S TOXIN H C AS A PROTEIN CARRIER TARGETING C N S

Tetanus toxin is made of FiC and LC. The LC component of tetanus toxin is a protease. The FiC component binds to its receptor GTlb and mediates the retrograde migration of the holotoxin. As early as 1980, Bizzini et al. (1980) did a very interesting experiment comparing the spinal cord motoneuron targeting via retrograde transport of the LC conjugated to the FiC and the LC alone following IM injection. Their experiment established that the FiC fragment was retrogradely transported from the axonal endings within the muscle to the motoneuronal perikarya in the spinal cord. The enzymatic part of the toxin (LC) alone was not detected in the spinal cord motor neurons. As tetanus toxin FiC has the specific affinity for motor neurons, it is therefore an excellent candidate to

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

260

20. VIRAL VECTOR AXONAL UPTAKE AND RETROGRADE TRANSPORT MECHANISMS AND APPLICATIONS

ferry therapeutic macromolecules into motoneurons. The HC of tetanus toxin has been used to carry ^-galactosidase and GFP as makers for motor neuron targeting (Kissa et al., 2002; Maskos et al., 2002; Miana-Mena et al., 2004). More importantly, the HC component has been utilized to deliver therapeutic agents such as superoxide dismutase 1 (SODl) and cardiotropin1 into motor neurons in an attempt to target MNDs (Figueiredo et al., 1997; Bordet et al., 2001; Francis et al., 1995, 2004). Using a different approach. Knight et al. (1999) developed a HC and DNA complex to deliver therapeutic DNA into motor neurons. The HC component of tetanus toxin has also been used in combination with an adenoviral vector. In this system, the HC of tetanus toxin w^as attached to the surface of an adenoviral vector either through polylysine or an antibody. These experiments show^ed that the HCmodified adenoviral vector could be efficiently transported into the hypoglossal nucleus after IM injection into the tongue of mice (Schneider et al., 2000).

VI.

VECTOR PSEUDOTYPING

Viral pseudotyping is a strategy that can be used to create viral vectors with new tropism and trafficking properties. Pseudotyping of envelope fusion proteins is a natural adaptive mechanism of viral evolution that has been described for certain viruses, such as m^embers of the Baculoviridae, Orthomyxoviridae and Metaviridae families (Pearson and Rohrmann, 2002). Capsid proteins and envelope glycoproteins are implicated in virus attachment and interactions with cellular receptors, determining cell tropism. Manipulation of these viral surface proteins therefore may improve the transduction capacity of these vectors, expanding or restricting their tropism. Furthermore, experiments with vector pseudotyping demonstrated that pseudotyped vectors could achieve higher transduction titers and increase transduction efficacy (Reiser et al., 1996; Mochizuki et al., 1998; Sanders, 2002). The CNS is a network with extensive cell-cell connections. Vectors that target a restricted ceU population might be particularly useful in CNS gene therapy As mentioned briefly above, many commonly used gene transfer vectors have been pseudotyped for CNS gene transfer application. A.

AAV Pseudotyping

Among all AAV serotypes that have been described to date, serotype 2 has been the most widely used in gene transfer studies. AAV2 vectors are also one of the

best characterized vectors in CNS gene transfer. All AAV serotypes share homologies in their genomes. However, various AAV serotypes possess different capsid proteins that in turn are required to attach to different cellular receptors for viral infection. Therefore, different AAV serotypes display distinct patterns of cell tropisms with some degrees of overlap. Cross-packaging of rAAV2 DNA with capsids of other serotypes is one of the strategies that have been tried to change tropism of AAV2-based vectors. Recombinant AAV2 vectors have been shown to predominantly transduce neurons in the brain and spinal cord (McCown et al., 1996; Peel et a l , 1997). Other AAV serotj^es have been reported to have transduction patterns different from that of the rAAV2-based vectors in the CNS. Recombinant AAVl, AAV4, and AAV5, for example, were reported to transduce astrocytes and ependymal cells in the CNS following injection into the lateral ventricle and striatum (Davidson et al, 2000; Tenenbaum et al., 2004). By mixing of different capsid proteins from different serotypes, novel AAV vectors were generated possessing unique tropism and transduction patterns. Burger et al. (2004) described the generation of hybrid vectors between rAAV2 and rAAVl, rAAV2 and rAAV5. These vectors were demonstrated to induce higher transduction efficiency and wider distribution of gene expression in CNS compared to rAAV2-based vectors. These hybrid vectors also showed neuronal-specific transduction preference. Furthermore, these vectors were shown to transduce projection neurons after injection into the hippocampus and striatum. Recombinant and pseudotyped AAV vectors have appealing characteristics that make them promising candidates for CNS gene delivery. However, their broad cell and tissue tropism may impair their binding affinity and limit their utility in CNS gene delivery. Despite the fact that different AAV serotypes undergo retrograde axonal transport to the spinal cord from a peripheral injection site, AAV vectors also have high affinity for skeletal muscle, an issue that must be considered for human application. In small animal models, rAAV2 vectors were shown to undergo retrograde migration and induce significant levels of gene expression in CNS motor neurons following muscle or nerve injection (Boulis et al., 2003a; Kaspar et al., 2003). However, in humans, the larger muscle mass and the considerable length of motor neuron axons are likely to impede neuronal uptake and the retrograde axonal transport of the vectors, confounding efforts to translate these techniques into clinical application. Strategies to enhance delivery of AAV vectors to spinal cord motor neurons and limit their muscle

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

APPLICATION OF RETROGRADE TRANSPORT OF VIRAL VECTORS

affinity are possible. Research showed that specific regions of AAV capsid proteins tolerate genetic manipulations and retain their functionality (Girod et al., 1999; Warrington et a l , 2004). The N terminals of AAV VP proteins have been inserted with single chain antibodies and ligand peptides to modify its tissue specificity (Buning et al., 2003). Strategies are also being developed to modify AAV2 capsid proteins by insertion of peptides that selectively bind to motor neurons for CNS targeting. As mentioned above, the C fragment of tetanus toxin binds to cell surface gangliosides (Gjib) of the nerve terminals in the neuromuscular junction and is actively transported back to the CNS via retrograde transport pathway. Short peptides that mimic C fragment binding properties therefore should possess neuron-specific binding capacity and could be used for CNS targeting. These short peptides can be isolated through phage display technique (Koivunen et al., 1999). Our laboratory has reported the isolation of short peptides with specific affinity for G^ib and showed the enhanced uptake of the peptides by differentiated PC12 neuronal cells (Federici et al., 2005; Liu et a l , 2003, 2004a, 2005). Recombinant AAV2 vectors with the insertion of these peptides into the capsid displayed significantly enhanced neuronal binding (Davis et al., 2005). Similarly, our laboratory and others are attempting to covalently link the whole C fragment to the AAV coat through a variety of strategies. B.

Lentipseudotyping

Lentiviral vectors for CNS gene transfer are usually based on the HIV-1. However, over the last few years, due to safety concerns, vectors derived from other lentiviruses, such as the non-primate equine infectious anemia virus (EIAV) and feline immunodeficiency virus (FIV), have been developed as gene transfer vectors to the CNS (Poeschla, 2003). The majority of these vectors are pseudotyped with the VSV-G glycoprotein envelope protein. More recently, rabies glycoproteins, as well as glycoproteins from other members of the Rhabdoviridae family have been utilized to pseudotype HIV vectors. Mitrophanous et al. (1999) developed an EIAV-based lentiviral vector pseudotyped with VSV-G and Rabies-G proteins for neuronal transduction. These EIAV pseudotypes were able to mediate stable and long-term transduction of neural cells in vitro and in vivo. Desmaris et al. (2001) described the retrograde axonal transport of HIV-1 lentiviral vectors pseudotyped with VSV-G and Mokola envelope glycoproteins. By intranasal instillation of the vectors, they

261

demonstrated that both pseudotypes were efficiently transported to olfactory neuron cell bodies. The retrograde axonal transport of lentiviral vectors pseudotyped with the VSV-G glycoprotein envelope had been previously reported by Blomer et al. (1997) and Alisky et al. (2000), using HIV- and FIV-based vectors, respectively. In our hands, however, delivery of VSV-G-pseudotyped vectors from intraneural and IM injection has been sporadic and limited. In fact, Kaspar et al. (2003) exploited the poor retrograde transport of these vectors to differentiate the effects of peripheral growth factor expression and motor neuron expression after IM vector injection. Because Rabies-G glycoprotein mediates the virus transport through axons via the neuromuscular junction (Etessami et al., 2000), the use of the Rabies-G envelope for vector pseudotyping may be especially useful to access motor neurons in the spinal cord, the target of MNDs. In this context, Mazarakis et al. (2001) have demonstrated that Rabies-G-pseudotyped EIAV-based lentiviral vectors can be retrogradely transported into motor neurons of lumbar spinal cord after peripheral delivery by IM injection, while VSV-G-pseudotyped vectors remain at the site of injection and transduce muscles exclusively. More recently, several novel lentipseudotyped vectors have been designed. HIV-1-based lentiviral vectors pseudotyped with glycoproteins from different members of the Rhabdoviridae family, Lyssavirus genus (Rabies virus, EBL, Lagos bat Lyssavirus and Duvenhage) have demonstrated neuronal tropism in vitro (Teng et al., 2005).

VII- APPLICATION OF RETROGRADE TRANSPORT OF VIRAL VECTORS Viral vectors possessing the retrograde transport capacity have the potential for application to the treatment of peripheral and CNS diseases. Gene transfer mediated by retrograde transport can bypass the blood-brain barrier, reach deep CNS structures that are difficult to target by conventional surgery, or induce targeted gene delivery through the injection of a remote site. As such, therapeutic application of peripheral vector injections include a variety of degenerative and functional disorders of the nervous system including the MNDs, peripheral neuropathy, pain, and spasticity. The treatment of other neurodegenerative disorders including Alzheimer's and Parkinson's may leverage retrograde axonal trafficking within the CNS.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

262 A.

20. VIRAL VECTOR AXONAL UPTAKE AND RETROGRADE TRANSPORT: MECHANISMS AND APPLICATIONS

Motor Neuron Diseases

MNDs, including amyotrophic lateral sclerosis (ALS) and spinal muscular atrophies (SMA), are neurodegenerative disorders. Despite their distinct etiologies and clinical variability, these diseases share a common progressive and fatal outcome, due to the death of upper a n d / o r lower motor neurons (Cifuentes-Diaz et al., 2002; Krivickas, 2003; Strong, 2003). In an attempt to avoid or limit neuronal damage, therapeutic strategies for MNDs are mainly focused on neuroprotection of motor neurons. Gene-based treatment using viral vectors that can be retrogradely transported to transduce motor neurons in CNS from muscles or peripheral nerves may prove to be an alternative treatment option (Alisky and Davidson, 2000). The retrograde transport of adenoviral vectors has been applied to a variety of animal models of motor neuron disease. Gravel et al. (1997) reported the rescue of axotomized motor neurons by intraneural injection of an adenoviral vector expressing brain-derived neurotrophic factor (BDNF) and ciliary neurotrophic factor (CNTF). The expression of BDNF and CNTF was able to maintain the survival of the axotomized motoneurons for at least 5 weeks. Using an IMinjection approach, Acsadi et al. (2002) applied an adenoviral vector containing the glial cell-derived neurotrophic factor (AdGDNF) to an ALS animal model. They reported that injection of AdGDNF into tibialis, gastrocnemius, quadriceps, and paraspinal muscles delayed disease onset and progression, and prolonged animal life span by 12%. Several other viral vector systems have also been designed to deliver genes that control the apoptotic mechanism, which is considered as one of the main causes of MND. Yamashita et al.

(2001, 2003) have successfully demonstrated neuronal protection with this approach, by remote gene delivery of the anti-apoptotic Bcl-2 gene to motor neurons of hypoglossal nuclei, after IM injection of Ad vectors into the tongue of ALS transgenic mice. Kaspar et al. (2003) have demonstrated neuroprotection in a mouse model of ALS by using an AAV vector encoding insulin-like growth factor I (IGF-I). The vector was injected intramuscularly and was transported into projection motor neurons via retrograde axonal transport. Gene expression of IGF-I in the transduced motor neurons demonstrated therapeutic benefits after injection into hind limb and intercostal muscles. The injection of AAV vectors before the symptoms developed not only delayed the disease onset, but also slowed the rate of disease progress. Furthermore, the injection of AAV vectors after the onset of symptoms still bore a significant therapeutic potential. When AAV IGF-I vector was injected after the symptoms developed, gene expression of IGF-I not only extended ALS mouse life expectancy, but also delayed their functional decline related to the disease process. This finding not only suggests that clinical translation is possible, but assuages concerns that impaired axonal transport and axonal loss in motor neuron disease will prevent the application of vector retrograde axonal transport. The authors reported that around 1.1% viral particles were transported to the lumbar spinal cord after IM viral vector injection (Figs. 5 and 6). On the other hand, in AAV-mediated gene transfer of GDNF (AAV-GDNF) into an ALS disease animal model, Wang et al. (2002b) reported that it was the GDNF protein that was retrogradely transported to the spinal cord motoneurons instead of AAV-GDNF vectors. They demonstrated that the retrogradely

I S

LI LL

1

\

J 110 120 i at omet (days)

VL^

r

Qisejo GFP GDNF IGF-1

1 —I-' 1 1 130 149 ISO IK} Agis at d«itli (da^)

Qtsal 91 107 t22

"1

Qeatbl 123 134 1 ^

t

FIGURE 5 Targeting the spinal cord by retrograde transport to MNs. Left: age of onset of disease symptoms in G93A SODl animals injected at 60 days of age with AAV-GFP (green), AAV-GDNF (blue), or AAV-IGF-1 (red). Right: survival analysis in animals injected at 60 days of age with AAV NTFs. Disease onset and mortality were significantly delayed in G93A SODl mice treated with AAV-GDNF (blue) and AAV-IGF-1 (red), compared to AAV-GFP-treated animals (green). Adapted from Kaspar et al. (2003, Fig. 1, p. 840)

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

APPLICATION OF RETROGRADE TRANSPORT OF VIRAL VECTORS

263

reported improvements in survival of motor neurons in the brainstem and spinal cord of ALS transgenic mice, leading to a considerable delay in the onset of ALS symptoms (Fig. 7). Despite the feasibility of all these approaches using remote delivery in animal models, several practical issues must be considered before translating to human clinical application. First, axonal transport may be impaired in MND, hindering an efficient delivery (Jablonka and Sendtner, 2003; Murakami et al., 2001). Second, humans have larger muscle mass and longer motor neurons, needing vectors in large scale to achieve efficient uptake, even by IM or intraneural injections. And finally, some vectors have high binding affinity for muscle, which can reduce their neuronal uptake and retrograde axonal transport.

B. 110 dayt

Neuropathy

End-stago

FIGURE 6 AAV-IGF-1 protects MNs and delays astroglial response in G93A SOD 1 mice. Top: Histological evaluation of 110day-old lumbar spinal cord in (left) AAV-GFP- and (right) AAV-IGF1-treated animals. Bottom: Quantification of surviving MNs in wild type (WT), AAV-GFP-, and AAV-IGF-1-injected animals at 110 days of age and end-stage classification of the disease. Adapted from Kaspar et al. (2003, Fig. 3, p. 841)

transported GDNF proteins delayed the disease onset and progression of the motor dysfunction, and prolonged the life span of the treated animals by 17 days. Using a similar IM injection approach, Azzouz et al. (2004b) applied a vascular endothelial growth factor (VEGF)-expression RabG pseudotyped EIAV vector to ALS mice. They reported that a single IM injection of VEGF-expressing EIAV virus delayed disease onset and slowed disease progression, even when the treatment was initiated after the symptoms developed. Gene replacement and gene silencing using small interfering RNA (siRNA) represent alternative strategies for the treatment of MNDs through retrograde axonal transport delivery of viral vectors to spinal cord motor neurons. The survival motor neuron (SMN) gene, which is mutated in SMA, was delivered into the motor neurons through multiple IM injections. It reduced motor neuron death, and increased the life expectancy by 20-38% (Azzouz et al., 2004a). By using siRNA targeting the mutated SODl gene, Ralph et al. (2005) significantly decreased the SODl expression by multiple IM injections of lentiviral vectors. They

Peripheral neuropathy is caused by degeneration of sensory axons resulting in sensory loss. The most common cause of peripheral neuropathy is diabetes mellitus (DM). Similar symptoms and pathophysiology are seen in complications of chemotherapy or toxin exposure. Treatment of neuro-pathy is largely palliative. Neurotropic factors such as neurotrophin-3 (NT-3) and NGF are effective in animal models of neuropathy. However, their clinical application is limited due to their short half-life after administration. Retrograde HSV delivery of these neurotropic factors has been proven to be effective in the delay of the disease progression in animal models of neuropathy. Gene expression of NT-3 (Fig. 8) and NGF (Fig. 9) in DRGs delivered by SC HSV inoculation has been demonstrated to be protective in neuropathy induced by pyridoxine (PDX) or cisplatin (Chattopadhyay et al., 2002, 2004).

C. Pain Pain, including both malignant and chronic pain syndromes, is another disorder that can be targeted by retrograde delivery of viral vectors. Clinically, pain can be effectively controlled by analgesic drugs through endogenous jti, 8, or K opioid peptide receptors. These pharmacological agents are unfortunately not without their adverse effects. HSV-mediated preproenkephalin (PPE), as well as glial cell-derived growth factor (GDNF) and glutamic acid decarboxylase (GAD) have been utilized to alter the excitatory/inhibitory pathways to control pain development. By SC inoculation

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

264

20. VIRAL VECTOR AXONAL UPTAKE AND RETROGRADE TRANSPORT MECHANISMS AND APPLICATIONS

*

Wild-type

EIAV-Emp

EIAV^OOIHPI

FIGURE 7 Motor neuron survival is improved by RNAi-mediated silencing of SODl. Lumbar spinal cord sections (top) from EIAV-injected mice at the end stage of disease (low magnification, 10 X; high magnification, 40X). Survival of motor neurons was higher in EIAV-SODlHPl-injected mice {top, right) compared with control-injected animals {top, left). {Bottom, left): Quantification of surviving spinal motor neurons at the end stage of disease. {Bottom, right): Long-term expression of j?-galactosidase in spinal motor neurons of EIAV-SODlHpl-treated mice at the end stage of disease. EIAV-SODIHPI transduction (blue staining) protects against neuronal degeneration (arrow), arrowhead indicates degenerating untransduced cell. *P<0.005, n = 3. Adapted from Ralph et al. (2005, Fig. 4, p. 4).

of an FiSV vector expressing PPE, researchers demonstrated that expression of PPE significantly reduced pain-related behavior during the delayed phase of formalin test in an inflammatory pain animal model (Wilson et al., 1999; Goss et al., 2001). Goss et al. (2002) also reported that this HSV-based vector was effective in the treatment of pain related to cancer metastasis, which bears features of both neuropathic pain and inflammatory pain. Furthermore, GDNF and GAD (Fig. 10) delivered by HSV-based vectors, have also been shown to be effective in the management of neuropathic pain after SC injection of the viruses in a spinal nerve ligation pain model (Flao et a l , 2003; Liu et a l , 2004b). More recently, PPE delivered by FiSV inoculation in the vibrissal pad evoked a potent anti-allodynic effect. This effect was mediated by high level of PPE-derived met-enkephalin in the trigeminal ganglion (Meunier et al., 2005). The authors injected 5 X 10^ pfu HSV PPE into the rat vibrissal pad and found that a ninefold

increase of proenkephalin A mRNA was detected in the trigeminal ganglion on the gene transfer side. Bilateral mechanical hypersensitiveness was significantly attenuated in animals overproducing proenkephalin A in the trigeminal ganglion. D.

Alzheimer's Disease and Parkinson's Disease

Besides the common characteristics of rAAV2 and CAV-2 vectors to transduce mostly neurons and induce low immunogenicity, it has been demonstrated that both vectors, when injected into the dentate gyrus (DG) of the hippocampus and striatum, spread to projection neurons into the layer II of the entorhinal cortex and the substantia nigra pars compacta, respectively (Kaspar et al., 2002; Peltekian et al., 2002) (Fig. 11). The loss of layer II entorhinal cortex neurons is associated with Alzheimer's disease and dopaminergic neurons of the substantia nigra pars compacta are affected in Parkinson's disease. The ability of these vectors to

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

APPLICATION OF RETROGRADE TRANSPORT OF VIRAL VECTORS

265

i r y^

FIGURE 8 HSV-mediated expression of NT-3 in vivo prevents axonal degeneration caused by PDX intoxication. Cross-section of sciatic nerve, 7 days after completion of PDX intoxication. Both the PDX-intoxicated (PDX) and PDX-intoxicated HSV-lacZtreated (lacZ) nerves show substantial loss of fibers and the presence of degenerating fibers. The PDX-intoxicated HSV-NT3-treated (NT-3) animals have substantial preservation of nerve fiber morphology, in agreement with the preservation of electrophysiological parameters and behavioral measures of nerve function. From Chattopadhyay et al. (2002, Fig. 4, p. 24), with permission.

1

30 20

L

g 10 \

8 5 O

i ^

tl n ^ J' ^

FIGURE 9 Release of NGF from transduced primary DRG neurons. Cells were infected with HSV vectors at an m.o.i. of 1 for 1 h, and, 72 h later, the supernatant was assayed by ELISA. QL2HNT3 contains NT-3 under the control of a fusion promoter consisting of LAP2 and the HCMV lEp. SHN contains NGF under the control of the human cytomegalovirus immediate-early promoter (HCMV lEp). And SLN contains NGF under the control of the HSV latency active promoter element (LAP2). From Chattopadhyay et al. (2004, Fig. 2, p. 932), with permission.

FIGURE 10 L5 DRG transduced with HSV-GAD by footpad inoculation shows increased GAD67-like immunoreactivity in a broad spectrum of DRG neurons of all sizes compared to the contralateral (control, vehicle-injected) DRG Adapted from Liu et al. (2004, Fig. 2, p. 59).

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

266

20. VIRAL VECTOR AXONAL UPTAKE AND RETROGRADE TRANSPORT: MECHANISMS AND APPLICATIONS

site of injection

(hippocampus)

faleral entortiinaf cortex

FIGURE 11. Retrograde transduction of the entorhinal cortex resulting from a single injection of 108 particles of CAV2(CMV-GFP) into the DC of the hippocampus. (Top left and right): Schematic representation of coronal sections, corresponding to the injection site into the hippocampus {left) and the afferent entorhinal cortex {right). (A) Pattern of GPP expression through the hippocampus 14 days after injection of 10^ CAV2(CMV-GFP) vector particles into the DG. GPP expression was scattered through the entire structure with sonne transduced cells, mainly interneurons, distant from the injection site. The axonal network is intensely labeled at the level of both DG and Ammons' horn (CAl, CA2, CAS). Scale bar, 100 |Lim. (B) Pattern of GPP expression through the entorhinal cortex after a single injection of 10^ CAV2(CMV-GPP) vector particles into the DG. Conversely to the injection site, transgene expression is precisely confined to layer II neurons, known to massively project to neurons of the dentate gyrus, and encompasses the rostro-caudal extend of the entorhinal cortex along almost 0.3 mm. Scale bar, 25 ^m. Adapted from Peltekian et al. (2002, Fig. 4, p. 28).

selectively transduce such structures reinforces the possibility for developing therapeutic gene delivery strategies via retrograde axonal transport. Kaspar et al. (2002) were successful in conferring neuroprotection to entorhinodentate projection neurons, using the AAV2 vector system to deliver an anti-apoptotic gene into the hippocampus of mice. A recombinant AAV vector harboring an anti-apoptotic gene, Bcl-xL cDNA, was injected into the hippocampus. Entorhinal layer II neurons showed gene expression within 2 weeks following gene transfer. Expression of Bcl-xL gene protected entorhinal layer II neurons from injuryinduced neuronal death. The authors also demon-

strated that the AAV particles, but not the transgene proteins, were retrogradely transported, hence implicating retrograde vector delivery in the mechanism of protection. Interestingly, in a Parkinson's disease animal model, Wang et al. (2002a) reported that the GDNF protein from an AAV vector encoding GDNF was retrogradely transported to the projection neurons and elicited a therapeutic effect. E,

Spasticity

Spasticity is a motor dysfunction characterized by increased muscular resistance to movement with

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

CONCLUSION

increased velocity of movement. Severe spasticity is often accompanied by spasms, which are involuntary muscle contractions that are often severely painful. Spasticity commonly occurs in the presence of upper motor neuron lesions or disconnection of the synaptic connection between the brain and spinal motor neurons. Spasticity is a disabling symptom of spinal cord injury, multiple sclerosis, and cerebral palsy. Spasms and spasticity are thought to result from a dysfunction of inhibitory signals within the spinal cord that depend on descending motor pathways. Treatments including anti-spasticity drugs, botulinum toxin, electrical stimulation, and surgery are being currently used to reduce spasticity (Das and Park, 1989; Simpson et al., 1996). In cases where patients do not respond to pharmacological and surgical interventions, gene therapy might be warranted, using a vector capable of neuronal tropism and targeted neuromodulation, in order to achieve focused synaptic inhibition. Combining the retrograde transport property of the viral vectors mentioned above with proteins that modulate neurotransmitter release (Johns et al., 1999; Teng et al., 2005; Yang et al., 2005), injection of these viral vectors into the spastic muscles would specifically disrupt acetycholine release at the neuromuscular junction, relieving the symptoms.

VIIL

CONCLUSION

Axonal transport is a physiological process involving material transfer between cell body and the axonal terminal of neurons. Depending upon the direction of the material flow, axonal transport is divided into anterograde and retrograde axonal transport. Retrograde axonal transport is not only used to transfer physiological materials back to the cell body from the periphery, but also is an important pathway for some neurotropic viruses to invade the CNS. Taking advantage of the retrograde transport property of the neurotropic viruses, genetically modified neurotropic and non-neurotropic viral vectors could be used to target a specific projection neuronal population from a peripheral site. These molecular trojan horses possess the ability to ferry therapeutic DNA molecules into a specific cell population in a non-invasive fashion. Gene delivery via retrograde transport pathway could serve as an important therapeutic option for the treatment of pain, neuropathy, spasticity, and neurodegenerative diseases including ALS, Alzheimer's disease and Parkinson's disease.

267

References Acsadi, G., Anguelov, R.A., Yang, H., Toth, G., Thomas, R., Jani, A., Wang, Y, lanakova, E., Mohammad, S., Lewis, R.A. and Shy, M.E. (2002) Increased survival and function of SODl mice after glial cell-derived neurotrophic factor gene therapy. Hum. Gene Ther., 13:1047-1059. AHsky, J.M. and Davidson, B.L. (2000) Gene therapy for amyotrophic lateral sclerosis and other motor neuron diseases. Hum. Gene Ther., 11: 2315-2329. Alisky, J.M., Hughes, S.M., Sauter, S.L., Jolly, D., Dubensky Jr., T.W., Staber, RD., Chiorini, J.A. and Davidson, B.L. (2000) Transduction of murine cerebellar neurons with recombinant FIV and AAV5 vectors. Neuroreport, 11: 2669-2673. Azzouz, M., Le, T., Ralph, G.S., Walmsley, L., Monani, U.R., Lee, D.C., Wilkes, R, Mitrophanous, K.A., Kingsman, S.M., Burghes, A.H. and Mazarakis, N.D. (2004a) Lentivector-mediated SMN replacement in a mouse model of spinal muscular atrophy. J. Clin. Invest., 114:1726-1731. Azzouz, M., Ralph, G.S., Storkebaum, E., Walmsley, L.E., Mitrophanous, K.A., Kingsman, S.M., Carmeliet, P. and Mazarakis, N.D. (2004b) VEGF delivery with retrogradely transported lentivector prolongs survival in a mouse ALS model. Nature, 429: 413-417. Bearer, E.L., Breakefield, X.O., Schuback, D., Reese, T.S. and LaVail, J.H. (2000) Retrograde axonal transport of herpes simplex virus: evidence for a single mechanism and a role for tegument. Froc. Natl. Acad. Sci. USA, 97: 8146-8150. Bisby, M.A. (1982) Functions of retrograde axonal transport. Fed. Froc, 41: 2307-2311. Bizzini, B., Grob, R, Glicksman, M. and Akert, K. (1980) Use of the B-IIb tetanus toxin derived fragment as a specific neuropharmacological transport agent. Brain Res., 193: 221-227. Blomer, U., Naldini, L., Kafri, T., Trono, D., Verma, I.M. and Gage, FH. (1997) Highly efficient and sustained gene transfer in adult neurons with a lentivirus vector. J. Virol., 71: 6641-6649. Borchelt, D.R., Koliatsos, V.E., Guamieri, M., Fardo, C.A., Sisodia, S.S. and Price, D.L. (1994) Rapid anterograde axonal transport of the cellular prion glycoprotein in the peripheral and central nervous systems. J. Biol. Chem., 269:14711-14714. Bordet, T., Castelnau-Ftakhine, L., Fauchereau, R, Friocourt, G., Kahn, A. and Haase, G. (2001) Neuronal targeting of cardiotrophin-1 by coupling with tetanus toxin C fragment. Mol. Cell Neurosci., 17: 842-854. Boulis, N.M., Noordmans, A.J., Song, D.K., Imperiale, M.J., Rubin, A., Leone, P., During, M. and Feldman, E.L. (2003a) Adeno-associated viral vector gene expression in the adult rat spinal cord following remote vector delivery. Neurobiol. Dis., 14: 535-541. Boulis, N.M., Turner, D.E., Dice, J.A., Bhatia, V. and Feldman, E.L. (1999) Characterization of adenoviral gene expression in spinal cord after remote vector delivery. Neurosurgery, 45: 131-137; discussion 137,138. Boulis, N.M., Turner, D.E., Imperiale, M.J. and Feldman, E.L. (2002) Neuronal survival following remote adenovirus gene delivery. J. Neurosurg. Spine, 96: 212-219. Boulis, N.M., Willmarth, N.E., Song, D.K., Feldman, E.L. and Imperiale, M.J. (2003b) Intraneural colchicine inhibition of adenoviral and adeno-associated viral vector remote spinal cord gene delivery Neurosurgery, 52: 381-387; discussion 387. Buning, H., Ried, M.U., Perabo, L., Gerner, KM., Huttner, N.A., Enssle, J. and Hallek, M. (2003) Receptor targeting of adeno-associated virus vectors. Gene Ther., 10:1142-1151.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

268

20. VIRAL VECTOR AXONAL UPTAKE AND RETROGRADE TRANSPORT MECHANISMS AND APPLICATIONS

Burger, C , Gorbatyuk, O.S., Velardo, MJ., Peden, C.S., Williams, P., Zolotukhin, S., Reier, PJ., Mandel, RJ. and Muzyczka, N. (2004) Recombinant AAV viral vectors pseudotyped with viral capsids from serotypes 1, 2, and 5 display differential efficiency and cell tropism after delivery to different regions of the central nervous system. Mol. Ther., 10: 302-317. Campenot, R.B. and Maclnnis, B.L. (2004) Retrograde transport of neurotrophins: fact and function. J. NeurobioL, 58: 217-229. Chattopadhyay M., Goss, J., Wolfe, D., Coins, W , Huang, S., Glorioso, J., Mata, M. and Fink, D. (2004) Protective effect of herpes simplex virus-mediated neurotrophin gene transfer in cisplatin neuropathy Brain, 127: 929-939. Chattopadhyay M., Wolfe, D., Huang, S., Goss, J., Glorioso, J.C., Mata, M. and Fink, D.J. (2002) In vivo gene therapy for pyridoxine-induced neuropathy by herpes simplex virus-mediated gene transfer of neurotrophin-3. Ann. Neurol., 51:19-27. Cifuentes-Diaz, C., Frugier, T. and Melki, J. (2002) Spinal muscular atrophy. Semin. Pediatr. Neurol., 9:145-150. Cook, M.L. and Stevens, J.G. (1973) Pathogenesis of herpetic neuritis and ganglionitis in mice: evidence for intra-axonal transport of infection. Infect. Immun., 7: 272-288. Das, T. and Park, D. (1989) Botulinum toxin in treating spasticity. Br. J. Clin. Pract., 43: 401-403. Davidson, B.L., Stein, C.S., Heth, J.A., Martins, I., Kotin, R.M., Derksen, T.A., Zabner, J., Ghodsi, A. and Chiorini, J.A. (2000) Recombinant adeno-associated virus type 2, 4, and 5 vectors: transduction of variant cell types and regions in the mammalian central nervous system. Proc. Natl. Acad. Sci. USA, 97: 3428-3432. Davis, A., Stachler, M., Shi, W., Boulis, N. and Bartlett, J. (2005). Manipulating AAV2 tropism for enhanced delivery of AAV vectors to the spinal cord. Mol. Ther., 11: S46. Desmaris, N., Bosch, A., Salaun, C , Petit, C , Prevost, M.C., Tordo, N., Perrin, P., Schwartz, O., de Rocquigny, H. and Heard, J.M. (2001) Production and neurotropism of lentivirus vectors pseudotyped with lyssavirus envelope glycoproteins. Mol. Ther., 4: 149-156. Dolivo, M., Beretta, E., Bonifas, V. and Foroglou, C. (1978) Ultrastructure and function in sympathetic ganglia isolated from rats infected with pseudorabies virus. Brain Res., 140: 111-123. Dolivo, M., Honegger, P., George, C , Kiraly, M. and Bommeli, W. (1979) Enzymatic activity, ultrastructure and function in ganglia infected with a neurotropic virus. Prog. Brain Res., 51: 51-57. Etessami, R., Conzelmann, K.K., Fadai-Ghotbi, B., Natelson, B., Tsiang, H. and Ceccaldi, PE. (2000) Spread and pathogenic characteristics of a G-deficient rabies virus recombinant: an in vitro and in vivo study J. Gen. Virol., 81: 2147-2153. Federici, T., Liu, J., Teng, Q., Garrity-Moses, M., Yang, J. and Boulis, N. (2005). In vivo spinal cord uptake and neuronal binding properties of Tetl: a potential means for enhanced spinal cord AAV delivery Mol. Ther., 11: S332. Figueiredo, D.M., Hallewell, R.A., Chen, L.L., Fairweather, N.F., Dougan, G., Savitt, J.M., Parks, D.A. and Fishman, PS. (1997) Delivery of recombinant tetanus-superoxide dismutase proteins to central nervous system neurons by retrograde axonal transport. Exp. Neurol., 145: 546-554. Finiels, R, Ribotta, M.G.Y., Barkats, M., Samolyk, M.-L., Robert, J.-J., Privat, A., Revah, F. and Mallet, J. (1995) Specific and efficient gene transfer strategy offers new potentialities for the treatment of motor neuron diseases. Neuroreport, 7: 373-378.

Francis, J.W., Bastia, E., Matthews, C.C, Parks, D.A., Schwarzschild, M. A., Brown Jr., R.H. and Fishman, PS. (2004) Tetanus toxin fragment C as a vector to enhance delivery of proteins to the CNS. Brain Res., 1011: 7-13. Francis, J.W., Hosier, B.A., Brown Jr., R.H., and Fishman, PS. (1995) CuZn superoxide dismutase (SOD-1): tetanus toxin fragment C hybrid protein for targeted delivery of SOD-1 to neuronal cells. J. Biol. Chem., 270:15434-15442. Geraghty, R.J., Krummenacher, C , Cohen, G.H., Eisenberg, R.J. and Spear, RG. (1998) Entry of alphaherpesviruses mediated by poliovirus receptor-related protein 1 and poliovirus receptor. Science, 280:1618-1620. Ghadge, G., Roos, R., Kang, U., Wollmann, R., Fishman, P., Kalynych, A., Barr, E. and Leiden, J. (1995) CNS gene delivery by retrograde transport of recombinant replication-defective adenoviruses. Gene Ther., 2:132-137. Girod, A., Ried, M., Wobus, C , Lahm, H., Leike, K., Kleinschmidt, J., Deleage, G. and Hallek, M. (1999) Genetic capsid modifications allow efficient re-targeting of adeno-associated virus type 2. Nat. Med., 5: 1052-1057. Glass, J.D. and Griffin, J.W (1994) Retrograde transport of radiolabeled cytoskeletal proteins in transected nerves. J. Neurosci., 14: 3915-3921. Glorioso, J.C. and Fink, D.J. (2004) Herpes vector-mediated gene transfer in treatment of diseases of the nervous system. Annu. Rev Microbiol., 58: 253-271. Goss, J.R., Harley, C.F., Mata, M., O'Malley, M.E., Coins, WF., Hu, X., Glorioso, J.C. and Fink, D.J. (2002) Herpes vector-mediated expression of proenkephaUn reduces bone cancer pain. Ann. Neurol., 52: 662-665. Goss, J.R., Mata, M., Coins, W.F., Wu, H.H., Glorioso, J.C. and Fink, D.J. (2001) Antinociceptive effect of a genomic herpes simplex virus-based vector expressing human proenkephalin in rat dorsal root ganglion. Gene Ther., 8: 551-556. Gosztonyi, G. and Ludwig, H. (2001) Interactions of viral proteins with neurotransmitter receptors may protect or destroy neurons. Curr. Top Microbiol. Immunol., 253:121-144. Gravel, C , Gotz, R., Lorrain, A. and Sendtner, M. (1997) Adenoviral gene transfer of ciliary neurotrophic factor and brain-derived neurotrophic factor leads to long-term survival of axotomized motor neurons. Nat. Med., 3: 765-770. Hao, S., Mata, M., Wolfe, D., Huang, S., Glorioso, J.C. and Fink, D.J. (2003) HSV-mediated gene transfer of the glial cell-derived neurotrophic factor provides an antiallodynic effect on neuropathic pain. Mol. Ther., 8: 367-375. Hirokawa, N. (1993) Axonal transport and the cytoskeleton. Curr. Opin. NeurobioL, 3: 724-731. Hirokawa, N. and Takemura, R. (2004) Molecular motors in neuronal development, intracellular transport and diseases. Curr. Opin. NeurobioL, 14: 564-573. Jablonka, S. and Sendtner, M. (2003) Molecular and cellular basis of spinal muscular atrophy Amyotroph. Lateral Scler. Motor Neuron Disord., 4: 144-149. Jacob, Y, Badrane, H., Ceccaldi, PE. and Tordo, N. (2000) Cytoplasmic dynein LC8 interacts with lyssavirus phosphoprotein. J. Virol, 74: 10217-10222. Johns, D.C., Marx, R., Mains, R.E., O'Rourke, B. and Marban, E. (1999) Inducible genetic suppression of neuronal excitability. J. Neurosci., 19: 1691-1697. Kaspar, B.K., Erickson, D., Schaffer, D., Hinh, L., Gage, RH. and Peterson, D.A. (2002) Targeted retrograde gene delivery for neuronal protection. Mol. Then, 5: 50-56.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

CONCLUSION Kaspar, B.K., Llado, J., Sherkat, N., Rothstein, J.D. and Gage, F.H. (2003) Retrograde viral delivery of IGF-1 prolongs survival in a mouse ALS model. Science, 301: 839-842. Kaspar, B.K., Vich, V., Christian, L., Rothsten, J.D. and Gage, RH. (2004) AAV retrograde transport potential and therapeutic approaches for ALS. Mol. Ther., 9: S18. Kissa, K., Mordelet, E., Soudais, C , Kremer, E.J., Demeneix, B.A., Brulet, R and Coen, L. (2002) In vivo neuronal tracing with GFPTTC gene delivery. Mol. Cell Neurosci., 20: 627-637. Knight, A., Carvajal, J., Schneider, H., Coutelle, C , Chamberlain, S. and Fairweather, N. (1999) Non-viral neuronal gene delivery mediated by the HC fragment of tetanus toxin. Eur. J. Biochem., 259: 762-769. Koivunen, E., Arap, W., Rajotte, D., Lahdenranta, J. and Pasqualini, R. (1999) Identification of receptor ligands with phage display peptide libraries. J. Nucl. Med., 40: 883-888. Kremer, E.J., Boutin, S., Chillon, M. and Danos, O. (2000) Canine adenovirus vectors: an alternative for adenovirus-mediated gene transfer. J. Virol., 74: 505-512. Krivickas, L.S. (2003) Amyotrophic lateral sclerosis and other motor neuron diseases. Phys. Med. Rehab. Clin. N. Am., 14: 327-345. Kwong, A.D. and Frer\kel, N. (1989) The herpes simplex virus virion host shutoff function. J. Virol, 63: 4834-4839. Lalli, G., Gschmeissner, S. and Schiavo, G. (2003) Myosin Va and microtubule-based motors are required for fast axonal retrograde transport of tetanus toxin in motor neurons. J. Cell Sci., 116: 4639^650. Langevin, C , Jaaro, H., Bressanelli, S., Fainzilber, M. and Tuffereau, C. (2002) Rabies virus glycoprotein (RVG) is a trimeric ligand for the N-terminal cysteine-rich domain of the mammalian p75 neurotrophin receptor. J. Biol. Chem., 277: 37655-37662. Lentz, T.L., Burrage, T.G., Smith, A.L., Crick, J. and Tignor, G.H. (1982) Is the acetylcholine receptor a rabies virus receptor? Science, 215:182-184. Li, Y., Wang, X., Guo, H. and Wang, S. (2004) Axonal transport of recombinant baculovirus vectors. Mol. Ther., 10: 1121-1129. Liu, J., Teng, Q., Garrity-Moses, M., Federici, T, Tanase, D., Imperiale, M. and Boulis, N. (2005) A novel peptide defined through phage display for therapeutic protein and vector neuronal targeting. Neurobiol. Dis., 19: 407-418 Liu, J., Teng, Q., Tanase, D., Garity-Moses, M. and Boulis, N. (2004a) Neuronal membrane specific binding peptides isolated via phage display biopanning against trisialogangliosides. Mol. Ther., 9: S167. Liu, J., Teng, Q., Tanase, D., Garrity-Moses, M., Baker, K. and Boulis, N. (2003) Short peptides with clostridial ganglioside binding properties defined through phage display. Soc. Neurosci. Abstr., 29: 495,496. Liu, J., Wolfe, D., Hao, S., Huang, S., Glorioso, J.C, Mata, M. and Fink, D.J. (2004b) Peripherally delivered glutamic acid decarboxylase gene therapy for spinal cord injury pain. Mol. Ther., 10: 57-66. Martin, X. and Dolivo, M. (1983) Neuronal and transneuronal tracing in the trigeminal system of the rat using the herpes virus suis. Brain Res., 273: 253-276. Martinat, C , Jarousse, N., Prevost, M.C. and Brahic, M. (1999) The GDVII strain of Theiler's virus spreads via axonal transport. J. Virol., 73: 6093-6098. Martinov, VN., Sefland, L, Walaas, S.I., Lomo, T, Nja, A. and Hoover, F. (2002) Targeting functional subtypes of spinal motoneurons and skeletal muscle fibers in vivo by intramuscular

269

injection of adenoviral and adeno-associated viral vectors. Anat. Embryol. (Berl.), 205: 215-221. Maskos, U., Kissa, K., Cloment, C.S. and Brulet, P (2002) Retrograde trans-synaptic transfer of green fluorescent protein allows the genetic mapping of neuronal circuits in transgenic mice. Proc. Natl. Acad. Sci. USA, 99:10129-10125. Mata, M., Zhang, M., Hu, X. and Fink, D.J. (2001) HveC (nectin-1) is expressed at high levels in sensory neurons, but not in motor neurons, of the rat peripheral nervous system. J. Neurovirol., 7: 476-480. Mazarakis, N.D., Azzouz, M., Rohll, J.B., EUard, EM., Wilkes, EJ., Olsen, A.L., Carter, E.E., Barber, R.D., Baban, D.E, Kingsman, S.M., Kingsman, A.J., O'Malley, K. and Mitrophanous, K.A. (2001) Rabies virus glycoprotein pseudotyping of lentiviral vectors enables retrograde axonal transport and access to the nervous system after peripheral delivery Hum. Mol. Genet., 10: 2109-2121. McCown, TJ., Xiao, X., Li, J., Breese, G.R. and Samulski, R.J. (1996) Differential and persistent expression patterns of CNS gene transfer by an adeno-associated virus (AAV) vector. Brain Res., 713: 99-107. McGeoch, D.J., Dalrymple, M.A., Davison, A.J., Dolan, A., Frame, M.C, McNab, D., Perry, L.J., Scott, J.E. and Taylor, P (1988) The complete DNA sequence of the long unique region in the genome of herpes simplex virus type 1. J. Gen. Virol., 69(Pt 7): 1531-1574. Meunier, A., Latremoliere, A., Mauborgne, A., Bourgoin, S., Kayser, V, Cesselin, R, Hamon, M. and Pohl, M. (2005) Attenuation of pain-related behavior in a rat model of trigeminal neuropathic pain by viral-driven enkephalin overproduction in trigeminal ganglion neurons. Mol. Ther., 11: 608-616. Miana-Mena, F.J., Munoz, M.J., Roux, S., Ciriza, J., Zaragoza, P., Brulet, P. and Osta, R. (2004) Anon-viral vector for targeting gene therapy to motoneurons in the CNS. Neurodegen.Dis., 1:101-108. Millecamps, S., Mallet, J. and Barkats, M. (2002) Adenoviral retrograde gene transfer in motoneurons is greatly enhanced by prior intramuscular inoculation with botulinum toxin. Hum. Gene Ther., 13: 225-232. Miranda-Saksena, M., Armati, P., Boadle, R.A., Holland, D.J. and Cuniungham, A.L. (2000) Anterograde transport of herpes simplex virus type 1 in cultured, dissociated human and rat dorsal root ganglion neurons. J. Virol., 74:1827-1839. Mitrophanous, K., Yoon, S., Rohll, J., Patil, D., Wilkes, R, Kim, V, Kingsman, S., Kingsman, A. and Mazarakis, N. (1999) Stable gene transfer to the nervous system using a non-primate lentiviral vector. Gene Ther., 6: 1808-1818. Mochizuki, H., Schwartz, J.P, Tanaka, K., Brady, R.O. and Reiser, J. (1998) High-titer human immunodeficiency virus type 1-based vector systems for gene delivery into nondividing cells. J. Virol., 72: 8873-8883. Moya, K.L., Hassig, R., Creminon, C , Laffont, I. and Di Giamberardino, L. (2004) Enhanced detection and retrograde axonal transport of PrPc in peripheral nerve. J. Neurochem., 88:155-160. Murakami, T, Nagano, I., Hayashi, T, Manabe, Y, Shoji, M., Setoguchi, Y and Abe, K. (2001) Impaired retrograde axonal transport of adenovirus-mediated E. coli LacZ gene in the mice carrying mutant SODl gene. Neurosci. Lett., 308:149-152. Ochs, S. (1972) Fast transport of materials in mammalian nerve fibers. Science, 176: 252-260. Ohka, S., Yang, WX., Terada, E., Iwasaki, K. and Nomoto, A. (1998) Retrograde transport of intact poliovirus through the axon via the fast transport system. Virology, 250: 67-75. Oztas, E. (2003) Neuronal tracing. Neuroanatomy, 2: 2-5.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

270

20. VIRAL VECTOR AXONAL UPTAKE AND RETROGRADE TRANSPORT MECHANISMS AND APPLICATIONS

Pasteur, L., Roux, E. and Chamberland, C. (1884) Nouvelle communication sur la rage. C. R. Acad. Sc., 98: 457-463. Pearson, M.N. and Rohrmann, G.F. (2002) Transfer, incorporation, and substitution of envelope fusion proteins among members of the baculoviridae, orthomyxoviridae, and metaviridae (insect retrovirus) families. J. Virol., 76: 5301-5304. Peel, A.L., Zolotukhin, S., Schrimsher, G.W., Muzyczka, N. and Reier, P.J. (1997) Efficient transduction of green fluorescent protein in spinal cord neurons using adeno-associated virus vectors containing cell type-specific promoters. Gene Ther., 4: 16-24. Peltekian, E., Garcia, L. and Danos, O. (2002) Neurotropism and retrograde axonal transport of a canine adenoviral vector: a tool for targeting key structures undergoing neurodegenerative processes. Mol. Ther., 5: 25-32. Poeschla, E.M. (2003) Non-primate lentiviral vectors. Curr. Opin. Mol. Ther., 5: 529-540. Poisson, N., Real, E., Gaudin, Y, Vaney M.C., King, S., Jacob, Y, Tordo, N. and Blondel, D. (2001) Molecular basis for the interaction between rabies virus phosphoprotein P and the dynein light chain LC8: dissociation of dynein-binding properties and transcriptional functionality of R J. Gen. Virol., 82: 2691-2696. Ralph, G.S., Mazarakis, N.D. and Azzouz, M. (2005) Therapeutic gene silencing in neurological disorders, using interfering RNA. J. Mol. Med., 83: 413-419. Raux, H., Flamand, A. and Blondel, D. (2000) Interaction of the rabies virus P protein with the LC8 dynein light chain. J. Virol., 74: 10212-10216. Reiser, J., Harmison, G., Kluepfel-Stahl, S., Brady, R.O., Karlsson, S. and Schubert, M. (1996) Transduction of nondividing cells using pseudotyped defective high-titer HIV type 1 particles. Proc. Natl. Acad. Sci. USA, 93:15266-15271. Reynolds, A.J., Bartlett, S.E. and Hendry, LA. (2000) Molecular mechanisms regulating the retrograde axonal transport of neurotrophins. Brain Res. Rev., 33:169-178. Ridoux, v., Robert, J.J., Zhang, X., Perricaudet, M., Mallet, J. and Le Gal La Salle, G. (1994) Adenoviral vectors as functional retrograde neuronal tracers. Brain Res., 648:171-175. Roizman, B. and Sears, A.E. (1993) Herpes simplex viruses and their replication. In: Roizman B., Whitney R.J. and Lopez C. (Eds.), the human herpeviruses. Raven Press, New York, pp. 11-68. Sanders, D.A. (2002) No false start for novel pseudotyped vectors. Curr. Opin. BiotechnoL, 13: 437-442. Schiavo, G. and Chao, M.V (2004) Motors, adaptors, and receptors: key elements of neuronal transport. J. NeurobioL, 58: 161-163. Schneider, H., Groves, M., Muhle, C , Reynolds, P.N., Knight, A., Themis, M., Carvajal, J., Scaravilli, R, Curiel, D.T., Fairweather, N.R and Coutelle, C. (2000) Retargeting of adenoviral vectors to neurons using the He fragment of tetanus toxin. Gene Ther., 7: 1584-1592. Schwab, M.E. and Thoenen, H. (1977) Retrograde axonal and transsynaptic transport of macromolecules: physiological and pathophysiological importance. Agents Actions., 7: 361-368. Simpson, D., Alexander, D., O'Brien, C , Tagliati, M., Aswad, A., Leon, J., Gibson, J., Mordaunt, J. and Monaghan, E. (1996) Botulinum toxin type A in the treatment of upper extremity spasticity: a randomized, double-blind, placebo-controlled trial. Neurology, 46: 1306-1310. Sinha, K., Box, M., LaUi, G., Schiavo, G., Schneider, H., Groves, M., Siligardi, G. and Fairweather, N. (2000) Analysis of mutants of tetanus toxin He fragment: ganglioside binding, cell binding and retrograde axonal transport properties. Mol. Microbiol., 37: 1041-1051.

Smith, G.A., Gross, S.P. and Enquist, L.W. (2001) Herpesviruses use bidirectional fast-axonal transport to spread in sensory neurons. Proc. Natl. Acad. Sci. USA., 98: 3466-3470. Sodeik, B., Ebersold, M.W. and Helenius, A. (1997) Microtubulemediated transport of incoming herpes simplex virus 1 capsids to the nucleus. J. Cell Biol., 136: 1007-1021. Soudais, C , Laplace-Builhe, C , Kissa, K. and Kremer, E.J. (2001) Preferential transduction of neurons by canine adenovirus vectors and their efficient retrograde transport in vivo. Faseb J., 15: 2283-2285. Spear, PG., Eisenberg, R.J. and Cohen, G.H. (2000) Three classes of cell surface receptors for alphaherpesvirus entry. Virology, 275:1-8. Strong, M.J. (2003) The basic aspects of therapeutics in amyotrophic lateral sclerosis. Pharmacol. Ther., 98: 379-414. Tenenbaum, L., Chtarto, A., Lehtonen, E., Velu, T., Brotchi, J. and Levivier, M. (2004) Recombinant AAV-mediated gene delivery to the central nervous system. J. Gene Med., 6(Suppl 1): S212-S222. Teng, Q., Garrity-Moses, M., Federici, T, Yang, J., Kutner, R., Tordo, N., Reiser, J. and Boulis, N. (2005) Neuronal gene transfer with HIV-1-based lentiviral vectors pseudotyped with lyssavirus glycoproteins. Mol. Ther., 11: S187. Teng, Q., Tanase, D., Liu, J., Garrity-Moses, M., Baker, K. and Boulis, N. (2005) Adenoviral clostridial light chain gene-based synaptic inhibition through synaptobrevin elimination. Gene Ther., 12:108-119. Thomas, C.E., Birkett, D., Anozie, I., Castro, M.G. and Lowenstein, PR. (2001) Acute direct adenoviral vector cytotoxicity and chronic, but not acute, inflammatory responses correlate with decreased vector-mediated transgene expression in the brain. Mol. Ther., 3: 36^6. Thoulouze, M.L, Lafage, M., Schachner, M., Hartmann, U., Cremer, H. and Lafon, M. (1998) The neural cell adhesion molecule is a receptor for rabies virus. J. Virol., 72: 7181-7190. Tsiang, H. (1979) Evidence for an intraaxonal transport of fixed and street rabies virus. J. Neuropathol. Exp. Neurol., 38: 286-299. Tsiang, H. (1993) Pathophysiology of rabies virus infection of the nervous system. Adv. Virus Res., 42: 375-412. Tuffereau, C , Desmezieres, E., Benejean, J., Jallet, C , Flamand, A., Tordo, N. and Perrin, P. (2001) Interaction of lyssaviruses with the low-affinity nerve-growth factor receptor p75NTR. J. Gen. Virol., 82: 2861-2867. Vallee, R.B. and Bloom, G.S. (1991) Mechaiusms of fast and slow axonal transport. Annu. Rev. Neurosci., 14: 59-92. VaUee, R.B., WilUams, J.C, Varma, D. and Bamhart, L.E. (2004) D)mein: an ancient motor protein involved in multiple modes of transport. J. NeurobioL, 58: 189-200. von Barfheld, C.S. (2004) Axonal trar\sport and neuronal transcytosis of trophic factors, tracers, and pathogens. J. NeurobioL, 58: 295-314. 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. and Ozawa, K. (2002a) Delayed delivery of AAV-GDNF prevents nigral neurodegeneration and promotes functional recovery in a rat model of Parkinson's disease. Gene Ther., 9: 381-389. Wang, L.J., Lu, YY, Muramatsu, S., Ikeguchi, K., Fujimoto, K., Okada, T., Mizukami, H., Matsushita, T., Hanazono, Y, Kume, A., Nagatsu, T., Ozawa, K. and Nakano, I. (2002b) Neuroprotective effects of glial cell line-derived neurotrophic factor mediated by an adeno-associated virus vector in a transgenic animal model of amyotrophic lateral sclerosis. J. Neurosci., 22: 6920-6928. Warrell, M.J. and Warrell, D.A. (2004) Rabies and other lyssavirus diseases. Lancet, 363: 959-969.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

CONCLUSION Warrington, K., Gorbatyuk, O., Harrison, J., Opie, S., Zolotukhin, S. and Muzycka, N. (2004) Adeno-associated virus type 2 VP2 capsid protein is nonessential and can tolerate large peptide insertions at its N terminus. J. Virol., 78: 6595-6609. Wilson, S., Yeomans, S., Bender, M., Lu, Y., Coins, W. and Glorioso, J. (1999) Antihyperalgesic effects of infection with a preproenkephalin-encoding herpes virus. Proc. Natl. Acad. Sci. USA., 96: 3211-3216. Yamamura, J., Kageyama, S., Uwano, T., Kurokawa, M., Imakita, M. and Shiraki, K. (2000) Long-term gene expression in the anterior horn motor neurons after intramuscular inoculation of a live herpes simplex virus vector. Gene Ther., 7: 934-941. Yamashita, S., Mita, S., Arima, T., Maeda, Y, Kimura, E., Nishida, Y, Murakami, T., Okado, H. and Uchino, M. (2001) Bcl-2 expression by retrograde transport of adenoviral vectors with Cre-loxP recom-

271

bination system in motor neurons of mutant SODl transgenic mice. Gene Ther., 8: 977-986. Yamashita, S., Mita, S., Kato, S., Okado, H., Ohama, E. and Uchino, M. (2003) Bcl-2 expression using retrograde transport of adenoviral vectors inhibits cytochrome c-release and caspase-1 activation in motor neurons of mutant superoxide dismutase 1 (G93A) transgenic mice. Neurosci. Lett., 350:17-20. Yang, J., Teng, Q., Garrity-Moses, M., Federici, T., Najm, I., Chabardes, S., Moffitt, M. and Boulis, N. (2005) Gene therapy of epilepsy by adenovirus-mediated tetanus toxin light chain gene transfer. Mol. Ther., 11: S168. Zemanick, M.C., Strick, PL. and Dix, R.D. (1991) Direction of transneuronal transport of herpes simplex virus 1 in the primate motor system is strain-dependent. Proc. Natl. Acad. Sci. USA., 88: 8048-8051.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

CHAPTER

21 Gene Therapy for Spinal Cord Injury Marc]. Ruitenberg, William TJ. Hendriks, Gerard]. Boer, ]oost Verhaagen

Abstract: Injuries to the adult mammalian spinal cord often lead to severe damage to both ascending (sensory) and descending (motor) nerve pathways with no or hardly any prospect of complete functional recovery. Future spinal cord repair strategies will most likely comprise a multi-factorial approach addressing several issues, including the optimization of survival and function of spared central nervous system neurons and the modulation of trophic and inhibitory influences to promote and guide axonal regrowth. Viral vector-mediated transfer of neurotrophic factor genes by direct vector injections is emerging as a novel and effective strategy to express neurotrophic proteins in the injured nervous system, including the spinal cord. Ex vivo transfer of neurotrophic factor genes is additionally explored as a way to more efficiently bridge lesion cavities with cellular implants for axonal regeneration. Several viral vector systems, based on herpes simplex virus, adenovirus, adeno-associated virus, lentivirus and moloney leukemia virus have been employed. The genetic modification of fibroblasts, Schwann cells, olfactory ensheathing glia cells and stem cells, prior to implantation to the injured spinal cord has resulted in improved cellular nerve guides. The steady advances that have been made in combining new viral vector systems with a range of promising cellular platforms for ex vivo gene transfer holds promising perspectives for the development of new neurotrophic factor-based therapies to repair the injured nervous system. Keywords: spinal cord injury; gene therapy; viral vectors; neuroregeneration; cellular implants; neurotrophic factors I.

is essential for the performance of numerous autonomous and voluntary tasks, including coordinated movements, bladder, bowel and sexual functions. Spinal cord injury (SCI) results in most instances in an immediate and severe disruption of both ascending and descending spinal pathways, leaving patients quadri- or paraplegic for the rest of their lives. Worldwide, thousands of people are victims of SCI and many more are added each year (McDonald, 1999). In the USA alone, the annual incidence of spinal injury is approximately 40 cases per million population, which is about 11,000 new cases each year (The National SCI Statistical Center, University of Alabama, USA). Trauma-related SCI primarily affects yoimg, healthy people between 21 and 30 years

INTRODUCTION

The mammalian spinal cord, located within the vertebral column, connects the brain with the peripheral organs and body tissue. The spinal cord contains large numbers of ascending (sensory) and descending (motor) nerve pathways carrying information to the brain and back. In a transverse section, the spinal cord can be divided into the butterfly-shaped gray matter, containing the cell bodies of motor neurons and intraspinal neurons, and white matter, which contains millions of myelinated axons traversing up and down the spinal cord. Proper functioning of the spinal cord

Gene Therapy of the Central Nervous System: From Bench to Bedside

273

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

274

21. GENE THERAPY FOR SPINAL CORD INJURY

of age (^25% of cases; Van Asbeck et al., 2000). Complete or incomplete quadriplegia due to sustained cervical cord injuries accounts for about half the SCI population. The rest, with either complete or incomplete paralysis, have injuries in thoracic, lumbar or sacral segments of the spinal cord. The functional outcome following SCI is dependent on the anatomical level, the nature (e.g. laceration, compression, contusion) and completeness of the injury. International comparisons of the prevalence rates of SCI range from 110 to 1120 cases per million inhabitants (Blumer and Quine, 1995). Although these numbers may seem rather low compared with incidence rates of other diseases, the emotional, social and economic costs are very high since most patients surviving the acute phase do have a life expectancy of several decades. Apart from experimental pharmacological treatment with antioxidants and anti-inflanunatory agents to mitigate secondary damage to the spinal cord (for review: Amar and Levy, 1999), no repair strategies for SCI are available leaving patients with very poor clinical prospects. In this chapter, we will review the steady advances that have been made in neuroregeneration research with the development of advanced gene and cell therapy to promote regeneration in animal models for SCI. This overview will be preceded by a brief description of the pathophysiology of SCI and the predominantly used animal models.

II.

THE PATHOPHYSIOLOGY OF SCI

The pathophysiological changes following neurotraumatic events are usually classified as 'primary' and 'secondary' in nature. Primary injury is the direct consequence of spinal cord contusion or laceration, causing disruption of cellular membranes and axonal pathways as well as microvascular damage. Spinal cord trauma rapidly leads to oxygen deprivation (ischaemia), shifts in electrolyte concentrations, and the release of excitatory amino acids (e.g. glutamate). These pathophysiological changes trigger a destructive cascade of events that lead to inflammation, progressive cell death, demyelination and a neural scar in the subacute and chronic phases, together defined as secondary damage (for reviews: Schwab and Bartholdi, 1996; McDonald and Sadowsky 2002). A.

The Acute Phase

Compression or laceration of the spinal cord due to displacement of bone and intervertebral disc fragments

results in acute and irreversible damage. Hemorrhages occur due to microvascular damage, in particular in the gray matter. Interruption of blood flow, hypotension and vasoconstriction are thought to contribute to hypoperfusion and subsequent ischaemia. These vascular changes are thought to be instrumental to the induction of a destructive signaling cascade, spreading to neighboring tissue that was initially spared (for review: Tator and Fehlings, 1991). The disruption of axonal and cellular membrane integrity leads to a dramatic imbalance in electrolyte concentrations, causing biochemical injury. Increased extracellular potassium concentrations disturb neuronal excitability and synaptic transmission, inducing a transient conduction block and subsequent primary paralysis (neurogenic shock). B.

The Subacute Phase

Intracellular calcium and sodium levels rise shortly after injury (Balentine and Spector, 1977; for review: Banik et al., 1987), leading to an uptake of interstitial fluid and subsequent cell edema. The accumulation of intracellular calcium is most likely caused by an influx of calcium into the cell via voltage-dependent calcium channels, N-methyl-D-aspartate (NMDA) receptor channels, release of calcium from cellular organelles, as well as the failure of ATP-mediated calcium extrusion (Agrawal and Fehlings, 1996; Leybaert and Hemptinne, 1996). The activation of calcium-dependent proteases and phospholipases results in breakdown of cellular membranes causing massive and predominantly necrotic cell death, free radical formation and inflammation. Apoptosis becomes more prominent at a later stage (for reviews: Braughler and Hall, 1992; Young, 1992; Hall, 2001). The excessive release of the excitotoxins glutamate and aspartate contributes to apoptosis (Faden and Simon, 1988; Panter et al., 1990; Liu et al., 1991; for reviews: Choi, 1992; Alessandri and Bullock, 1998). In more chronically injured spinal cord, apoptotic oligodendrocytes have been found in the rim of spared white matter surrounding the necrotic lesion (Crowe et al., 1997; Liu et al., 1997), explaining the progressive demyelination (Crowe et al., 1997; Shuman et al., 1997). Selective inhibition of NMDA receptor channels (e.g. with dextrorphan, dizocilpine [MK-801] or gacyclidine) can reduce secondary loss of neural tissue and improve functional outcome (Haghighi et al., 1996, 2000; Wada et a l , 1999; Feldblum et al., 2000; Gaviria et al., 2000a, b; Terada et al., 2001). Similarly, blocking of non-NMDA glutamate receptors reduced oxidative stress following spinal cord contusion (Mu et al., 2002). Side effects have so far

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

RAT MODELS FOR EXPERIMENTAL SCI

prevented an anti-excitatory pharmacotherapy from being clinically applicable. The activation and recruitment of endogenous microglial cells and astrocytes as well as blood-derived immune cells that induce a classical inflammatory response is characteristic for the subacute phase. The onset of reactive gliosis is already visible shortly after the initial injury. Astroglial cells upregulate glial fibrillary acidic protein (GFAP) expression and accumulate at the lesion borders (lizuka et al., 1987). Activation of monocyte-derived microglial cells becomes apparent by the upregulation of several cell surface proteins involved in immune system activation. Elevated expression of major histocompatibility complex class I and II antigens (Popovich et al., 1993; Koshinaga and Whittemore, 1995), complement factor C3 receptor (Watanabe et al., 1999) and macrophage marker ED-1 (Carlson et al., 1998) has been reported in microglia that surround the injury site, indicating a transformation from a quiescent state into fully active phagocytic cells. In addition, pro-inflammatory chemokines and cytokines such as tumor necrosis factor-a (TNF-a), macrophage colony stimulating factor and interleukins IL-ljS and IL-6, are dramatically upregulated in reactive glial cells within days after injury (Bartholdi and Schwab, 1997; Streit et al., 1998). These molecules play a key role in the recruitment of inflammatory cells from the blood stream (Klusman and Schwab, 1997; Leskovar et al., 2000). Substantial evidence is available that inflammation after SCI directly contributes to secondary cell loss. Elevated expression of TNF-a (Lee et al., 2000; Hermann et al., 2001) and IL-ljg (Nesic et al., 2001) after spinal trauma has been directly related to injury-induced apoptotic death, which in case of IL1/? could be completely abolished by local infusion of IL-1 receptor antagonist (Nesic et al., 2001). Systemic administration of anti-inflammatory cytokine interleukin-10 (IL-10) attenuates the early inflammation and increases neuronal survival (Brewer et al., 1999). In the USA, prompt administration of methylprednisolone (within 8 h after injury) is currently the standard of care in an attempt to reduce swelling during the acute and subacute phases thereby containing secondary damage (Bracken, 2002; Bracken and Holford, 2002). In animal models, methylprednisolone-mediated tissue preservation has been correlated with reduction of oxidative damage (Hall and Braughler, 1982; Topsakal et al., 2002) as well as suppression of inflammation (Bartholdi and Schwab, 1995; Xu et al., 1998; Oudega et al., 1999). Unfortunately, the clinical efficacy of methyl-prednisolone is rather ambiguous (Hurlbert, 2000).

C.

275

The Chronic Phase

While the impact zone is cleared from cellular debris by phagocytosis, leaving one or more fluid-filled cysts, meningeal fibroblasts and Schwann cells invade the injury site (Matthews et al., 1979; Krikorian et al., 1981; Pasterkamp et al., 1999; De Winter et al., 2002). In particular, after penetrating injuries that cause disruption of the dural membranes, massive proliferative and migratory responses of these cells occur, starting at the end of the first week and continuing for several weeks. Together with several layers of hypertrophic astrocytes that delineate the lesion borders, meningeal fibroblasts become the main constituent of the neural scar. These meningeal cells have been implicated in the deposition of extracellular matrix (ECM) components at the injury site, which subsequently induces the formation of basement membrane (Stichel et al., 1999). The formation of a scar can be regarded as a neural wound healing response that prevents further spread of damage to the uninjured cord. Apart from being a physical barrier, however, several putative inhibitors for axon growth are expressed in the neural scar (Fawcett and Asher, 1999; Pasterkamp and Verhaagen, 2001; Niclou et al., 2005). The neural scar is therefore considered a major molecular obstacle for CNS regeneration. Injured axons form typical swollen endings ('endbulbs') usually at a short distance from the fibrotic core of the neural scar. This is indicative for the non-permissive nature of the neural scar (Shearer and Fawcett, 2001; Niclou et al., 2005).

IIL

RAT MODELS FOR EXPERIMENTAL SCI

Laboratory animal models for SCI are crucial to obtain more insight in the pathophysiological changes that occur after injury, and to develop strategies to augment regeneration and functional recovery. Animal models for SCI should mimic human pathology as accurately as possible. Several experimental models for SCI have been developed in the rat, including injury types of both a closed (contusive) and penetrating (transection) nature (for review, see Kwon et al., 2002). 'Weight-drop' or contusion injury and spinal cord transection, either complete or incomplete, are most commonly used. Here the advantages and disadvantages of these lesion models are briefly discussed. Most traumas to the human spinal cord are the so-called 'closed' injuries that are characterized by bruising or compression of the spinal cord but not laceration, i.e. the integrity of the dural membranes is not compromised. The pathology that is induced by this

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

276

21. GENE THERAPY FOR SPINAL CORD INJURY

type of injury, in particular the formation of a central necrotic core that eventually evolves into a cystic cavity as well as a peripheral rim of spared spinal cord tissue, also occurs in the weight-drop impactor lesion model (Grimer, 1992). The similarities with human pathology (Metz et al., 2000) make this model clinically relevant and particularly useful to study pathophysiological changes after the initial trauma. The main drawback, however, is that it is difficult to distinguish axonal sparing from regrowth, which makes it very difficult to assess possible relations between functional recovery and the anatomical correlate following experimental intervention, e.g. cell or gene therapy. In contrast to contusion lesions, surgical transections of the spinal cord, either partial or complete, are very well suited to study axonal tract regeneration. Although these types of injuries also induce tissue necrosis, the formation of cysts is often less prominent. Axonal regrowth in these models is relatively simple to assess via anterograde a n d / o r retrograde tracing techniques (Vercelli et al., 2000), as well as electrophysiological analysis Muir and Webb, 2000). For complete spinal cord transections, possible relationships between anatomical pathway regeneration and recovery of function can be examined by re-transection of the spinal cord lesion site, thereby eliminating all axon input from the proximal stump that may have contributed to the functional recovery Partial transections of specific spinal cord tracts can also be used to study anatomical regeneration of these targeted pathways, and have the advantage of a lower incidence of death as well as being less traumatic to experimental animals. However, the relationship between regenerating axons and functional recovery can be more difficult to study due to residual locomotor function, compensatory mechanisms and unwanted incomplete lesioning resulting in sparing of a proportion of the targeted nerve pathway (Steward et al., 2003). Although still a matter of some controversy (Levi et al., 1996), persistent impairments in voluntary motor function following human SCI have mostly been attributed to the disruption of integrated signaling between descending spinal motor pathways (Nathan, 1994) and ascending sensory tracts. Therefore, most gene and cell therapy studies in animal models have focused on promoting regrowth of main motor pathways, i.e. corticospinal tract (CST), rubrospinal tract (RST) and ascending sensory spinal tract (for review, see Blits et al., 2002). The CST is derived from layer V pyramidal neurons in the sensorimotor cortex and the RST originates from the magnocellular part of the red nucleus in the mesencephalon. The ascending fibers are localized

in the dorsal funiculus and originate from primary sensory neurons in the dorsal root ganglia. IV.

W H A T IS REQUIRED TO ELICT SUCCESSFUL REGENERATION AFTER SCI?

The pathophysiological changes that occur in the injured spinal cord dictate the design of therapeutic intervention strategies, including cell and gene therapy. It is apparent from the previous paragraphs that the following issues should be addressed specifically: (1) Minimalization of the deleterious effects of early trauma and inflammation, (2) Improving the long-term survival and maintenance of spared neurons and fibers in partial lesions of the spinal cord, (3) Replacement of lost neural tissue at the impact site, (4) Integrated exploitation of trophic and inhibitory influences to guide axonal regrowth, and finally, (5) Remyelinization of compromised and regenerating axons as well as the establishment of appropriate synaptic connections by regenerating axons to restore damaged neural circuits. To date, several experimental treatments, dealing with each of these points, have been employed to augment functional recovery after SCI (reviewed in Blits et al., 2002). In general, these approaches have followed one out of two main strategies: (a) reduction of secondary damage to the spinal cord and (b) stimulation of axonal regeneration to reconnect the brain with local motor and sensory circuitry in the spinal cord. Future spinal cord repair strategies will most likely be multidisciplinary in nature. Here, we will specifically focus on the progress that has been made with cell and gene therapy to promote repair of the injured spinal cord. V.

A.

DIRECT VIRAL VECTOR^MEDIATED GENE TRANSFER IN THE INJURED SPINAL CORD Herpse Simplex Viral Vectors

The first viral vectors used for direct genetic manipulation of the nervous system were based on herpes simplex virus (HSV; Kaplitt et al., 1991). Most in vivo studies on HSV vectors have focused on the peripheral nervous system (PNS) since wild-type HSV naturally targets sensory neurons (Leib and Olivo, 1993). Upon administration, HSV vectors are retrogradely transported from the site of injection (Keir et al., 1995). This tropism of HSV for sensory (ORG) neurons is due to the high level of its preferred high-affinity receptor

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

DIRECT VIRAL VECTOR-MEDIATED GENE TRANSFER IN THE INJURED SPINAL CORD

nectin-1 on peripheral sensory axons (Mata et al., 2001). These characteristics make HSV-based vectors well suited for gene therapeutic applications in conditions that affect the PNS. In a pioneering study by Federoff and cov^orkers, HSV-mediated expression of NGF in the axotomized rat superior cervical ganglion prevented the decline in tyrosine hydroxylase expression that otherwise would have occurred in these neurons (Federoff et al., 1992). In a different injury model, HSV-1mediated expression of the anti-apoptotic protein Bcl-2 protected motoneurons from subsequent degeneration following ventral root avulsion. Injection of the vector into the spinal cord 1 week before lumbar root avulsion prolonged the lifespan of lesioned motoneurons but did not preserve choline acetyl transferase (ChAT) expression (Yamada et al., 2001). In a different study, HSV vectors encoding Bcl-2 and GDNF were combined in a similar ventral root avulsion model. Transduction of motoneurons with either one of the vectors resulted in a substantial increase in the number of surviving motoneurons but again not in the preservation of ChAT expression. The combination of both HSV-Bcl-2 and HSV-GDNF vectors, however, resulted in significantly increased numbers of ChAT-positive motor neurons in the ventral horn (Natsume et al., 2002). B,

Adenoviral Vectors

In 1993 there were four independent reports on the generation of first generation adenoviral (AdV) vectors for gene transfer to the nervous system (Akli et al., 1993; Bajocchi et al., 1993; Davidson et al., 1993; Le Gal La Salle et al., 1993). Adenoviral vectors efficiently transduce astrocytes, oligodendrocytes, ependymal cells and neurons (Hermens et al., 1997). These first generation AdV vectors have a deleted El-region, resulting in a significantly diminished albeit not complete deactivation of the expression of other early (E) and late (L) genes needed for adenoviral replication (Hermens and Verhaagen, 1998). Characterizing in vitro experiments have demonstrated that high amounts of biologically active neurotrophic factor is produced by AdV vector-transduced cells (Smith et a l , 1996; Dijkhuizen et al., 1997). In vivo, Baumgartner and Shine (1998) applied AdV vectors for GDNF, BDNF, NGF, NT-3 and CNTF in a sciatic nerve transection model. AdV vectors were injected into the hind limb muscles of newborn rats, and the vector retrogradely transported to the lumbar spinal cord directing transgene expression in motoneurons. AdV-mediated expression of GDNF resulted in the protection of a significant number of

277

sciatic nerve motoneurons against axotomy-induced cell death at 1 and 3 weeks post-lesion compared with AdV-LacZ-treated controls. At 6 weeks post lesion this neuroprotective effect did not persist (Baumgartner and Shine, 1998). This is in line with earlier studies on AdV-CNTF and -BDNF injections into the facial nerve 1 week after unilateral axotomy of the facial nerve. Rescue of motoneurons was also achieved following direct injection in the facial nerve (Gravel et al., 1997). Dijkhuizen et al. (1998) studied the possibility to transduce Schwann cells distal from a sciatic nerve crush with the potential aim of neurotrophin gene transfer to promote peripheral nerve regeneration. Direct injection of AdV-LacZ into the uninjured sciatic nerve resulted in high levels of transgene expression lasting for 12 days. The number of transduced Schwann cells gradually declined, with only a few transduced Schwann cells present at 45 days. Infusion of the vector at the time of, or immediately after a crush failed to significantly transduce Schwann cells in the distal nerve portion (Dijkhuizen et a l , 1998). However, large numbers of transduced Schwann cells distal from the lesion site could be obtained when the vector was administered 1 day post-injury. Thus, AdV vectors can transduce Schwann cells in an injured peripheral nerve but transgene expression is rather short lived. Following direct injection of AdV vector in the lesioned spinal cord, the timing of vector injection in the post-lesion period appears to be a critical parameter. Direct gene transfer using AdV vectors has been applied in a small number of spinal cord repair studies (Zhang et al., 1998; Ruber et a l , 2000; Romero et al., 2001). In the dorsal root injury model, in which the lumbar dorsal roots of adult rats were severed and reanastomosed, AdV vectors have been used to transgenically express the neurotrophins NT-3 and NGF in an attempt to promote regrowth of the central branch of primary sensory neurons (Zhang et al., 1998; Romero et al., 2001). In the study by Zhang et a l , AdV-LacZ or AdV-NT-3 were injected into the ventral horn of the lumbar spinal cord 2 weeks post-lesion. Transgene expression was found in glial cells as well as motoneurons for up to 40 days. AdV vector-derived NT-3 attracted injured sensory axons into the spinal cord and regrowing sensory axons were mainly found in the gray matter, penetrating as deep as lamina V In a similar study, Romero et al. (2001) applied AdV-NGF to the lumbar spinal cord 2 weeks after dorsal root avulsion. Robust axonal regeneration into denervated as well as ectopic locations was observed within the dorsal spinal cord with a near-normal recovery of sensory function. The observed functional recovery was

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

278

21. GENE THERAPY FOR SPINAL CORD INJURY

completely abolished after retransection of the dorsal roots. An important observation was the extensive sprouting of non-injured sensory axons in AdV-NGF injected animals, which may have contributed to symptoms of hyperalgesia and chronic pain (Romero et al., 2000, 2001). Adenoviral vector-mediated expression of the chemorepulsive protein semaphorin3A (sema3A) counteracts the NGF-induced sprouting of sensory fibers, indicating that this chemorepellent is capable of preventing detrimental sprouting of NGFsensitive fibers (Tang et al., 2004). Gene transfer in the dorsohemisected rat spinal cord was investigated by injecting AdV-LacZ (Blits, 2002) or an AdV vector encoding green fluorescent protein (AdV-GFP) (Huber et al., 2000). Neurons, astrocytes, oligodendrocytes and peripheral cells infiltrating the lesion site were transduced. In the study by Blits (2002). transduction was inefficient when the injection was given immediately following the lesion, but more robust transgene expression was observed when injected a few hours prior to the lesion or 1 week postlesion. Following similar procedures, vector injections were given one day prior to the lesion in the study by Huber et al. (2000). AdV-GFP-transduced cells were found within 3 mm on each side of the lesion. Both studies showed high levels of transgene expression between 3 and 8 days, which had ceased by 21 days. Both virusmediated inflammation, direct toxicity, promotor down regulation and loss of the episomally present transgene DNA may have contributed to the transient nature of AdV-directed transgene expression (see below). More recently, spinal motoneurons were transduced via retrograde transport of AdV-NT3 following injection of the vector into the sciatic nerve, 2 weeks after a contralateral GST lesion above the pyramidal decussation (Zhou et al., 2003). Expression of NT-3 was detected up to 3 weeks post-injection (longest time point studied) but the amount of NT-3 had decreased compared to the 1-week time point. No effect on sprouting of CST fibers was seen in unlesioned animals. Interestingly, following injury, AdV vectorniediated NT-3 expression in ventral motor neurons induced sprouting of CST fibers from the unlesioned side, across the midline, into the denervated area. C.

Adeno'Associated and Lentiviral Vectors

Problems with immunogenicity and direct toxicity of HSV and AdV vectors inspired the development of adeno-associated viral (AAV) and lentiviral (LV) vectors (Kaplitt et al., 1994; Naldini et al., 1996). Following injections of AAV-GFP injections into the rat

cervical spinal cord, numerous GFP-positive cells were observed, mainly intraspinal neurons that expressed the transgene for months without signs of toxicty (Peel et al., 1997; Dijkhuizen et al., 1999; Ruiten-berg et al., 2002a; Blits et al., 2003). In recent years, a number of studies have used AAV vector-niediated expression of neurotrophic factors in spinal cord lesion models, either at the level of the soma of axotomized neurons or distally from the axon stump. No papers have as yet been published on the application of LV vectors in SCI. After a cervical lesion of the RST, injections of AAV-BDNF into the red nucleus counteracted lesioninduced atrophy of rubrospinal neurons (Ruitenberg et al., 2004). Although the effect was not yet visible at 2 weeks after axotomy and vector administration, AAVBDNF treatment fully reversed atrophy at 5 weeks post-lesion. The lack of an effect at the 2-week time point was most likely caused by a delay in the onset of expression after AAV injection and the relatively slow build up of therapeutic BDNF levels. Interestingly, AAV-BDNF injections in the red nucleus 18 months following the lesion resulted in partial but significant reversal of the atrophy in chronically lesioned neurons (Ruitenberg et al., 2004). In an attempt to foster axonal growth from a Schwann cell bridge into the caudal spinal cord, intraspinal neural cells caudal to the implant were transduced with AAV vectors encoding for BDNF and NT-3 (Blits et al., 2003). Although the distal application of AAV-BDNF and AAV-NT3 led to a modest improvement in hind limb function, no evidence was found for growth of regenerated axons from the Schwann cell implants into the caudal spinal cord. Interestingly, retrograde tracing demonstrated that twice as many neurons extended processes toward the distal end of the Schwann cell graft, suggesting that AAV-mediated expression of BDNF and NT3 modified the local circuitry in the lumbar spinal cord, caudal to the Schwann cell bridge, but had no effect on the growth of axons in the graft itself. The studies summarized above provide proof of principle that cell survival and nerve fiber regeneration can be stimulated by viral vector-mediated neurotrophic factor gene transfer. Although encouraging results have been achieved, the usefulness of HSV vectors for in vivo CNS gene therapy to date has been limited by transient expression of transgenes due to latencyassociated shut-off of promoters and most importantly toxicity and immunogenicity of HSV vectors (Wood et al., 1994a, b; Keir et al., 1995; Kennedy 1997; Lilley et al., 2001). Similarly, after the direct application of AdV vectors, transgene expression is transient due to

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

COMBINING TRANSPLANTATION AND GENE THERAPY

an immunological response and direct toxicity (Zhang and Schneider, 1994; Byrnes et al., 1995, 1996; Wood et al., 1996), that occurs in particular following injection of higher dosages of vector particles (Hermens and Verhaagen, 1998; Huber et al., 2000). AdV vectors were initially rendered replication-deficient through deletions in the El-region, at high multiplicities of infection (MOI) the El region becomes dispensable for replication, in particular in vivo (Jones and Shenk, 1979; Nevins, 1981). As a result El-independent activation of the E2 promoter can lead to replication of wild-type adenoviral DNA, followed by accumulation of viral gene products that lead to cytopathic effects (Zhang and Schneider, 1994) and activation of the host immune system (Byrnes et al., 1995; Hermens et al., 1997). Also direct cytotoxic and immunogenic effects of AdV vector-mediated gene transfer have been shown in a number of studies. Injection of AdV-LacZ into the striatum of two different inbred rat strains led to inflammations in this brain area. Severity of the inflammation seemed to depend on the rat strain used (Byrnes et al., 1995). Inflammation was not evident at 2 months after injection but at that time point transgene expression was greatly reduced. Similar inflammatory responses were found after injection of the vector in the brain of adult mice (Kajiwara et al., 1997). Re-exposure to AdV vector leads to severe inflammation and microglial activation, resulting in local demyelination and further decrease in transgene expression (Byrnes et al., 1996). With the development of safe, non-toxic vector systems, i.e. AAV and LV vectors, long-term studies on the efficacy of neurotrophin gene transfer for the promotion of neuroregeneration are now feasible and have opened opportunities to further develop direct gene transfer as a potential treatment for neurotraumatic lesions. VI.

COMBINING TRANSPLANTATION A N D GENE THERAPY

Schwann cells and fetal spinal cord tissue were the first transplants used in attempts to repair the injured spinal cord (Richardson et al., 1980; David and Aguayo, 1981; Bregman and Reier 1986; Bregman et al., 1997). The hypothesis was that neurons in the CNS fail to regenerate because of an inhibitory environment and that changing this environment would elicit a regenerative response in CNS neurons. The promising results obtained with these transplants, increased understanding of the molecular factors that promote neurite outgrowth, and the development of viral vectors inspired the idea of combining transplantation with

279

gene transfer (Blits et al., 2002). During the last decade, a number of genetically engineered cellular implants have been transplanted into the injured CNS to study their neuroprotective and neurite regrowth-promoting effects, including primary fibroblasts, Schwann cells, peripheral nerve tissue bridges, olfactory ensheating glia cells and neural stem cells. A»

Fibroblasts

Fibroblasts are easily sustainable in vitro, they survive implantation to the spinal cord, and they secrete molecules such as collagen and fibronectin that can reconstitute an ECM at the lesion site along which neurons can extend axons (Lacroix and Tuszynski, 2000). Primary, retrovirally transduced NGF-secreting fibroblasts implanted in the uninjured adult rat spinal cord, survived for up to 1 year and induced a robust ingrowth of primary sensory neurites (Tuszynski et al., 1994). Control implants and basic fibroblast growth factor (bFGF)-producing implants promoted significantly less neurite growth. To examine if NGFsecreting fibroblasts could induce sprouting from other axotomized spinal neurons, these cells were implanted in midthoracic dorsal hemisection lesions (Tuszjmski et a l , 1996, 1997). Recipients of NGF-secreting grafts showed deficits in locomotion over a wire mesh not different from control animals after three months. NGF-secreting implants elicited sprouting from spinal primary sensory afferents, local motor axons, and putative coerulospinal axons, but no regeneration of corticospinal axons was observed. Axons responding to NGF penetrated into the implant but did not exit into distal spinal cord tissue (Tuszjmski et al., 1997). Implantation of these cells into the chronically injured dorsohemisected rat spinal cord (Grill et al., 1997b), at 1-3 months after lesioning, resulted in robust growth of coerulospinal and primary sensory axons of the dorsolateral funiculus into the implant and lesion site. This indicates that chronically lesioned axons retain their responsiveness to neurotrophins (Grill et a l , 1997b). In another study by Grill et al. (1997a), primary fibroblasts genetically modified to secrete NT-3 were grafted to acute dorsal hemisection lesion cavities. At 3 months post-implantation, significant partial functional recovery occurred in grafted animals, but most importantly, corticospinal axons did grow for up to 8 mm distal to the injury site. CST axons extended through the spinal cord gray matter but not through the implant nor into white matter (Grill et al., 1997a). Implantation of primary fibroblasts secreting BDNF into the lesion cavity after a partial cervical hemisection

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

280

21. GENE THERAPY FOR SPINAL CORD INJURY

reportedly also elicited a significant growth response of CNS motor axons, i.e. RST fibers (Liu et al., 1999b). At both 1 and 2 months post-lesion and implantation, anterograde and retrograde tracing showed RST axons regenerating through and around the implants. These axons grew for long distances within white matter caudal to the implant, and terminated in spinal cord gray matter regions that are the normal target areas of RST axons. Significant recovery of forelimb function occurred in animals that received BDNF-producing implants and this recovery was abolished after a second lesion transecting the regenerated axons (Liu et al., 1999b, 2002; Kim et al., 2001). Besides inducing axonal regrowth of the RST, implantation of BDNFsecreting fibroblasts prevented neuronal atrophy and death in the red nucleus, indicating beneficial retrograde effects (Liu et al., 2002). Following implantation into the contused spinal cord, BDNF and NGF secreting fibroblasts were reported to promote improved functional recovery in rats after implantation of these cells into the lesion cavity (Kim et al., 1996). In a different study, primary fibroblasts expressing the reporter gene LacZ, or the neurotrophic factors NT-3, BDNF, CNTF, NGF or bFGF were grafted to the contused adult rat spinal cord (McTigue et al., 1998). Ten weeks after injury, NT3 and BDNF implants contained significantly more axons than control or other growth factor-producing implants. In addition, more myelin basic protein-positive profiles were detected in these implants, indicating increased myelination of ingrowing axons. BDNF and NT-3 secreting implants did indeed contain more BrdU-positive oligodendrocytes than control LacZ implants, suggesting proliferative effects of these neurotrophins on oligodendrocytes (McTigue et al., 1998). Primary fibroblasts have also been genetically modified by AdV vector-mediated gene transfer to express the reporter genes GFP and LacZ (Liu et al., 1998). Cell density and morphology of the implants were comparable to those observed in retrovirally transduced implants, suggesting similar long-term cell survival. However, the level of transgene expression rapidly decreased with time. A mild immunological response to the grafted, transduced cells occured, which together with the loss of the episomal AdV genome during cell division, did likely account for the decline in transgene expression (Liu et al., 1998). B,

Schwann Cells

Schwann cells govern peripheral nerve regeneration by secreting trophic factors and ECM molecules

like laminin and collagen as well as cell adhesion molecules (Snyder and Senut, 1997). In addition, Schwann cells provide physical support to injured peripheral axons by their alignment into the so-called bands of Biingner that form pathways for axonal regrowth (Ide, 1996). These properties make Schwann cells attractive candidates for transplantation to the injured spinal cord (Xu et a l , 1995; Bamber et al., 2001). To test if Schwann cell-induced axonal growth in the spinal cord (Xu et al., 1995) could be optimized using genetic engineering, cultured primary adult rat Schwann cells were genetically engineered to secrete NGF (Tuszynski et al., 1998). Following implantation into the midthoracic spinal cord of adult rats, these cells survived for up to 1 year and became densely penetrated by primary sensory nociceptive axons originating from the dorsolateral funiculus compared to control implants. Axons in both NGF-secreting and control implants became myelinated by Schwann cells (Tuszynski et al., 1998). When applied in the lesion cavity of a dorsal hemisection of the rat spinal cord, a significant increase in growth of spinal cord axons into the implant was observed (Weidner et al., 1999). A more dense network of coerulospinal axons and central processes of primary sensory afferents was found in the transduced implants as compared to untransduced implants. In addition, these axons became ensheathed and in some instances remyelinated by Schwann cells. In this study, the implanted Schwann cells exhibited a phenotypic and temporal course of differentiation into a myelinating state while aligning spontaneously. At 3 days after implantation, Schwann cells were still in an undifferentiated or nonmyelinating state. At 2 weeks they had upregulated the cell adhesion molecule LI, a marker for differentiated non-myelinating Schwann cells. At 3 weeks, PO glycoprotein, a major constituent of peripheral myelin, was detected, indicating that some Schwann cells may have adopted a myelinating phenotype. As no differences for Schwann cell markers were found between NGF-secreting and control implants NGF itself did not appear to modulate the Schwann cell myelinating phenotype. The observed time course of Schwann cell differentiation after grafting to a CNS injury site is a recapitulation of the redifferentiation also occurring after PNS injury. The physiological response of Schwann cells to injury appears to be retained following transplantation at an ectopic site, i.e. in the injured spinal cord (Weidner et al., 1999). Schwann cells modified by retroviral vectors to secrete BDNF and implanted as trails in and distal to the transection site of the adult rat spinal cord attracted

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

COMBINING TRANSPLANTATION AND GENE THERAPY

more dorsal root ganglion and spinal and supraspinal fibers than control Schwann cell implants (Menei, et al., 1998). The Schwann cell trail was maintained for at least 1 month and most fibers sprouted at the level of the transection. Many fibers crossing the transection site and growing into the BDNF Schwann cell trail were CGRP-positive sensory fibers. The BDNFsecreting Schwann cells, however, appeared not to myelinate regenerating axons based on the absence of PO expression, which was detected in normal and NGF-transduced Schwann cells. These findings therefore suggest that Schwann cells in the presence of BDNF remain in a dedifferentiated state promoting fiber regeneration but not myelination (Menei et a l , 1998). Together, these studies demonstrate that supraphysiological levels of neurotrophic factor production by genetically modified Schwann cells can augment the regenerative potential of injured spinal axons but that other Schwann cell characteristics such as axon myelination may be affected depending on what neurotrophic factor was expressed. C,

Peripheral Nerve Bridges

As with Schwann cells, significant functional recovery and axonal growth into peripheral nerve pieces has been reported following implantation into the lesioned spinal cord. In one study, regenerating CST fibers did exit the graft and were observed in the distal cord (Cheng et al., 1996). In general, ingrowth of CNS nerve fibers in peripheral nerve implants does occur but not all neuronal populations are equally responsive to their microenvironment (Anderson et al., 1998). Therefore, genetic modification of peripheral nerve bridges to overexpress neurotrophic proteins was investigated as a way to improve the permissive properties of peripheral nerve implants. Microinjection of AdV into pieces of intercostals nerve results in transduced Schwann cells expressing the transgene for up to 30 days in culture (Blits et al., 1999). Following implantation of these ex vivo AdV-LacZ-transduced autologous intercostal nerve pieces as bridges in the hemisected adult rat spinal cord (Blits et al., 1999), numerous j6-galactosidase-expressing cells were observed at 1 week after implantation. At 2 weeks, however, the nun\ber of cells expressing the transgene had declined and at 4 weeks hardly any transduced cells were present anymore. CST regeneration and motor recovery was studied following implantation of intercostal nerves genetically modified with AdV-NT-3 (Blits et al., 2000) or AdV-LacZ as a control. Neurofilament staining revealed ingrowth

281

of many fibers from the host spinal cord into mock-, AdV-LacZ- or AdV-NT-3-transduced implants, but no CST axons were seen to regenerate into the intercostal nerve implants. Instead, CST axons grew ventral from the implants in the spared gray matter. Three to fourfold more CST fibers were observed ventral from the transplanted nerve bridge transduced with AdV-NT-3 as compared to control bridges. Regrowth and sprouting of CST axons occurred over more than 8 mm distal to the lesion site, similar to the observations that were described following NT-3-producing fibroblasts (Grill et al., 1997a). Animals implanted with Ad-NT-3-transduced peripheral nerves also exhibited improved hind limb performance after 8 weeks. Thus, transient overexpression of NT-3 in peripheral nerve tissue bridges during the first 2 weeks after implantation is sufficient to stimulate regrowth of CST fibers and to promote partial recovery of hind limb function. D.

Olfactory Ensheathing Glia Cells

Recently, there has been increasing interest to use olfactory ensheathing glia cells (OEGs) as cellular conduits to stimulate regeneration in the damaged CNS. In the primary olfactory system, OEGs have been implicated in guiding new primary olfactory axons growing from the olfactory epithelium in the periphery toward their CNS target area, the olfactory bulb glomeruli (Williams et al., 2004). This interesting property of OEGs, i.e. guidance of growing axons in mature CNS tissue, has turned them into attractive candidates for neural transplantation purposes in the injured spinal cord. At present, a significant number of studies have demonstrated the potential of OEGs to support regenerative growth of selective axon populations in the lesioned spinal cord (Ramon-Cueto and Nieto-Sampedro, 1994; Li et al., 1998; Ramon-Cueto et al., 1998), often accompanied with a certain degree of functional recovery (Li et a l , 1997; Ramon-Cueto et al., 2000). An important advantage of OEG implants over Schwann cell grafts is that they do not trigger upregulation of inhibitory proteoglycans at the graft-host interface (Lakatos et al., 2000,2003; Plant et a l , 2001). Also these cells do not entrap regenerating axons preventing them from entering the distal spinal cord (Ramon-Cueto et al., 1998). Although the results obtained with OEG implants were encouraging, observed regenerative growth through these grafts was still far from optimal. The growth-promoting effect of OEGs is probably not solely caused by secretion of neurotrophins as they do express only low levels of NGF and BDNF, while NT-3 could not be detected in these cells (Boruch

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

282

21. GENE THERAPY FOR SPINAL CORD INJURY

et al., 2001). Therefore the first studies combining gene therapy and OEG implantation have focused on creating OEGs that secrete high levels of BDNF, NT-3 and GDNF (Ruitenberg et al., 2003, 2005; Cao et al., 2004). This could potentially enhance the capacity of these cells to promote regeneration. First, AdV, AAV-2 and LV vectors have been tested to study the possibility of transducing OEGs prior to implantation in the lesioned spinal cord (Ruitenberg et al., 2002b). Ex vivo transduction of primary OEG with AdV vectors resulted in transient transgene expression following implantation in the lesioned spinal cord whereas persistent transgene expression was observed in LV vector or retrovirally transduced OEG implants. Genetic modification of OEGs to make them secrete neurotrophic factors was achieved (Ruitenberg et al., 2003, 2005; Cao et al., 2004). AAV-2 vectors did not transduce OEGs (Ruitenberg et a l , 2002b). Transgenic OEG, engineered with BDNF and NT-3 encoding AdV vectors do stimulate regrowth of lesioned RST and CST axons, respectively and diminishes size of the primary lesion site following acute implantation at the injury site (Ruitenberg et al., 2003, 2005). Similar results have been obtained with OEGs stably transduced to secrete large amounts of GDNF (Cao et al., 2004). Robust axonal ingrowth was observed in areas containing GDNF-secreting OEGs compared to control cell implants. Analysis of anterograde CST tracing showed augmented CST axon sprouting at the lesion and transplantation area in animals that received GDNF-secreting OEG implant but no significant differences in CST axon numbers were found further distally. These observations suggest that chronic expression of GDNF from OEG implants entrapped regenerating axons at the site of engraftment due to high neurotrophic factor content at the site of engraftment. A similar phenomenon was recently described in the spinal cord after direct AAV vector-mediated gene transfer of GDNF to ventral motoneurons (Blits et al., 2004). These observations underline the importance of vectors with regulatory transgene expression (Blesch et a l , 2001).

E-

oligodendrocytes. Only a handful of studies on the use of stem cells for spinal cord repair have been performed. Mouse ES cells have been implanted into the rat spinal cord 9 days after a contusive injury of the spinal cord (McDonald et al., 1999). Implant-derived cells survived and differentiated into astrocytes, oligodendrocytes and neurons, and migrated for as far as 8 m m from the lesion site. Hind limb weight support and partial hind limb coordination was improved. Although this study did not investigate the effect of the implant on axonal regeneration, the observed functional recovery was probably partly due to enhanced myelination as most stem cells differentiated into oligodendrocytes. Functional recovery of weight bearing and hind limb stepping is also reported for neural stem cell-containing scaffolds implanted in the hemisected spinal cord (Teng et al., 2002). Anterograde tracing showed CST fibers passing through the injury epicenter to the caudal cord, which was not seen in control animals. Using in vitro oligodendrocyte-differentiated ES cells in chemically demyelinated rat spinal cord or the spinal cord of myelin-deficient Shiverer mouse, large numbers of cells survived and remyelinated host axons (Liu et al., 2000). The fate of a retrovirally transduced neural stem cell line (CI7) has been described by Liu et al. following implantation in the spinal cord (Liu et al., 1999a). The cells survived and expressed NT-3 for at least 2 months and the cells differentiated into both neurons and glia. On the basis of nestin staining, many C17.NT3 cells incompletely differentiated or remained quiescent. In the C17.NT-3 grafts, axon ingrowth from the host spinal cord was reported based on neurofilament staining, whereas no axons were found in the control grafts. In a first step to create new synaptic targets for regenerating spinal axons following injury, neural stem cells genetically engineered to overexpress trkC neurotrophin receptor were transplanted in intact rat spinal cord (Castellanos et al., 2002). Combined with in vivo treatment with NT-3, nearly all cells survived. Astrocytic differentiation was reduced and most trkCtransduced cells remain in an undifferentiated state following treatment with NT-3.

Neural Stem Cells

Stem cells are widely investigated for the development of cell therapies in a variety of diseases. Embryonic stem (ES) cells as well as somatic neural stem cells have been used to repair the injured CNS. A neural stem cell has the potential to generate all cell types in the CNS, including neurons, astrocytes and

VIL

F U T U R E PERSPECTIVE

Approximately 2 decades of research by numerous laboratories have resulted in the creation of viral vectors that enable long term, local expression of putative therapeutic proteins in the injured spinal cord.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

FUTURE PERSPECTIVE

These advanced viral vectors (based on adeno-associated virus and lentivirus) are devoid of the immunogenic and toxic side effects observed v\^ith the initially employed vectors (based on herpes virus and adenovirus). The effects of neurotrophic factors secreted by implanted genetically modified cellular implants (fibroblast, Schwann cells, OEGs and peripheral nerve tissue bridges) or overexpressed in neural cells that w^ere transduced by viral vectors directly injected in the spinal cord has nov^ been studied in some detail. Neurotrophic factors are capable of preventing neuronal atrophy, counteracting secondary damage at lesion sites and promoting a certain extend of axon regrowth of injured spinal nerve tracts. Limited functional recovery has been observed foUov^ing neurotrophic factor gene therapy. The development of more effective gene therapy for SCI is now hampered by our limited understanding of the molecular mechanisms that underlie the failure of regeneration. New strategies to identify other factors involved in the neural repair process include the use of microarrays and proteomics (Bonilla et al., 2002; Costigan et al., 2002; Jimenez et al., 2005). These new technologies allow the study of global changes in gene and protein expression. Several laboratories have started comparative analysis of gene and protein expression profiles in injured central and peripheral neurons. Since peripheral neurons do regenerate successfully these studies should reveal key molecular differences between regenerating and nonregenerating neurons (Costigan et al., 2002; Jimenez et al., 2005). We anticipate that these studies will lead to the identification of new targets for gene therapy. It is clear that a single intervention (for instance, gene therapy for a neurotrophic factor) will not be sufficient to repair the spinal cord. A combination of treatments based on our increasing knowledge of the pathophysiology of SCI will be required to promote more significant repair. A therapy will most likely include a combination of growth-promoting molecules either delivered via genetically modified cells or by direct gene transfer as discussed in this review and the neutralization of outgrowth inhibitory factors present in CNS myelin and in the neural scar (Fouad et al., 2001, Hermanns et al., 2001, Bradbury et al., 2002, Niclou et al., 2005). References Agrawal, S.K. and Fehlings, M.G. (1996) Mechanisms of secondary injury to spinal cord axons in vitro: role of N a + , Na(+)K(+)-ATPase, the Na(+)-H+ exchanger, and the Na(+)-Ca2+ exchanger. J. Neurosci., 16: 545-552.

283

Alessandri, B. and Bullock, R. (1998) Glutamate and its receptors in the pathophysiology of brain and spinal cord injuries. Prog. Brain Res., 116: 303-330. Akli, S., Caillaud, C , Yigne, E., Stratford-Perricaudet, L.D., Poenaru, L., Perricaudet, M., Kahn, A. and Peschanski, M.R. (1993) Transfer of a foreign gene into the brain using adenovirus vectors. Nat. Genet., 3: 224-228. Amar, A.P. and Levy, M.L. (1999) Pathogenesis and pharmacological strategies for mitigating secondary damage in acute spinal cord injury Neurosurgery, 44: 1027-1039; discussion 1039-1040. Anderson, PN., Campbell, G., Zhang, Y. and Lieberman, A.R. (1998) Cellular and molecular correlates of the regeneration of adult mammalian CNS axons into peripheral nerve grafts. Prog. Brain Res., 117: 211-232. Bajocchi, G., Feldman, S.H., Crystal, R.G. and Mastrangeli, A. (1993) Direct in vivo gene transfer to ependymal cells in the central nervous system using recombinant adenovirus vectors. Nat. Genet., 3: 229-234. Balentine, J.D. and Spector, M. (1977) Calcification of axons in experimental spinal cord trauma. Ann. Neurol., 2: 520-523. Bamber, N.L, Li, H., Lu, X., Oudega, M., Aebischer, P and Xu, X.M. (2001) Neurotrophins BDNF and NT-3 promote axonal re-entry into the distal host spinal cord through Schwann cell-seeded mini-channels. Eur. J. Neurosci., 13: 257-268. Banik, N.L., Hogan, E.L. and Hsu, C.Y. (1987) The multimolecular cascade of spinal cord injury. Studies on prostanoids, calcium, and proteinases. Neurochem. Pathol., 7: 57-77. Bartholdi, D. and Schwab, M.E. (1995) Methylprednisolone inhibits early inflammatory processes but not ischemic cell death after experimental spinal cord lesion in the rat. Brain Res., 672:177-186. Bartholdi, D. and Schwab, M.E. (1997) Expression of pro-inflammatory cytokine and chemokine mRNA upon experimental spinal cord injury in mouse: an in situ hybridization study. Eur. J. Neurosci., 9:1422-1438. Baumgartner, B.J. and Shine, H.D. (1998) Neuroprotection of spinal motoneurons following targeted transduction with an adenoviral vector carrying the gene for glial cell line-derived neurotrophic factor. Exp. Neurol., 153: 102-112. Blesch, A., Conner, J.M. and Tuszynski, M.H. (2001) Modulation of neuronal survival and axonal growth in vivo by tetracyclineregulated neurotrophin expression.Gene Ther., 8: 954-960. Blits, B. (2002) Cell and gene therapy experimental strategies for spinal cord regeneration. Thesis, Vrije Universiteit Amsterdam, 175p. Blits, B., Boer, G.J. and Verhaagen, J. (2002) Pharmacological, cell, and gene therapy strategies to promote spinal cord regeneration. Cell Transplant., 11: 593-613. Blits, B., Carlstedt, T.P, Ruitenberg, M.J., De Winter, R, Hermens, W.T., Dijkhuizen, PA., Claassens, J.W., Eggers, R., Van Der Sluis, R., Tenenbaum, L., Boer, G.J. and Verhaagen, J. (2004) Rescue and sprouting of motoneurons following ventral root avulsion and reimplantation combined with intraspinal adeno-associated viral vector-mediated expression of glial cell line-derived neurotrophic factor or brain-derived neurotrophic factor. Exp. Neurol., 189: 303-316. Blits, B., Dijkhuizen, PA., Boer, G.J. and Verhaagen, J. (2000) Intercostal nerve implants transduced with an adenoviral vector encoding neurotrophin-3 promote regrowth of injured rat corticospinal tract fibers and improve hind limb function. Exp. Neurol, 164: 25-37. Blits, B., Dijkhuizen, PA., Carlstedt, T.P, Poldervaart, H., Schiemanck, S., Boer, G.J. and Verhaagen, J. (1999) Adenoviral vector-mediated

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

284

21. GENE THERAPY FOR SPINAL CORD INJURY

expression of a foreign gene in peripheral nerve tissue bridges implanted in the injured peripheral and central nervous system. Exp. Neurol., 160: 256-267. Blits, B., Oudega, M., Boer, G.J., Bartlett-Bunge, M. and Verhaagen, J. (2003) Adeno-associated viral vector-mediated neurotrophin gene transfer in the injured adult rat spinal cord improves hindlimb function. Neuroscience, 118: 271-281. Blumer, C.E. and Quine, S. (1995) Prevalence of spinal cord injury: an international comparison. Neuroepidemiology, 14: 258-268. Bonilla, I.E., Tanabe, K. and Strittmatter, S.M. (2002) Small prolinerich repeat protein lA is expressed by axotomized neurons and promotes axonal outgrowth. J. Neurosci., 22: 1303-1315. Boruch, A.V., Conners, J.J., Pipitone, M., Deadwyler, G., Storer, P.D., Devries, G.H. and Jones, K.J. (2001) Neurotrophic and migratory properties of an olfactory ensheathing cell line. Glia, 33: 225-229. Bracken, M.B. (2002) Steroids for acute spinal cord injury. Cochrane Database Syst. Rev., CD001046. Bracken, M.B. and Holford, T.R. (2002) Neurological and functional status 1 year after acute spinal cord injury: estimates of functional recovery in National Acute Spinal Cord Injury Study II from results modelled in National Acute Spinal Cord Injury Study III. J. Neurosurg., 96: 259-266. Bradbury, E.J., Moon, L.D., Popat, R.J., King, V.R., Bennett, G.S., Patel, PN., Fawcett, J.W. and McMahon, S.B. (2002) Chondroitinase ABC promotes functional recovery after spinal cord injury. Nature, 416: 636-640. Braughler, J.M. and Hall, E.D. (1992) Involvement of lipid peroxidation in CNS injury. J. Neurotrauma, 9(Suppl 1): S1-S7. Bregman, B.S. and Reier, P.J. (1986) Neural tissue transplants rescue axotomized rubrospinal cells from retrograde death. J. Comp. Neurol., 244: 86-95. Bregman, B.S., McAtee, M., Dai, H.N. and Kuhn, PL. (1997) Neurotrophic factors increase axonal growth after spinal cord injury and transplantation in the adult rat. Exp. Neurol., 148: 475-494. Brewer, K.L., Bethea, J.R. and Yezierski, R.P (1999) Neuroprotective effects of interleukin-10 following excitotoxic spinal cord injury. Exp. Neurol., 159: 484-493. Byrnes, A.P, MacLaren, R.E. and Charlton, H.M. (1996) Immunological instability of persistent adenovirus vectors in the brain: peripheral exposure to vector leads to renewed inflammation, reduced gene expression, and demyelination. J. Neurosci., 16: 3045-3055. Byrnes, A.P, Rusby J.E., Wood, M.J. and Charlton, H.M. (1995) Adenovirus gene transfer causes inflammation in the brain. Neuroscience, 66: 1015-1024. Cao, L., Liu, L., Chen, Z.Y., Wang, L.M., Ye, J.L., Qiu, H.Y, Lu, C.L. and He, C. (2004) Olfactory ensheathing cells genetically modified to secrete GDNF to promote spinal cord repair. Brain, 127: 535-549 Carlson, S.L., Parrish, M.E., Springer, J.E., Doty, K. and Dossett, L. (1998) Acute inflammatory response in spinal cord following impact injury. Exp. Neurol., 151: 77-88. Castellanos, D.A., Tsoulfas, P, Frydel, B.R., Gajavelli, S., Bes, J.C. and Sagen, J. (2002) TrkC overexpression enhances survival and migration of neural stem cell transplants in the rat spinal cord. Cell. Transplant., 11: 297-307. Cheng, H., Cao, Y. and Olson, L. (1996) Spinal cord repair in adult paraplegic rats: partial restoration of hind limb function. Science, 273: 510-513. Choi, D.W (1992) Excitotoxic ceU death. J. NeurobioL, 23:1261-1276. Costigan, M., Befort, K., Karchewski, L., Griffin, R.S., D'Urso, D., Allchome, A., Sitarski, J., Mannion, J.W, Pratt, R.E. and Woolf, C.J.

(2002) Replicate high-density rat genome oligonucleotide microarrays reveal hundreds of regulated genes in the dorsal root ganglion after peripheral nerve injury BMC Neurosci., 3:16. Crowe, M.J., Bresnahan, J.C, Shuman, S.L., Masters, J.N. and Beattie, M.S. (1997) Apoptosis and delayed degeneration after spinal cord injury in rats and monkeys. Nat. Med., 3: 73-76. David, S. and Aguayo, A.J. (1981) Axonal elongation into peripheral nervous system ''bridges'' after central nervous system injury in adult rats. Science, 214: 931-933. Davidson, B.L., Allen, E.D., Kozarsky, K.F., Wilson, J.M.and Roessler, B.J. (1993) A model system for in vivo gene transfer into the central nervous system using an adenoviral vector. Nat. Genet., 3, 219-223. De Winter, R, Oudega, M., Lankhorst, A.J., Hamers, E P , Blits, B., Ruitenberg, M.J., Pasterkamp, R.J., Gispen, W.H. and Verhaagen, J. (2002) Injury-induced class 3 semaphorin expression in the rat spinal cord. Exp. Neurol., 175: 61-75. Dijkhuizen, PA., Hermens, W.T., Teunis, M.A. and Verhaagen, J. (1997) Adenoviral vector-directed expression of neurotrophin-3 in rat dorsal root ganglion explants results in a robust neurite outgrowth response. J. NeurobioL, 33: 172-184. Dijkhuizen, PA., Pasterkamp, R.J., Hermens, W.T., de Winter, R, Giger, R.J. and Verhaagen, J. (1998) Adenoviral vector-mediated gene delivery to injured rat peripheral nerve. J. Neurotrauma, 15: 387-397. Faden, A.I. and Simon, R.P. (1988) A potential role for excitotoxins in the pathophysiology of spinal cord injury. Ann. Neurol., 23: 623-626. Fawcett, J.W and Asher, R.A. (1999) The glial scar and central nervous system repair. Brain Res. Bull., 49: 377-391. Federoff, H.J., Geschwind, M.D., Geller, A.I. and Kessler, J.A. (1992) Expression of nerve growth factor in vivo from a defective herpes simplex virus 1 vector prevents effects of axotomy on sympathetic ganglia. Proc. Nafl. Acad. Sci. USA, 89:1636-1640. Feldblum, S., Amaud, S., Simon, M., Rabin, O. and D'Arbigny, P. (2000) Efficacy of a new neuroprotective agent, gacyclidine, in a model of rat spinal cord injury J. Neurotrauma, 17:1079-1093. Fouad, K., Dietz, V. and Schwab, M.E. (2001) Improving axonal growth and functional recovery after experimental spinal cord injury by neutralizing myelin associated inhibitors. Brain Res. Rev, 36: 204-212. Gaviria, M., Privat, A., d'Arbigny, P., Kamenka, J.M., Haton, H. and Ohanna, F. (2000a) Neuroprotective effects of gacyclidine after experimental photochemical spinal cord lesion in adult rats: dosewindow and time-window effects. J. Neurotrauma, 17:19-30. Gaviria, M., Privat, A., d'Arbigny, P , Kamenka, J., Haton, H. and Ohanna, F. (2000b) Neuroprotective effects of a novel NMDA antagonist, Gacyclidine, after experimental contusive spinal cord injury in adult rats. Brain Res., 874: 200-209 Gravel, C , Gotz, R., Lorrain, A. and Sendtner, M. (1997) Adenoviral gene transfer of ciliary neurotrophic factor and brain-derived neurotrophic factor leads to long-term survival of axotomized motor neurons. Nat. Med., 3: 765-770. Grill, R., Murai, K., Blesch, A., Gage, RH. and Tuszynski, M.H. (1997a) Cellular delivery of neurotrophin-3 promotes corticospinal axonal growth and partial functional recovery after spinal cord injury. J. Neurosci., 17: 5560-5572. Grill, R.J., Blesch, A. and Tuszynski, M.H. (1997b) Robust growth of chronically injured spinal cord axons induced by grafts of genetically modified NGF-secreting cells. Exp. Neurol., 148: 444-452. Gruner, J.A. (1992) A monitored contusion model of spinal cord injury in the rat. J. Neurotrauma, 9:123-126; discussion 126-128.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

FUTURE PERSPECTIVE Haghighi, S.S., Agrawal, S.K., Surdell, D. Jr., Plambeck, R., Agrawal, S., Johnson, G.C. and Walker, A. (2000) Effects of methylprednisolone and MK-801 on functional recovery after experimental chronic spinal cord injury. Spinal Cord, 38: 733-740. Haghighi, S.S., Johnson, G.C, de Vergel, C.F. and Vergel Rivas, B.J. (1996) Pretreatment with NMDA receptor antagonist MK801 improves neurophysiological outcome after an acute spinal cord injury Neurol. Res., 18: 509-515. Hall, E.D. (2001) Pharmacological treatment of acute spinal cord injury: how do we build on past success? J. Spinal Cord Med., 24: 142-146. Hall, E.D. and Braughler, J.M. (1982) Effects of intravenous methylprednisolone on spinal cord lipid peroxidation and N a + + K + ATPase activity. Dose-response analysis during 1st hour after contusion injury in the cat. J. Neurosurg., 51:1^-15?>. Hermann, G.E., Rogers, R.C., Bresnahan, J.C. and Beattie, M.S. (2001) Tumor necrosis factor-alpha induces cFOS and strongly potentiates glutamate-mediated cell death in the rat spinal cord. Neurobiol. Dis., 8: 590-599. Hermanns, S., Klapka, N. and Muller, H.W. (2001) The collagenous lesion scar - an obstacle for axonal regeneration in brain and spinal cord injury. Restor. Neurol. Neurosci., 19:139-148. Hermens, W.T., Giger, R.J., Holtmaat, A.J., Dijkhuizen, P.A., Houweling, D.A. and Verhaagen, J. (1997) Transient gene transfer to neurons and glia: analysis of adenoviral vector performance in the CNS and PNS. J. Neurosci. Methods., 71: 85-98. Hermens, W.T. and Verhaagen, J. (1998) Viral vectors, tools for gene transfer in the nervous system. Prog. Neurobiol., 55: 399^32. Huber, A.B., Ehrengruber, M.U., Schwab, M.E. and Brosamle, C. (2000) Adenoviral gene transfer to the injured spinal cord of the adult rat. Eur. J. Neurosci., 12: 3437-3442. Hurlbert, R.J. (2000) Methylprednisolone for acute spinal cord injury: an inappropriate standard of care. J. Neurosurg., 93: 1-7. Ide, C. (1996) Peripheral nerve regeneration. Neurosci. Res., 25: 101-121. lizuka, H., Yamamoto, H., Iwasaki, Y., Yamamoto, T. and Konno, H. (1987) Evolution of tissue damage in compressive spinal cord injury in rats. J. Neurosurg., 66: 595-603. Jimenez, C.R., Stam, F.J., Li, K.W., Gouwenberg, Y, Homshaw, M.P., De Winter, R, Verhaagen, J. and Smit, A.B. (2005) Proteomics of the injured rat sciatic nerve reveals protein expression dynamics during regeneration. Mol. Cell Proteomics, 4: 120-132. Jones, N. and Shenk, T. (1979) An adenovirus type 5 early gene function regulates expression of other early viral genes. Proc. Natl. Acad. Sci. USA, 76: 3665-3669. Kajiwara, K., Bymes, A.P, Charlton, H.H., Wood, M.J., Wood, K.J. (1997) Immune responses to adenoviral vectors during gene transfer in the brain. Hum. Gene Ther., 8: 253-265. Kaplitt, M.G., Leone, R, Samulski, R.J., Xiao, X., Pfaff, D.W., O'Malley, K.L. and During, M.J. (1994) Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat. Genet., 8:148-154. KapHtt, M.G., Pfaus, J.G., Kleopoulos, S.R, Hanlon, B.A., Rabkin, S.D. and Pfaff, D.W (1991) Expression of a functional foreign gene in adult mammalian brain following in vivo transfer via a herpes simplex virus type 1 defective viral vector. Mol. Cell. Neurosci., 2: 320-330. Keir, S.D., Mitchell, W.J., Feldman, L.T. and Martin, J.R. (1995) Targeting and gene expression in spinal cord motor neurons following intramuscular inoculation of an HSV-1 vector. J. NeuroviroL, 1: 259-267.

285

Kennedy, PG. (1997) Potential use of herpes simplex virus (HSV) vectors for gene therapy of neurological disorders. Brain, 120: 1245-1259. Kim, D.H., Gutin, PH., Noble, L.J., Nathan, D., Yu, J.S. and Nockels, R.P. (1996) Treatment with genetically engineered fibroblasts producing NGF or BDNF can accelerate recovery from traumatic spinal cord injury in the adult rat. Neuroreport, 7: 2221-2225. Kim, D., Schallert, T., Liu, Y, Browarak, T., Nayeri, N., Tessler, A., Fischer, I. and Murray, M. (2001) Transplantation of genetically modified fibroblasts expressing BDNF in adult rats with a subtotal hemisection improves specific motor and sensory functions. Neurorehab. Neural Re., 15:141-150. Klusman, I. and Schwab, M.E. (1997) Effects of pro-inflammatory cytokines in experimental spinal cord injury Brain Res., 762:173-184. Koshinaga, M. and Whittemore, S.R. (1995) The temporal and spatial activation of microglia in fiber tracts undergoing anterograde and retrograde degeneration following spinal cord lesion. J. Neurotrauma, 12: 209-222. Krikorian, J.G., Guth, L. and Donati, E.J. (1981) Origin of the connective tissue scar in the transected rat spinal cord. Exp. Neurol., 72: 698-707. Kwon, B.K., Oxland, T.R. and Tetzlaff, W (2002) Animal models used in spinal cord regeneration research. Spine, 27: 1504-1510. Lakatos, A., Barnett, S.C. and Franklin, R.J. (2003) Olfactory ensheathing cells induce less host astrocyte response and chondroitin sulphate proteoglycan expression than Schwann cells following transplantation into adult CNS white matter. Lxp. Neurol, 184: 237-246. Lacroix, S. and Tuszjmski, M.H. (2000) Neurotrophic factors and gene therapy in spinal cord injury. Neurorehab. Neural Re., 14: 265-275. Lakatos, A., Franklin, R.J. and Barnett, S.C. (2000) Olfactory ensheathing cells and Schwann cells differ in their in vitro interactions with astrocytes. Glia, 32: 214-225. Lee, Y.B., Yune, T.Y, Baik, S.Y, Shin, YH., Du, S., Rhim, H., Lee, E.B., Kim, Y.C., Shin, M.L., Markelonis, G.J. and Oh, T.H. (2000) Role of tumor necrosis factor-alpha in neuronal and glial apoptosis after spinal cord injury. Exp. Neurol., 166,190-195. Le Gal La Salle, G., Robert, J.J., Berrard, S., Ridoux, V., StratfordPerricaudet, L.D., Perricaudet, M. and Mallet, J. (1993) An adenovirus vector for gene transfer into neurons and glia in the brain. Science, 259: 988-990. Leib, D.A. and Olivo, PD. (1993) Gene delivery to neurons: is herpes simplex virus the right tool for the job? Bioessays, 15: 547-554. Leskovar, A., Moriarty, L.J., Turek, J.J., Schoenlein, LA. and Borgens, R.B. (2000) The macrophage in acute neural injury: changes in cell numbers over time and levels of cytokine production in mammalian central and peripheral nervous systems. J. Exp. Biol, 203:1783-1795. Levi, A.D., Tator, C.H. and Bunge, R.P (1996) Clinical syndromes associated with disproportionate weakness of the upper versus the lower extremities after cervical spinal cord injury. Neurosurgery, 38:179-183; discussion 183-175. Leybaert, L. and De Hemptirme, A. (1996) Changes of intracellular free calcium following mechanical injury in a spinal cord slice preparation. Exp. Brain Res., 112: 392-402 Li, Y, Field, P.M. and Raisman, G. (1997) Repair of adult rat corticospinal tract by transplants of olfactory ensheathing cells. Science, 277: 2000-2002. Li, Y, Field, RM. and Raisman, G. (1998) Regeneration of adult rat corticospinal axons induced by transplanted olfactory ensheathing cells. J. Neurosci., 18:10514-10524.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

286

21. GENE THERAPY FOR SPINAL CORD INJURY

Lilley, C.E., Groutsi, R, Han, Z., Palmer, J.A., Anderson, P.N., Latchman, D.S. and Coffin, R.S. (2001) Multiple immediate-early genedeficient herpes simplex virus vectors allowing efficient gene delivery to neurons in culture and widespread gene delivery to the central nervous system in vivo. J. Virol, 75: 4343-4356. Liu, Y, Himes, B.T., Murray, M., Tessler, A. and Fischer, I. (2002) Grafts of BDNF-producing fibroblasts rescue axotomized rubrospinal neurons and prevent their atrophy. Exp. Neurol., 178: 150-164. Liu, Y, Himes, B.T., Solowska, J., Moul, J., Chow, S.Y, Park, K.I., Tessler, A., Murray M., Snyder, E.Y and Fischer, I. (1999a) Intraspinal delivery of neurotrophin-3 using neural stem ceUs genetically modified by recombinant retrovirus. Exp. Neurol., 158: 9-26. Liu Y, Himes B.T., Tryon B., Moul J., Chow S.Y, Jin, H., Murray, M., Tessler, A. and Fischer, 1. (1998) Intraspinal grafting of fibroblasts genetically modified by recombinant adenoviruses. Neuroreport, 9: 1075-1079. Liu, Y, Kim, D., Himes, B.T., Chow, S.Y, SchaUert, T., Murray M., Tessler, A. and Fischer, I. (1999b) Transplants of fibroblasts genetically modified to express BDNF promote regeneration of adult rat rubrospinal axons and recovery of forelimb function. J. Neurosci., 19: 4370-4387. Liu, D., Thangnipon, W. and McAdoo, D.J. (1991) Excitatory amino acids rise to toxic levels upon impact injury to the rat spinal cord. Brain Res., 547: 344-348. Liu, S., Qu, Y, Stewart, T.J., Howard, M.J., Chakrabortty, S., Holekamp, T.F. and McDonald, J.W. (2000) Embryonic stem cells differentiate into oligodendrocytes and myelinate in culture and after spinal cord transplantation. Proc. Nati. Acad. Sci. USA, 97: 6126-6131. Liu, X.Z., Xu, X.M., Hu, R., Du, C , Zhang, S.X., McDonald, J.W., Dong, H.X., Wu, Y.J., Fan, G.S., Jacquin, M.R, Hsu, C.Y and Choi, D.W. (1997) Neuronal and glial apoptosis after traumatic spinal cord injury. J. Neurosci., 17: 5395-5406. Mata, M., Zhang, M., Hu, X. and Fink, D.J. (2001) HveC (nectin-1) is expressed at high levels in sensory neurons, but not in motor neurons, of the rat peripheral nervous system. J. NeuroviroL, 7: 476-480. Matthews, M.A., St Onge, M.R and Faciane, C.L. (1979) An electron microscopic analysis of abnormal ependymal cell proliferation and envelopment of sprouting axons following spinal cord transection in the rat. Acta. Neuropathol. (Berl.), 45: 27-36. McDonald, J.W. (1999) Repairing the damaged spinal cord. Sci. Am., 281: 64-73. McDonald, J.W., Liu, X.Z., Qu, Y, Liu, S., Mickey S.K., Turetsky D., Gottlieb, D.I. and Choi, D.W. (1999) Transplanted embryonic stem cells survive, differentiate and promote recovery in injured rat spinal cord. Nat. Med., 5: 1410-1412. McDonald, J.W. and Sadowsky, C. (2002) Spinal-cord injury. Lancet, 359: 417-425. McTigue, D.M., Homer, PJ., Stokes, B.T. and Gage, RH. (1998) Neurotrophin-3 and brain-derived neurotrophic factor induce oligodendrocyte proliferation and myelination of regenerating axons in the contused adult rat spinal cord. J. Neurosci., 18: 5354-5365. Menei, P., Montero-Menei, C , Whittemore, S.R., Bunge, R.P and Bunge, M.B. (1998) Schwann cells genetically modified to secrete human BDNF promote enhanced axonal regrowth across transected adult rat spinal cord. Eur. J. Neurosci., 10: 607-621. Metz, G.A., Curt, A., van de Meent, H., Klusman, I., Schwab, M.E. and Dietz, V. (2000) Validation of the weight-drop contusion model in rats: a comparative study of human spinal cord injury. J. Neurotrauma, 17: 1-17.

Mu, X., Azbill, R.D. and Springer, J.E. (2002) NBQX treatinent improves mitochondrial function and reduces oxidative events after spinal cord injury. J. Neurotrauma, 19: 917-927. Muir, G.D. and Webb, A.A. (2000) Mini-review: assessment of behavioural recovery following spinal cord injury in rats. Eur. J. Neurosci., 12: 3079-3086. Naldini, L., Blomer, U., Gallay, P , Ory, D., Mulligan, R., Gage, RH., Verma, I.M. and Trono, D. (1996) In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science, 272: 263-267. Nathan, P.W. (1994) Effects on movement of surgical incisions into the human spinal cord. Brain, 117: 337-346. Natsume, A., Mata, M., Wolfe, D., Oligino, T., Goss, J., Huang, S., Glorioso, J. and Fink, J. (2002) Bcl-2 and GDNF delivered by HSV-mediated gene transfer after spinal root avulsion provide a synergistic effect. J. Neurotrauma,. 19: 61-68. Nesic, O., Xu, G.Y, McAdoo, D., High, K.W, Hulsebosch, C. and Perez-Pol, R. (2001) IL-1 receptor antagonist prevents apoptosis and caspase-3 activation after spinal cord injury. J. Neurotrauma, 18: 947-956. Nevins, J.R. (1981) Mechanism of activation of early viral transcription by the adenovirus ElA gene product. Cell, 26: 213-220. Niclou, S.P, Ehlert, E.M. and Verhaagen, J. (2005) Chemorepellent axon guidance molecules in spinal cord injury. J. Neurotrauma, In press. Oudega, M., Vargas, C.G., Weber, A.B., Kleitman, N. and Bunge, M.B. (1999) Long-term effects of methylprednisolone following transection of adult rat spinal cord. Eur. J. Neurosci., 11: 2453-2464. Ranter, S.S., Yum, S.W and Faden, A.I. (1990) Alteration in extracellular amino acids after traumatic spinal cord injury. Ann. Neurol., 27: 96-99. Pasterkamp, R.J., Verhaagen, J. (2001) Emerging roles for semaphorins in neural regeneration. Brain. Res. Rev., 35: 36-54. Pasterkamp, R.J., Anderson, P.N. and Verhaagen, J. (2001) Peripheral nerve injury fails to induce growth of lesioned ascending dorsal column axons into spinal cord scar tissue expressing the axon repellent Semaphorin3A. Eur. J. Neurosci., 13: 457-471. Pasterkamp, R.J., Giger, R.J., Ruitenberg, M.J., Holtmaat, A.J., De Wit, J., De Winter, F. and Verhaagen, J. (1999) Expression of the gene encoding the chemorepellent semaphorin III is induced in the fibroblast component of neural scar tissue formed following injuries of adult but not neonatal CNS. Mol. Cell. Neurosci., 13: 143-166. Peel, A.L., Zolotukhin, S., Schrimsher, G.W., Muzyczka, N. and Reier, P.J. (1997) Efficient transduction of green fluorescent protein in spinal cord neurons using adeno-associated virus vectors containing cell type-specific promoters. Gene. Then, 4: 16-24. Popovich, RG., Streit, WJ. and Stokes, B.T. (1993) Differential expression of MHC class II antigen in the contused rat spinal cord. J. Neurotrauma, 10: 37-46. Plant, G.W, Bates, M.L. and Bunge, M.B. (2001) Inhibitory proteoglycan immunoreactivity is higher at the caudal than the rostral Schwann cell graft-transected spinal cord interface. Mol. Cell. Neurosci., 17: 471^87. Ramon-Cueto, A., Cordero, M.L, Santos-Benito, F.R and Avila, J. (2000) Functional recovery of paraplegic rats and motor axon regeneration in their spinal cords by olfactory ensheathing glia. Neuron, 25: 425-435. Ramon-Cueto, A. and Nieto-Sampedro, M. (1994) Regeneration into the spinal cord of transected dorsal root axons is promoted by ensheathing gha taransplants. Exp. Neurol., 127: 232-244.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

FUTURE PERSPECTIVE Ramon-Cueto, A., Plant, G.W., Avila, J. and Bunge, M.B. (1998) Long-distance axonal regeneration in the transected adult rat spinal cord is promoted by olfactory ensheathing glia transplants. J. Neurosci., 18: 3803-3815. Richardson, RM., McGuinness, U.M. and Aguayo, A.J. (1980) Axons from CNS neurons regenerate into PNS grafts. Nature, 284: 264-265. Romero, M.I., Rangappa, N., Garry, M.G. and Smith, G.M. (2001) Functional regeneration of chronically injured sensory afferents into adult spinal cord after neurotrophin gene therapy. J. Neurosci., 21: 8408-8416. Romero, M.I., Rangappa, N., Li, L., Lightfoot, E., Garry, M.G. and Smith, G.M. (2000) Extensive sprouting of sensory afferents and hyperalgesia induced by conditional expression of nerve growth factor in the adult spinal cord. J. Neurosci., 20: 4435-4445. Ruitenberg, M.J., Blits, B., Dijkhuizen, RA., te Beek, E.T., Bakker, A., van Heerikhuize, J.J., Pool, C.W., Hermens, W.T., Boer, G.J. and Verhaagen, J. (2004) Adeno-associated viral vector-mediated gene transfer of brain-derived neurotrophic factor reverses atrophy of rubrospinal neurons following both acute and chronic spinal cord injury. Neurobiol. Dis., 15: 394-406. Ruitenberg, M.J., Eggers, R., Boer, G.J. and Verhaagen, J. (2002a) Adeno-associated viral vectors as agents for gene delivery: application in disorders and trauma of the central nervous system. Methods, 28: 182—194, Review. Ruitenberg, M.J., Levison, D.B., Lee, S.V., Verhaagen, J., Harvey, A.R. and Plant, G.W. (2005) NT-3 expression from engineered olfactory ensheathing glia promotes spinal sparing and regeneration. Brain: 128: 839-853. Ruitenberg, M.J., Plant, G.W., Christensen, C.L., Blits, B., Niclou, S.P, Harvey, A.R., Boer, G.J. and Verhaagen, J. (2002b) Viral vectormediated gene expression in olfactory ensheathing glia implants in the lesioned rat spinal cord. Gene Ther., 9:135-146. Ruitenberg, M.J., Plant, G.W., Hamers, F.P, Wortel, J., Blits, B., Dijkhuizen, P.A., Gispen, W.H., Boer, G.J. and Verhaagen, J. (2003) Ex vivo adenoviral vector-mediated neurotrophin gene transfer to olfactory ensheathing glia: effects on rubrospinal tract regeneration, lesion size, and functional recovery after implantation in the injured rat spinal cord. J. Neurosci., 23: 7045-7058. Schwab, M.E. and Bartholdi, D. (1996) Degeneration and regeneration of axons in the lesioned spinal cord. Physiol. Rev., 76: 319-370. Shearer, M.G. and Fawcett, J.W. (2001) The astrocyte/meningeal cell interface - a barrier to successful nerve regeneration? Cell Tissue Res., 305: 267-273. Shuman, S.L., Bresnahan, J.C. and Beattie, M.S. (1997) Apoptosis of microglia and oligodendrocytes after spinal cord contusion in rats. J. Neurosci. Res., 50: 798-808. Smith, G.M., Hale, J., Pasnikowski, E.M., Lindsay, R.M., Wong, V., Rudge, J.S. (1996) Astrocytes infected with replication-defective adenovirus containing a secreted form of CNTF of NT3 show enhanced support of neuronal populations in vitro. Exp. Neurol., 139:156-166. Snyder, E.Y. and Senut, M.C. (1997) The use of normeuronal cells for gene delivery. Neurobiol. Dis., 4: 69-102. Steward, O., Zheng, B., and Tessier-Lavigne, M. (2003) False resurrections: distinguishing regenerated from spared axons in the injured central nervous system. J. Comp. Neurol., 459:1-8. Stichel, C.C, Hermarms, S., Luhmann, H.J., Lausberg, R, Niermann, H., D'Urso, D., Servos, G., Hartwig, H.G. and MuUer, H.W. (1999) Inhibition of collagen IV deposition promotes regeneration of injured CNS axons. Eur. J. Neurosci., 11: 632-646.

287

Streit, W.J., Semple-Rowland, S.L., Hurley, S.D., Miller, R.C., Popovich, PG. and Stokes, B.T. (1998) Cytokine mRNA profiles in contused spinal cord and axotomized facial nucleus suggest a beneficial role for inflammation and gliosis. Exp. Neurol., 152: 74-87. Teng, YD., Lavik, E.B., Qu, X., Park, K.I., Ourediuk, J., Zurakowski, D., Langer, R. and Snyder, E.Y. (2002) Functional recovery following traumatic spinal cord injury mediated by a unique polymer scaffold seeded with neural stem cells. Proc. Natl. Acad. Sci. USA, 99: 3024-3029. Tang, X.Q., Tanelian, D.L. and Smith, G.M. (2004) Semaphorin3A inhibits nerve growth factor-induced sprouting of nociceptive afferents in adult rat spinal cord. J. Neurosci. 24: 819-827. Tator, C.H. and Fehlings, M.G. (1991) Review of the secondary injury theory of acute spinal cord trauma with emphasis on vascular mechanisms. J. Neurosurg., 15'. 15-26. Terada, H., Kazui, T., Takinami, M., Yamashita, K., Washiyama, N. and Muhammad, B.A. (2001) Reduction of ischemic spinal cord injury by dextrorphan: comparison of several methods of administration. J. Thorac. Cardiovasc. Surg., 122: 979-985. Topsakal, C , Erol, RS., Ozveren, M.R, Yilmaz, N. and Ilhan, N. (2002) Effects of methylprednisolone and dextromethorphan on lipid peroxidation in an experimental model of spinal cord injury. Neurosurg. Rev, 25: 258-266. Tusz5mski, M.H., Gabriel, K., Gage, F.H., Suhr, S., Meyer, S. and Rosetti, A. (1996) Nerve growth factor delivery by gene transfer induces differential outgrowth of sensory, motor, and noradrenergic neurites after adult spinal cord injury. Exp. Neurol., 137: 157-173. Tuszynski, M.H., Murai, K., Blesch, A., Grill, R. and Miller, I. (1997) Functional characterization of NGF-secreting cell grafts to the acutely injured spinal cord. Cell Transplant., 6: 361-368. Tuszynski, M.H., Peterson, D.A., Ray, J., Baird, A., Nakahara, Y and Gage, EH. (1994) Fibroblasts genetically modified to produce nerve growth factor induce robust neuritic ingrowth after grafting to the spinal cord. Exp. Neurol., 126:1-14. Tuszynski, M.H., Weidner, N., McCormack, M., Miller, I., Powell, H. and Conner, J. (1998) Grafts of genetically modified Schwann cells to the spinal cord: survival, axon growth, and myelination. Cell Transplant., 7:187-196. Van Asbeck, RW., Post, M.W. and Pangalila, R.R (2000) An epidemiological description of spinal cord injuries in The Netherlands in 1994. Spinal Cord, 38: 420-424. Vercelli, A., Repici, M., Garbossa, D. and Grimaldi, A. (2000) Recent techniques for tracing pathways in the central nervous system of developing and adult mammals. Brain Res. Bull., 51:11-28. Wada, S., Yone, K., Ishidou, Y, Nagamine, T, Nakahara, S., Niiyama, T. and Sakou, T. (1999) Apoptosis following spinal cord injury in rats and preventative effect of N-methyl-D-aspartate receptor antagonist. J. Neurosurg., 91: 98-104. Watanabe, T., Yamamoto, T, Abe, Y, Saito, N., Kumagai, T. and Kayama, H. (1999) Differential activation of microglia after experimental spinal cord injury. J. Neurotrauma, 16: 255-265. Weidner, N., Blesch, A., Grill, R.J. and Tuszynski, M.H. (1999) Nerve growth factor-hypersecreting Schwann cell grafts augment and guide spinal cord axonal growth and remyelinate central nervous system axons in a phenotypically appropriate manner that correlates with expression of LI. J. Comp. Neurol., 413: 495-506. Williams, S.K., Franklin, R.J., Bemett, S.C. (2004) Response of olfactory ensheeting cells to the degeneration and regeneration of the peripheral olfactory system and the involvement of the neuroregulins. J. Comp. Neurol, 470: 50-62.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

288

21. GENE THERAPY FOR SPINAL CORD INJURY

Wood, M.J., Byrnes, A.R, Pfaff, D.W., Rabkin, S.D. and Charlton, H.M. (1994a) Inflammatory effects of gene transfer into the CNS with defective HSV-1 vectors. Gene. Ther., 1: 283-291. Wood, M.J., Byrnes, A.R, Rabkin, S.D., Rfaff, D.W and Charlton, H.M. (1994b) Immunological consequences of HSV-1-mediated gene transfer into the CNS. Gene. Ther., l(Suppl. 1): S82. Wood, M.J., Charlton, H.M., Wood, K.J., Kajiwara, K. and Byrnes, A.R (1996) Immune responses to adenovirus vectors in the nervous system. Trends Neurosci., 19: 497-501. Xu, J., Fan, G., Chen, S., Wu, Y, Xu, X.M. and Hsu, C.Y. (1998) Methylprednisolone inhibition of TNF-alpha expression and NF-kB activation after spinal cord injury in rats. Brain Res. Mol. Brain Res., 59:135-142. Xu, X.M., Guenard, V., Kleitman, N., Aebischer, P. and Bunge, M.B. (1995) A combination of BDNF and NT-3 promotes supraspinal axonal regeneration into Schwann cell grafts in adult rat thoracic spinal cord. Exp. Neurol, 134: 261-272.

Yamada, M., Natsume, A., Mata, M., Oligino, T., Goss, J., Glorioso, J. and Fink, D.J. (2001) Herpes simplex virus vector-mediated expression of Bcl-2 protects spinal motor neurons from degeneration following root avulsion. Exp. Neurol., 168: 225-230. Young, W (1992) Role of calcium in central nervous system injuries. J. Neurotrauma., 9(Suppl. 1): S9-S25. Zhang, Y, Dijkhuizen, P.A., Anderson, P.N., Lieberman, A.R. and Verhaagen, J. (1998) NT-3 delivered by an adenoviral vector induces injured dorsal root axons to regenerate into the spinal cord of adult rats. J. Neurosci. Res., 54: 554-562. Zhang, Y and Schneider, R.J. (1994) Adenovirus inhibition of cell translation facilitates release of virus particles and enhances degradation of the cytokeratin network. J. Virol., 68: 2544-2555. Zhou, L., Baumgartner, B.J., Hill-Felberg, S.J., McGowen, L.R. and Shine, H.D. (2003) Neurotrophin-3 expressed in situ induces axonal plasticity in the adult injured spinal cord. J. Neurosci., 23: 1424-1431.

IV. GENE THERAPY FOR PAIN AND SPINAL CORD DISEASES

C H A P T E R

22 Prodrug-Activation Gene Therapy Kaveh Asadi-Moghaddarriy E. Antonio Chiocca

Abstract: The gene therapy approach most commonly used in clinical trials for brain tumors is gene-directed enzyme-prodrug therapy also known as suicide gene therapy. This approach is comprised of three components; the prodrug to be activated, the enzyme used for activation, and the delivery system for the corresponding gene (Anderson, 2000). With this strategy, the systemically administered prodrug is ideally converted to the active chemotherapeutic agent only in cancer cells, thereby allowing a maximal therapeutic effect while limiting systemic toxicity. Several suicide gene therapy approaches are being explored: (i) herpes simplex virus type 1 thymidine kinase/ganciclovir; (ii) cytosine deaminase/5-fluorocytosine; (iii) cytochrome P450/cyclophosphamide or ifosfamide; (iv) guanine phosphoribosyl-transferase/6-thioxantine; (v) nitroreductase/CB1954; (vi) carboxylesterase/CPT-11; (vii) Escherichia coli purine nucleoside phosphorylase/purine analogs. Keywords: suicide gene therapy; HSV thymidine kinase; XGPRT; PNP; cytosine deaminase; cytochrome P450; nitroreductase; carboxylesterase

L

an acyclic analog of the natural nucleoside 2'-deoxyguanosine (Faulds and Heel, 1990) and is a specific substrate of the HSVtk, which is multiple magnitudes more efficient than human nucleoside kinase at monophosphorylating GCV (see Fig. 1) (Elion et al., 1977). The resulting GCV-monophosphate is then converted by cellular kinases into the toxic GCV-triphosphate. GCV-triphosphate's structural resemblance to 2'-deoxyguanosine triphosphate makes it a substrate for DNA polymerase. Once bound to DNA polymerase, GCVtriphosphate inhibits the enzyme or is incorporated into DNA, causing DNA chain elongation to terminate. This causes cell death by inhibition of incorporation of dGTP into DNA, and also by prevention of chain elongation (Mesnil and Yamasaki, 2000). GCV metabolites target replicating cells much like the S-phase specific chemotherapeutics. Glioma cells, which were transduced and selected to express HSVtk were 5000 times more sensitive to GCV than nontransduced cells (Shewach et al., 1994), and

HERPES SIMPLEX VIRUS TYPE 1 THYMIDINE KINASE (HSVtk)/ GANCICLOVIR

The most widely studied suicide gene strategy is the herpes simplex virus thymidine kinase (HSVtk) enzyme approach. The HSVtk system was developed in 1986 (Moolten, 1986) and was the first approach used in patients with malignant brain tumors in 1992 (Oldfield et al., 1993). This approach has been conducted in combination with guanosine-based prodrugs, such as ganciclovir and acyclovir, originally developed as antiviral agents (De Clercq, 2000). HSVtk converts these nontoxic nucleoside analog prodrugs into phosphorylated compounds. Consequently, these compounds directly inhibit DNA polymerase and render the formed DNA molecule unstable, leading to DNA synthesis arrest and cell death. The most commonly used prodrug is ganciclovir (GCV). GCV is

Gene Therapy of the Central Nervous System: From Bench to Bedside

291

Copyright© 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

293

HERPES SIMPLEX VIRUS TYPE 1 THYMIDINE KINASE (HSVtk)/GANCICLOVIR

TABLE 1 Selection of Closed HSVtk/GCV Gene Therapy Trials in Human Brain Tumors Phase

Tumor type

Application

Vector

Patients, N

Median survival

I

Recurrent GBM or metastasis

Stereotactic injection into tumor of RV-producer cells

RV

15

8 mo[Ram et aL, 1997 ]

I

Recurrent GBM

Stereotactic injection of adenoviral vector

AV

13

4 mo[Trask et aL, 2000]

I/II

Recurrent GBM

Ommaya reservoir injection of RV-producer cells

RV

30

8 mo[Pardos et a l , 2003]

I/II

Recurrent GBM

Freehand injection into resected tumor cavity of RV-producer cells

RV

12

7 mo[Klatzmann et aL, 1998]

I/II

Recurrent GBM

Freehand injection into resected tumor cavity of RV-producer cells

RV

48

9 mo[Shand et aL, 1999]

I/II

Primary/recurrent GBM

Freehand injection into resected tumor cavity of RV-producer cells or adenoviral vector

RVorAV

21 (7 RV, 7AV, 7 LacZ)

7 mo (RV) 15 mo (AV) 8 mo (LacZ)[Sandmair et aL, 2000]

II/III

Primary / recurrent GBM

Freehand injection into resected tumor cavity of adenoviral vector

AV

36(17AV, 19 control)

16 mo (AV) 9 mo (control) [Immonen et aL, 2004]

III

Primary GBM

Freehand injection into resected tumor cavity, followed by radiotherapy

RV

248 (124 RV, 124 control)

365 d (RV) 354 d (control)[Rainov, 2000]

Abbreviations: AV, adenovirus; d, days; GBM, glioblastoma multiforme; GCV, ganciclovir; HSVtk, herpes simplex virus thymidine kinase; LacZ, E. coli jS-galactosidase gene; mo, months; RV, retrovirus.

Amulticenter uncontrolled study including 48 patients with recurrent GBIVI showed a median survival time of 8.6 months. Tumor recurrence was absent on IVIRI in seven patients for at least 6 months, in two patients for at least 12 months, and one patient remained recurrence free at 24 months (Shand et al., 1999). A similar phase-I study with 12 children between ages 2 and 15 years with recurrent malignant supratentorial brain tumors showed a disease progression at a median time of 3 months after treatment (Packer et al., 2000). A large controlled phase-III study was conducted for an ultimate confirmation of the efficacy of the retroviral HSVt/c/GCV approach. This study used an adjuvant gene therapy protocol to the standard therapy of maximum surgical resection and irradiation for newly diagnosed GBIVI. After 4 years of follow-up of 248 patients, who were divided in a gene therapy and a control arm, survival analysis showed no advantage of gene therapy in terms of tumor progression and overall survival (Rainov, 2000). Several Phase-I trials were conducted in order to identify the effectiveness and safety of adenoviral vectors bearing the HSVtk gene. In one study, 13 malignant brain tumor patients were treated with a single intratumoral injection of a replication-defective adenoviral vector, followed by GCV treatment. Patients who received the highest vector dose showed central nervous system toxicity (confusion, seizures). Two patients survived for 2 years before lethal tumor

progression occurred and one patient survived 2.5 years after treatment and remained in stable condition (Trask et al., 2000). Retroviruses were compared with adenoviruses in another phase-I/II trial, including 21 patients with primary or recurrent GBJVI. Seven randomly selected patients with a total of eight tumors were treated with the retrovirus vector, and adenoviruses were used for seven tumors in seven patients. At the time of surgical resection, the two experimental groups were treated with either retrovirus VCP or adenoviruses expressing HSVtk (AdvHSVffc), whereas the control group received either adenovirus or retrovirus VPC expressing the marker gene lacZ (E. coli jS-galactosidase). Four patients with adenovirus injections had a significant increase in anti-adenovirus antibodies and two of them had a short-term fever reaction. Frequency of epileptic seizures increased in two patients. After subsequent GCV treatment, the adenovirus group had significant improvement in mean survival time, with 15 month compared to 7.4 (retrovirus group), and 8.3 month (control group). The assumed mechanism contributing to the greater efficacy of the adenovirus group were greater titer, a benefit from the inflammatory reaction to adenoviruses, and the ability of adenovirus to infect nonreplicating cells. In the retrovirus group, all treated gliomas showed progression by MRl at the 3-month time point, whereas three of the seven patients treated with AdvHSVfA: remained stable (Sandmair et al., 2000). On the basis of these results, a

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

294

22. PRODRUG'ACTIVATION GENE THERAPY

5-FC

5-FU

OH 5-FdUMP

FIGURE 2 Cytosine deaminase (CD); 5-fluorocytosinde (5-FC); 5-fluorouracil (5-FU); 5-fluorodeoxyuridine monophosphate (5-FdUMP).

randomized controlled trial involving 36 patients with operable primary or recurrent malignant glioma was conducted. Seventeen patients received AdvHSVffc by local injection into the wound bed after tumor resection, followed by intravenous GCV administration. The control group of 19 patients received standard care consisting of radical tumor resection followed by radiotherapy in those patients with primary tumors. AdvHSVfA: treatment produced a significant increase in mean survival from 39 to 70.6 weeks. The median survival time increased from 37.7 to 62.4 weeks. Six patients had increased anti-adenovirus antibody titers, without adverse effects (Immonen et al., 2004).

II.

CYTOSINE DEAMINASE (CD)/ S^FLUOROCYTOSINE (5^FC)

The cytosine deaminase (CD)/5-fluorocytosine (5-FC) approach is the next most widely studied suicide gene therapy approach. CD is uniquely expressed in certain fungi and bacteria and it converts the prodrug 5-FC (used to treat infections by fungi such as Candida albicans and Cryptococcus neoformans) into the active agent 5-fluorouracil (5-FU) (see Fig. 2). While 5-FC is nontoxic to human cells because of the lack of CD, 5-FU is used to treat cancers like colon, pancreatic, and breast cancer. The cytotoxic effects of 5-FU occur following its conversion to 5-fluoro-2'-deoxyuridine-5'-monophosphate (5-FdUMP). 5-FdUMP is an irreversible inhibitor of thymidylate synthase and thus inhibits DNA synthesis by deoxythymidine triphosphate (dTTP) deprivation and causes DNA strand breakage, leading to cell death (Grem, 1996). Rodent gliosarcoma cells expressing the E. coli CD gene in vitro become 77 times more sensitive to 5-FC (Aghi et a l , 1998). In addition, tumor cells expressing CD may present CD peptides on MHC class I,

where they could lead to an immune response (Mullen et al., 1996). In contrast to the HSVffc/GCV approach, 5-FU metabolites do not require cell-cell contact for a bystander effect. On cell lysis, 5-FU is released into the medium and is thus likely to be responsible for the bystander effect, and, indeed, the 5-FU levels in the medium correlated well with the degree of cytotoxicity (Kuriyama et al., 1998). Animal studies of C D / 5-FC using adenoviral vectors for rodent and human glioma cell lines showed an increase in survival time compared to controls (Miller et al., 2002). In order to improve tumor cell killing a therapeutic vector has been developed that encompasses two suicide genes to sensitize cells doubly to GCV and 5-FC (Aghi et a l , 1998; Desaknai et al., 2003). From the CD/5-FC clinical trials underway none is for brain tumors.

III. CYTOCHROME P450 (CYP)/ CYCLOPHOSPHAMIDE (CPA) OR IFOSFAMIDE (IFA) The large number of different cytochrome P450 (CYP) isozymes, and the fact that many drugs are metabolized by them, makes the choice of prodrugs quite wide (Waxman et al., 1999). But to date, this area has been dominated by two prodrugs, cyclophosphamide (CPA) and ifosfamide (IFA). Most of CYPs are expressed in the liver rather than in tumor cells, so the goal of this strategy is to selectively increase tumor cell exposure to cytotoxic drug metabolites by targeting expression of enzymes to tumor cells. CPA is a prodrug that is activated by liver-specific enzymes of the CYP family (see Fig. 3). The active form of CPA, phosphoramide mustard, is an alkylating agent that generates DNA cross-links and consecutively DNA strand breaks and cell death. The efficacy of CPA in treating brain tumors

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

CYTOCHROME P450 (CYP)/CYCLOPHOSPHAMIDE (CPA) OR IFOSFAMIDE (IFA)

295

HO.

.OH CYP

spontaneous — >-

O

CI CI

| N.

CI

CI

4-hydroxy CPA

Phosphoramide Mustard

FIGURE 3 Cytochrome P450 (CYP); cyclophosphamide (CPA),

has been limited by the fact that although CPA crosses the blood-brain barrier, its active metabolites can be generated only by liver P450, and these metabolites are poorly transported across the blood-brain barrier (Wei et al., 1994). Gene therapy using the rat cytochrome P450 2B1 (CYP2B1) which activates CPA with high efficiency (Clarke and Waxman, 1989) was designed primarily for the use in brain tumors since other malignancies have already access to CPA's active metabolites. The implantation of CYP2B1 expressing retroviral vectors was shown to induce regression of intracerebral rat glioma cells after intratumoral or intrathecal CPA administration (Manome et al., 1996). Like the CD/5-FC approach, CPA metabolites do not require cell-cell contact for a bystander effect, distributing by passive diffusion (Wei et al., 1995). Clinical trials of cancer gene therapy commonly involve inoculation of a replication-defective vector. When inoculated tissue was studied, very little diffusion away from the needle tract was noted (Puumalainen et al., 1998). In order to improve this hurdle several strategies have been used. One approach uses an oncolytic virus (OV) to deliver the CYP2B1 cDNA into the tumor (Chase et al., 1998). OVs are genetically altered viruses with deletions that restrict viral replication in normal cells but permit it in tumor cells (Smith and Chiocca, 2000). It has been reported that nucleoside analogs are synergistic in their anticancer action with alkylating agents (Andersson et al., 1996). While CPA metabolites are alkylating agents, GCV metabolites are nucleoside analogs. One approach uses the replacement of the large subunit of the HSV-1 genome with the CYP2B1 gene to generate a HSV-1 vector (rRp450) that is able to kill tumor cells through three modes: (1) using viral oncolysis and rendering infected cell sensitive to (2) CPA and (3) GCV (Chase et al..

1998). Subcutaneous tumors established from glioma cell lines in immunodeficient mice regress only when they are treated with rRp450, CPA, and GCV (Aghi et al., 1998). This has led to the hypothesis that after DNA chain alkylation by CPA metabolites, DNA repair mediated by DNA polymerases, is affected by GCV metabolites. The transient immunosuppression provided by activated CPA metabolites has also been shown to favor viral replication and anticancer effects in vivo (Ikeda et al., 1999). The cytochrome P450 system actually comprises two polypeptide components, the P450 and the P450 reductase (RED). RED expression is required to provide full catalytic activity of the rat CYP2B1 for gene therapy. Studies performed with rat gliosarcoma cells stably transfected with CYP2B1, a n d / o r RED cDNA, showed that further supplementation of RED by gene transfer enhanced the CYP/CPA efficiency, although tumor cells express enough RED to fulfill the transferred CYP gene's capability of CPA conversion (Chen et al., 1997). The addition of RED not only improves CYP2Bl-mediated conversion of CPA, but also provides the ability to convert other prodrugs such as tirapazamine into active anticancer agents (Jounaidi and Waxman, 2000). Another way to improve the therapeutic index of CYP/CPA is the inhibition of hepatic metabolism of CPA (Huang et al., 2000). This might decrease systemic toxicity of CPA metabolites as well as increasing prodrug availability to tumor cells expressing CYP. One of the essential steps in tumorigenesis is the active recruitment of a neovascular supply by the neoplasm. A modified CPA regime showed an anti-angiogenic effect, which also increased the therapeutic index of the CYP/CPA approach (Browder et al., 2000; Jounaidi and Waxman, 2001). Three CYP/CPA clinical trials are underway, none of them on brain tumors.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

296

22. PRODRUG'ACTIVATION GENE THERAPY

IV. G U A N I N E PHOSPHORIBOSYL^ TRANSFERASE/6^THIOXANTHINE The E. coli gpt gene encodes for the enzyme xanthine/guanine phosphoribosyl-transferase (XGPRT). Mammalian cells do not efficiently use xanthine for purine nucleotide synthesis. Therefore, cells producing XGPRT after transfection can be selectively grown with xanthine as the sole precursor for guanine nucleotide formation in a medium containing inhibitors (aminopterin and mycophenolic acid) that block de novo purine nucleotide synthesis (Mulligan and Berg, 1980, 1981). XGPRT transforms a xanthine analog, 6-thioxanthine (6-TX) to a toxic form for mammalian cells (Besnard et al., 1987). The weakly toxic purine analog 6-TX is phosphorylated to 6-thioxanthine monophosphate (6-XMP) by XGPRT (see Fig. 4). 6-XMP is subsequently converted to the highly toxic 6-thioguanine monophosphate (6-GMP). Rat glioma cells were infected with a retroviral vector expressing the gpt gene and a clonal line exhibited significant 6-TX susceptibility in vitro. In a 'l^ystander"

6-TX 6-TXRMP FIGURE 4 Xanthine / guanine phosphoribosyl-transferase (XGPRT); 6-thioxanthine (6-TX); 6-thioxanthine riboso-monophosphate (6-TXRMP).

assay, tumor cells from the clonal line efficiently transferred 6-TX sensitivity to uninfected tumor cells. This in vitro bystander effect was abrogated when transduced and untransduced cells were separated by a microporous membrane, suggesting that it was not mediated by highly diffusible metabolites. In vivo both 6-TX and 6-thioguanine (6-TG) significantly inhibited the growth of subcutaneously transplanted XGPRT expressing clonal tumor cells. In an intracerebral model, both 6-TX and 6-TG exhibited significant antiproliferative effects against transduced clonal tumors cells (Tamiya et al., 1996). In a nude mouse model retrovirus-mediated transfer of the gpt gene into rat glioma cells without subsequent selection still inhibited the proliferation of this mixed polyclonal population upon treatment with 6-TX (Ono et al., 1997). There is no GPT/6-TX clinical trial for brain tumors underway.

V.

NITROREDUCTASE/CB1954

The E. coli enzyme nitroreductase (NTR) is used in combination with the prodrug CB1954 [5-(aziridinl-yl)-2,4-dinitrobenzamide] as a suicide gene therapy approach. CB1954 is a synthesized, weak alkylating agent (Khan and Ross, 1969, 1971). The activating enzyme for CB1954 in mammals is DT-diaphorase (NAD(P)H dehydrogenase). This enzyme converts CB1954 to its 4-hydroxylamino derivative (see Fig. 5) (Knox et a l , 1988). After acetylation via thioesters such as acetyl coenzyme A (CoA), an alkylating agent is produced in a further activation step. The activated prodrug is then capable of forming poorly reparable DNA cross-links. The E. coli ISTTR is sensitized to CB1954 whereas human DT-diaphorase is poorly capable of performing this conversion, thereby limiting toxicity to transformed cells (Boland et al., 1991). The advantage of the NTR/CB1954 approach is that killing mediated by activated CB1954 is not dependent on the cell

CONH

CONH, NTR HO—NH

DNA CB1954 FIGURE 5

Nitroreductase (NTR).

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

E. COLI PURINE NUCLEOSIDE PHOSPHORYLASE (PNP)/PURINE ANALOGS

CPT-11

297

SN-38

FIGURE 6 Carboxylesterase (CE).

cycle phase, potentially allowing quiescent tumor cells to be killed. In vitro studies using a retroviral vector expressing the E. coli NTR gene in human colorectal and pancreatic cancer cell lines showed selective killing of NTR-expressing cells following CB1954 administration (Green et al., 1997). To improve the delivery of the NTR gene a replication-defective adenoviral vector expressing NTR was constructed (Grove et al., 1999). In vivo studies with the adenoviral vector in nude mice bearing human ovarian carcinoma showed promising results (Weedon et al., 2000). One study also demonstrated a synergistic effect when cells expressing both NTR and HSVtk were treated with a combination of CB1954 and GCV (Bridgewater et al., 1995). Furthermore, a significant bystander effect was seen with the NTR/CB1954 approach analogous to the HSVtk/GCV system (Grove et a l , 1999). A very effective NTR vector is an oncolytic adenovirus. The combination of viral oncolysis and NTR expression resulted in significantly greater sensitization of colorectal cancer cells to the prodrug CB1954 in vitro. In vivo, the oncolytic adenoviral vector was shown to replicate in subcutaneous colorectal cancer tumor xenografts in immunodeficient mice, resulting in more NTR expression and greater sensitization to CB1954 than with replicationdefective viruses (Chen et al., 2004). From the four NTR/CB1954 clinical trials underway, none is for brain tumors.

VL

CARBOXYLESTERASE (CE)/CPT^ 11

The enzyme-prodrug combinations described so far illustrate the concept of introducing a viral or bacterial enzyme to provide an activity that is almost absent in mammalian cells, whereas in the carboxylesterases (CE) approach a mammalian enzyme is used. CE converts the prodrug CPT-11 (irinotecan.

7-ethyl-10-[4-(l-piperidino)-l-piperidino] carbonyloxycamptothecin) to its active moiety, SN-38 (see Fig. 6). In this process, CPT-11 undergoes hydrolysis or deesterification to form the active metabolite SN-38, which is 100-1,000 times as potent as CPT-11 as an inhibitor of topoisomerase I (Rothenberg, 1997). In addition, SN38 freely passes though cell membranes, increasing the likelihood for a bystander effect. The activation of CPT-11 in humans has been thought to be mediated by the hepatic (hCE2) and human intestinal (hiCE) CEs. Evidence of in vivo activation of CPT-11 by endogenous human CEs of <5% has led to the search of more efficient CEs from mammalian sources. In vitro assays with human tumor cell lines expressing different mammalian CEs found the rabbit CE (rCE) to be the most efficient CE so far. CPT-11 has demonstrated remarkable antitumor activity both in animal models and in phase-II/III trials. Moreover, investigations with more than 40 drugs in a CNS xenograft model using multiple adult and pediatric glioma cell lines found CPT-11 to be the most efficient agent tested (Hare et al., 1997). An initial clinical trial of CPT-11 without viral-directed enzyme delivery in 60 patients with recurrent high-grade gliomas showed partial response in 10/49 (20%) GBM patients and 1/8 (12.5%) AA patients (Colvin et al., 1998). Currently, experiments are ongoing to compare efficacy and toxicity of hiCE and rCE with various viral constructs, including adenoviruses and retroviruses in which expression of intracellular or secreted forms of each enzyme are regulated by different tumorcell-specific promoters.

VII. E. COLI PURINE NUCLEOSIDE PHOSPHORYLASE (PNP)/PURINE ANALOGS The E. coli purine nucleoside phosphorylase (PNP) is involved in purine metabolism. The prodrugs used with

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

298

22. PRODRUG'ACTIVATION GENE THERAPY

HO

1

!^\

cellular kinases

OH

OH

F-dAdo

2-F-dAde

F-dAdo-MP

FIGURE 7 E. coli purine nucleoside phosphorylase (PNP); 6-methyl purine-deoxyribose (6-MeP-dR); 6-methyl purine (6-MeP); fluorodeoxyadenosine (F-dAdo); fluoroadenine (2-F-dAde); monophosphate (MP).

this strategy are 6-methyl purine-deoxyribose (6-MePdR), fluoro-deoxyadenosine (F-dAdo), and flouro-arabinosyl adenosine monophosphat (F-araAMP). E. coli PNP cleaves these prodrugs to methyl purine (6-MeP) and 2fluoro-adenine (2-F-Ade), respectively (see Fig. 7). These two agents are converted to ATP analogs, which inhibit RNA and/or protein synthesis. The specific mechanism of action of these two agents is not known. It is possible that these agents inhibit dsRNA deaminases. liowever, RNA and/or protein synthesis inhibition makes these agents effective against proliferating and nonproliferating tumor cells (Parker et al., 1998). Furthermore, the active metabolites have a high potency and a high bystander activity which does not require cell-to-cell contact, distributing by passive diffusion (Fiughes et al., 1998). Only a few (0.1-1%) expressing cells are needed to kill the entire cell population (Parker et al., 1998). Once metabolites such as 6-MeP are generated within a solid tumor, they appear to have a long half-life within the

tumor (greater than 24 h) and are very slowly released from tumor masses (Gadi et al., 2003). This long tumor half-life may be one explanation for the ability of these drugs to ablate tumors without profound systemic toxicity. Although, 6-MeP and 2-F-Ade are known to be too toxic for systemic administration (i.e., animals die at drug doses below those that could safely cause tumor regression) (Philips et al., 1954). Several in vitro studies showed profound tumor cell killing across several cancer cell types by the PNP strategy (Da Costa et al., 1996; Lockett et al., 1997; Nestler et al., 1997). A direct comparison with HSVtk/ GCV or CD/5-FC demonstrated faster and more effective tumor cell killing by PNP/purine analogs (Lockett et al., 1997; Nestler et al., 1997). Strong in vivo antitumor effects in mouse models have also been shown (Martiniello-Wilks et al., 1998; Mohr et al., 2000). In vivo studies with human glioma cell lines developed for the E. coli PNP approach demonstrated strong antitumor effects

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

CONCLUSIONS

after prodrug therapy with 6-MeP-dR and 2-F-araA (Gadi et al, 2003). There are no clinical trials for brain tumors underway using the E. coli PNP approach. VIII.

CONCLUSIONS

The completed clinical trials in brain tumor gene therapy have offered some promising results. However, delivering genetic vectors into solid brain tumors, and efficient in situ gene transfer remains one of the most significant hurdles in gene therapy Inefficient transduction needs to be improved before any largerscale clinical trials are conducted. The efficiency of transduction could be improved by modifying VPC, which have the ability to track even single tumor cells invading the surrounding brain tissue (Herrlinger et al., 2000). The currently used manual injection of VPCs, might be improved by the use of three-dimensional neuronavigation techniques and automated slow-speed injection devices (Rutka et al., 2000). The next step in the developraent of brain tumor gene therapy will also require a multimodal approach that combines several ideas in order to improve the poor outcome of malignant brain tumors. References Aghi, M., Kramm, CM., Chou, T.C. et al. (1998) Synergistic anticancer effects of ganciclovir/thymidine kinase and 5-fluorocytosine/cytosine deaminase gene therapies. J. Natl. Cancer Inst., 90: 370-380. Anderson, W.F. (2000) Gene therapy scores against cancer. Nat. Med., 6: 862-863. Andersson, B.S., Sadeghi, T., Siciliano, M.J. et al. (1996) Nucleotide excision repair genes as determinants of cellular sensitivity to cyclophosphamide analogs. Cancer Chemother. Pharmacol., 38: 406-416. Barba, D., Hardin, J., Sadelain, M. et al. (1994) Development of antitumor immunity following thymidine kinase-mediated killing of experimental brain tumors. Proc. Natl. Acad. Sci. USA, 91: 4348-4352. Besnard, C , Monthioux, E. and Jami, J. (1987) Selection against expression of the Escherichia coli gene gpt in hprt+ mouse teratocarcinoma and hybrid cells. Mol. Cell Biol., 7: 4139-4141. Boland, M.P, Knox, R.J. and Roberts, J.J. (1991) The differences in kinetics of rat and human dt diaphorase result in a differential sensitivity of derived cell lines to cb 1954 (5-(aziridin-l-yl)-2,4dinitrobenzamide). Biochem. Pharmacol., 41: 867-875. Bridgewater, J.A., Springer, C.J., Knox, R.J. et al. (1995) Expression of the bacterial nitroreductase enzyme in mammalian cells renders them selectively sensitive to killing by the prodrug cbl954. Eur. J. Cancer, 31A: 2362-2370. Browder, T., Butterfield, C.E., Kraling, B.M. et al. (2000) Antiangiogenic scheduling of chemotherapy improves efficacy against experimental drug-resistant cancer. Cancer Res., 60: 1878-1886.

299

Chase, M., Chung, R.Y. and Chiocca, E.A. (1998) An oncolytic viral mutant that delivers the cyp2bl transgene and augments cyclophosphamide chemotherapy. Nat. Biotechnol., 16: 444-448. Chen, M.J., Green, N.K., Reynolds, G.M. et al. (2004) Enhanced efficacy of Escherichia coli nitroreductase/cbl954 prodrug activation gene therapy using an elb-55k-deleted oncolytic adenovirus vector. Gene Ther., 11: 1126-1136. Chen, L., Yu, L.J. and Waxman, D.J. (1997) Potentiation of cytochrome p450/cyclophosphamide-based cancer gene therapy by coexpression of the p450 reductase gene. Cancer Res., 57: 4830^837 Clarke, L. and Waxman, D.J. (1989) Oxidative metabolism of cyclophosphamide: identification of the hepatic monooxygenase catalysts of drug activation. Cancer Res., 49: 2344-2350. Colvin, O.M., Cokgor, I., Ashley, D.M. et al. (1998) Irinotecan treatment of adults with recurrent or progressive malignant glioma. Proc. Am. Soc. Clin. Oncol., [Abstract 1493]: 387a. Culver, K.W., Ram, Z., Wallbridge, S. et al. (1992) In vivo gene transfer with retroviral vector-producer cells for treatment of experimental brain tumors. Science, 256: 1550-1552. Da Costa, L.T., Jen, J. He, T.C. et al. (1996) Converting cancer genes into killer genes. Proc. Natl. Acad. Sci. USA, 93: 4192-4196. De Clercq, E. (2000) Guanosine analogues as anti-herpesvirus agents. Nucleosides Nucleotides Nucleic Acids, 19:1531-1541. Desaknai, S. Lumniczky, K. Esik, O. et al. (2003) Local tumour irradiation enhances the anti-tumour effect of a double-suicide gene therapy system in a murine glioma model. J. Gene Med., 5: 377-385. Elion, G.B., Furman, PA., Fyfe, J.A. et al. (1977) Selectivity of action of an antiherpetic agent, 9-(2-hydroxyethoxymethyl) guanine. Proc. Natl. Acad. Sci. USA, 74: 5716-5720. Faulds, D. and Heel, R.C. (1990) Ganciclovir. A review of its antiviral activity, pharmacokinetic properties and therapeutic efficacy in cytomegalovirus infections. Drugs, 39: 597-638. Pick, J., Barker, KG., 2nd, Dazin, R et al. (1995) The extent of heterocellular communication mediated by gap junctions is predictive of bystander tumor cytotoxicity in vitro. Proc. Natl. Acad. Sci. USA, 92: 11071-11075. Gadi, V.K., Alexander, S.D., Waud, W.R. et al. (2003) A long-acting suicide gene toxin, 6-methylpurine, inhibits slow growing tumors after a single administration. J. Pharmacol. Exp. Ther., 304: 1280-1284. Green, N.K., Youngs, D.J., Neoptolemos, J.P et al. (1997) Sensitization of colorectal and pancreatic cancer cell lines to the prodrug 5-(aziridin-l-yl)-2,4-dinitrobenzamide (cbl954) by retroviral transduction and expression of the E. coli nitroreductase gene. Cancer Gene Ther., 4: 229-238. Grem, J.L. (1996) 5-fluoropyrimidines. In: Chabner B.A. and Longo D.L. (Eds.), Cancer Chemotherapy and Biotherapy: Principles and Practice. Lippincott, Philadelphia, pp. 149-212. Grove, J.L, Searle, PR, Weedon, S.J. et al. (1999) Virus-directed enzyme prodrug therapy using cbl954. Anticancer Drug Des., 14: 461-472. Hare, C.B., Elion, G.B., Houghton, P.J. et al. (1997) Therapeutic efficacy of the topoisomerase i inhibitor 7-ethyl-10-(4-[l-piperidino]-l-piperidino)-carbonyloxy-camptothecin against pediatric and adult central nervous system tumor xenografts. Cancer Chemother. Pharmacol., 39: 187-191, Herrlinger, U., Woiciechowski, C , Sena-Esteves, M. et al. (2000) Neural precursor cells for delivery of replication-conditional hsv-1 vectors to intracerebral gliomas. Mol. Ther., 1: 347-357. Huang, Z., Raychowdhury, M.K. and Waxman, D.J. (2000) Impact of liver p450 reductase suppression on cyclophosphamide

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

300

22. PRODRUG-ACTIVATION GENE THERAPY

activation, pharmacokinetics and antitumoral activity in a cytochrome p450-based cancer gene therapy model. Cancer Gene Ther., 7: 1034-1042. Hughes, B.W., King, S.A., Allan, P.W. et al. (1998) Cell to cell contact is not required for bystander cell killing by Escherichia coli purine nucleoside phosphorylase. J. Biol. Chem., 273: 2322-2328 Ikeda, K., Ichikawa, T, Wakimoto, H. et al. (1999) Oncolytic virus therapy of multiple tumors in the brain requires suppression of innate and elicited antiviral responses. Nat. Med., 5: 881-887. Immonen, A., Vapalahti, M., Tyynela, K. et al. (2004) Advhsv-tk gene therapy with intravenous ganciclovir improves survival in human malignant glioma: a randomised, controlled study. Mol. Ther., 10: 967-972. Ishii-Morita, H., Agbaria, R., Mullen, C.A. et al. (1997) Mechanism of 'bystander effect' killing in the herpes simplex thymidine kinase gene therapy model of cancer treatment. Gene Ther., 4: 244-251. Jacobs, A., Voges, J., Reszka, R. et al. (2001) Positron-emission tomography of vector-mediated gene expression in gene therapy for gHomas. Lancet, 358: 727-729. Jounaidi, Y. and Waxman, D.J. (2000) Combination of the bioreductive drug tirapazamine with the chemotherapeutic prodrug cyclophosphamide for p450/p450-reductase-based cancer gene therapy Cancer Res., 60: 3761-3769. Jounaidi, Y. and Waxman, D.J. (2001) Frequent, moderate-dose cyclophosphamide administration improves the efficacy of cytochrome p-450/cytochrome p-450 reductase-based cancer gene therapy. Cancer Res., 61: 4437-4444. Khan, A.H. and Ross, W.C. (1969) Tumour-growth inhibitory nitrophenylaziridines and related compounds: structure-activity relationships. Chem. Biol. Interact., 1: 27-47. Khan, A.H. and Ross, W.C. (1971) Tumour-growth inhibitory nitrophenylaziridines and related compounds: structure-activity relationships. li. Chem. Biol. Interact., 4: 11-22. Kim, J.H., Kim, S.H., Kolozsvary, A. et al. (1995) Selective enhancement of radiation response of herpes simplex virus thymidine kinase transduced 91 gliosarcoma cells in vitro and in vivo by antiviral agents. Int. J. Radiat. Oncol. Biol. Phys., 33: 861-868. Klatzmann, D., Valery, C.A., Bensimon, G. et al. (1998) A phase i/ii study of herpes simplex virus type 1 thymidine kinase "suicide" gene therapy for recurrent glioblastoma. Study group on gene therapy for glioblastoma. Hum. Gene Ther., 9: 2595-2604. Knox, R.J., Friedlos, R, Jarman, M. et al. (1988) A new cytotoxic, DNA interstrand crosslinking agent, 5-(aziridin-l-yl)-4-hydroxylamino-2-nitrobenzamide, is formed from 5-(aziridin-l-yl)2,4-dinitrobenzamide (cb 1954) by a nitroreductase enzyme in walker carcinoma cells. Biochem. Pharmacol., 37: 4661^669. Kuriyama, S., Masui, K., Sakamoto, T. et al. (1998) Bystander effect caused by cytosine deaminase gene and 5-fluorocytosine in vitro is substantially mediated by generated 5-fluorouracil. Anticancer Res., 18: 3399-3406. Lockett, L.J., Molloy PL., Russell, P.J. et al. (1997) Relative efficiency of tumor cell killing in vitro by two enzyme-prodrug systems delivered by identical adenovirus vectors. Clin. Cancer Res., 3: 2075-2080. Manome, Y, Wen, P.Y, Chen, L. et al. (1996) Gene therapy for mahgnant gliomas using replication incompetent retroviral and adenoviral vectors encoding the cytochrome p450 2b 1 gene together with cyclophosphamide. Gene Ther., 3: 513-520. Martiniello-Wilks, R., Garcia-Aragon, J., Daja, M.M. et al. (1998) In vivo gene therapy for prostate cancer: preclinical evaluation of two different enzyme-directed prodrug therapy systems delivered by identical adenovirus vectors. Hum. Gene Ther., 9: 1617-1626.

Mesnil, M. and Yamasaki, H. (2000) Bystander effect in herpes simplex virus-thymidine kinase/ganciclovir cancer gene therapy: role of gap-junctional intercellular communication. Cancer Res., 60: 3989-3999. Miller, C.R., WilHams, C.R., Buchsbaum, D.J. et al. (2002) Intratumoral 5-fluorouracil produced by cytosine deaminase/5fluorocytosine gene therapy is effective for experimental human glioblastomas. Cancer Res., 62: 773-780. Miura, P., Moriuchi, S., Maeda, M. et al. (2002) Sustained release of low-dose ganciclovir from a silicone formulation prolonged the survival of rats with gliosarcomas under herpes simplex virus thymidine kinase suicide gene therapy. Gene Ther., 9: 1653-1658. Mohr, L., Shankara, S., Yoon, S.K. et al. (2000) Gene therapy of hepatocellular carcinoma in vitro and in vivo in nude mice by adenoviral transfer of the Escherichia coli purine nucleoside phosphorylase gene. Hepatology, 31: 606-614. Moolten, F.L. (1986) Tumor chemosensitivity conferred by inserted herpes thymidine kinase genes: paradigm for a prospective cancer control strategy. Cancer Res., 46: 5276-5281. Mullen, C.A., Petropoulos, D. and Lowe, R.M. (1996) Treatment of microscopic pulmonary metastases with recombinant autologous tumor vaccine expressing interleukin 6 and Escherichia coli cytosine deaminase suicide genes. Cancer Res., 56: 1361-1366. Mulligan, R.C. and Berg, P. (1980) Expression of a bacterial gene in mammalian cells. Science, 209: 1422-1427. Mulligan, R.C. and Berg, P (1981) Selection for animal cells that express the Escherichia coli gene coding for xanthine-guanine phosphoribosyltransferase. Proc. Natl. Acad. Sci. USA, 78: 2072-2076. Nestler, U., Heinkelein, M., Lucke, M. et al. (1997) Foamy virus vectors for suicide gene therapy. Gene Ther., 4: 1270-1277. Niranjan, A., Wolfe, D., Tamura, M. et al. (2003) Treatment of rat gliosarcoma brain tumors by hsv-based multigene therapy combined with radiosurgery. Mol. Ther., 8: 530-542. Oldfield, E.H., Ram, Z., Culver, K.W. et al. (1993) Gene therapy for the treatment of brain tumors using intra-tumoral transduction with the thymidine kinase gene and intravenous ganciclovir. Hum. Gene Ther., 4: 39-69. Ono, Y, Ikeda, K., Wei, M.X. et al. (1997) Regression of experimental brain tumors with 6-thioxanthine and Escherichia coli gpt gene therapy Hum. Gene Ther., 8: 2043-2055. Packer, R.J., Raffel, C , Villablanca, J.G. et al. (2000) Treatment of progressive or recurrent pediatric malignant supratentorial brain tumors with herpes simplex virus thymidine kinase gene vectorproducer cells followed by intravenous ganciclovir administration. J. Neurosurg., 92: 249-254. Parker, W.B., Allan, P W , Shaddix, S.C. et al. (1998) Metabolism and metabolic actions of 6-methylpurine and 2-fluoroadenine in human cells. Biochem. Pharmacol., 55:1673-1681. Prados, M.D., McDermott, M., Chang, S.M. et al. (2003) Treatment of progressive or recurrent glioblastoma multiforme in adults with herpes simplex virus thymidine kinase gene vectorproducer cells followed by intravenous ganciclovir administration: a phase i/ii multi-institutional trial. J. NeurooncoL, 65: 269-278. Philips, F.S., Sternberg, S.S., Hamilton, S. et al. (1954) The toxic effects of 6-mercaptopurine and related compounds. Ann. NY Acad. Sci., 60: 283-296. Puumalainen, A.M., Vapalahti, M., Agrawal, R.S. et al. (1998) Betagalactosidase gene transfer to human malignant glioma in vivo using replication-deficient retroviruses and adenoviruses. Hum. Gene Ther., 9: 1769-1774.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

CONCLUSIONS Rainov, N.G. (2000) 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., 11: 2389-2401. Ram, Z., Culver, K.W., Oshiro, E.M. et al. (1997) Therapy of malignant brain tumors by intratumoral implantation of retroviral vector-producing cells. Nat. Med., 3:1354-1361. Ram, Z., Culver, K.W., Walbridge, S. et al. (1993) In situ retroviralmediated gene transfer for the treatment of brain tumors in rats. Cancer Res., 53: 83-88. Ram, Z., Walbridge, S., Shawker, T. et al. (1994) The effect of thymidine kinase transduction and ganciclovir therapy on tumor vasculature and growth of 91 gliomas in rats. J. Neurosurg., 81: 256-260. Rothenberg, M.L. (1997) Topoisomerase i inhibitors: review and update. Ann. Oncol., 8: 837-855. Rutka, J.T., Taylor, M., Mainprize, T. et al. (2000) Molecular biology and neurosurgery in the third millennium. Neurosurgery, 46: 1034-1051. Sandmair, A.M., Loimas, S., Puranen, R et al. (2000) Thymidine kinase gene therapy for human malignant glioma, using replication-deficient retroviruses or adenoviruses. Hum. Gene Ther., 11: 2197-2205. Shand, N., Weber, R, Mariani, L. et al. (1999) A phase 1-2 clinical trial of gene therapy for recurrent glioblastoma multiforme by tumor transduction with the herpes simplex thymidine kinase gene followed by ganciclovir. Gli328 european-canadian study group. Hum. Gene Ther., 10: 2325-2335. Shewach, D.S., Zerbe, L.K., Hughes, TL. et al. (1994) Enhanced cytotoxicity of antiviral drugs mediated by adenovirus directed transfer of the herpes simplex virus thymidine kinase gene in rat glioma cells. Cancer Gene Ther., 1: 107-112.

301

Smith, E.R. and Chiocca, E.A. (2000) Oncolytic viruses as novel anticancer agents: turning one scourge against another. Expert Opin. Investig Drugs, 9: 311-327. Tamiya, T., Ono, Y., Wei, M.X. et al. (1996) Escherichia coli gpt gene sensitizes rat glioma cells to killing by 6-thioxanthine or 6-thioguanine. Cancer Gene Ther., 3:155-162. Trask, T.W, Trask, R.R, Aguilar-Cordova, E. et al. (2000) Phase i study of adenoviral delivery of the hsv-tk gene and ganciclovir administration in patients with current malignant brain tumors. Mol. Ther., 1:195-203. Waxman, D.J., Chen, L., Hecht, J.E. et al. (1999) Cytochrome p450based cancer gene therapy: recent advances and future prospects. Drug Metab. Rev, 31: 503-522. Weedon, S.J., Green, N.K., McNeish, LA. et al. (2000) Sensitisation of human carcinoma cells to the prodrug cbl954 by adenovirus vector-mediated expression of E. coli nitroreductase. Int. J. Cancer, 86: 848-854. Wei, M.X., Tamiya, T., Chase, M. et al. (1994) Experimental tumor therapy in mice using the cyclophosphamide-activating cytochrome p450 2bl gene. Hum. Gene Ther., 5: 969-978. Wei, M.X., Tamiya, T., Rhee, R.J. et al. (1995) Diffusible cytotoxic metabolites contribute to the in vitro bystander effect associated with the cyclophosphamide/cytochrome p450 2b 1 cancer gene therapy paradigm. Clin. Cancer Res., 1: 1171-1177. Wiewrodt, R., Amin, K., Kiefer, M. et al. (2003) Adenovirus-mediated gene transfer of enhanced herpes simplex virus thymidine kinase mutants improves prodrug-mediated tumor cell killing. Cancer Gene Ther., 10: 353-364. Wolfe, D., Niranjan, A., Trichel, A. et al. (2004) Safety and biodistribution studies of an hsv multigene vector following intracranial delivery to non-human primates. Gene Ther., 11: 1675-1684.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

C H A P T E R

23 Clinical Trials of Gene Therapy for Canavan Disease Matthias Klugmann^, Claudia B. Leichtlein

Abstract: Canavan disease (CD) is an autosomal-recessive early onset leukodystrophy caused by loss of function mutations in the gene for aspartoacylase (ASPA). ASPA deficiency results in severe psychomotor retardation, seizures and premature death. The histopathology includes spongiform degeneration of the brain, dilated ventricles and dysmyelination, while the periphery is not affected. Under normal conditions, ASPA hydrolyzes N-acetyl aspartate (NAA) into aspartate and acetate, the latter representing the building block for lipid synthesis required for myelination. The NAA metabolism is highly segmented in that NAA production occurs in neurons, while its degradation is restricted to myelinating glia. The cellular pathology in CD is not fully understood but the loss of ASPA function leads to increased NAA levels in brain and body fluids and reduction of acetate in the CNS. The etiology of CD is thought to relate to cytotoxic and osmotic effects of excess NAA, but also to potential hypomyelination due to an undersupply with acetate. CD is a monogenic disease with a pathology confined to the brain, and no treatment exists. Therefore, CD has been a candidate for gene replacement utilizing ASPA gene transfer to the CNS and human experimentation started even before animal models of CD became available, and an ongoing phase 1 trial is the most advanced gene therapy for a non-malignant CNS disorder. ASPA deficiency in the mouse or the rat causes a CD-like pathology and these models have been used for gene therapy approaches, and to study disease mechanisms. This review will focus on latest results in the context of ASPA gene transfer in human, mouse and the rat. Keywords: Canavan disease; aspartoacylase (ASPA); seizure; axon-glia signaling; N-acetyl aspartate (NAA); AAV; leukodystrophy; dysmyelination

L

that this clinical entity, which became known as Canavan disease (CD) or Van Bogaert-Bertrand syndrome, followed an autosomal recessive trait in three Jewish patients with brain vacuolization (Van Bogaert and Bertrand, 1949). CD is a fatal, childhood leukodystrophy. The nosology includes mental retardation, head lag, hypotonia, macrocephaly, disability to attain developmental milestones, seizures and death usually before the age of 10 (Janson et al., 2002). Depending on the extent of medical care and nursing available, patients may survive into the second decade of life in a vegetative state (Zafeiriou et al., 1999).

THE NOSOLOGY OF CANAVAN DISEASE

When in 1931, Myrtelle May Canavan described a case of progressive spongy degeneration of the cerebral white matter wrongly diagnosed with Schilder's disease (Canavan, 1931), little did she know she identified a childhood disease that should become the subject of the world's first gene therapy for a neurogenetic disorder. Later Van Bogaert and Bertrand recognized

"Corresponding author. Tel: +49 6221 548200; Fax: +49 6221 544496; E-mail: [email protected]

Gene Therapy of the Central Nervous System: From Bench to Bedside

303

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

304

23. CLINICAL TRIALS OF GENE THERAPY FOR CANAVAN DISEASE

Brain pathology of CD patients shows a progressive vacuolization of the white matter, edema, astrocytic swellings and elongated mitochondria (Globus and Strauss, 1928; Canavan, 1931; Van Bogaert and Bertrand, 1949; Adachi et al, 1972; Adomato et al., 1972). While the dysmyelination is caused by death of the myelinforming cells, the oligodendrocytes, there is no loss of neurons. The underlying biochemical defect in CD is the lack of the enzyme aspartoacylase (ASPA) (Matalon et al., 1988). As a consequence, there is a build-up of the ASPA substrate N-acetyl-L-aspartate (NAA) in brain, blood and urine. Analysis of the gene encoding ASPA in CD patients revealed more than 40 different mutations resulting in a loss of ASPA function (Kaul et al., 1993; Surendran et al., 2003). Yet, only two different mutations (Tyr231X and Glu285Ala) account for 98% of affected Ashkenazi Jews. The carrier frequency of the two most common ASPA mutations among Jews is approximately 1:40, and genetic carrier testing is in place in this population (Leib et al., 2005). CD is a rare disease with about 1000 affected children in the USA. However, its monogenetic nature, severity and no existing treatment provide a strong rationale for a gene replacement strategy (see Section IV). To be able to interpret the data on ASPA gene transfer approaches in animal models and humans it is necessary to discuss the proposed function of ASPA and its substrate N-acetyl-aspartate (NAA) in normal and diseased individuals.

11.

THE N A A METABOLIC CYCLE

N-acetyl-aspartate is the most abundant free amino acid in the central nervous system (CNS), second only to glutamate (Tallan et al., 1956; Baslow, 1997). Due to its high concentration (5-10 mM) in the CNS (Michaelis et al., 1993; Wang and Zimmerman, 1998) and its well-characterized profile in proton magnetic resonance spectroscopy (MRS), NAA is used as a surrogate marker for the non-invasive determination of neuronal integrity. This approach is used to monitor the degenerative changes of a number of neurological disorders (Arnold et al., 1990; Klunk et al., 1992; Reynolds et al., 2005). The evaluation of these MRS studies is based on the idea that the loss of neurons or axonal function correlates with diminished NAA levels. The physiological roles of NAA are not well understood. NAA has been implied in osmoregulation (Baslow, 2002) or as acetate source for oligodendrocytes (see Section II.A) during myelination (Chakraborty

et al., 2001; Madhavarao et al., 2005). NAA production and degradation are remarkably compartmentalized (Fig. 1; see Section II.A, C) suggesting that it may play a role in neuron-glial signaling and communication during early postnatal stages of normal CNS development (Baslow, 2000). The lack of NAA is not well tolerated. A 3-year-old child that did not show any NAA signal in magnetic resonance spectograms of the brain, probably due to a defect in acetyl CoA/L-aspartate N-acetyltransferase (ANAT), was described with neurodevelopmental retardation, microcephaly and a mild hypomyelination (Martin et al., 2001). This case supports the idea that NAA plays a role in early stages of myelination. If NAA had an additional metabolic supportive function in maintenance of mature myelin, the condition of this patient could be proposed to worsen. However, the study by Martin and co-workers has not been followed up to date. The effects of increased NAA in the CD brain are not fully understood. NAA has been shown to induce seizures after intracerebroventricular administration to normal rats, probably by neuronal overexcitation through metabotropic glutamate receptors (Akimitsu et al., 2000; Yan et al., 2003). These data suggest that increased NAA levels (15-20 mM) in the ASPA-deficient brain (Leone et al., 2000) greatly participate in the development of the seizure phenotype observed in 50% of all CD patients. Disturbed neural osmoregulation or glia-neuronal signaling has also been proposed (Baslow, 2000). For a better understanding of the many facets of the complex pathology of CD, it is important to dissect the different parts of the NAA metabolic cycle illustrated in Fig. 1.

A.

N A A Production

Within the cell, 15% of the NAA is found in the cytosol, and 25% is located in the mitochondria of the nerve endings (Mcintosh and Cooper, 1965). The concentration of NAA increases considerably in the developing brain (Tallan, 1957). NAA synthesis occurs in mitochondria (Clark, 1998) and probably endoplasmic reticulum (Lu et al., 2004) of neurons. Immunohistochemical studies confirmed NAA predominantly in neurons but also to a lesser extent in oligodendrocytes (Bhakoo and Pearce, 2000; Madhavarao et a l , 2005). The enzyme responsible for NAA anabolism is ANAT (EC 2.3.1.17). The ANAT gene is yet to be mapped and the primary structure of the enzyme is unknown but partial biochemical purification suggested a single 670 kDa protein (Madhavarao et al..

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

THE NAA METABOLIC CYCLE

305

FIGURE 1 Model of NAA-dependent oligodendrocyte function, (a). Normal myelin (green) is generated in the presence of ASPA. NAA (circle and green dot) is synthesized from L-aspartate (circle) and acetyl CoA by ANAT in mitochondria and possibly ER of neurons, and trafficks into the cytosol (1). Via an unknown transport mechanism or diffusion, NAA is exported into the extracellular space (2). NAA import to oligodendrocytes may be partially mediated by the transporter NaDC3 or by yet unknown mechanisms (3). ASPA hydrolyzes NAA into free acetate (green dot) and L-aspartate in the oligodendrocyte cytosol (4). Acetate serves as building block for lipids (5) that are recruited into the myelin membrane (green), (b). ASPA-deficiency leads to abnormal myelin (red). Synthesis and oligodendrocyte-import of NAA (1-3) occur despite loss-of-function mutations in the ASPA gene but NAA-degradation does not. The NAA build-up in the extracellular space causes secondary overstimulation of metabotropic glutamate receptors (mGluR) and eventually seizures (6). The undersupply of acetate in the oligodendrocyte results in dysmyelination (7), and eventually oligodendrocyte death. The myelin defect might cause secondary axonopathy (8).

2003) catalyzing the synthesis of NAA from acetyl CoA and aspartate. NAA does not enter the neuronal metabolism except its conversion into N-acetylaspartylglutamate (NAAG) from NAA and L-glutamate

by N-acetyl glutamate synthetase (EC 2.3.1.1) (Caldovic et al., 2002). NAAG is the most abundant brain dipeptide and was proposed to possess neuromodulatory potential (Tsai and Coyle, 1995; Neale et al., 2000).

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

306

23. CLINICAL TRIALS OF GENE THERAPY FOR CANAVAN DISEASE

Interstitial NAA levels can also increase via the hydrolysis of NAAG by the astrocyte-specific plasma membrane enzyme glutamate carboxypeptidase II (EC 3.4.17.21) whose active site faces the extracellular space (Berger et al., 1999). The by-product glutamate is then metabolized by astrocytes and recycled to neurons (reviewed by Baslow, 2000). B.

N A A Shuttling

Interestingly, neither NAA nor NAAG can be hydrolyzed in neurons, the cell type in which they are predominantly produced. For hydrolysis, NAA is shuttled to oligodendrocytes, while NAAG is processed by astrocytes. This means that there must be transport mechanisms in place for shuttling NAA between the neuronal, extracellular and the oligodendroglial compartment. It is not clear however, how NAA is exported from the neuron into the extracellular space for subsequent uptake by the oligodendrocyte. At physiological pH, NAA is present as a bivalent anion. Due to the intraneuronal-extracellular (10-15 mM and 80-100 ]iM, respectively) concentration gradient (Michaelis et al., 1993; Taylor et al., 1994; Sager et al., 1997) and the inside-negative membrane potential, diffusion of NAA into the interstitial space is possible. A transport system for the export of NAA from neurons might exist but has not been identified so far. However, diffusion from the outside to the inside of the oligodendrocyte seems unlikely since a passive transport would have to overcome the existing intracellular negative membrane potential. Recently, the Na+-dependent high-affinity dicarboxylate transporter NaDC3 was shown to transport NAA in heterologous systems and in primary astrocytes (Huang et al., 2000). Coupling of NaDC3 to the catabolic compartment of the NAA cycle would require similar expression profiles of both the transporter and ASPA. It appears counterintuitive though that NaDC3 is expressed predominantly in meninges, cortical astrocytes and cerebellar Bergmann glia (Kekuda et al., 1999), while the NAA-degrading enzyme ASPA (see Section II.C) is almost exclusively expressed in oligodendrocytes (Klugmann et al., 2003; Madhavarao et al., 2004). Moreover, the affinity of NaDC3 for NAA is comparably low (George et al., 2004). However, there is some coexpression of ASPA and NaDC3 in the eye (George et al., 2004). If NaDC3 was the main facilitator for oligodendroglial uptake of NAA, the loss of NaDC3 function could be predicted to result in a CD-like pathology. This idea awaits experimental proof since no natural or engineered NaDC3

mutants are known to date. Taken together, the data available on NaDC3 suggest that the major NAA transporter in oligodendroglia is yet to be identified. C.

N A A Degradation

After uptake by the oligodendrocyte, NAA is specifically hydrolyzed by aspartoacylase (amidohydrolase II; EC 3.5.1.15) to produce acetate and L-aspartate (Birnbaum, 1952). ASPA is the best characterized among all enzymes involved in the NAA metabolic cycle. When NAA cannot be hydrolyzed as a result of ASPA deficiency in CD patients, levels of NAA in urine are increased more than 50-fold (Matalon and Michals-Matalon, 1999). In order to identify the ASPA expression domains in rodents, earlier studies utilized biochemical assays to determine ASPA activity in protein lysates from different tissues or cell types (Bhakoo et a l , 2001). Before genetic testing became available both approaches, determination of NAA and enzyme activity, were employed for diagnostic purposes (Hagenfeldt et a l , 1987; Matalon et al., 1988). Different ASPA expression domains have been reported depending on the detection method. While ASPA enzyme activity was most prominent in white matter tracts in vivo, and in oligodendrocytes and astrocytes in vitro (Bhakoo et al., 2001), histological investigations on ASPA mRNA and protein expression in the postnatal rodent brain proved the oligodendrocyte to be the sole source of this enzyme with no expression in neurons or astrocytes (Kirmani et al., 2003; Klugmann et al., 2003; Madhavarao et al., 2004). The latter studies indicate that detection of ASPA enzyme activity in cultured astrocytes is an artefact. Interestingly, ASPA activity is also found in peripheral tissues albeit the presence of its substrate, NAA, is limited to the CNS. In particular, ASPA activity found in kidney of both rat and mouse is even 2-3 times higher than in brain (Kitada et al., 2000; Matalon et a l , 2000). This notion was confirmed by Western blot analysis of protein lysates from kidney and brain using a rat anti-ASPA polyclonal serum (Fig. 2). There is currently no explanation for the expression of ASPA outside the CNS. One could assume that the enzyme specificity is broader than previously thought and that in the periphery, ASPA may hydrolyze substrates other than NAA. This, however, is contradicted by the fact that the loss of ASPA function causes a disease phenotype that is clearly restricted to the CNS while peripheral organs are not affected. In rats and mice, ASPA is a 312-amino-acid protein with the predicted size of 36.7 kDa. The high degree

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

ANIMAL MODELS OF CD

o^O <'S'

^.^ #

FIGURE 2 Western blot analysis of ASPA expression in brain and kidney of normal rats. 20 ]ig of total protein was separated by 12% PAGE and ASPA detection was achieved using a polyclonal rat-anti ASPA antibody (Klugmann et al., 2003). Note that the enzyme levels are higher in kidney despite the lack of the ASPA substrate NAA.

of homology between the mouse, rat and man is 87% indicating an important biological role of this enzyme. Yet biochemical or immunohistochemical analysis of ASPA was prevented by the absence of a specific antiserum. Only recently, Klugmann and co-workers produced an antibody against ASPA and detected the rat antigen specifically in the soma of oligodendrocytes in the postnatal brain (Klugmann et al., 2003). The same study showed that ASPA is a cytosolic protein of 37 kDa with no expression in myelin. An independent study confirmed these findings but detected some additional ASPA immunoreactivity in microglia (Madhavarao et al., 2004). In the rat, the temporal expression profile of both enzyme activity and protein expression correlate well with the time course of myelination, which is characterized by a logarithmic increase around postnatal day 14 to peak and a plateau after 21 days of age (Bhakoo et al., 2001; Klugmann et a l , 2003, 2005). In adult animals (>6 months), ASPA expression declines (Bhakoo et al., 2001; Klugmann et al., 2005) coinciding with the low turnover of myelin following the early active phase of myelin production (Norton and Poduslo, 1973). Moreover, the loss of ASPA activity results in decreased levels of lipids and acetate in CD and its mouse model (see Section III.A) (Madhavarao et al., 2005). Hence, ASPA-mediated degradation of NAA appears to be important to contribute acetate moieties for lipid synthesis required during CNS myelination.

IIL

ANIMAL MODELS OF CD

Two independent animal models of CD are now available reprising most part of the pathology of the human condition by means of histological and biochemical abnormalities (outlined in Section I). The ASPA gene has been targeted to engineer a knockout (KO) mouse (Matalon et al., 2000). In addition, a natural rat model.

307

termed tremor (tm) rat, was identified to carry a genomic deletion spanning the entire ASPA gene locus (Kitada et al., 2000). Both models have been used to study the etiology of CD and for the attempt of preclinical interventions in the context of a gene replacement strategy aiming to translate possible findings into the clinic. Regardless of the type of ASPA mutation, the unifying principle in all three species is the loss of ASPA function. Despite the high overall interspecies accordance with most aspects of the pathology in CD, the pathomechanisms are not understood. Table 1 shows an overview of the most important pathological features in CD and its models. A*

The ASPA Knockout (KO) Mouse

A targeted inactivation of the ASPA gene in the mouse was accomplished by introduction of a stop codon in exon 4 (Matalon et al., 2000). The presumed product of the targeted allele is an aberrant protein lacking any biological activity. Compared to wild type or heterozygous littermates, ASPA-KO mice present with reduced body weight and enlarged head. Neurological deficits include seizures, tremor, ataxia and spasticity NAA levels in ASPA-deficient mouse brains are increased about 10 times compared to normal littermates, while in Canavan patients NAA levels rise 50-fold over controls. It is of note that while urinary excretion of the NAA derivate, NAAG, is a prominent feature in CD (Burlina et al., 1994,1999), it is not known if it is preceded by elevation of NAAG levels in the CNS. Analyses of NAAG brain levels and activity of the NAAG hydrolyzing enzyme, glutamate carboxypeptidase II (see Section II.A), in the ASPA-deficient mouse showed no differences to wild-type controls (Surendran et al., 2004a). However, it is not clear if NAAG levels in urine of ASPA-KO mice are increased. Enlarged ventricles and white matter vacuolization, most prominently in subcortical areas, cerebellum and spinal cord has been described in ASPA-deficient mice (Matalon et al., 2000, 2003; Surendran et al., 2005a) and MRI analyses showed an increase in water content similar to patients with CD (Matalon et al., 2000), while neurons have not been reported to be affected. The life expectancy of ASPA-KO mice is 1.5-9 months (Matalon et al., 2000) while human patients die at around 10 years of age (see Section I and Table 1). Considering the shorter absolute life span and the faster development of mice, the relative survival times are comparable between the two species. Yet, it remains unknown why the loss of ASPA function causes early death and secondary effects are likely

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

308

23. CLINICAL TRIALS OF GENE THERAPY FOR CANAVAN DISEASE

TABLE 1

ASPA' deficiency in Different Species

Gene product

NAA levels'* brain/urine

Seizure type/ prevalence

Life span

Axonal swellings

Reproduction

Missense^

Misfolded ASPA

1.5-3.5/50^

Convulsions/>50%'^

3-20 years

No^

N.a.

Mouse

Deletion downstream exon 3""

Truncated ASPA

2.8/10^

Convulsions/low^

1.5-9 months'^

No

Fertile but reproduction impaired-'^

Tremor

Rat

Complete deletion^

None ^

Absence-like and convulsions/100%^

3-9 months^

Yes^

Both sexes infertile^

Tremor carrier

Rat

Deletion of one allele^

ASPA'

Absence-like/100%^

Normal

No

Normal

Syndrome

Species

Canavan disease

Human

ASPA-knockout

ASPA mutation

Normar

Note: Overview of the exact genetic defect and the corresponding pathology in Canavan disease and its mouse and rat models. All homozygous tremor rats exhibit petit mal epilepsy while this seizure type has not been reported in CD or the KO mouse. Convulsive seizures occur in the latter two species albeit with lower prevalence. Carriers do not show phenotypic abnormalities save tm/+ rats develop late onset seizures in the absence of any additional deficits. " Fold increase relative to controls. ^ Most frequently; Matalon and Michaelis-Matalon (1999). ' Traeger and Rapin (1998). ^ Baslow (2000). ' Matalon et al. (2000). /Surendran et al. (2005b). ^ Kitada et al. (2000). '' Klugmann et al. (2005). ' McPhee et al. (2005). Enzyme activity only 50% compared to wild type. ' T. Serikawa (personal communication). >" Higashiguchi et al. (1991). N. a. = not assessed.

to have an impact on survival. To that end expression profiling of ASPA-KO brains revealed a robust downregulation of genes involved in signal transduction, such as GABA-A receptor 6, neurogenic differentiation factor and the glutamate transporter EAAT4, while a group of cell death genes, such as serine protease inhibitor (Spi2), caspase-11 and iL-ljS-converting enzyme were upregulated (Surendran et a l , 2003). In addition, glutamate levels were reduced to five-fold. The significance of these findings in the etiology of CD is currently unclear. Another enigmatic result of ASPA deficiency in mice (and rats, see Section III.B) is an impaired reproduction (Surendran et al., 2005b). Offspring from heterozygous ASPA ( + / - ) mice showed the expected Mendelian ratio for wild-type, carriers and homozygous littermates. However, mating homozygous females with heterozygous males produced fetal death and fewer offspring compared to heterozygous mothers. It is not clear how the lack of ASPA might affect reproduction, but the proposed role of NAA and NAAG as glial target-specific signaling molecules might suggest that increased levels of these substances in the developing embryo might compromise the delicate glial-neuronal communication (Baslow, 2000).

Importantly, a recent study showed that the synthesis of several major myelin-associated lipids is impaired in the absence of ASPA in both mouse and human and levels of free acetate were reduced in the brain but not in peripheral tissues of ASPA-KO mice (Madhavarao et al., 2005). These results suggest that the impaired myelination observed in CD is caused, at least in part, by a deficiency of NAA-derived acetate. B.

The Tremor Rat

The tremor (tm) rat is a natural ASPA-deficient mutant identified in the KyotorWistar colony (Yamada et al., 1985; Kitada et al., 2000). Homozygous tm/tm rats present with movement tremors, muscle weakness and seizures. Like its mouse and human counterparts, the ASPA-deficient rat shows spongiform degeneration of the brain and enlarged ventricles (Kitada et al., 2000; Klugmann et al., 2005). Positional cloning of the tm locus revealed a genomic deletion on rat chromosome 10q24 encompassing the genes for ASPA, an olfactory receptor and vanilloid receptor subtype I, the latter two being sensory receptors (Kitada et al., 2000). Although formally the tremor rat is a triple mutant it is unlikely that the lack of sensory receptors contributes

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

309

ASPA GENE THERAPY

to the CD-like pathology, because the phenotype of the tremor rat is very similar to the genetically engineered ASPA-KO mouse (see Section III.A). NAA levels in brains of tm/tm animals are increased about two-fold depending on the brain area examined (Kitada et al., 2000). The life expectancy is between 3 and 9 months (T. Serikawa, personal communication), and both sexes display dysgenesis of gonads (Kitada et al., 2000). Homozygous tm/tm breeding pairs do not produce offspring (our unpublished observations). It appears possible that the loss of ASPA directly interferes with fertility since the ASPA gene is active in testis and ovary of normal rats (Kitada et al., 2000). However, the impaired reproduction might also reflect a compromised development of the embryo as observed in the ASPA-KO mouse (see Section III. A). There are also some phenotypic differences between tm/tm rats and ASPA-deficient mice, the most obvious being mutant rats exhibit curled whiskers and waved hair (Kitada et al., 2000). Furthermore, while the lack of ASPA in the mouse only occasionally leads to seizures (Matalon et al., 2000), all homozygous tm/tm rats show absence-like seizures (see Table 1) which will be complemented with convulsive seizures later in life (Yamada et a l , 1985; Serikawa et al., 1987; Sasa et a l , 1988). Tremor rats display various neurological deficits during disease progression. Young tm/tm rats exhibit movement tremors at 2 weeks of age (Higashiguchi et al., 1991). At 8 weeks tremors become less prominent but absence-like seizures, characterized by staring and 5-7 Hz spike-wave complexes in hippocampal EEC, occur paralleled by severe impairment of motor coordination and locomotion at later stages (Yamada et al., 1985; Serikawa et al., 1987; Klugmann et a l , 2005;). The seizures can be inhibited by drugs that repress absence seizures in humans. Consequently, the tm rat has served as a model of petit mal epilepsy (Hanaya et al., 1995). Like CD patients, tm/tm rats show CNS dysmyelination while the periphery appears normal. Spongy degeneration occurs predominantly in white matter of pons, cerebellum, thalamus and spinal cord, spreading to the caudate-putamen, hypothalamus and cerebral cortex at later stages (Kondo et al., 1991; Kitada et al., 2000 ). The vacuoles mainly consist of swollen astrocytic processes and enlargement of the extracellular space, as well as occasional enlargements of the periaxonal space (Kondo et al., 1991). The profound morphological and behavioral deficits should be reflected in damages to oligodendrocytes and hence to myelin. In that regard it is surprising that the quantification of purified myelin

membranes showed a mere 20% reduction of brain myelin in tm/tm animals for up to 6 months of age (Klugmann et al., 2005), which confirmed the relatively moderate extent of demyelination in aged tm/ tm rats assessed by histological analyses (Kondo et al., 1991). Neuron numbers are not compromised but a recent study suggested the presence of axonal spheroids in homozygous tm/tm rats (Klugmann et al., 2005). The nature of these swellings is uncertain and the authors did not provide a comprehensive analysis of this observation. In addition, axonal involvement has not yet been described for CD or its models. Yet, it is widely accepted that axonal abnormalities results from defects in the myelinating cells whether myelin irregularities are apparent or not (Lappe-Siefke et al., 2003; Popko, 2003), and a failure in myelin compaction prevents maturation of the axonal cytoskeleton (Brady et a l , 1999; Popko, 2003). Reduction of myelin lipids in ASPA-deficient mice has been reported recently (Madhavarao et al., 2005). Changes in the composition of myelin proteins have not yet been addressed in any species lacking ASPA but may cause axonopathy rather than reduced absolute myelin levels. C.

Phenotypes of ASPA Null Carriers

No abnormalities have been reported for human CD carriers. Heterozygous ASPA-KO mice appear behaviorally and histologically indistinguishable from wild-type animals. Their life expectancy is not different from wild-type littermates and NAA levels are not increased although the aspartoacylase activity is reduced to 31% of wild-type controls (Matalon et al., 2000). Likewise, tm/-\- carriers do not exhibit any brain abnormalities and NAA levels are normal in the presence of 50% ASPA activity (McPhee et al., 2005). Notably, heterozygote tm/-\- rats exhibit the same type of epilepsy as their homozygous littermates less frequently and at a later stage, after 26 weeks (Higashiguchi et al., 1991). The occurrence of lateonset absence-like seizures in tm carriers is uncoupled from obvious pathological changes in the CNS since it is not accompanied by vacuolization, demyelination or excess NAA. This suggests that spongy degeneration is not a prerequisite for absence-like seizures in tm rats. IV.

ASPA GENE THERAPY

CD has been considered a prototype for human gene therapy of neurogenetic disorders because of its

V GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

310

23. CLINICAL TRIALS OF GENE THERAPY FOR CANAVAN DISEASE

monogenic nature, a pathology confined to the CNS, and the expectation that even low levels of transduction could ameliorate the disease phenotype via metabolic co-operativity (During and Ashenden, 1998). In 1996, these reasons provided the rationale to attempt ASPA gene therapy in humans (During, 1996), years before animal models became available for testing. Since the initial human experimentation, both viral and non-viral delivery systems and different routes of administration have been used for introducing ASPA to the brain (Table 2). Most studies utilized in vivo gene transfer, in which a shuttle vector carrying the ASPA gene is administered directly to the host (Table 2). Only one report described the ex vivo approach, in which DNA is delivered to cells in vitro for subsequent transplantation into a target tissue (see Section IV.B.2). The pros and cons of the various shuttle systems have been reviewed elsewhere (Janson et al., 2001; Hsich et al., 2002). An ideal vector for the treatment of ASPA deficiency should direct the therapeutic gene to oligodendrocytes throughout the CNS and should be non-toxic to the host. A liposomal formulation used in the initial clinical interventions proved to be safe but inefficient by means of its potential to transduce brain cells in vivo (Leone et al., 2000). Instead, viral vectors

TABLE 2 Date

Event

Subjects

were used for subsequent approaches (Table 2). There is currently no viral vector that preferentially targets oligodendrocytes and gene therapy interventions for CD or its models employed recombinant adeno-associated virus (rAAV) despite the well-known neuronal tropisms of this vector. The rationale for this controversial approach will be explained in Section IV.A.2. A.

Clinical Trials

In a groundbreaking work, Matthew During and Paola Leone pioneered the worldwide first attempt to apply gene therapy for a non-malignant CNS disorder (During, 1996; Leone et al., 1999, 2000). Results from this work provided the basis for FDA approvals on follow up studies using improved vectors and delivery protocols. The ongoing phase 1 trial for Canavan (see Section IV.A.2) is currently the most advanced human trial using AAV. 1 • The Pioneering Study In 1996, two Canavan children were administered, a non-viral lipid-entrapped, polycation-condensed delivery system (LPD) in conjunction with a recombinant plasmid-encoding ASPA, to the cerebroventricular

Chronicles of ASPA Gene Therapy

Probants

Age at treatment

Vector

Delivery

1996

In vivo

CD"

2

19-24 months

Liposome-polymer

Intracerebroventricular

1998

FDA approval ^

CD

14

N. a.

Liposome-polymer

N. a.

1998-2001

In vivo

CD

14

9 months-7 years

Liposome-polymer

Intracerebroventricular

2001

FDA approval ^

CD^

21

N. a.

AAV2

N. a.

2001-2004

In vivo

CD

12

1-5 years

AAV2

Intraparenchymal

2002

In vivo

Tremor rat''

N. a.

6 weeks

Ad

Intracerebroventricular

2003

In vivo

KO mouse -^

N. a.

12 weeks

AAV2

Intraparenchymal

2004

Ex vivo

KO mouse ^

N. a.

'Juvenile'

Neural progenitors

Intraparenchymal

2005

In vivo

Tremor rat''

N. a.

30 weeks

AAV2

Intraparenchymal

2005

In vivo

Tremor r a t '

N. a.

3 weeks

AAVl/2

Intraparenchymal

Note: ASPA gene replacement strategies were realized starting in 1996 marking the first gene therapy for a neurogenetic disease. An extended phase 1 trial proposing to employ AAV for ASPA gene transfer started in 2001, and is the most advanced AAV-based trial to date. Several years after clinical experimentation started, in 2003 and in 2005 reports were published on AAV-based ASPA gene therapy in the mouse and rat models, respectively. « During, 1996; Leone et al. (1999, 2000). ^ I.N.D.-7307. ' I.N.D.-9119. '^ Janson et al. (2002). ' Seki et al. (2003). / M a t a l o n e t a l . (2003). ^ Surendran et al. (2004b). '' McPhee et al. (2005). ' Klugmann et al. (2005). N. a. = not applicable.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

ASPA GENE THERAPY

space. The LPD was composed of a synthetic liposomepolymer complex. Prior to its application in humans, the delivery system was tested to be safe in normal rats and cynomologous monkeys (Leone et al., 2000). Likewise, no adverse effects were observed in the patients. In addition, mild neurological and radiological (regionally lowered NAA) improvements could be determined. Two years later, a second clinical trial with more participants using the same approach showed similar results. Again, most likely due to inadequacies of the delivery system, clinical changes were transient and not pronounced Qanson et al., 2002). 2.

rAAV'-Mediated Qene Therapy for CD

Given the encouraging results from the pilot studies, a new clinical protocol implemented an improved vector system and a new method of delivery (Janson et al., 2002). The protocol proposed to administer 900 billion genomic particles of rAAV serotype 2 carrying the ASPA cDNA (rAAV-ASPA), spread over six subcortical delivery sites Qanson et al., 2002). rAAV is superior over other viral systems because it does not induce an immune response in the host and confers sustained long-term expression in the CNS in vivo (Kaplitt et al., 1994; Xu et al., 2001; Lowenstein, 2002), and rAAV2 is well characterized and widely used as a gene transfer vector to the mammalian brain (for review see Janson et al., 2001). However, its internalization properties (tropism) are strictly neuron-specific (Xu et al., 2001), and it is not evident to attempt ASPA gene transfer to neurons when ASPA is normally expressed in oligodendrocytes. Yet, based on the assumption that excess NAA exerts toxic effects that cause the neuropathology in CD, it appeared justified to attempt AAV-ASPA transfer into neurons of CD patients in order to induce the hydrolysis of NAA in the compartment of its S3mthesis (Janson et al., 2002). The efficacy and safety of viral-mediated gene transfer and transgene expression was tested in normal rats and monkeys and no adverse effects could be observed upon the presence of ASPA in neurons (Janson et al., 2002). Stereotaxic injections started in 2001 and no results have been published in the medical literature to date but the most recent CD progress report (2004) on the web page of 'The Myelin Project' (http://www.myelin.org/) claims marked biochemical and neurological improvements in the four youngest patients of a cohort aged between 2 and 5 years when the treatment began, while the oldest probants experienced rather mild improvements. As CD is a disorder that manifests in early childhood, it seems self-evident that the success of a gene replacement strategy depends on the timing of the intervention.

B.

311

Preclinical Trials

Although human experimentation has commenced (see Section IV.A) only recently, information on the effects of rAAV-mediated ASPA gene transfer in both rodent models of CD has become available (Matalon et al., 2003; Klugmann et al., 2005; McPhee et al., 2005). Importantly, together these studies show that neuronal expression of ASPA does not reverse the CD-like neuropathology (see Section IV.B.l). !•

rAAV'Mediated Qene Therapy for CD Models

In the absence of a functional anti-ASPA antibody required for determination of introduced ASPA, Matalon and co-workers used a total dose of 3.2 X 10^ infectious rAAV2 particles to simultaneously express ASPA and the green fluorescent protein (GFP) bilaterally in the striatum and thalamus of 3-month-old ASPA-KO mice (Matalon et al., 2003). The expression cassette contained the chicken j] actin-CMV immediate early enhancer hybrid promoter and the Woodchuck postregulatory element (WPRE) for strong sustained transgene expression (Fitzsimons et al., 2002). Histological data on rAAV-mediated ASPA expression, assessed indirectly via numbers of GFP-labelled cells, revealed a low extent of transduction in vivo, but quantification of ASPA protein levels was not provided. Moreover, there was no comprehensive examination of ASPA enzyme activity or myelin content. A focal reduction of brain NAA in rAAV-ASPA-GFP-treated KO-mice was suggested by in vivo MRI/MRS examination. The most significant finding of this study was the complete absence of vacuoles in the plane of the injection sites while the enlargement of the ventricles was not resolved (Matalon et a l , 2003). Taken together, the interpretation of these data is hampered by the poor transduction efficiency and the lack of information on transgene expression and neurological findings. In a similar study, an rAAV2-ASPA vector identical to the one used for the clinical trial (Janson et al., 2002) was injected bilaterally into the striatum and thalamus of aged (30-week-old) tm/tm rats (McPhee et a l , 2005). The total dose injected was 3.2 X 10^° genomic particles per animal and analyses at 40 weeks of age included ASPA immunohistochemistry, NAA and enzyme activity measurements of tissue samples and testing of motor behavior but quantitative data on ASPA protein expression was lacking. As expected, AAV2-mediated transgene expression was exclusively observed in neurons. Due to retrograde axonal transport of rAAV (Kaspar et al., 2002) vector spread extended into distant sites like the cortex. ASPA activity in selected tissue samples

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

312

23. CLINICAL TRIALS OF GENE THERAPY FOR CANAVAN DISEASE

of ASPA-treated mutants reached approximately 15% of wild-type controls and NAA levels in tissue lysates showed a significant reduction compared with rAAVGFP treated controls but not with untreated tm/tm controls. Likewise, the motor performance assessed in a specialized behavioral test was significantly improved for rAAV-ASPA injected rats over rAAV-GFP treated controls but not over untreated tm/tm rats. Wild-type controls were not included in this experiment. These results point toward a detrimental effect of GPP rather than a therapeutic effect by ASPA. Hence, the significance of the results is questionable and the relevance of this study in the context of a gene therapy for a childhood disorder is not evident since the authors infused the vectors into aged animals. Finally, an independent study used a superior rAAV vector system for gene therapy in juvenile tm/tm rats (Klugmann et al., 2005). This report described the use of rAAVl / 2 , a novel chimeric serotype that has several fold improved transduction efficiency over rAAV2 (Richichi et al., 2004). Bilateral stereotaxic injections of a total dose of 5.2 XlO^ genomic rAAVl/2-ASPA particles into striatum and thalamus of 3-week-old tm/tm rats resulted in transgene expression in neuronal cell bodies and projections throughout the brain (Klugmann et al., 2005). Significantly, the timing of intervention, immediately post-weaning was consistent with the timing of diagnosis for CD and the robust expression of transgenic protein (Fig. 3) allowed correlations between biochemical findings and neurological changes. Myelination, NAA metabolism, motor behavior and seizure activity of untreated wild-type animals, rAAVl/2-GFP treated tm/tm animals, and

rAAVl/2-ASPA treated tm/tm littermates was examined between 5 and 26 weeks of age. Biochemical analyses performed on whole brain lysates. Levels of introduced ASPA as well as the corresponding enzyme activity were restored or exceeded those in wild-type controls. Yet, while reduced compared to AAV-GFP injected mutants, NAA levels in AAV-ASPA injected tm/tm rats remained elevated compared with wild-type controls. This result might suggest that neurons are not the sole locus of NAA synthesis (see Section II.A) and in the normal brain oligodendrocyte progenitors have low equilibrium concentrations of NAA (Urenjak et al., 1992,1993; Madhavarao et al., 2005). It is unclear if the low steady-state levels found in oligodendrocytes is due to dampened NAA synthesis or to rapid turnover by ASPA. This could be examined by NAA immunohistochemistry in ASPA-deficient tissue as the prediction would be that NAA synthesized in oligodendrocytes accumulates in this cell type when the catabolic path is disrupted. However, NAA biosynthesis appears to be downregulated in CD patients as a result of feedback inhibition (Moreno et al., 2001). Finally, although there was a dramatic positive effect on seizure activity upon AAV-ASPA transfer but dysmyelination, motor deficits were unchanged, and the extent of vacuolization was indistinguishable between AAV-ASPA and AAV-GFP treated mutants (Klugmann et a l , 2005). The results described by Klugmann and co-workers may have implications for human gene therapy applications since they showed that transfer of ASPA to neurons in vivo does not rescue the CD-like pathology. Instead, vectors with specific tropism for myelinating

FIGURE 3 Immunohistochemical detection of transgenic ASPA 5 months after stereotaxic delivery of 5.2 X 10^ vector genomes to striatum (str) and thalamus (th) of homozygous tm/tm rats. Note the enlarged ventricles (ve), a typical feature of the CD-like histopathology. (a). The rostro-caudal extent of ASPA expression in a sagittal section. ASPA immunoreactivity is highest near the injection sites but also spreads to hindbrain (hb), olfactory bulb (ob), hippocampus (he) and cortex (ctx) while the cerebellum (cb) is spared, (b). A coronal section shows the medio-lateral extent of transduction. The transgene is expressed in grey matter of both entire hemispheres although at different levels.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

313

ACKNOWLEDGMENTS

glia will be needed to provide a gene therapy for leukodystrophies (see Section VI). 2* Other Approaches of Qene Therapy for CD Models Adenovirus has a tropism for both neurons and glia. The effects of adenovirus-mediated ASPA gene transfer administered unilaterally into the lateral ventricle of 6-week-old tm/tm rats was examined with respect to seizure activity (Seki et al., 2002). Data presented in this report was limited to the quantification of EEG seizures over a period of 6 weeks post vector infusion while no information on transgene expression or NAA levels was provided. A significant reduction of seizure activity was only found 1 week after gene transfer probably reflecting the poor vector spread by the intracerebroventricular route or the known immunogenic potential of adenovirus (Seki et al., 2002). In an ex vivo approach neural precursor cells were infected with a retrovirus expressing ASPA for subsequent unilateral transplantation into striatum and cerebellum of ASPA-deficient mice (Surendran et al., 2004b). The report showed that implanted cells differentiated into oligodendrocytes and fibrous astrocytes. Although this technology is still in its infancy, these results show that stem cell therapy might be an alternative route for future treatments of CD.

to efficiently reduce NAA brain levels (Klugmann et a l , 2005). The results may have implications for both current and future human clinical trials employing neurotropic viral vectors to replace a missing protein that is normally expressed in oligodendrocytes. For CD in particular, and other leukodystrophies in which the deficient protein is specifically expressed in glia, novel vectors with tropism for such cell types are required for an effective treatment. No such vector is currently available despite attempts of capsid engineering to alter the natural tropism of existing AAV serotypes (Muzyczka and Warrington, 2005). Recently, several naturally occurring AAV serotypes have been isolated in monkeys (Gao et a l , 2003). Once characterized they might be used as vectors suited for the gene therapy of white matter diseases given they are safe and show a specific tropism for oligodendrocytes.

ACKNOWLEDGMENTS This work was supported by an EMBO fellowship to MK and the Neurological Foundation of New Zealand. The authors thank C. Wymond Symes for excellent technical help. References

V.

F U T U R E PROSPECTS

Although considered a rare disease, developing a cure and treatment for CD has the potential to aid in the treatment of other leukodystrophies or even in more common neurodegenerative diseases such as Alzheimer's, ALS and Parkinson's. In fact, the same vector system used for the ongoing Canavan trial is currently being used for the first human gene therapy trial of Parkinson's disease (During et al., 2001). Based on reduced brain acetate levels and myelin lipid synthesis in CD and its murine model, acetate deficiency has been proposed as the etiological mechanism of CD (Madhavarao et al., 2005). Consequently, increasing brain acetate levels by dietary supplementation of newborns has been suggested (Madhavarao et al., 2005). However, an acetate therapy would not address pathological effects mediated by excess NAA or its derivative NAAG. On the other hand, latest preclinical gene therapy approaches using animal models of CD indicate that ASPA ectopically expressed at high levels in neurons does not prevent CD pathogenesis and is not sufficient

Adachi, M., Torii, J., Schneck, L. and Volk, B.W. (1972). Electron microscopic and enzyme histochemical studies of the cerebellum in spongy degeneration (van Bogaert and Bertrans type). Acta. NeuropathoL, 20: 22-31. Adornato, B.T., O'Brien, J.S., Lampert, P.W., Roe, T.R and Neustein, H.B. (1972). Cerebral spongy degeneration of infancy. A biochemical and ultrastructural study of affected twins. Neurology, 22: 202-210. Akimitsu, T., Kurisu, K., Hanaya, R., lida, K., Kiura, Y, Arita, K., Matsubayashi, H., Ishihara, K., Kitada, K., Serikawa, T. and Sasa, M. (2000). Epileptic seizures induced by N-acetyl-L-aspartate in rats: in vivo and in vitro studies. Brain Res., 861: 143-150. Arnold, D.L., Matthews, RM., Francis, G. and Antel, J. (1990). Proton magnetic resonance spectroscopy of human brain in vivo in the evaluation of multiple sclerosis: assessment of the load of disease. Magn. Reson. Med., 14: 154-159. Baslow, M.H. (1997). A review of phylogenetic and metabolic relationships between the acylamino acids, N-acetyl-L-aspartic acid and N-acetyl-L-histidine, in the vertebrate nervous system. J. Neurochem., 68: 1335-1344. Baslow, M.H. (2000). Functions of N-acetyl-L-aspartate and N-acetylL-aspartylglutamate in the vertebrate brain: role in glial cell-specific signaling. J. Neurochem., 75\ 453-459. Baslow, M.H. (2002). Evidence supporting a role for N-acetyl-Laspartate as a molecular water pump in myelinated neurons in the central nervous system. An analytical review. Neurochem. Int., 40: 295-300.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

314

23. CLINICAL TRIALS OF GENE THERAPY FOR CANAVAN DISEASE

Berger, U.V., Luthi-Carter, R., Passani, L.A., Elkabes, S., Black, L, Konradi, C. and Coyle, J.T. (1999). Glutamate carboxypeptidase II is expressed by astrocytes in the adult rat nervous system. J. Comp. Neurol., 415: 52-64. Bhakoo, K.K., Craig, T.J. and Styles, P. (2001). Developmental and regional distribution of aspartoacylase in rat brain tissue. J. Neurochem., 79: 211-220. Bhakoo, K.K. and Pearce, D. (2000). In vitro expression of N-acetyl aspartate by oligodendrocytes: implications for proton magnetic resonance spectroscopy signal in vivo. J. Neurochem., 74: 254-262. Birnbaum, S.M., Levintow, L., Kingsley, R.B. and Greenstein, J.P. (1952). Specificity of amino acid acylases. J. Biol. Chem., 194: 455. Brady S.T., Witt, A.S., Kirkpatrick, L.L., de Waegh, S.M., Readhead, C , Tu, PH. and Lee, V.M. (1999). Formation of compact myelin is required for maturation of the axonal cytoskeleton. J. Neurosci., 19: 7278-7288. Burlina, A.P., Ferrari, V., Divry, P., Gradowska, W., Jakobs, C , Bennett, M.J., Sewell, A.C., Dionisi-Vici, C. and Burlina, A.B. (1999). N-acetylaspartylglutamate in Canavan disease: an adverse effector? Eur. J. Pediatr., 158: 406-409. Burlina, A.P., Skaper, S.D., Mazza, M.R., Ferrari, V., Leon, A. and Burlina, A.B. (1994). N-acetylaspartylglutamate selectively inhibits neuronal responses to N-methyl-D-aspartic acid in vitro. J. Neurochem., 63: 1174-1177. Caldovic, L., Morizono, H., Yu, X., Thompson, M., Shi, D., Gallegos, R., Allewell, N.M., Malamy, M.H. and Tuchman, M. (2002). Identification, cloning and expression of the mouse N-acetylglutamate synthase gene. Biochem. J., 364: 825-831. Canavan, M.M. (1931). Schilder's encephalitis periaxialis diffusa. Arch. Neurol. Psychiatr., 25: 299-308. Chakraborty G., Mekala, P, Yahya, D., Wu, G. and Ledeen, R.W. (2001). Intraneuronal N-acetylaspartate supplies acetyl groups for myelin lipid synthesis: evidence for myelin-associated aspartoacylase. J. Neurochem., 78: 736-745. Clark, J.B. (1998). N-acetyl aspartate: a marker for neuronal loss or mitochondrial dysfunction. Dev. Neurosci., 20: 271-276. During, M. (1996). Gene therapy in New Zealand. Science 272: 467. During, M.J. and Ashenden, L.M. (1998). Towards gene therapy for the central nervous system. Mol. Med. Today, 4: 485^93. During, M.J., Kaplitt, M.G., Stern, M.B. and Eidelberg, D. (2001). Subthalamic GAD gene transfer in Parkinson disease patients who are candidates for deep brain stimulation. Hum. Gene Then, 12: 1589-1591. Fitzsimons, H.L., Bland, R.J. and During, M.J. (2002). Promoters and regulatory elements that improve adeno-associated virus transgene expression in the brain. Methods, 28: 227-236. Gao, G., Alvira, M.R., Somanathan, S., Lu, Y, Vandenberghe, L.H., Rux, J.J., Calcedo, R., Sanmiguel, J., Abbas, Z. and Wilson, J.M. (2003). Adeno-associated viruses undergo substantial evolution in primates during natural infections. Proc. Natl. Acad. Sci. USA, 100: 6081-6086. George, R.L., Huang, W, Naggar, H.A., Smith, S.B. and Ganapathy V. (2004). Transport of N-acetylaspartate via murine sodium/ dicarboxylate cotransporter NaDC3 and expression of this transporter and aspartoacylase II in ocular tissues in mouse. Biochim. Biophys. Acta., 6: 63-69. Globus, J.H. and Strauss, I. (1928). Progressive degenerative subcortical encephalopathy (Schilder's disease). Arch. Neurol. Psychiatry 20: 1190-1228. Hagenfeldt, L., Bollgren, I. and Venizelos, N. (1987). N-acetylaspartic aciduria due to aspartoacylase deficiency - a new aetiology of childhood leukodystrophy. J. Inherit Metab. Dis., 10: 135-141.

Hanaya, R., Sasa, M., Ujihara, H., Fujita, Y, Amano, T., Matsubayashi, H., Serikawa, T. and Uozumi, T. (1995). Effect of antiepileptic drugs on absence-like seizures in the tremor rat. Epilepsia, 36: 938-942. Higashiguchi, T., Serikawa, T., Yamada, T., Kogishi, K., Kondo, A. and Yamada, J. (1991). Semidominant expression of absence-like seizure in tremor rats. Behav. Genet., 21: 537-545. Hsich, G., Sena-Esteves, M. and Breakefield, X.O. (2002). Critical issues in gene therapy for neurologic disease. Hum. Gene Ther., 13: 579-604. Huang, W, Wang, H., Kekuda, R., Fei, YJ., Friedrich, A., Wang, J., Conway, S.J., Cameron, R.S., Leibach, RH. and Ganapathy, V. (2000). Transport of N-acetylaspartate by the Na^+^-dependent high-affinity dicarboxylate transporter NaDC3 and its relevance to the expression of the transporter in the brain. J. Pharmacol. Exp. Ther., 295: 392-403. Janson, C , McPhee, S., Bilaniuk, L., Haselgrove, J., Testaiuti, M., Freese, A., Wang, D.J., Shera, D., Hurh, P, Rupin, J., Saslow, E., Goldfarb, O., Goldberg, M., Larijani, G., Sharrar, W, Liouterman, L., Camp, A., Kolodny E., Samulski, J. and Leone, P. (2002). Clinical protocol. Gene therapy of Canavan disease: AAV-2 vector for neurosurgical delivery of aspartoacylase gene (ASPA) to the human brain. Hum. Gene Ther., 13: 1391-1412. Janson, C.G., McPhee, S.W., Leone, P., Freese, A. and During, M.J. (2001). Viral-based gene transfer to the mammalian CNS for functional genomic studies. Trends Neurosci., 24: 706-712. KapHtt, M.G., Leone, P , Samulski, R.J., Xiao, X., Pfaff, D.W., O'Malley, K.L. and During, M.J. (1994). Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain. Nat. Genet., 8: 148-154. Kaspar, B.K., Erickson, D., Schaffer, D., Hinh, L., Gage, RH. and Peterson, D.A. (2002). Targeted retrograde gene delivery for neuronal protection. Mol. Ther., 5: 50-56. Kaul, R., Gao, G.P, Balamurugan, K. and Matalon, R. (1993). Cloning of the human aspartoacylase cDNA and a common missense mutation in Canavan disease. Nat. Genet., 5: 118-123. Kekuda, R., Wang, H., Huang, W , Pajor, A.M., Leibach, F.H., Devoe, L.D., Prasad, PD. and Ganapathy, V. (1999). Primary structure and functional characteristics of a mammalian sodium-coupled high affinity dicarboxylate transporter. J. Biol. Chem., 274: 3422-3429. Kirmani, B.F., Jacobowitz, D.M. and Namboodiri, M.A. (2003). Developmental increase of aspartoacylase in oligodendrocytes parallels CNS myelination. Brain Res. Dev Brain Res., 140:105-115. Kitada, K., Akimitsu, T., Shigematsu, Y, Kondo, A., Maihara, T., Yokoi, N., Kuramoto, T, Sasa, M. and Serikawa, T. (2000). Accumulation of N-acetyl-L-aspartate in the brain of the tremor rat, a mutant exhibiting absence-like seizure and spongiform degeneration in the central nervous system. J. Neurochem., 74: 2512-2519. Klugmann, M., Leichtlein, C.B., Symes, C.W., Serikawa, T., Young, D. and During, M.J. (2005). Restoration of aspartoacylase activity in CNS neurons does not ameliorate motor deficits and demyelination in a model of Canavan disease. Mol. Ther., 11: 745-753. Klugmann, M., Symes, C.W, Klaussner, B.K., Leichtlein, C.B., Serikawa, T., Young, D. and During, M.J. (2003). Identification and distribution of aspartoacylase in the postnatal rat brain. Neuroreport, 14: 1837-1840. Klunk, W.E., Panchalingam, K., Moossy, J., McClure, R.J. and Pettegrew, J.W. (1992). N-acetyl-L-aspartate and other amino acid metabolites in Alzheimer's disease brain: a preliminary proton nuclear magnetic resonance study. Neurology, 42: 1578-1585. Kondo, A., Nagara, H., Akazawa, K., Tateishi, J., Serikawa, T. and Yamada, J. (1991). CNS pathology in the neurological mutant rats

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

ACKNOWLEDGMENTS zitter, tremor and zitter-tremor double mutant (spontaneously epileptic rat, SER). Brain, 114: 979-999. Lappe-Siefke, C , Goebbels, S., Gravel, M., Nicksch, E., Lee, J., Braun, RE., Griffiths, I.R. and Nave, K.A. (2003). Disruption of Cnpl uncouples oligodendroglial functions in axonal support and myelination. Nat. Genet., 33: 366-374. Leib, J.R., GoUust, S.E., Hull, S.C. and Wilfond, B.S. (2005). Carrier screening panels for Ashkenazi Jews: is more better? Genet. Med., 7:185-190. Leone, P., Janson, C.G., Bilaniuk, L., Wang, Z., Sorgi, R, Huang, L., Matalon, R., Kaul, R., Zeng, Z., Freese, A., McPhee, S.W., Mee, E. and During, M.J. (2000). Aspartoacylase gene transfer to the mammalian central nervous system with therapeutic implications for Canavan disease. Ann. Neurol., 48: 27-38. Leone, R, Janson, C.G., McPhee, S.J. and During, M.J. (1999). Global CNS gene transfer for a childhood neurogenetic enzyme deficiency: Canavan disease. Curr. Opin. Mol. Ther., 1: 487-492. Lowenstein, PR. (2002). Immunology of viral-vector-mediated gene transfer into the brain: an evolutionary and developmental perspective. Trends Immunol., 23: 23-30. Lu, Z.H., Chakraborty G., Ledeen, R.W., Yahya, D. and Wu, G. (2004). N-Acetylaspartate synthase is bimodally expressed in microsomes and mitochondria of brain. Brain Res. Mol. Brain Res., 122: 71-78. Madhavarao, C.N., Arun, P., Moffett, J.R., Szucs, S., Surendran, S., Matalon, R., Garbem, J., Hristova, D., Johnson, A., Jiang, W. and Namboodiri, M.A. (2005). Defective N-acetylaspartate catabolism reduces brain acetate levels and myelin lipid synthesis in Canavan's disease. Proc. Natl. Acad. Sci. USA, 102: 5221-5226. Madhavarao, C.N., Chinopoulos, C , Chandrasekaran, K. and Namboodiri, M.A. (2003). Characterization of the N-acetylaspartate biosynthetic enzyme from rat brain. J. Neurochem., 86: 824-835. Madhavarao, C.N., Moffett, J.R., Moore, R.A., Viola, R.E., Namboodiri, M.A. and Jacobowitz, D.M. (2004). Immunohistochemical localization of aspartoacylase in the rat central nervous system. J. Comp. Neurol, 472: 318-329. Martin, E., Capone, A., Schneider, J., Hennig, J. and Thiel, T. (2001). Absence of N-acetylaspartate in the human brain: impact on neurospectroscopy? Ann. Neurol., 49: 518-521. Matalon, R. and Michals-Matalon, K. (1999). Biochemistry and molecular biology of Canavan disease. Neurochem. Res., 24: 507-513. Matalon, R., Michals, K., Sebesta, D., Deanching, M., Gashkoff, P and Casanova, J. (1988). Aspartoacylase deficiency and N-acetylaspartic aciduria in patients with Canavan disease. Am. J. Med. Genet., 29: 463-471. Matalon, R., Rady PL., Piatt, K.A., Skinner, H.B., Quast, M.J., Campbell, G.A., Matalon, K., Ceci, J.D., Tyring, S.K., Nehls, M., Surendran, S., Wei, J., Ezell, E.L. and Szucs, S. (2000). Knock-out mouse for Canavan disease: a model for gene transfer to the central nervous system. J. Gene Med., 2:165-175. Matalon, R., Surendran, S., Rady PL., Quast, M.J., Campbell, G.A., Matalon, K.M., Tyring, S.K., Wei, J., Peden, C.S., Ezell, E.L., Muzyczka, N. and Mandel, R.J. (2003). Adeno-associated virusmediated aspartoacylase gene transfer to the brain of knockout mouse for Canavan disease. Mol. Ther., 7: 580-587. Mcintosh, J.C. and Cooper, J.R. (1965). Studies on the function of Nacetyl aspartic acid in brain. J. Neurochem., 12: 825-835. McPhee, S.W, Francis, J., Janson, C.G., Serikawa, T, Hyland, K., Ong, E.O., Raghavan, S.S., Freese, A. and Leone, P. (2005). Effects of AAV-2-mediated aspartoacylase gene transfer in the tremor rat model of Canavan disease. Brain Res. Mol. Brain Res., 135: 112-121.

315

Michaelis, T, Merboldt, K.D., Bruhn, H., Hanicke, W and Frahm, J. (1993). Absolute concentrations of metabolites in the adult human brain in vivo: quantification of localized proton MR spectra. Radiology 187: 219-227. Moreno, A., Ross, B.D. and Bluml, S. (2001). Direct determination of the N-acetyl-L-aspartate synthesis rate in the human brain by (13)C MRS and [l-(13)C]glucose infusion. J. Neurochem., 77: 347-350. Muzyczka, N. and Warrington, K.H., Jr. (2005). Custom adeno-associated virus capsids: the next generation of recombinant vectors with novel tropism. Hum. Gene Ther., 16: 408^16. Neale, J.H., Bzdega, T. and Wroblewska, B. (2000). N-Acetylaspartylglutamate: the most abundant peptide neurotransmitter in the mammalian central nervous system. J. Neurochem., 75: 443-452. Norton, W.T. and Poduslo, S.E. (1973). Myelination in rat brain: method of myelin isolation. J. Neurochem., 21: 749-757. Popko, B. (2003). Myelin: not just a conduit for conduction. Nat. Genet., 33: 327-328. Reynolds, L.M., Cochran, S.M., Morris, B.J., Pratt, J.A. and Rejmolds, G.P (2005). Chronic phencyclidine administration induces schizophrenia-like changes in N-acetylaspartate and N-acetylaspartylglutamate in rat brain. Schizophr. Res., 73:147-152. Richichi, C , Lin, E.J., Stefanin, D., Colella, D., Ravizza, T, Grignaschi, G., Veglianese, P, Sperk, G., During, M.J. and Vezzani, A. (2004). Anticonvulsant and antiepileptogenic effects mediated by adenoassociated virus vector neuropeptide and expression in the rat hippocampus. J. Neurosci., 24: 3051-3059. Sager, T.N., Fink-Jensen, A. and Hansen, A.J. (1997). Transient elevation of interstitial N-acetylaspartate in reversible global brain ischemia. J. Neurochem., 68: 675-682. Sasa, M., Ohno, Y, Ujihara, H., Fujita, Y, Yoshimura, M., Takaori, S., Serikawa, T. and Yamada, J. (1988). Effects of antiepileptic drugs on absence-like and tonic seizures in the spontaneously epileptic rat, a double mutant rat. Epilepsia, 29: 505-513. Seki, T, Matsubayashi, H., Amano, T, Kitada, K., Serikawa, T, Sakai, N. and Sasa, M. (2002). 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., 328: 249-252. Serikawa, T, Ohno, Y, Sasa, M., Yamada, J. and Takaori, S. (1987). A new model of petit mal epilepsy: spontaneous spike and wave discharges in tremor rats. Lab. Anim., 21: 68-71. Surendran, S., Campbell, G.A., Tyring, S.K. and Matalon, R. (2005a). Aspartoacylase gene knockout results in severe vacuolation in the white matter and gray matter of the spinal cord in the mouse. Neurobiol. Dis., 18: 385-389. Surendran, S., Ezell, E.L., Quast, M.J., Wei, J., Tyring, S.K., MichalsMatalon, K. and Matalon, R. (2004a). Aspartoacylase deficiency does not affect N-acetylaspartylglutamate level or glutamate carboxypeptidase II activity in the knockout mouse brain. Brain Res., 6: 268-271. Surendran, S., Michals-Matalon, K., Quast, M.J., Tyring, S.K., Wei, J., Ezell, E.L. and Matalon, R. (2003). Canavan disease: a monogenic trait with complex genomic interaction. Mol. Genet. Metab., 80: 74-80. Surendran, S., Shihabuddin, L.S., Clarke, J., Taksir, T V , Stewart, G.R., Parsons, G., Yang, W , Tyring, S.K., Michals-Matalon, K. and Matalon, R. (2004b). Mouse neural progenitor cells differentiate into oligodendrocytes in the brain of a knockout mouse model of Canavan disease. Brain Res. Dev. Brain Res. 153; 19-27. Surendran, S., Szucs, S., Tyring, S.K. and Matalon, R. (2005b). Aspartoacylase gene knockout in the mouse: Impact on reproduction. Reprod. Toxicol., 20: 281-283.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

316

23. CLINICAL TRIALS OF GENE THERAPY FOR CANAVAN DISEASE

Tallan, H.H. (1957). Studies on the distribution of N-acetyl-L-aspartic acid in brain. J. Biol Chem., 224: 4 1 ^ 5 . Tallan, H.H., Moore, S. and Stein, W.H. (1956). N-Acetyl-L-aspartic acid in brain. J. Biol. Chem., 219: 257-264. Taylor, D.L., Davies, S.E., Obrenovitch, T.P., Urenjak, J., Richards, D.A., Clark, J.B. and Symon, L. (1994). Extracellular N-acetylaspartate in the rat brain: in vivo determination of basal levels and changes evoked by high K+. J. Neurochem., 62: 2349-2355. Traeger, E.C. and Rapin, I. (1998). The clinical course of Canavan disease. Pediatr. Neurol, 18: 207-212. Tsai, G. and Coyle, J. T. (1995). N-acetylaspartate in neuropsychiatric disorders. Prog. NeurobioL, 46: 531-540. Urenjak, J., Williams, S.R., Gadian, D.G. and Noble, M. (1992). Specific expression of N-acetylaspartate in neurons, oUgodendrocytetype-2 astrocyte progenitors, and immature oligodendrocytes in vitro. J. Neurochem., 59: 55-61. Urenjak, J., Williams, S.R., Gadian, D.G. and Noble, M. (1993). Proton nuclear magnetic resonance spectroscopy unambiguously identifies different neural cell types. J. Neurosci., 13: 981-989. Van Bogaert, L. and Bertrand, I. (1949). Sur une idiotie familiale avec degeresce spongUeuse de neuraxe (note preliminaire). Acta. Neurol. Belg., 49: 572-587.

Wang, Z.J. and Zimmerman, R.A. (1998). Proton MR spectroscopy of pediatric brain metabolic disorders. Neuroimaging Clin. N. Am., 8: 781-807. Xu, R., Janson, C.G., Mastakov, M., Lawlor, P., Young, D., Mouravlev, A., Fitzsimons, H., Choi, K.L., Ma, H., Dragimow, M., Leone, P , Chen, Q., Dicker, B. and During, M.J. (2001). Quantitative comparison of expression with adeno-associated virus (AAV-2) brain-specific gene cassettes. Gene Ther., 8:1323-1332. Yamada, J., Serikawa, T., Ishiko, J., Inui, T., Takada, H., Kawai, Y. and Okaniwa, A. (1985). Rats with congenital tremor and curled whiskers and hair. Jikken Dobutsu, 34:183-188. Yan, H.D., Ishihara, K., Serikawa, T. and Sasa, M. (2003). Activation by N-acetyl-L-aspartate of acutely dissociated hippocampal neurons in rats via metabotropic glutamate receptors. Epilepsia, 44: 1153-1159. Zafeiriou, D.I., Kleijer, W.J., Maroupoulos, G., Anastasiou, A.L., Augoustidou-Sawopoulou, P., Papadopoulou, P., Kontopoulos, E.E., Pagan, E. and Payne, S. (1999). Protracted course of N-acetylaspartic aciduria in two non-Jewish siblings: identical clinical and magnetic resonance imaging findings. Brain Dev, 21: 205-208.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

C H A P T E R

24 Gene Therapy for the Late Infantile Form of Batten Disease Dolan Sondhi, Neil R. Hoc/cett, Stephen M. Kaminsky, Mark M. Souweidane, Michael G. Kaplitt, Ronald G. Crystal

Abstract: This chapter describes a program to assess gene transfer as a therapeutic approach to delay the neurological decline in children with the late infantile form of neuronal ceroid lipofuscinosis (LINCL). The disease arises from autosomal recessive inheritance of rare mutations in the CLN2 gene leading to a deficiency in the lysosomal protease tripeptidyl peptidase I (TPP-I). The challenge for a potential treatment is to obtain a therapeutic level of the target protein throughout the brain over the long term. Direct injection into the brain of a gene transfer vector derived from AAV serotype 2, AAV2(^uhCLN2, was chosen as the most easily implemented approach to begin a human clinical study. Limited pre-clinical efficacy studies were performed in rats and monkeys to demonstrate feasibility. Upon proof of concept, a toxicology study and a manufacturing program were executed providing the supporting data for commencing a clinical study in June 2004. This ongoing study is providing an insight on the feasibility of this approach in slowing the neurodegeneration in children with LINCL as well as potentially using similar approaches to treat other neurodegenerative diseases of lysosomal storage. Keywords: AAV2; intracranial gene therapy; clinical trial

1.

IL

INTRODUCTION

There are approximately 200 children in the developed countries in various stages of LINCL. The disease manifests itself at age 2-4 (Kurachi et al., 2000). Although there is variability among individuals as to the time of onset, the major symptoms that bring children to medical attention are seizures, ataxia, myoclonus, impaired speech and developmental regression. Diagnosis is usually made by electron microscopy (EM) of lymphocytes, skin, conjunctiva or rectal tissue, which demonstrate the lysosomal storage bodies that are the hallmark of the disease. Characteristic electroencephalograms (EEG), visual evoked potentials (VEP) and electroretinograms (ERG) confirm the diagnosis (Boustany, 1996; Goebel et a l , 1999; Mole, 1999). Genetic testing is available to determine the specific genes and mutations involved (Dawson and

The neuronal ceroid lipofuscinoses are rare, autosomal recessive genetic lysosomal storage diseases with progressive neurological degeneration leading to death (Boustany, 1996; Goebel et a l , 1999; Wisniewski and Zhong, 2001; Haltia, 2003). While originally classified on the basis of age of onset and histopathology, the neuronal ceroid lipofuscinoses can now be divided into at least eight separate diseases based on molecular genetic studies (Goebel et al., 1999; Wisniewski et al., 2001a, b). In 2001, we irutiated a program to assess gene transfer as a therapeutic approach to delay the neurological decline in children with the late infantile form of neuronal ceroid lipofuscinosis (LINCL), commonly referred to as the late infantile form of Batten disease. This chapter describes the pre-clinical development pathway and the design of the clinical trial.

Gene Therapy of the Central Nervous System: From Bench to Bedside

LINCL

317

Copyright © 2006, Elsevier, Inc. All rights of reproduction in any form reserved.

318

24. GENE THERAPY FOR THE LATE INFANTILE FORM OF BATTEN DISEASE

Cho, 2000). There is evidence that more than one locus may be associated with LINCL (Sleat et a l , 1997). Less definitive diagnostic methods include other neurophysiological studies, observation of a loss of cells within the retina, blood or urine tests (e.g., these patients show elevated levels of dolichol) and brain scans, such as computed tomography (CT) and magnetic resonance imaging (MRI), which reveal the loss of brain tissue. MRI studies show a marked diffuse parenchymal volume loss both infratentorially and supratentorially (Brockmann, 1996; Jarvela et al., 1997; Seitz et a l , 1998; Vanhanen et al., 2004). The ventricles are

FIGURE 1

enlarged, likely because of dilation due to parenchymal volume loss and there is also mild white matter FLAIR hyperintensity (Brockmann et al., 1996; Jarvela et al., 1997; Seitz et al., 1998; Vanhanen et al., 2004; Fig. 1). A gradual decline follows and afflicted children generally become wheelchair bound and blind between 4 and 6 years, with death occurring by ages 8-12 (Williams et al., 1999). At a morphologic level, the disease is characterized by CNS atrophy, with progressive loss of neurons and retinal cells (Boustany, 1996; Birch, 1999; Williams et a l , 1999; Haltia, 2003). Affected cells show characteristic autofluorescent, curvilinear

Magnetic resonance imaging of the brain of a 6-year-old child with LINCL.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

319

CLN2 GENE AND PROTEIN

lysosomal storage bodies. The main component of this storage material is the subunit c of mitochondrial ATP synthetase, suggesting a defect in the turnover of this protein (Palmer et a l , 1992; Umehara et al., 1997; Ezaki et al., 1999). Even though the storage bodies are present in cells throughout the body, the primary pathology and the cause of morbidity is associated with the pathology in the brain. At present, the only treatments available for this devastating disease are palliative, such as anticonvulsants, physical and occupational therapy, adequate feeding and sleep and prevention of physical discomfort (Wisniewski and Zhong, 2001; Haltia, 2003). There are two early stage experimental therapies for Batten disease: the gene therapy clinical trail described in this chapter and a stem cells study that is in the early stages of obtaining clinical approval (http://www. stemcellsinc.com/clinicaltrials/clinicaltrials.html). IIL

The gene is mapped to chromosomal locus l l p l S , is 6.65 kb in length and consists of 13 exons and 12 introns (Sleat et al., 1997). There are at least 24 identified mutations of the CLN2 gene that are associated with LINCL (Sleat et al., 1999). There is a large variability in the progress of the disease in different individuals and this is dependent on the specific mutation. For example, some mutations of the CLN2 gene result in a low but detectable level of enzyme activity, resulting in a delay of the onset of manifestations compared to other CLN2 mutations (Sleat et a l , 1999). The primary translation product of the CLN2 gene is a 563 residue "pre-pro" form of TPP-I that includes a 16 residue signal sequence, a 180 residue propeptide and a 367 residue active mature form (Fig. 2; Lin et al., 2001). Following cleavage of the signal peptide, the 547 amino acid "pro" form is secreted into the endoplasmic reticulum with simultaneous addition of carbohydrates including mannose-6-phosphate (Lin and Lobel, 2001). From the endoplasmic reticulum it is trafficked primarily to the lysosome where the acidic p H induces autolytic cleavage of the propeptide at residues 196-197, resulting in a 367 residue, 46 kDa, active, mature form

CLN2 GENE A N D PROTEIN

LINCL is caused by the autosomal recessive inheritance of mutations in the CLN2 gene (Sleat et al., 1997).

1000

CLN2 gene Basepairs Exons 1 2

3

4

5

6

7

8

9

10 11

13

12

-s-s-a-H-B-+ 11 • *^ * G3556C\R208Stop (splicing) \

Pre-pro-TPP-l

i^^^^^^ Signal sequence

I

(aa1-16)

I

Q422H Glycosylation sites

Y

Y pH 4.0 in lysosome

Propeptide (aa 17-195)

Mature TPP-I Mature TPP-I (aa 196-563) FIGURE 2 Schematic of the CLN2 gene, mutations and the TPP- I protein. The CLN2 gene consists of 13 exons (rectangles) and 12 introns. The locations of the three most common mutations in children with LINCL are shown. The gene is transcribed and spliced onto an mRNA that encodes a 563 amino acid (aa) product (pre-proTPP-I). The signal sequence (aa 1-16) directs the nascent protein to the secretory pathway with insertion into the endoplasmic reticulum and glycosylation at the sites indicated. Pre-pro TPP-I is cleaved to form the 547 residue proTPP-I; this form is inactive until exposed to low p H in the lysosomes which results in proteolytic cleavage at aa 196-197 to yield the active mature product of 367 residues (46 kDa).

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

320

24. GENE THERAPY FOR THE LATE INFANTILE FORM OF BATTEN DISEASE

of the TPP-I protein. But, a small fraction of the newly synthesized pro-TPP-I is trafficked out of the cell (Lin and Lobel, 2001). With a pi of 6.5, it diffuses in the local milieu until endocytosed via the mannose-6-phosphate receptor of nearby cells, delivered to their lysosomes and activated. The mature form of TPP-I functions as a tripeptidyl peptidase, which is presumed to have a role in protein turnover due to the accumulation of subunit c of mitochondrial ATP synthetase in the lysosome of subjects with no TPP-I activity (Palmer et al., 1992; Umehara et al., 1997; Ezaki et al., 1999). It may also have a role in the protection of cells from apoptosis (Lane et a l , 1996; Dhar et a l , 2002). IV. THERAPEUTIC OPTIONS FOR TREATING THE CNS MANIFESTATIONS OF LINCL In a previous review, we proposed that development of gene transfer with AAV2 vectors provided the best option for clinical advancement (Sondhi et al., 2001). This argument is summarized here. Treatment of LINCL presents special challenges due to the fact that the primary pathology of the disease is centered in the brain (Boustany, 1996; Williams et al., 1999; Haltia, 2003). These challenges include: (1) it is assumed that the therapeutic protein, fimctional TPP-I, will need to be provided on a long-term and preferably permanent basis; (2) proteins are unable to cross the blood-brain barrier, which is formed in the vertebrate brain by tight junctions between capillary endothelial cells that isolate the CNS from most circulating macromolecules which may be administered via the bloodstream (Kniesel and Wolburg, 2000); (3) if the TPP-I protein (or CLN2 cDNA) were delivered directly to the CNS to circumvent the blood-brain barrier, it would require the same broad coverage throughout the CNS as the disease pathology; and (4) as with any drug, potential therapies for LINCL therapy will require a satisfactory toxicological profile at the effective dose and this will need to be determined in clinical studies. There are three general therapeutic strategies that have been seriously considered for the treatment of this disorder: (1) allogeneic stem cell therapy; (2) enzyme replacement therapy; and (3) gene therapy. Each option presents unique challenges for delivery, distribution and duration with respect to the known pathology of LINCL and the potential risk/benefit ratio. Within the limits of the current biological understanding, the technology available and the uncertain regulatory barriers, we have concluded that allogenic

stem cell therapy is unlikely to be a viable therapeutic option in the short term. It is unknown if such cells can persist and migrate in the human brain and achieve the levels and spatial distribution of TPP-I required for therapy. Despite these concerns, there is an investigational new drug application currently submitted to the FDA for stem cell therapy for the late-infantile and infantile forms of Batten disease using primary neuronal precursors (http://www.stemcellsinc.com/clinicaltrials/clinicaltrials.html). The safety profile of stem cells is also unknown and clinical study will require cautious dose escalation a n d / o r gradual expansion of the numbers of administration sites with assessment of long-term safety. Enzyme replacement therapy has the significant, but theoretically solvable, challenge of producing the recombinant protein. The extensive resources required to produce the large amounts of recombinant protein required per patient would require the manpower and resources of the commercial biotechnology/pharmaceutical industry for this commercially unviable orphan disease target. Enzyme replacement therapy has been successfully used to treated systemic lysosomal storage diseases such as Fabry's disease (loannou et al., 2001), mucopolysaccharidosis type I (Turner et al., 2000) in animals and the non-neuropathic forms of Gaucher's disease in humans (www.genzyme.com/cerezyme/) with a therapeutic level being attained by weekly intravenous infusions of the recombinant enzyme. For neurological diseases such as LINCL, the blood-brain barrier precludes this approach unless a method could be devised to transiently permeabilize the blood-brain barrier to re-establish sufficient cellular TPP-I levels in the brain to reduce the aberrant lysosomal storage and halt progressive loss of neurons. While this would be a great breakthrough, to develop and validate a safe mechanism to penetrate the blood-brain barrier seems unlikely in the short term. By contrast, gene therapy is a method that has a proven safety profile in humans (Crystal et al., 2002; Harvey et al., 2002) and can provide long-term production of the therapeutic protein in the brain of experimental animals (Frisella et al., 2001; Sondhi et al., 2005). The major challenge of gene therapy is to demonstrate that the currently available vectors are capable of producing sufficient amounts of TPP-I in the appropriate anatomic regions to achieve therapeutic levels to stabilize a n d / o r reverse the progressive CNS deterioration associated with LINCL. The extensive animal data on gene transfer to the brain and a consideration of the biology of LINCL lead us to conclude that gene therapy is a viable option in

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

CHALLENGES FOR EFFECTIVE GENE THERAPY OF LINCL

the short term. Because LINCL is a progressive, fatal disease of children, there is justification for experimental gene therapy for this disorder and a strong impetus to pursue the transition from basic research to clinical study as rapidly as possible. Since the level and spatial requirements for TPP-I that would make an impact on disease in children with LICNL are largely unknown, a decision was made to proceed with the best candidate vector available at the time. These considerations eliminate gene-delivery strategies such as non-viral strategies (due to the low efficiency and transience of gene transfer), and Antivirus and herpes simplex virus (due to the safety concerns in humans). Although, these other approaches may provide effective and safe long-term gene transfer to the brain, there is insufficient safety a n d / o r efficacy data for proceeding with the pre-clinical development of these vectors for LINCL without long development timelines. Since the vast majority of the target cells in the CNS do not proliferate, the gene transfer vector must be capable of effectively transferring a gene to quiescent cells. This constraint eliminates conventional retrovirus vectors based on the Maloney murine leukemia virus, which requires proliferating cells for efficient gene transfer. Retroviruses have been used in human trials with ex vivo transduction of hematologic stem cells, and there has been success with this approach when there is a selective advantage to the transduced cells (Cavazzana-Calvo et al., 2000). Because neither of these options are available to LINCL there is no basis for proceeding with a retroviral gene transfer study for LINCL in humans. Once non-viral, retroviral, lentiviral and herpes viral vectors have been eliminated, there remain two possible gene transfer vectors for direct administration strategies: adenovirus (Ad) and adeno-associated virus (AAV). V.

CHALLENGES FOR EFFECTIVE GENE THERAPY OF LINCL

Constrained to the use of Ad or AAV as the candidate gene transfer vector, there are three challenges that were identified as critical to development for a treatment for LINCL: providing therapeutic levels of TPP-I protein, maintaining this level for sufficient duration and distributing therapeutically relevant levels throughout the brain. The TPP-I protein concentration target is based on the delayed onset of the symptoms of LINCL in subjects, who have CLN2 mutations that reduce but do not eliminate TPP-I activity. There are rare examples

321

of people with 5-10% of normal TPP-I levels that have greatly delayed appearance of symptoms (Sleat et al., 1999). Therefore, we have posited that an effective therapy will achieve 5-10% or more of the normal endogenous TPP-I level. The caveat is that, while this level may be enough to prevent further damage in an initially healthy neuron, there is no evidence to gauge what TPP-I level is required to clear the pre-existing storage defect in a TPP-I-deficient neuron with inclusion bodies. In general, an Ad vector achieve higher expression levels than AAV vectors, especially in the first 2 weeks post-administration (Hackett et a l , 2000; Haskell et a l , 2003). With respect to duration, since TPP-I activity is observed in animals of all ages it is intuitive that persistent expression of the CLN2 gene is required following gene transfer. Therefore, AAV, which provides long-term transgene expression in most tissues including brain, is a prime candidate vector. While Ad is a generally excluded therapeutic option for genetic diseases due to transient expression, it is not clear that it cannot play an important therapeutic role. In this context, we have developed the concept called "setting back the clock," which will be tested in the future in LINCL knockout mice but is summarized here. This concept is based on the knowledge that, in LINCL, inclusion bodies are proposed to slowly accumulate in the neurons and at some threshold neuronal damage results. In this context, theoretically, gene transfer with Ad, which provides high-level TPP-I levels for a few weeks, may be sufficient to clear the inclusion bodies and thereby provide a period of protection from neurological damage. But the kinetics of inclusion body formation are unknown and there may be a burden of protein recycling in the developing brain with a slower rate thereafter. If gene transfer could reduce the amount of inclusion bodies early in development, even by transient expression of a therapeutic protein delivered by an Ad, vector the levels may be brought under the threshold for pathology and restore normal function to the brain in the long term. Despite the appeal of the "setting back the clock" concept with Ad, the primary consideration in choice of vectors is safety and this finally dictated our choice of AAV2 for pre-clinical development. Ad is known to induce a strong anti-vector immune response and injection of adenoviral vectors into brain is known to be inflammatory (Smith et al., 1997; Bohn et al., 1999). By contrast, AAV vectors have consistently given rise to long-term transgene expression in experimental animals and have an excellent safety record in humans, including a few individuals who have received administration to the

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

322

24. GENE THERAPY FOR THE LATE INFANTILE FORM OF BATTEN DISEASE

brain (During et al., 2001; Janson et al., 2002). Although other AAV serotypes yield higher transgene expression level or more widespread transgene expression following administration to the brain, the pre-clinical development pathway, especially for a non-him\an-derived AAV would be much more complex and the danger of unanticipated side effects would be greater. Another important consideration in developing gene therapy for LINCL is how to distribute the vector throughout the CNS. LINCL has a diverse set of symptoms that is not attributable to a focal neurological loss in the brain (Williams et al., 1999; Wisniewski et al., 2001b). Moreover, inspection of MRI images of children with LINCL suggest widespread neural atrophy (Fig. 1). It is known that AAV2 vectors do not spread widely after injection into the brain. A number of methods have been suggested to enhance vector distribution including the use of simultaneous heparin infusion or intra-ventricular vector injection with mannitol to permeabilize the blood-brain barrier (Nguyen et al., 2001; Mastakov et al., 2002a). While demonstration of success in experimental animals is important, the potential for a medical complication confounds pre-clinical development and clinical trails with two simultaneous experimental reagents. Further, the effective distribution of TPP-I in the brain following gene transfer does not require transduction of every cell because the expressed protein provides cross-correction to neighboring cells. As described above, addition of the pro-TPP-I to cells derived from individuals with LINCL results in a reduction in the accumulation of the abnormal storage products (Lin and Lobel, 2001). Therefore, cells that are not corrected through vector transduction can nevertheless acquire wild-type TPP-I protein from a neighboring transduced cell — a process referred to as cross-correction (Sondhi et al., 2005). Further, axonal transport of either vector or pro-TPP-I taken up at the distal end of axons can correct storage disorder at the cell body, which may be several millimeters away (ChamberUn et al., 1998; Sondhi et al., 2001; Kaspar et al., 2002; Passini et al., 2002). Therefore, injection of one region with vector results in TPP-I in distant cell bodies that have axonal projection to the injection site. With extensive knowledge of circuitry of the brain, sites of injection can be chosen to optimize distribution via axonal transport. VL

PRE-CLINICAL EFFICACY STUDIES

Studies of AAV2-mediated gene transfer in the treatment of an animal model of mucopolysaccaharidosis

VII (MPS VII (jS-glucuronidase deficiency or Sly syndrome), a lysosomal storage disorder that bears many similarities to LINCL (Skorupa et al., 1999; Frisella et al., 2001), have established several principles relevant to the development of gene therapy for LINCL. MPS VII mice are characterized by the accumulation of storage granules in CNS neurons (untreated mice live up to 5 months) similar to LINCL. ^-Glucuronidase, the deficient lysosomal enzyme, is normally secreted and cross-corrects neighboring cells via mannose-6phosphate receptor-mediated uptake. Studies with MPSVn mice (ElUger et al., 1999; Skorupa et al., 1999; Stein et al., 1999; Bosch et al., 2000; Davidson et al., 2000; Sferra et al., 2000; Frisella et al., 2001) have established: (1) direct gene transfer to the CNS provides correction of storage defect in the adult brain; (2) transplantation of wild-type cells can correct mutant cells over an area much greater than the region of transplantation thus validating that cross-correction occurs; and (3) direct CNS administration of AAV delivered jS-glucuronidase can reverse storage and behavioral defects and extend survival. Incorporating the considerations for gene transfer for MPS VII, an AAV2 vector-based therapy was investigated for the delivery of therapeutic levels and distribution of TPP-I to the brain (Sondhi et al., 2005). Approval of a human clinical trial was based entirely on pre-clinical efficacy studies in wild-type (CLN2 + / + ) animals (Crystal et al., 2004). A knockout mouse was reported late in the development process (see below), but was never considered critical to the clinical path by the regulatory groups (Sleat et al., 2004). Although scientifically interesting, there is no guarantee that the mouse model would provide an approximation of the human disease due to parallel or alternate pathways that exist in different species. Further, achieving therapeutic benefit in the mouse model would not necessarily predict the same outcome in humans due to the substantial differences in brain size and complex immunological differences between inbred mice and humans. Studies in CLN2 + / + animals required analytical methods that distinguished vector-derived TPP-I from the endogenous levels. To resolve this problem, a vector, AAV2cuhCLN2, with the human cDNA for CLN2 and an optimized Kozak sequence for translation initiation was used in conjunction with a monoclonal antibody specific to the human TPP-I protein. These reagents enabled the detection of vector-derived TPP-I in the context of normal endogenous TPP-I levels of the rat, but not of the monkey. Histology of rat brain using this monoclonal antibody provided evidence

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

PRE'CLINICAL EFFICACY STUDIES

that vector-derived TPP-I is present, however, quantitative measure of TPP-I was not possible nor was the evaluation of the limits of detection possible. In addition, a quantitative, enzymatic activity assay for TPP-I was used that was not species-specific, therefore, only TPP-I levels substantially above background could be reliably detected. These methods can distinguish those areas with TPP-I levels that are well above endogenous (which refers to the levels in normal individuals and is designated as 100%), but does not readily predict the volume around the injection site that will surpass the 5% target for therapeutic benefit. It is reasonable to assume that achieving the therapeutic levels is much broader than the tip of the iceberg seen with this assay of low sensitivity. The AAV2 expression cassette included the CAG promoter on the basis of published data that it produced long-lasting gene expression in the brain via AAV2-mediated gene transfer (Niwa et al., 1991; Daly et al., 1999a, b). The CAG hybrid promoter consists of the enhancer of the cytomegalovirus lEl gene, the chicken j5-actin promoter, splice donor and intron and the splice acceptor of rabbit jS-globin. The polyadenylation/transcription stop signals are also derived from rabbit jS-globin. These signals controlled the expression of the human cDNA for CLN2 with an optimized Kozak context around the start codon. When packaged into the genome of AAV2, this vector is referred to as AAV2cuhCLN2 (Crystal et al., 2004; Sondhi et al., 2005). AAV2-mediated CLN2 gene transfer to the brain of rats was used to establish the general characteristics of TPP-I expression (Sondhi et al., 2005). With respect to time course, TPP-I was first detected by immunohistochemistry at ~4 weeks following gene transfer and extended up to the last time-point studied, which was 18 months (Fig. 3A, B). TPP-I was produced locally after injection of vector into a number of structures within the brain including cerebellum, frontal cortex, parietal cortex and the striatum. In each injected structure, triple immunofluorescence indicated that TPP-I-positive cells were also neuN positive (neurons) and not GFAP-positive cells (astrocytes). With respect to levels, injection of 2.5 X10^ particle units of AAV2cuhCLN2 into striatum resulted in a TPP-I activity 1.5-fold the endogenous level at 4 weeks and remained stable in that range for at least 12 weeks. Therefore, AAV2cuhCLN2 can achieve greater than the therapeutic target, at least locally at the site of injection. TPP-I neurons positive for human TPP-I were detected a significant distance from the vector injection site indicating a mechanism of spread of the vector a n d / or TPP-I protein through the brain possibly by axonal

323

transport and/or cross correction of cells with extended processes (Fig. 3C, D). This even included neurons as far away as the hemisphere contralateral to the injection site (Fig. 3D). It was estimated that -^50% of the striatum and ~5% of the hemisphere contained TPP-I-positive neurons following a single injection of 10^° particle units into the striatum. Finally, while the literature indicates that immunity against a xenotropic gene delivered by AAV depends on the route of administration (Herzog et al., 2002; Song et al, 2002; Arruda et al, 2004; Chenuaud et al., 2004; Couto, 2004; Flotte, 2004; Gao et al., 2004), our studies in rats show that for intracranial injection of human TPP-I, this is not a problem, likely due to the immuno-privileged nature of the brain (Mastakov et al., 2002b). The rat studies indicated that functional TPP-I could be delivered specifically to neurons, even at locations distant from the site of injection without stimulating immunity. However, it remained to be established if the extent of TPP-I activity spread seen in the rat brain, would translate to effective distribution in larger brains such as in non-human primates and ultimately in humans. It is possible that the spread is limited by distance per se or by anatomical barriers that are further apart in larger brains. Therefore, the brain of African green monkeys were injected with AAV2cuhCLN2 and TPP-I expression was assessed by the enzymatic activity assay and immunohistochemistry. The dose was scaled with respect to brain size and, as with rats, TPP-I levels were significantly above the endogenous background at the site of injection suggesting that therapeutic levels could be obtained locally. By immunohistochemistry, TPP-I expression was detected in neurons around the injection site (Sondhi et al., 2005). Since the TPP-I protein in African green monkeys is 93% identical to the human protein, immunohistochemistry reagents required to distinguish vector-encoded human TPP-I from endogenous monkey TPP-I were not as sensitive. Thus, the possible spread of TPP-I from the injection site to neurons in distant brain structures could not be assessed. While our data focused on the use of AAV serotype 2, which facilitated clinical development because of extensive safety data, other investigators have explored the use of adenoviral, AAV serotype 5 and lentiviral vectors for CLN2 gene transfer (Haskell et al., 2003). Similar to our experiments, a single injection of vector encoding the CLN2 gene into the cerebellum or striatum produced TPP-I expression detectable by immunofluorescence. TPP-I enzymatic activity was also detected at levels somewhat higher than those achieved with AAV2 vectors. Use of a jS-galactosidase

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

324

24. GENE THERAPY FOR THE LATE INFANTILE FORM OF BATTEN DISEASE

i:

A. Naive

f^vnimwi

1t.iiiim

iamt

%

.*

C. Axonal transport tosubsantia nigra

D. Axonal transport to contra^teral iiemisptiere

FIGURE 3 TPP-I protein accumulation following AAV2cuhCLN2 gene transfer and expression of TPP-I in structures far from the injection site. Rats were injected into the striatum with 1 |Lil (10^° particle units) of AAV2cuhCLN2 and TPP-I distribution was assessed at 18 months by anti-TPP-I immunoperoxidase staining on sagittal sections of the striatum of the injected hemisphere. (A) naive; (B) AAV2cuhCLN2 at 18 months TPP-I expression was observed in various regions far from the injection site; (C) TPP-I-positive cells in the substantia nigra at 18 months; and (D) TPP-I-positive cells in the contralateral hemisphere. Magnification bar = 1 mm for panels A and B and 50 |LiM for panels C and D.

and TPP-I double expression vector provided evidence for cross-correction through the appearance of TPP-I-positive, j5-galactosidase-negative cells. VII.

PRE^CLINICAL TOXICOLOGY STUDIES

With a complete set of data supporting efficacy, using an AAV2 vector our group proceeded to translate

this toward a clinical study. A decision to proceed with a pre-clinical program involves great cost and commitment of time. Therefore, development of a clinical candidate often occurs with a technology that is no longer state-of-the-art (in this case, AAV2 rather than newer serotypes that may offer better performance) as the various stages of development proceed. We decided to lock in the AAV2 platform with evidence of transgene expression, but not functional clearance of lysosomal storage in a knock out mouse (model available late in

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

PRE'CLINICAL TOXICOLOGY STUDIES

our development process). This decision was made for the following reasons: (1) there is no evidence that the storage defect phenotype is the primary cause of disease therefore establishing a litmus test based on this attribute may be misleading; (2) although alternate serotypes of AAV have shown greater magnitude of transgene expression, the extent of this expression varies from species to species and may not be relevant to humans; (3) the use of AAV2 in clinical studies is well documented with a good safety profile thus improving the likelihood of "doing no harm'' in a clinical trial; (4) therapeutic levels of TPP-I are significantly below the sensitivity of assays in our experimental animal studies, i.e., even if alternative serotypes yield higher levels of expression this may not be an advantage in the context of 5-10% of a normal level as target and the potential for adverse events due to overexpression may become an issue; and (5) the requirements for toxicology and development program for manufacturing and quality control for human use of a new serotype of AAV, for which there is no previous clinical precedence, is likely to be more stringent than that for AAV2. The combination of evidence for long-term, moderate-level TPP-I production following AAV2cuhCLN2 gene transfer along with the demonstration of axonal transport to widen the range of TPP-I distribution beyond the immediate injection site produced justification to proceed to the clinic (Sondhi et al., 2005). Moreover, the demonstration of efficacy in animal models in other lysosomal storage diseases such as MPSVII suggested at least a possibility of a successful outcome of a clinical study for LINCL (Stein et al., 1999; Frisella et al., 2001). Finally, clinical development is a process of iterative improvement and establishing the safety of an AAV2 platform for gene transfer would provide valuable data applicable to other gene transfer vectors for this and related diseases. In light of the limitations of animal models to mimic all aspects of human biology and the urgent medical need for treatments for LINCL, we initiated a pre-clinical development program. This started with a consultation with the FDA for product design, proposals for a toxicology study and for the drug manufacturing process. The supportive response from the FDA arose at least in part from the severity of the disease in combination with an emerging picture of the expected toxicity profile of AAV2 vectors. For example, many groups have studied biodistribution of AAV vectors, the potential spread to gonads as well as the insertion of the vector DNA into the chromosome in testes (Kho et al., 2000; Arruda et al., 2001; Monahan et al., 2002; Couto et a l , 2004; Pachori et al., 2004). These considerations are largely irrelevant

325

for a fatal childhood disease such as LINCL where death is inevitable in the absence of a successful therapy. Therefore, the stringency for toxicological para-meters for a clinical study for a therapeutic for LINCL is reduced, which is analogous to developing a new drug for cancer where a degree of uncertainty on outcorae and even the possibility of side effects are acceptable. Pre-clinical studies to support an investigational new drug application included standard safety and toxicology assessment in rodents. No biodistribution study was required although samples of tissue were retained for future studies. The study involved comparing male and female Fisher 344 rats injected bilaterally into the striatum with AAV2cuhCLN2 vs. phosphate-buffered saline-injected control rats over 18 months. The dose was 10^^ particle units delivered directly to the rat brain and translates, on a weight basis to a dose of approximately 5 X 10^^ in humans. The rats were observed three times a week and morbidity and mortality was noted. At specified times, a subset of rats were sacrificed with assessment of hematological and histopathological parameters. For all timepoints assessed there was no abnormality in the AAV2(^u'^CLN2-treated group for any parameter of complete blood count or serum chemistry and the histopathological data for both groups were the same. Behavioral assessment of rats is relatively unsophisticated and anomalies resulting from vector administration or transgene expression may be difficult to observe. Therefore, non-human primates were chosen as the second species for toxicology studies. Two groups received CNS administration of 3.6 X 10^° and 3.6 X 10^^ particle units of AAV2cuhCLN2 in parallel with various control (AAV2Null with no transgene, phosphatebuffered saline and sham-injected) groups. As planned for the human study, vector was administered to 12 locations through 6 burr holes. As with rats, a subset was sacrificed at intervals up to 1 year for assessment of hematological and histopathological parameters. In addition, the non-human primates in the study were videotaped with a prescribed list of challenges and the responding behaviors were recorded and analyzed via a standardized scoring algorithm. All remaining monkeys were also bled periodically to assess differential and serum chemistry As with the rat toxicology study, there were no abnormalities that were attributable to AAV2cuhCLN2 gene transfer (for an example, see Fig. 4A demonstrating serum liver enzyme values over time). However, there was a minor evidence that the procedure of direct vector administration into 12 sites in the brain could cause local trauma. In this context, histopathological

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

326

24. GENE THERAPY FOR THE LATE INFANTILE FORM OF BATTEN DISEASE

B, AAV2cuhCLN2 Controls 150

AAV2cuhCLN2, 3.6x10^° pu AAV2cuhCLN2, 3.6x10^^ pu

0

S

100

c (0

C. AAV2Nuil

o c E (Q

O

*-•

50

r(0 a

-y/-

0 Pre 0 3

14 28 56 Time (days)

-//90

*mi 180

360

FIGURE 4 Acute effects of direct CNS administration of AAV2cyhCLN2 on serum liver enzyme levels and on histopathology in African green monkeys. African green monkeys (n == 12 per group) were injected with 3 X 10" or 3 X IQio particle units of AAV2cuhCLN2 and controls were injected with AAV2Null, PBS or sham injected. At various time points, all monkeys were sedated and blood was drawn for complete blood count and serum chemistry. (A) Aspartate aminotransferase for all surviving monkeys (mean ± standard error) is shown with the shaded area representing the normal range. At weeks 1, 13, 26 and 52, a subset was sacrificed and histopathology analysis was performed. Examples of the injection site in the caudate are shown for monkeys sacrificed at 1 week with the area of gliosis shown by the dotted circle and the hemosiderin shown with an arrow. (B) AAV2cuhCLN2; (C) AAVNull.

examination showed mild gliosis and hemosiderin in the brain around the injection sites (Fig. 4B, C). This was injection related, was most pronounced at 1 week post-injection, and was resolved at the longer timepoints. There was a similar level of severity in the control and AAV2cuhCLN2 groups. VIIL

M A N U F A C T U R I N G THE CLINICAL GRADE AAV2^nHCLN2 VECTOR

The fact that LINCL is a fatal disease with a very small number of patients also impacts the requirements for manufacturing. In standard drug development, which anticipates large phase III trials and downstream development toward a marketed product, a scalable process and infrastructure development is critical. In the case of LINCL, these issues are moot because there

are only about 200 LINCL patients in the developed world. Therefore, the small-scale production process transferred from the research laboratory with transfection of adherent 293 cells in 10 stack Cell Factory was sufficient to make all of the vector necessary for the phase I/II trial and in the event of clinical benefit may be scaled up with minimal changes. The cells were transfected with two plasmids, the first containing the genome of the AAV2cuhCLN2 vector and the second plasmid pPAK-MA2 (similar to the pDG plasmid used in research laboratories containing the Ad E2, E4, VA) and AAV2 (rep and cap) helper functions (Grimm et al., 1998; Qui et al., 2002). The lysate was collected after 48 h and the vector purified by a iodixanol step gradient density centrifugation followed by heparin affinity chromatography (Zolotukhin et al., 1999). Prior to manufacture of clinical lots, a development program was followed (Fig. 5) to ensure that

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

MANUFACTURING THE CLINICAL GRADE AAV2p,,HCLN2 VECTOR

Certified Cell line

QC Regimen 22 assays for sterility, adventitious agents, identity

Optimize reagents, conditions, timing, and equipment

Develop SOP's and batch records, training

Harvest and purification

Establish methods, incl. gradient & affinity methods and optimize yield

In Process QC, develop SOP's and batch records, training

Product formulation and fill

Develop in house testing for identity, purity, characterization stability protocol

Develop SOP's and batch record, training

Development of double transfection

327

Define lot release criteria

Practice in GMP facility

FIGURE 5 Development of a GMP production scheme for AAV2cuhCLN2. The left vertical path indicates steps in the manufacture of AAV. Boxes to the right of each step are the process development activities, assays and documentation requirements. QC equality control; SOP = standard operating practice; GMP = good manufacturing practice.

a reproducible, well-controlled manufacturing process was available. The purpose of this program is to identify reagents compatible with the requirements of current good manufacturing practice (GMP), to develop standard operating procedures (SOPs), to train personnel and to understand critical aspects of the manufacturing process. Initially, the performance of a working cell bank from a certified human embryonic kidney 293 cell line was assessed to ensure yields were in line with those obtained in the research laboratory. The plasmids needed to make the AAV2cuhCLN2 vector were reengineered to confer kanamycin resistance on the host bacterium to eliminate the use of ampicillin, a potential allergen, which would require residual testing in the final product. The ability to transfect the 293 cells with different reagents and with different amounts of plasmid was assessed and a procedure using the Polyfect reagent (QIAGEN) and 1 mg of Ad/AAV helper plasmid and 0.5 mg of the vector plasmid was found optimal. Various aspects of the purification procedure were also optimized such as the final concentration and buffer exchange for

which a reverse dialysis procedure was devised. This process was sufficient to bring the 200 ml eluted from the heparin column in the last step of the purification to --10 ml of the required concentration. In addition, quality control assays including in vitro gene transfer, replication-competent AAV titer and infectious titer were developed and validated. On the basis of this type of process development, a complete set of SOPs were drafted covering all steps of manufacturing and practice batches were performed under GMP conditions by GMP personnel for the purpose of training and in order to refine and fine tune these SOPs. Following the development phase, seven clinical batches of vector have been purified (Table 1). In general, each batch was sufficient for only one or two patients, but yields have been improved following batch 5 by forced aeration of the cell factory following transfection. The gene transduction titers have been reproducible and stability of this parameter to storage has been demonstrated. In addition to in-house assessment of potency and purity, safety parameters have been assessed by contract laboratories. As for

V GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

328

24. GENE THERAPY FOR THE LATE INFANTILE FORM OF BATTEN DISEASE TABLE 1

Batch

Properties of Clinical Batches of AAV2cuhCLN2« Genomic DNA contamination (ng/dose)*^

Total yield (particle units)

Physical titer by ELISA(particle units/ml)

Genome titer (genome copies/ml)

Physical titer/genome copies

In vitro gene transfer^

Lotl

L3X1012

1.0X1012

3.0 X 1010

35

85

160

Lot 2

5.0X1012

1.0X1012

7.4 X 1010

13.5

250

65

Lots

4.8X10^2

1.2X1012

6.4 X 1010

18.7

297

83

Lot 4

7.5 X 10^2

1.5X1012

7.1 X 1010

21.3

734

60

Lots

2.1 X 10^2

3.0 X 1010

NCy

ND

ND

ND

Lot 6

1.6X1013

1.8X1012

1.1X1010

180

387

55

Lot 7

5.2X10^2

1.1X1012

ND

ND

247

ND

"The characteristics of the seven candidate clinical batches of AAV2(;uhCLN2 are tabulated. ''In a standardized assay, vector (10^° pu) was used to infect 293 cells and the TPP-I activity of the media was assessed after 72 h. "^Genomic DNA contamination was determined by Taqman realtime PCR using the rRNA gene as a target. '^ND = not determined.

toxicology, the fact that we are treating a fatal childhood disease and performing only a phase I study reduces the stringency of testing. For example, the lot release for the AAV2(^uhCLN2 vector contained testing for only a subset of adventious agents compared to the lot release criteria used in applications relevant to a less severely diseased population. The critical drug control parameters that remain relevant are the identity, purity and potency of the active ingredient and the control and stability of this formulated reagent during delivery to the clinical trial subjects. The presence of adventitious agents and residuals of manufacture, likely to be relevant for long-term safety would come to the forefront when a therapeutic benefit was demonstrated. As mentioned above, studies of vector stability under conditions for delivery to the patient remains a critical component of drug development. Under the proposed clinical study, a catheter composed of glass capillary is used for intracranial delivery. It was therefore essential to show that functional gene transfer by AAV2cuhCLN2 vector was not diminished by the injection catheter. Under conditions mimicking the surgery, the vector was recovered from the catheter and the concentration and biological activity was assessed (Fig. 6). The concentrations of AAV2 in the first 50 )il was significantly diminished, possibly due to adherence to the walls of the catheter. The next 300 |Lil was recovered from the catheter with minimal loss in titer and potency. Therefore, during surgery the first 50 jLil of vector is discarded prior to inserting the catheters into the brain, thereby assuring appropriate delivery of functional vector.

IX. CLINICAL PROTOCOL DESIGN The design of clinical studies to assess safety and efficacy of AAV2(3uhCLN2 gene transfer as well as the anticipation of product development was complicated by the small number of patients available. According to the Batten disease registry, there are approximately 200 subjects in the USA with LINCL which is insufficient for traditional phases I, II, III studies consisting of initial safety, dose-escalation and a multi-center, randomized placebo-controlled double blind evaluation of efficacy. Instead, the current clinical study combines safety and efficacy with a single administration at a dose with the potential for therapeutic value (Crystal et al., 2004). The single dose used for the study is a total of 3.6 X 10^^ particle units divided among 12 locations. This is based on extrapolation of TPP-I activity in rats to the desired therapeutic levels and on practical surgical considerations. Injection of 10^° particle units into the brain of rat gave an excess TPP-I activity of 6000 units per gm of brain wet weight compared to the endogenous activity of 20000 units. Assuming linear scaling of expression with dose, the target of 10% endogenous TPP-I activity (i.e., 2000 units) over the whole human brain (1,000 g) would be derived from the injection of 3.3 X 10^^ particle units. An additional constraint is related to the risk of direct surgical injection into the brain. Greater than six burr holes were deemed inadvisable for safety reasons and a slow infusion rate of 2 |Lil/min through each of the six catheters was believed necessary to minimize local damage to the brain. Therefore, limiting time of administration to

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

329 10' j2

o n

o

E I-

E >> (/)

>»

± : T3 * -
c 0 a o

10'

c

I

Jl

I

s

10

3

10

>i •>

U3 O (U "T CL Q.

o Q.

lO""

50

100

150

200

250

300

350

Volume of AAV2hcuCLN2 ejected from catheter (^il) FIGURE 6 Stability of vector in the injection catheter. AAV2cuhCLN2 (2 X IQii p u / m l by ELISA) was loaded into a syringe and the needle/catheter assembly was attached. This was mounted on the delivery pump and vector was collected from the end of the catheter in 25 |il aliquots were collected at a flow rate of 2 jxl per minute. Fractions were assayed in triplicate for physical titer by ELISA (closed symbols) and in duplicate by in vitro functional gene transfer assay (open symbols).

150 min (to limit time under anesthesia) we decided to deliver no more than 300 fil per burr hole (150 |il at each of two depths per burr hole). Since AAV2 vectors tend to aggregate at concentrations of >2 X 10^^ p u / m l we were limited to a total dose of 3.6 X 10^^ particle units. All clinical trials have many challenging issues related to ethics and design. A particularly difficult decision for this clinical trial design was whether to treat LINCL patients with AAV2cuhCLN2 at the mild or severe phase of the disease. For gene therapy this issue is an ongoing dilemma. Due to the risk of direct surgical administration of vector into the brain, we decided to first treat patients with advanced LINCL for whom the impact of possible adverse events would have less impact on life expectancy. This decision is complicated by the fact that advanced LINCL patients have impairment in several organ systems and therefore there is a chance of serious adverse events possibly related to underlying disease but indistinguishable from effects the drug or surgery. In addition, it may be argued that the potential for benefit of gene transfer diminishes as the neurodegeneration in the brain increases. As a compromise, the trial was designed to start with five subjects with advanced disease (designated Group A) and, pending review by the regulatory

agencies, to proceed to six subjects with moderate disease (designated Group B; Crystal et al., 2004). Administration of AAV2(^uhCLN2 to subjects with advanced LINCL also complicates the choice of safety endpoints. The toxicology studies in rats and nonhuman primates provided no evidence to anticipate likely adverse effects. Therefore, aside from the brain, there is no obvious anatomical site of interest for safety assessments. Moreover, the neurological tests and EEC normally used as sensitive tests of drug safety in normal children provide no insight in LINCL patients with severely degraded neurological function. Therefore, a set of standard medical and hematological assessment parameters were used to assess vector safety with the caveat that their significance is better understood in normal subjects than in children with LINCL. Secondary to safety assessment are the efficacy endpoints, which provide additional challenges. The LINCL clinical rating scale in which the broad loss of brain function is the primary readout provided a useful starting point. This clinical rating scale, modified from an original model by Steinfeld et al. (2002), was constructed to provide an integrated sense of each patient's disease severity. The scale is comprised of three functional categories: motor function, seizures and language, each of which are rated

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

330

24. GENE THERAPY FOR THE LATE INFANTILE FORM OF BATTEN DISEASE

on a scale of 0-3 allowing for a total CNS disability score between 0 and 9. Patients receiving total disability ratings between 0 and 4 are categorized in the gene transfer protocol as "severe" or Group A, while those who receive total disability scores of 5-6 are classified as "moderate" or Group B. This scale was modified from an original design by Steinfeld et al. (2002), which included a categorization for vision. We eliminated this category for the following reasons: (1) most patients are blind by the time of enrollment in the trial, largely due to the rapid progression of the disease; and (2) the central nervous system vector delivery route selected for this trial proscribes the possibilities for visual improvement in subjects post vector administration, since retinal cells are not targeted for gene transfer. In addition, the Steinfeld scale categorized patients as "severe" if they received total CNS disability scores between 0 and 3. This was broadened in our modification to account for the fact that for most LINCL patients, seizure activity is adequately controlled by anti-convulsants, and subjects are therefore expected to receive seizure scores of " 3 " (indicating no grand mal seizures in the last 3 months). Unfortunately, the modified LINCL clinical rating scale is relatively insensitive as an efficacy parameter. A favorable outcome for this trial might simply be a decrease in the rate of deterioration of this parameter but given variability among subjects and the relative insensitivity of the scale, it may take a long time and multiple subjects to see this impact. Although the rating scale is the most important tool, we also attempted to use magnetic resonance spectroscopy (MRS) to assess the status of the brain in LICNL patients. Previous studies have shown that there are decreases in N-acetylaspartate (NAA) concentration and increase in lactate concentration in the brains of LINCL patients (Brockmann et al., 1996; Jarvela et al., 1997; Seitz et al., 1998; Vanhanen et al., 2004). These assessments were made by the use of single voxel magnetic resonance spectrometers, but the advent of high magnetic field spectrometers has improved the resolution of these metabolites, thus local effects due to gene therapy may be evident from a lack of decrease in NAA concentration by the expected rate/amount. Therefore, the clinical protocol included both pre-surgical MRS spectroscopy as well as repeated assessments at 6 and 18 months post-therapy. A full clinical protocol based on the considerations above was accepted by the local regulatory agencies (Institutional Review Board and Institutional Biosafety Committee). In conjunction with the efficacy data and toxicology data presented above, it was also reviewed

by the National Institutes of Health Recombinant DNA Advisory Committee, who opted for no detailed review and the Food and Drug Administration. Therefore, subject accrual and gene transfer in those enrolled in the trial is ongoing.

X.

CONCLUSION

The program described above for commencing a clinical trial for LINCL by direct administration of AAV2 vector expressing CLN2 to the brain is a typical example of translational research in the academic setting. While missing some of the stringency that may be applied to a commercial entity in the development of a marketed drug, the program still consumed a substantial amount of resources and time. The overall cost is several million dollars and identification of support on the scale needed for such pre-clinical programs is difficult even in the context of the NIH roadmap for translational research. To maximize the return on the effort invested, other neurological lysosomal storage diseases may be subsequently identified that may benefit from a similar approach and the knowledge learned from the manufacturing program and toxicology studies for LINCL could be applied to these other diseases.

ACKNOWLEDGMENTS We thank N. Mohamed for help in preparing this manuscript. These studies were supported, in part, by UOl NS04758, NIH M01RR00047 from The NIH and The Will Rogers Memorial Foundation, Los Angeles, CA, and Nathan's Battle Foundation, Greenwood, IN. References Arruda, V.R., Fields, RA., Milner, R., Wainwright, L., De Miguel, M.R, Donovan, RJ., Herzog, R.W., Nichols, T.C., Biegel, J.A., Razavi, M., Dake, M., Huff, D., Flake, A.W., Couto, L, Kay, M.A. and High, K.A. (2001) Lack of germline transmission of vector sequences following systemic administration of recombinant AAV-2 vector in males. Mol. Ther., 4: 586-592. Arruda, V.R., Schuettrumpf, J., Herzog, R.W., Nichols, T.C., Robinson, N., Lotfi, Y., Mingozzi, R, Xiao, W., Couto, L.B. and High, K.A. (2004) Safety and efficacy of factor IX gene transfer to skeletal muscle in murine and canine hemophilia B models by adeno-associated viral vector serotype 1. Blood, 103: 85-92. Birch, D.G. (1999) Retinal degeneration in retinitis pigmentosa and neuronal ceroid lipofuscinosis: an overview. Mol. Genet. Metab., 66: 356-366. Bohn, M.C., Choi-Lundberg, D.L., Davidson, B.L., Leranth, C , Kozlowski, D.A., Smith, J.C, O'Banion, M.K. and Redmond, D.E.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

ACKNOWLEDGMENTS (1999) Adenovirus-mediated transgene expression in nonhuman primate brain. Hum. Gene Then, 10: 1175-1184. Bosch, A., Ferret, E., Desmaris, N. and Heard, J.M. (2000) Long-term and significant correction of brain lesions in adult mucopolysaccharidosis type VII mice using recombinant AAV vectors. Mol. Ther., 1: 63-70. Boustany, R.M. (1996) Batten disease or neuronal ceroid lipofuscinosis. In: Moser H.W., Ed., Handbook of Clinical Neurology, Vol. 22, No. 66. Neurodystrophies and Neurolipidoses. Elsevier Science B.V., New York, pp. 671-700. Brockmann, K., Pouwels, P.J., Christen, H.J., Frahm, J. and Hanefeld, F. (1996) Localized proton magnetic resonance spectroscopy of cerebral metabolic disturbances in children with neuronal ceroid lipofuscinosis. Neuropediatrics, 27: 242-248. Cavazzana-Calvo, M., Hacein-Bey, S., de Saint, B.C., Gross, R, Yvon, E., Nusbaum, R, Selz, F , Hue, C , Certain, S., Casanova, J.L., Bousso, R, Deist, FL, and Fischer, A. (2000) Gene therapy of human severe combined immunodeficiency (SCID)-Xl disease. Science, 288: 669-672. Chamberlin, N.L., Du, B., de Lacalle, S. and Saper, C.B. (1998) Recombinant adeno-associated virus vector: use for transgene expression and anterograde tract tracing in the CNS. Brain Res., 793: 169-175. Chenuaud, P., Larcher, T., Rabinowitz, J.E., Rrovost, N., Cherel, Y, Casadevall, N., Samulski, R.J. and Moullier, P. (2004) Autoimmune anemia in macaques following erythropoietin gene therapy Blood, 103: 3303-3304. Crystal, R.G., Harvey, B.-G., Wisnivesky, J.R, O'Donoghue, K.A., Chu, K.W., Maroni, J., Muscat, J.C, Pippo, A.L., Wright, C.E., Kaner, R.K., Leopold, PL., Kessler, P.D., Rasmussen, H.S., Rosengart, T.K. and Hollmann, C. (2002) Analysis of risk factors for local delivery of low and intermediate dose adenovirus gene transfer vectors to individuals with a spectrum of comorbid conditions. Hum. Gene Ther., 13: 65-100. Crystal, R.G., Sondhi, D., Hackett, N.R., Kaminsky, S.M., Worgall, S., Stieg, P., Souweidane, M., Hosain, S., Heier, L., Ballon, D., Dinner, M., Wisniewski, K., Kaplitt, M.G., Greenwald, B.M., Howell, J.D., Strybing, K., Dyke, J. and Voos, H. (2004) Administration of a replication-deficient adeno-associated virus gene transfer vector expressing the human CLN2 cDNA to the brain of children with late infantile neuronal ceriod lipofuscinosis. Hum. Gene Ther., 15: 1131-1154. Couto, L., Parker, A. and Gordon, J.W. (2004) Direct exposure of mouse spermatozoa to very high concentrations of a serotype-2 adeno-associated virus gene therapy vector fails to lead to germ cell transduction. Hum. Gene Ther., 15: 287-291. Couto, L.B. (2004) Preclinical gene therapy studies for hemophilia using adeno-associated virus (AAV) vectors. Semin. Thromb. Hemostat., 30: 161-171. Daly, T.M., Okuyama, T., Vogler, C , Haskins, M.E., Muzyczka, N. and Sands, M.S. (1999a) Neonatal intramuscular injection with recombinant adeno-associated virus results in prolonged betaglucuronidase expression in situ and correction of liver pathology in mucopolysaccharidosis type VII mice. Hum. Gene Ther., 10: 85-94. Daly T.M., Vogler, C , Levy B., Haskins, M.E. and Sands, M.S. (1999b) Neonatal gene transfer leads to widespread correction of pathology in a murine model of lysosomal storage disease. Proc. Natl. Acad. Sci. USA, 96: 2296-2300. Davidson, B.L., Brooks, A.I., Stein, C.S., Heth, J.A., Dubernsky, T.W., Sauter, S.L., Cory-Slechta, D.A. and Federoff, H.J. (2000) Correction of cellular pathology and behavioral deficits in adult

331

b-glucuronidase-deficient mice after FIV vector-mediated gene transfer to brain. Mol. Ther., 1: A688. Dawson, G. and Cho, S. (2000) Batten's disease: clues to neuronal protein catabolism in lysosomes. J. Neurosci. Res., 60: 133-140. Dhar, S., Bitting, R.L., Rylova, S.N., Jansen, P.J., Lockhart, E., Koeberl, D.D., Amalfitano, A. and Boustany, R.M. (2002) Flupirtine blocks apoptosis in batten patient lymphoblasts and in human postmitotic CLN3- and CLN2-deficient neurons. Ann. Neurol., 51: 448-466. During, M.J., Kaplitt, M.G., Stern, M.B. and Eidelberg, D. (2001) Subthalamic GAD gene transfer in Parkinson disease patients who are candidates for deep brain stimulation. Hum. Gene Ther., 12:1589-1591. Elliger, S.S., EUiger, C.A., Aguilar, C.R, Raju, N.R. and Watson, G.L. (1999) Elimination of lysosomal storage in brains of MPS VII mice treated by intrathecal administration of an adeno-associated virus vector. Gene Ther., 6: 1175-1178. Ezaki, J., Tanida, I., Kanehagi, N. and Kominami, E. (1999) A lysosomal proteinase, the late infantile neuronal ceroid lipofuscinosis gene (CLN2) product, is essential for degradation of a hydrophobic protein, the subunit c of ATP synthase. J. Neurochem., 72: 2573-2582. Flotte, T.R. (2004) Immune responses to recombinant adeno-associated virus vectors: putting preclinical findings into perspective. Hum. Gene Ther., 15: 716-717. Frisella, WA., O'Connor, L.H., Vogler, C.A., Roberts, M., Walkley, S., Levy, B, Daly, T.M. and Sands, M.S. (2001) Intracranial injection of recombinant adeno-associated virus improves cognitive function in a murine model of mucopolysaccharidosis type VII. Mol. Ther., 3: 351-358. Gao, G., Lebherz, C , Weiner, D.J., Grant, R., Calcedo, R., McCuUough, B., Bagg, A., Zhang, Y. and Wilson, J.M. (2004) Erythropoietin gene therapy leads to autoimmune anemia in macaques. Blood, 103:3300-3302. Grimm, D., Kern, A., Rittner, K. and Kleinschmidt, J.A. (1998) Novel tools for production and purification of recombinant adenoassociated virus vectors. Hum. Gene Ther., 9: 2745-2760. Goebel, H.H., Mole, S.E. and Lake, B.D. (1999) The Neuronal Ceroid Lipofuscinoses (Batten Disease). lOS Press, Amsterdam. Hackett, N.R. and Crystal, R.G. (2000) Adenovirus vectors for gene therapy. In: Smyth Tempelton N. and Lasic D.D., (Eds.), Gene Therapy. Marcel Dekker, New York, pp. 17-40. Haltia, M. (2003) The neuronal ceroid-lipofuscinoses. J. Neuropathol. Exp. Neurol., 62: 1-13. Harvey, B.-G., Maroni, J., O'Donoghue, K.A., Chu, K.W, Muscat, J.C, Pippo, A.L., Wright, C.E., HoUmarm, C , Wisnivesky, J.R, Kessler, R, Rasmussen, H., Rosengart, T.K. and Crystal, R.G. (2002) Safety of local delivery of low and intermediate dose adenovirus gene transfer vectors to individuals with a spectrum of comorbid conditions. Hum. Gene Ther., 13; 15-63. Haskell, R.E., Hughes, S.M., Chiorini, J.A., Alisky J.M. and Davidson, B.L. (2003) Viral-mediated delivery of the late-infantile neuronal ceroid lipofuscinosis gene, TPP-I to the mouse central nervous system. Gene Ther. 10: 34-42. Herzog, R.W., Fields, PA., Arruda, V.R., Brubaker, J.O., Armstrong, E., McClintock, D., Bellinger, D.A., Couto, L.B., Nichols, T.C. and High, K.A. (2002) Influence of vector dose on factor IX-specific T and B cell responses in muscle-directed gene therapy. Hum. Gene Ther., 13: 1281-1291. loannou, Y.A., Zeidner, K.M., Gordon, R.E. and Desnick, R.J. (2001) Fabry disease: preclinical studies demonstrate the effectiveness of alpha-galactosidase. A replacement in enzyme-deficient mice. Am. J. Hum. Genet., 68:14-25.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

332

24. GENE THERAPY FOR THE LATE INFANTILE FORM OF BATTEN DISEASE

Janson, C , McPhee, S., Bilaniuk, L., Haselgrove, J., Testaiuti, M., Freese, A., Wang, D.J., Shera, D., Hurh, P., Rupin, J., Saslow, E., Goldfarb, O., Goldberg, M., Larijani, G., Sharrar, W., Liouterman, L., Camp, A., Kolodny, E., Samulski, J. and Leone, P. (2002) Clinical protocol. Gene therapy of Canavan disease: AAV-2 vector for neurosurgical delivery of aspartoacylase gene (ASPA) to the human brain. Hum. Gene Then, 13: 1391-1412. Jarvela, I., Autti, T., Lamminranta, S., Aberg, L., Raininko, R. and Santavuori, P. (1997) Clinical and magnetic resonance imaging findings in Batten disease: analysis of the major mutation (1.02kb deletion). Ann. Neurol., 42: 799-802. Kaspar, B.K., Erickson, D., Schaffer, D., Hinh, L., Gage, EH. and Peterson, D.A. (2002) Targeted retrograde gene delivery for neuronal protection. Mol. Then, 5: 50-56. Kho, S.T., Pettis, R.M., Mhatre, A.N. and Lalwani, A.K. (2000) Safety of adeno-associated virus as cochlear gene transfer vector: analysis of distant spread beyond injected cochleae. Mol. Then, 2: 368-373. Kniesel, U. and Wolburg, H. (2000) Tight junctions of the bloodbrain barrien Cell. Mol. NeurobioL, 20: 57-76. Kurachi, Y, Oka, A., Mizuguchi, M., Ohkoshi, Y, Sasaki, M., Itoh, M., Hayashi, M., Goto, Y and Takashima, S. (2000) Rapid immunologic diagnosis of classic late infantile neuronal ceroid lipofuscinosis. Neurology, 54: 1676-1680. Lane, S.C., Jolly, R.D., Schmechel, D.E., Alroy, J. and Boustany, R.M. (1996) Apoptosis as the mechanism of neurodegeneration in Batten's disease. J. Neurochem., 67: 677-683. Lin, L. and Lobel, P. (2001) Production and characterization of recombinant human CLN2 protein for enzyme-replacement therapy in late infantile neuronal ceroid lipofuscinosis. Biochem. J., 357: 49-55. Lin, L., Sohar, I., Lackland, H. and Lobel, P (2001) The human CLN2 protein/tripeptidyl-peptidase I is a serine protease that autoactivates at acidic pH. J. Biol. Chem., 276: 2249-2255. Mastakov, M.Y, Baer, K., Kotin, R.M. and During, M.J. (2002a) Recombinant adeno-associated virus serotypes 2- and 5-mediated gene transfer in the mammalian brain: quantitative analysis of heparin co-infusion. Mol. Then, 5: 371-380. Mastakov, M.Y, Baer, K., Symes, C.W., Leichtlein, C.B., Kotin, R.M. and During, M.J. (2002b) Immunological aspects of recombinant adeno-associated virus delivery to the mammalian brain. J. Virol., 76: 8446-8454. Mole, S.E. (1999) Batten's disease: eight genes and still counting? Lancet, 354: 443-445. Monahan, P.E., Jooss, K. and Sands, M.S. (2002) Safety of adenoassociated virus gene therapy vectors: a current evaluation. Expert Opin. Drug Safety, 1: 79-91. Nguyen, J.B., Sanchez-Pemaute, R., Cunningham, J. and Bankiewicz, K.S. (2001) Convection-enhanced delivery of AAV-2 combined with heparin increases TK gene transfer in the rat brain. Neuroreport, 12: 1961-1964. Niwa, H., Yamamura, K. and Miyazaki, J. (1991) Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene, 108: 193-199. Pachori, A.S., Melo, L.G., Zhang, L., Loda, M., Pratt, R.E. and Dzau, VJ. (2004) Potential for germ line transmission after intramyocardial gene delivery by adeno-associated virus. Biochem. Biophys. Res. Commun., 313: 528-533. Palmer, D.N., Fearnley, I.M., Walker, J.E., Hall, N.A., Lake, B.D., Wolfe, L.S., Haltia, M., Martinus, R.D. and Jolly R.D. (1992) Mitochondrial ATP synthase subunit c storage in the ceroid-lipofuscinoses (Batten disease). Am. J. Med. Genet., 42: 561-567. Passini, M.A., Lee, E.B., Heuer, G.G. and Wolfe, J.H. (2002) Distribution of a lysosomal enzyme in the adult brain by axonal transport

and by cells of the rostral migratory stream. J. Neurosci., 22: 6437-6446. Qui, J.P, Mendez, B.S., Crystal, R.G. and Hackett, N.R. (2002) Construction and verification of an Ad/AAV helper plasmid designed for manufacturing recombinant AAV vectors for human administration. Mol. Then, 5: S47-S48. Seitz, D., Grodd, W, Schwab, A., Seeger, U., Klose, U. and Nagele, T. (1998) MR imaging and localized proton MR spectroscopy in late infantile neuronal ceroid lipofuscinosis. Am. J. NeuroradioL, 19: 1373-1377. Sferra, T.J., Qu, G., McNeely D., Rennard, R., Clark, K.R., Lo, W D . and Johnson, PR. (2000) Recombinant adeno-associated virusmediated correction of lysosomal storage within the central nervous system of the adult mucopolysaccharidosis type VII mouse. Hum. Gene Then, 11: 507-519. Skorupa, A.E, Fisher, K.J., Wilson, J.M., Parente, M.K. and Wolfe, J.H. (1999) Sustained production of beta-glucuronidase from localized sites after AAV vector gene transfer results in widespread distribution of enzyme and reversal of lysosomal storage lesions in a large volume of brain in mucopolysaccharidosis VII mice. Exp. Neurol., 160: 17-27. Sleat, D.E., Donnelly, R.J., Lackland, H., Liu, C.G., Sohar, I., Pullarkat, R.K. and Lobel, P. (1997) Association of mutations in a lysosomal protein with classical late-infantile neuronal ceroid lipofuscinosis. Science, 277: 1802-1805. Sleat, D.E., Gin, R.M., Sohar, I., Wisniewski, K., Sklower-Brooks, S., Pullarkat, R.K., Palmer, D.N., Lerner, T.J., Boustany, R.M., Uldall, P, Siakotos, A.N., Donnelly, R.J. and Lobel, P (1999) Mutational analysis of the defective protease in classic late-infantile neuronal ceroid lipofuscinosis, a neurodegenerative lysosomal storage disorden Am. J. Hum. Genet., 64: 1511-1523. Sleat, D.E., Wiseman, J.A., El Banna, M., Kim, K.H., Mao, Q., Price, S., Macauley, S.L., Sidman, R.L., Shen, M.M., Zhao, Q., Passini, M.A., Davidson, B.L., Stewart, G.R. and Lobel, P (2004) A mouse model of classical late-infantile neuronal ceroid lipofuscinosis based on targeted disruption of the CLN2 gene results in a loss of tripeptidyl-peptidase I activity and progressive neurodegeneration. J. Neurosci., 24: 9117-9126. Smith, J.G., Raper, S.E., Wheeldon, E.B., Hackney, D., Judy, K., Wilson, J.M. and Eck, S.L. (1997) Intracranial administration of adenovirus expressing HSV-TK in combination with ganciclovir produces a dose-dependent, self-limiting inflammatory response. Hum. Gene Then, 8: 943-954. Sondhi, D., Hackett, N.R., Apblett, R.L., Kaminsky, S.M., Pergolizzi, R.G. and Crystal, R.G. (2001) Feasibility of gene therapy for late neuronal ceroid lipofuscinosis. Arch. Neurol., 58:1793-1798. Sondhi, D., Peterson, D.A., Giannaris, E.L., Sanders, C.T., Mendez, B.S., Bishnu, D., Rostkowski, A., Blanchard, B., Bjugstad, K., Sladek, J.R.J., Redmond, D.E., Leopold, PL., Kaminsky, S.M., Hackett, N.R. and Crystal, R.G. (2005) AAV2-mediated CLN2 gene transfer to rodent and non-human primate brain results in long-term TPP-I expression compatible with therapy for LINCL. Gene Then, 12: doi: 10.1038/sj.gt.3302549. Song, S., Scott-Jorgensen, M., Wang, J., Poirier, A., Crawford, J., Campbell-Thompson, M. and Flotte, T.R. (2002) Intramuscular administration of recombinant adeno-associated virus 2 alpha-1 antitrypsin (rAAV-SERPINAl) vectors in a nonhuman primate model: safety and immunologic aspects. Mol. Then, 6: 329-335. Stein, C.S., Ghodsi, A., Derksen, T. and Davidson, B.L. (1999) Systemic and central nervous system correction of lysosomal storage in mucopolysaccharidosis type VII mice. J. Virol., 73: 3424-3429.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

ACKNOWLEDGMENTS Steinfeld, R., Heim, P., von Gregory, H., Meyer, K., Ullrich, K., Goebel, H.H. and Kohlschutter, A. (2002) Late infantile neuronal ceroid lipofuscinosis: quantitative description of the clinical course in patients with CLN2 mutations. Am. J. Med. Genet., 112: 347-354. Turner, C.T., Hopwood, J.J. and Brooks, D.A. (2000) Enzyme replacement therapy in mucopolysaccharidosis I: altered distribution and targeting of alpha-L-iduronidase in immunized rats. Mol. Genet. Metab., 69: 277-285. Umehara, R, Higuchi, I., Tanaka, K., Niiyama, T., Ezaki, J., Kominami, E. and Osame, M. (1997) Accumulation of mitochondrial ATP S5mthase subunit c in muscle in a patient with neuronal ceroid lipofuscinosis (late infantile form). Acta Neuropathol., 93: 628-632. Vanhanen, S.L., Puranen, J., Autti, T., Raininko, R., Liewendahl, K., Nikkinen, P., Santavuori, P., Suominen, P., Vuori, K. and Hakkinen, A.M. (2004) Neuroradiological findings (MRS, MRI, SPECT) in infantile neuronal ceroid-lipofuscinosis (infantile CLNl) at different stages of the disease. Neuropediatrics, 35: 27-35.

333

Williams, R.E., Gottlob, I., Lake, B.D., Goebel, H.H., Winchester, E.G. and Wheeler, R.B., CLN 2. (1999) Classic late infantile NCL. In: Goebel H.H., Mole S.E. and Lake B.D., (Eds.), The Neuronal Ceroid Lipofuscinoses (Batten Disease). lOS Press, Amsterdam, pp. 37-55. Wisniewski, K.E., Kida, E., Golabek, A.A., Kaczmarski, W , Connell, F. and Zhong, N. (2001a) Neuronal ceroid lipofuscinoses: classification and diagnosis. Adv. Genet., 45:1-34. Wisniewski K. and Zhong N. (Eds.) (2001) Batten Disease: Diagnosis, Treatment and Research. Academic Press, New York. Wisniewski, K.E., Zhong, N. and Philippart, M. (2001b) Pheno/genotypic correlations of neuronal ceroid lipofuscinoses. Neurology, 51: 576-581. Zolotukhin, S., Byrne, B.J., Mason, E., Zolotukhin, I., Potter, M., Chesnut, K., Summerford, C , Samulski, R.J. and Muzyczka, N. (1999) Recombinant adeno-associated virus purification using novel methods improves infectious titer and yield. Gene Ther., 6: 973-985.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

C H A P T E R

25 Molecular Imaging of Gene Therapy for Neurogenetic Diseases ]ohn H. Wolfe, Paul D. Acton, Harish Poptani, Charles H. Vite

Abstract: Gene transfer and stem cell transplantation hold great promise for treating neurogenetic diseases. Animal studies rely on post-mortem analysis of experimentally treated brains at intervals after the treatment is started. For human clinical application, it will be necessary to develop non-invasive methods to monitor the function of the transferred gene or stem cells, as well as the therapeutic effects, longitudinally in individual patients. Imaging modalities being developed to monitor gene therapy in the central nervous system include raagnetic resonance imaging and spectroscopy, radioisotope probes, optical methods, and others. Each modality has specific strengths and limitations imposed by the biological properties or processes being monitored, as well as those imposed by the chemistry and physics associated with the instrumentation and the functions being measured. Rapid progress is being made on developing novel reagents, which promises to expand the applicability for imaging modalities for neurogenetic disease studies. Keywords: non-invasive imaging; nuclear magnetic resonance (NMR); magnetic resonance imaging (MRI); magnetic resonance spectroscopy (MRS); diffusion-weighted imaging (DwI); diffusion-weighted spectroscopy (DwS); magnetization transfer imaging (MTI); radiotracer; positron emission tomography (PET); single photon emission computed tomography (SPECT); stem cell tracking; superparamagnetic iron oxide (SPIO) particles; reporter genes; pathologic imaging; optical imaging; bioluminescence; nearinfrared; animal models

L

blood-brain barrier (BBB), and the regional specialization of functions within the CNS. Experiments in mouse models have shown that if the gene is delivered to the target cells it can mediate the needed change. Treating the human brain is, however, a much more formidable challenge because it is about 2000 times larger in a 1-year-old child than in an adult mouse. Studies in some large animal models (cats, dogs, and non-human primates) have shown that it should be possible to scale up treatments to childrens' brains for certain types of diseases. This must be accompanied by the ability to monitor the function of the transferred gene and the therapeutic effects over the course of the disease, in individual patients using non-invasive methods.

G E N E T I C DISEASES OF THE C N S

A large number of single-gene deficiency diseases involve the central nervous system (CNS) and have generally been refractory to treatment (Scriver et a l , 2001). Most genetic diseases involve metabolic deficiencies, which typically affect cells throughout the brain, thus they require global correction. Effective treatment will need either widespread delivery of gene transfer vectors, widespread dissemination of the therapeutic gene product, or a combination. Gene therapy approaches for the brain require special methods due to the physical inaccessibility imposed by the skull, the isolation from the circulation by the

Gene Therapy of the Central Nervous System: From Bench to Bedside

335

Copyright © 2006, Ebevier, Inc. All rights of reproduction in any form reserved.

336

25. MOLECULAR IMAGING OF GENE THERAPY FOR NEUROGENETIC DISEASES

IL

IMAGING MODALITIES

A number of imaging modalities are being used and developed for monitoring gene therapy in the CNS, including nuclear magnetic resonance (NMR), positron emission tomography (PET), and optical methods. Each modality has specific strengths and limitations imposed by the biological properties or processes being monitored, as well as imposed by the chemistry and physics associated with each method (also see the following recent reviews, Gelovani, Tjuvajev and Blasberg, 2003; Massoud and Gambhir, 2003; Herschman, 2004; Min and Gambhir, 2004). Imaging modalities are generally used in three ways: to monitor activity of the gene transfer vector; to track transplanted cells; or to assess progression or resolution of neuropathology. The detection method must be sensitive enough to distinguish between normal and diseased brain, a n d / o r to identify transferred cells or gene activity against a background of endogenous tissue. The most advanced imaging methods for detecting reporter gene activity at the present time use radioisotope tracers. PET and the closely related single photon emission computed tomography (SPECT) measure the three-dimensional distribution of a radioactively labeled tracer in vivo. The main advantage of radioisotope detection is its high sensitivity, which can measure nanomolar (<10~^ M) concentrations of a tracer. The mass of radiotracer injected into the subject is extremely small and the radioactive half-life is very short (minutes to hours), thus the tracers generally do not have adverse physiological effects on the subject being studied. The major drawback of radiotracer methods is that the spatial resolution is poor compared to magnetic resonance imaging (MRI). A number of PET reporter genes have been developed, which use one of two biochen\ical principles: binding of a radioactive tracer to the reporter gene-encoded protein, usually a modified receptor; or accumulation of a metabolite of the radioactive molecule within cells expressing a gene that modifies the tracer, such as an enzyme. Not all of the reporter genes can be used in the brain, due to the inability of many probes to cross the intact BBB or due to the presence of endogenous receptors and neurotransmitters. NMR methods have been used extensively for the non-invasive study of intracranial disease. NMR methods detect changes in the spins of protons in the presence of magnetic fields. The method produces high-resolution images of brain anatomy, known as MRL Specialized applications of MRI can detect interactions between protons in water and protons in

structural macromolecules, such as myelin, thus they can be used to examine changes in myelination. The concentrations of specific metabolites can be determined using magnetic resonance spectroscopy (MRS). However, the number of different molecules that can be detected by this method is very limited. A technique called diffusion weighting (Dw), which can be used with imaging or spectroscopy, maps the microscopic motion of water protons. The motion can be distinguished between extra- and intracellular microenvironments, which provides a way to measure physiologic changes in the diseased brain. Combining different NMR methods can be used to study the location, anatomy, and biochemistry of a disease process. Optical methods have also been used for reporter gene activity. The most widely used are proteins that emit a fluorescent signal (e.g. gfp) or enzymes that cleave substrates into molecules that emit light in the visible spectrum (e.g. luciferase). The light path traveled by the emitted signal is very short and bone is a significant barrier, consequently they are generally limited to use outside the CNS. Near-infrared spectroscopy has the potential benefit of penetrating the skull, but to date no reporter gene has been devised to work in this spectrum. Other biological imaging methods, such as x-ray, computer-assisted tomography, or ultrasound, appear to have less promise for gene therapy applications. However, these may be combined with other modalities in cross-registration uses to take advantage of the inherent properties of more than one modality. This may be especially useful for neurosurgeons in achieving optimal injection strategies to deliver vectors or cells to the best target structures. III.

ANIMAL MODELS

The amount of imaging data on human patients for specific genetic disorders is relatively small due to the rarity of each disease and the variability of brain involvement. This problem can be addressed by studying animal disease models, where sufficient number of animals with a specific mutation, causing relatively uniform disease characteristics, can be produced and studied over time. The availability of large animal models of human diseases allows one to examine both regional and whole brain disease progression using NMR methods (MRI, magnetization transfer imaging (MTI), and MRS, including Dw studies) that are technically difficult to perform in the small brain of mouse models. A large brain may also allow sufficient

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

GENE TRANSFER IN THE CNS

separation of PET reporter genes from endogenous positive areas to overcome the limitation of low spatial resolution inherent in this modality. Animal studies can most effectively assess potential therapeutic efficacy if the animal model is a true homologue of the human disease. An animal homologue in this context is not merely a "model'' of the human disease, but has a mutation in the orthologous gene producing biochemical, pathological, and clinical abnormalities that are essentially the same as those in human patients. Many models have been discovered in large mammals, mostly in dogs and cats, and the mutations captured in breeding colonies (reviewed in Patterson et al., 1988; Wolfe, 1994; Raskins et a l , 1997; Watson and Wolfe, 2003; EUinwood et al., 2004). Large animal homologues of human genetic diseases are particularly useful for evaluating methods and procedures prior to use in human patients because the clinical logistics required for autologous transplantation of cells or direct transduction of cells are similar to the requirements in children. Large animals such as cats and dogs live long lives compared to rodents, which permits repeated biochemical, pathological, and clinical evaluation to assess the efficacy and safety of the treatment. The clinical veterinary medical knowledge about these species rivals that which is available in human medicine. There also is a large body of knowledge on the neurophysiology and neuroanatomy of the cat brain in particular, due to decades of use as one of the standard animal models for neuroscience. The large animal models are particularly important for translational studies of treatment for the CNS, because they have brains that are much closer in size as well as physical organization to humans. Most genetic diseases affect children in the first year of life and treatments are generally more effective the earlier they are initiated. The human infant brain is about 350 g at birth and grows to about 800 g at 1 year of age. The adult cat brain is about 35 g and adult dog brain is about 50-80 g, but the adult mouse brain is only 400 mg. Thus, there is a 2000-fold difference between the brains of a mouse and a 1-year-old child, while the scale up is only 10-20-fold from a dog or cat brain. The brain size also has a significant effect on the kinds of imaging studies and analyses that can be done in the CNS. The instrumentation used for patients can often be applied to cats and dogs, whereas the small size of the mouse brain mandates the use of specialized instrumentation to achieve the resolution needed. The most important limitation of going from rodent models to human brains for neurogenetic diseases

337

is the need to achieve widespread distribution of the therapeutic protein, because lesions are usually present throughout the CNS. Studying gene transfer under actual disease conditions can reveal not only the potential for therapeutic efficacy, but also the limitations that are likely to be encountered in the human brain. In certain types of diseases, e.g. in lysosomal storage disorders, murine models have shown that very low levels of the normal protein (enzymes) can correct large numbers of cells, because the enzyme is secreted by a genetically corrected cell and taken up by adjacent and even distal cells (Taylor and Wolfe, 1997; Passini et al., 2002; Hennig et a l , 2003). Similarly studies in a feline model of a-mannosidosis showed that approximately 4% of normal enzyme expressed after adeno-associated virus (AAV)-mediated gene transfer to multiple sites in the brain resulted in widespread reduction in storage lesions and there was improvement, not just stabilization, of the clinical signs (Vite et al., 2005). Thus, the solutions devised for the problems encountered in treating large animals should be applicable to human patients. IV-

GENE TRANSFER IN THE CNS

Reporter genes, such as jS-galactosidase and luciferase, have played vital roles in the understanding of the molecular mechanisms of gene expression. The concept of a reporter gene provides a much simpler method for genetic analysis than the conventional hybridization techniques. A defined nucleotide sequence is introduced into a cell, whose expression results in a well-defined protein, which can be measured using the reporter probe. These reporter genes also have provided important information in the design of delivery systems in gene transfer into somatic tissue. However, monitoring these reporter systems in living animals requires either highly invasive tissue biopsy or post-mortem analysis of the animal under study. Optical imaging techniques, using GFP and luciferase, have been developed to provide non-invasive monitoring of gene expression in vivo, although these techniques are limited to either organisms which are transparent to light, or, in larger animals, to detection of gene expression near the surface of the skin (Bhaumik and Gambhir, 2002). In addition, accurate quantification of the emitted light is difficult due to absorption of the light by the animal. Detection in the CNS is further compromised by the inability of light to penetrate the skull. Therefore, techniques are required which enable repetitive, non-invasive, quantitative imaging

V GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

338

25. MOLECULAR IMAGING OF GENE THERAPY FOR NEUROGENETIC DISEASES

of reporter genes in living animals, which should be equally applicable to human studies. v.

RADIOISOTOPE METHODS (PET A N D SPECT)

PET and SPECT measure the three-dimensional distribution of a radioactively labeled tracer in vivo. PET imaging uses radioisotopes which decay with the emission of a positron, which travels a short distance in tissue before annihilating with an electron to form two colinear photons, each with 511 keV energy Radioisotopes used with SPECT emit just a single y-ray, with energies dependent on the specific radioisotope used. The technology for detecting these y-rays differs between the two systems. PET generally utilizes a ring of radiation detectors surrounding the subject, which measure the two colinear y-rays in coincidence using electronic collimation. SPECT detectors utilize a mechanical collimator to form an image of the single y-rays. In general, clinical PET has superior spatial resolution and system sensitivity, although SPECT is considerably less expensive, and much more widely available. In routine clinical practice, spatial resolutions of 4-6 mm are achievable. SPECT resolution is limited by the design of the collimator, and the distance from the subject to the detector. Most clinical systems can attain a resolution of 6-10 mm, though better resolution is possible with more specialized collimators. Technological developments of both PET and SPECT have led to the implementation of specialized systems for animal imaging, with much higher spatial resolution (<2 mm) (Ishizu et a l , 1995; Weber and Ivanovich, 1995; Cherry et a l , 1997; Chatziioannou et a l , 1999; Green et al., 2001; MacDonald et a l , 2001; McElroy et al., 2002; Acton and Kung, 2003; Surti et a l , 2003). The temporal resolution of both systems is poor, as it can take several minutes of scanning to acquire sufficient counts to form an image. However, despite the poorer spatial and temporal resolution of PET and SPECT compared with other modalities, such as MRI, the sensitivity of radionuclide imaging is very high. PET and SPECT can measure nanomolar (< 10"^ M) concentrations of a tracer, many orders of magnitude smaller than can be detected with MRI. The mass of radiotracer injected is extremely small, and generally does not affect the biological system under study. Many isotopes that exist for PET imaging also occur naturally in living organisms and are amenable to labeling biologically important molecules. These

include carbon (^^C), nitrogen (^^N), oxygen (^^O), and fluorine (^^F), with half-lives in the range from 2 min to 2 h. These short half-lives generally require a local production facility, with an on-site cyclotron. SPECT radioisotopes include technetium (^^"^Tc) and iodine (^^^I) and generally have longer half-lives, thus they are much more widely available than PET radioisotopes. Both PET and SPECT are now capable of imaging biological systems with unprecedented sensitivity, due to the tremendous advances in the development of novel radiopharmaceuticals. The first use of these techniques in imaging gene expression in an animal model was performed with the herpes simplex virus type 1 thymidine kinase (HSVl-tk) as the marker gene and a reporter probe, 5-iodo-2' -fluoro-2' -deoxy-1 - j8-D-arabinof uranosyluracil (FIAU) labeled with a y-emitting isotope (Tjuvajev et al., 1996). Similar probes, such as 8-fluorogancyclovir (FGCV) can be labeled with positron-emitting isotopes, permitting the repetitive, non-invasive imaging of the expression of the HSVl-tk gene in living animals using PET (Gambhir et al., 1998, 1999, 2000; Haubner et al., 2000; Herschman et a l , 2000; Iyer et a l , 2001; Tjuvajev et al., 1998, 1999). The HSVl-TK enzyme phosphorylates radiolabeled nucleoside analogs, such as [i^FJFGCV and [^^FJFIAU. Upon phosphorylation, the PET tracers become trapped in the cell, and can be measured over time as an increased accumulation of activity by the PET scanner. These tracers are not phosphorylated by endogenous human thymidine kinase, so retention of the probe occurs only in those cells expressing the HSVl-tk variant (Fig. 1). After sufficient time, the tracer washes out from nonHSVl-tk expressing cells, leading to an increasing signal-to-background in the target region. Recently, a mutant form of the HSVl-tk gene has been developed (Gambhir et al., 2000) that increases the sensitivity of this imaging technique. Quantification of the amount of HSVl-tk reporter gene expression is complicated by the fact that the amount of tracer in a cell represents not only HSVl-TK enzyme activity, but also transport of the probe across the cell membrane (MacLaren et a l , 2000). Recently, however, kinetic modeling of the dynamic uptake of [^^F]fluorohydroxymethylbutylguanine (FHBG) in tissue expressing the mutant HSVl-tk reporter gene was performed, indicating that the rate of phosphorylation gives an accurate measure of the levels of reporter protein (Green et al., 2004). Despite the promising results obtained using this reporter system, application in the brain is hampered by the fact that the PET and SPECT reporter probes for HSVl-tk expression cannot

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

RADIOISOTOPE METHODS (PET AND SPECT)

RG2TK+

RG2TK-

FIGURE 1 Radioisotope Imaging in vivo showing the high degree of sensitivity. Small animal pinhole SPECT image of [I-123]FIAU uptake in rat glioma (RG2) tumor xenografts in a mouse. The tumors either were engineered to express the HSVl-tk reporter gene (RG2TK+) or were the wild-type glioma (RG2TK-) as a Control. (From Choi et al., 2005; with permission.)

penetrate the intact BBB (Yaghoubi et al., 2001; Nanda et al., 2002). Although HSVl-tk suicide gene therapy for brain tumors has been demonstrated (Nanda et al., 2001; Fecci et al., 2002), imaging with this reporter system in the brain currently is not possible. However, disruptions in the BBB, such as those resulting from cerebral hemorrhage or tumors, may allow influx of the tracer into the brain. Alternative methods for imaging gene expression utilize the expression of receptor or transporter proteins. One such system, using the dopamine D2 receptor (D2R), has been developed which uses the widely available PET D2R tracer [i^F]fluoroethylspiperone (FESP) as a reporter probe. FESP binds with high affinity to the D2R, and has been shown to provide a quantitative measure of D2R expression in living animals (MacLaren et a l , 1999). However, this technique also has potential problems, such as occupancy of the ectopic D2R by the endogenous agonist. When a ligand activates the D2R, cellular levels of cyclic adenosine monophosphate (cAMP) may be affected which could lead to physiologic consequences on the tissue under study (MacLaren et al., 2000). However, mutant strains of the D2R which do not activate the signal pathway have been developed and studied with PET (Liang et a l , 2001). In the brain, the presence of endogenous D2R, particularly in the striatum, would provide a large background signal, and make it difficult to resolve the presence of the D2R reporter

339

system. The same problem applies to the dopamine transporter (DAT), which has been used as a reporter gene for SPECT in tissue outside the brain (Auricchio et a l , 2003). The sodium-iodide symporter (NIS) has been suggested as a possible reporter gene for the brain (Chung, 2002; Shin et al., 2004), and has been used to image neural stem cell trafficking in vivo (Kim et al., 2005). The NIS occurs naturally in high concentrations in the thyroid, and provides an active transport mechanism for sodium and iodide ions into cells. The expression of the symporter can be imaged with radioactive iodine, such as ^^^I for SPECT or ^^^I for PET, or other tracers which bind to the same site, such as ^^"^Tc-pertechnetate. Unfortunately, like the HSVl-tk system, the reporter probes for imaging the NIS do not cross the intact BBB, and rely on disruptions in the barrier to allow imaging to occur (Cho et al., 2002). Recent studies have attempted to use gene therapy to treat neurodegenerative diseases, such as Parkinson's disease (PD). The focal implantation of therapy genes in regions of the brain affected by PD could restore the function of dopaminergic cells. These cells synthesize dopamine from tyrosine using two enzymes, tyrosine hydroxylase (TH) and aromatic-L-amino acid decarboxylase (AADC). Restoration of dopaminergic function in a non-human primate model of PD has been observed recently using an adeno-associated viral vector carrying the gene for AADC into the putamen (Tsukada et a l , 2005). An important feature of the AADC system is that it has a dual role as both a therapeutic gene, and a reporter gene for PET imaging. The PET tracers [i^FJEDOPA and [ ^ - 1 I C ] L - D O P A are decarboxylated by AADC in a similar manner to L-DOPA, becoming trapped in cells in proportion to the concentration of the enzyme. Imaging [JS-^^C]L-DOPA revealed an increase in tracer uptake resulting from the expression of the AADC therapy gene, which lasted many months, and correlated with improvements in motor function (Tsukada et al., 2005). In contrast, other imaging markers of the dopaminergic system, such as dopamine transporters and postsynaptic receptors, remained unaffected. This technique offers a promising method for gene therapy in the brain, in which expression of the therapy gene itself can be monitored directly with PET imaging. Imaging gene expression with PET or SPECT reporter probes has been achieved using a variety of reporter systems, including HSVl-tk, D2R, and NIS. Imaging reporter gene expression in the brain presents a much more difficult challenge, due to the inability of many probes to cross the intact BBB, or from

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

340

25. MOLECULAR IMAGING OF GENE THERAPY FOR NEUROGENETIC DISEASES

the presence of endogenous neurotransmitters. While many of these issues remain to be solved, other novel systems, such as AADC, could provide an important role in both gene therapy and reporter gene imaging. VL

MAGNETIC RESONANCE METHODS

NMR methods are proven modalities for the noninvasive study of intracranial disease. NMR methods detect changes in the spins of protons in the presence of magnetic fields. These methods can localize changes in proton spin in three-dimensional space providing a non-invasive, sensitive method for producing high-resolution images of the brain's anatomy, known as MRI. MRI is routinely used in the clinic for diagnostic imaging and is the technique of choice for anatomic imaging of the brain. For gene therapy experiments in the CNS, NMR methods have been used to study changes in pathology, track transplanted cells, and detect a reporter gene. The contrast in MR images depends on proton spin density and on the T^ and T2 relaxation times of the tissue. T^ is called the spin-lattice or longitudinal relaxation time and is the characteristic time for magnetization to return to the main magnetic field (2) axis after the application of radiofrequency pulses or gradients. T2 is defined as the spin-spin or transverse relaxation time and is the time for spin magnetization to dephase in the transverse (xy) plane. By systematically varying acquisition parameters such as the repetition time (TR) or the echo time (TE) in an MRI experiment, one can generate contrast that is weighted either toward T^ or T2. Short TR and TE experiments lead to images with T^-weighted contrast, whereas T2-weighted images result from sequences employing longer TR and TE. The result of this is that tissues with higher water content (e.g. synovial fluid, cerebrospinal fluid) appear hypointense (dark) on T^-weighted images and hyperintense (bright) on T2-weighted images relative to surrounding tissues (Delikatny and Poptani, 2005). Additional contrast on MRI may be generated by imaging the effect on signal intensity of the exchange of magnetization between protons in structural macromolecules, such as myelin, and protons in water. MTI is an MRI method whereby contrast is generated by imaging the effect on signal intensity of the exchange of magnetization between protons in structural macromolecules and protons in water (Wolff and Balaban, 1989; Wolff et a l , 1991). The contrast present in MT images may be used quantitatively, via the magnetization transfer ratio (MTR), to examine the effect of

pathology on magnetization transfer contrast (MTC). The MTR has been used to examine the maturation of white matter tracts (Engelbrecht et al., 1998; Rademacher et al., 1999), to detect and monitor the progression of demyelinating disorders (Dousset, 1993; Lexa and Grossman, 1994; Filippi et al., 1995; Loevner et al., 1995; Grossman, 1999; McGowan et al., 2000), and to detect Wallerian degeneration prior to its appearance on conventional MRI (Lexa and Grossman, 1994). While the T^- and T2-weighted images provide anatomical information of the brain, diffusion-weighted imaging (DwI) provides information about diffusion of water molecules in restricted environments. DwI can non-invasively quantify the apparent diffusion coefficient (ADC), a measure of water molecule random motion that is sensitive to pathological changes in a tissue (Le Bihan, 1992). Diffusion of water molecules in the white matter tracts is anisotropic, as axonal membranes and myelin sheaths present barriers to the motion of water molecules. The observed diffusivity is generally much higher in directions along fiber tracts than perpendicular to them (Chenevert et a l , 1990; Beaulieu, 2002). Fractional anisotropy (FA) is a measure of directionality of diffusion, whereas ADC is a measure of magnitude of diffusion (Basser and Pierpaoli, 1996). Patients with Canavan disease (CD) exhibit a decrease in ADC probably due to the degeneration of the white matter tract leading to a gelatinous-type brain tissue that enhances restricted motion resulting in a reduction in ADC (Engelbrecht et al., 2002; Sener, 2003). On the other hand, patients suffering with Huntington disease (HD) showed an increase in the ADC of white matter and caudate nucleus (Mascalchi et al., 2004). A. MRI in Stem Cell Tracking and Reporter Gene Expression Monitoring stem and progenitor cell transplantation in the CNS involves labeling the cells ex vivo with a contrast agent that distinguishes them from host cells. To enhance the sensitivity of detection, neural stem cells (NSCs) have been labeled with MR contrast agents that substantially decrease the observed MR signal intensity. Superparamagnetic iron oxide (SPIO) particles have been shown to be effective for this purpose (Bulte et a l , 1999; Dodd et al., 1999; Sipe et al., 1999; Hoehn et al., 2002; Arbab et a l , 2003; Frank et a l , 2003; Jendelova et al., 2003; Zelivyanskaya et al., 2003). Due to their small crystal size (7-10 nm), SPIO particles align in an applied magnetic field to create extremely large microscopic field gradients

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

MAGNETIC RESONANCE METHODS

that cause a substantial decrease in the MRI signal facilitating the detection of the labeled cells. Recently, the use of micron size magnetite particles (MPIO) have led to the detection of single transplanted cells in vivo (Shapiro et al., 2004). Several cell-labeling techniques have been used, including attachment of iron oxide particles to the cell surface (Safarik and Safarikova, 1999), internalization into the cells by fluid-phase or receptor-mediated endocytosis (Bulte et al., 2001; Hinds et a l , 2003; Jendelova et al., 2003), and internalization by lipofection (Hoehn et al., 2002; Frank et al., 2003; Zelivyanskaya et al., 2003). Another approach is to cross-link the SPIO particles with dextran and conjugate them to short HIV-tat peptides (Josephson et al.,1999; Lewin et al., 2000; Dodd et al., 2001). Magnetic susceptibility contrast induced by ironoxide particles has successfully been used for highresolution single cell imaging (Dodd et al., 1999; Hinds et al., 2003), lymphocytic T cell (Yeh et al., 1995; Dodd et al., 2001) and stem cell (Bulte et al., 2001) tracking, and for monitoring gene expression in vivo (AUport and Weissleder, 2001; Weissleder et al., 2000). The applicability of these iron oxide particles as MR contrast agents has been demonstrated in phantom studies (Fleige et al., 2002), cells (Dodd et al., 1999, 2001; Sipe et a l .

341

1999; Hinds et al., 2003), and in ex vivo tissue samples (Bulte et al., 1999; Modo et al., 2002). These strategies have been used to demonstrate the ability of neural stem cells to migrate toward lesions that express chemo-attractants for the NSCs to migrate (Bulte et al., 2001; Hoehn et al., 2002; Jendelova et al., 2003; Zelivyanskaya et a l , 2003). Metabolic and degenerative brain disorders might lack these kinds of attractants for the transplanted NSCs and as such it is necessary to evaluate the extent of NSC migration using non-invasive methods. When NSCs are transplanted neo-natally into the ventricles, widespread migration occurs that can be detected using MRI (Magnitsky et a l , 2005) (Fig. 2). An MRI-detectable reporter gene system has recently been developed based on the superparamagnetic properties of the ferritin protein (Genove et al., 2005). Ferritin occurs naturally as heteromultimers of a and jS chain and belongs to a family of metalloproteins. When ferritin binds iron in cells, it forms a superparamagnetic compound that appears as a hypodensity in MR images. This can potentially be used for long-term detection as a reporter gene, keeping the advantage of high resolution provided by MRI. It also does not require loading cells with iron particles or giving a subject a radioactive compound. However,

FIGURE 2 Detection of dispersed neural stem cells labelled with superparamagnetic iron oxide (SPIO) particles in the mouse brain. Neural stem cells (NSCs) (C17.2) were labelled with SPIO and transplanted into the ventricles of newborn mice. The analysis was performed at 7 weeks of age. Regions of MRI hypo-intensity observed Ex Vivo (9.4 T) correlate with Prussian blue staining in histological sections of post-mortem, illustrating widespread migration of the NSCs. (From Magnitsky et al., 2005; with permission.)

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

342

25. MOLECULAR IMAGING OF GENE THERAPY FOR NEUROGENETIC DISEASES

the effects of expressing this type of gene for long periods of time in CNS cells are not known. B, MRI and Spectroscopy of Pathologic Changes in the CNS An approach to monitoring gene transfer in the CNS includes documenting changes in neuropathology in response to therapy This is a direct measure of efficacy and an indirect measure of gene transfer and expression. Since animal models can be analyzed in statistically significant cohorts at multiple time points after therapeutic intervention, they provide a means to assess the fidelity between the imaging findings, the distribution of gene transfer and level of gene expression, and the histopathological changes in the disease. A large number of lysosomal storage diseases affect the CNS but relatively little data exist on MR studies in patients, due to the infrequency of each type of disease and the variability of brain involvement in affected patients. Many animal models of these diseases exist which provide sufficient numbers of affected and control subjects with a specific mutation, resulting in relatively uniform disease characteristics that can be studied longitudinally. Increases in ADC (Sener, 2004), increases in white matter abnormalities, and cortical atrophy have all been observed in patients with mucopolysaccharidosis (MPS) diseases (Zafeiriou et a l , 2001; Barone et al., 2002; Matheus et al., 2004), which are defects in glycosaminoglycan catabolism.

Normal.

AMD.

The disease a-mannosidosis (AMD) has been studied in a cat model by serial MRI at 8,12, and 16 weeks of age, which showed qualitative differences in the white matter between affected and normal cats (Vite et a l , 2001). In general, the white matter in affected cats was isointense to gray matter while that of normal cats was hypointense to gray matter on T2-weighted fast spin echo images. Post-mortem histopathologic analysis of the brain showed these MR signal intensity changes to be associated with decreased myelin. Cats with AMD showed significant differences in MTR of all white matter tracts examined, with an 8-15% decrease in MTR, which is similar in magnitude to other demyelinating diseases (Dousset, 1993). Remarkably, in affected cats treated with a viral vector carrying a copy of the wild-type a-mannosidase transgene, white matter signal abnormalities were corrected (Fig. 3) and MTR values were statistically the same as normal cats in injected white matter tracts (Vite et al., 2005). These imaging findings were accompanied by improvement in myelination found at post-mortem analysis. In vivo MR spectroscopy (MRS) is a well-developed technique that has been used extensively in the detection of neurodegenerative diseases and in monitoring response to therapy. MRS detects changes in the local distribution of chemical compounds or metabolites other than water. Proton MRS can detect lactate, creatine (Cr), choline (Cho), phosphocreatine, myoinositol (ml), N-acetylaspartate (NAA), glutamate, and other metabolites in the brain (Lenkinski and Schnall, 1996).

Treated.

•'^^y " ^^p'" FIGURE 3 MRI analysis of AAV vector-mediated gene therapy in the a-raannosidosis cat. The myelin tracks in the normal cat appear as hypo-intense (dark) areas (arrow: A, E). Affected cats have significant hyperintensity of white matter tracks (B, F). The treated cats show a pattern similar to normal in the areas of the forebrain (C), but less improvement in the cerebellum (G). (Panels A-C Reprinted from Vite et al., 2005; with permission.)

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

343

MAGNETIC RESONANCE METHODS

These metabolites are involved in cellular energy metabolism, cell membrane synthesis, or they serve as neuronal markers. Specifically, NAA is a marker of mature neurons (decreases in NAA are indicative of neuronal death); Cho is an important precursor of cell membrane synthesis (increases may reflect membrane damage to myelin); and ml is believed to be a glial cell marker (increases may reflect underlying gliosis) (Lenkinski and Schnall, 1996). Concentrations of these metabolites can be used to identify the biochemistry of specific disease processes including neoplasia (Fulham et al., 1992; Barker et al., 1994; Kinoshita et a l , 1994; Poptani et a l , 1995; Negendank et al., 1996), metabolic encephalopathy (Ross et a l , 1994), seizures (Ng et a l , 1994), and neuro-degenerative disorders (Tzika, 1993; van der Knapp, 1994; Castillo et al., 1996; Lenkinski and Schnall, 1996). Monitoring of therapy of various disease processes has also been performed using MRS. Canavan disease (CD) is an autosomal recessive leukodystrophy caused by deficiency of aspartoacylase (ASPA).DeficiencyofASPAleadstoelevationofN-acetylL-aspartic acid (NAA) in the brain and urine. Increased NAAle vels have been found in pediatric patients (Grodd et a l , 1991; Aydinli et al., 1998) as well as mouse models of this disease (Leone et al., 2000; Matalon et al., 2000, 2003). In addition to increased NAA, a decrease in choline-containing compounds has been observed in patients (Grodd et a l , 1991) along with reduced glutamate and GABA in mouse models of Canavan's (Surendran et a l , 2003). Using ^^C MRS, a reduction in the rate of NAA synthesis was observed in the brain of patients with CD (Moreno et al., 2001). Steady-state measurements of metabolites using ^^C MRS also showed an increase in NAA along with reduced glutamate in patients (Bluml, 1999). Gene therapy using AAVs has been successful in alleviating the disease in the mouse model to some extent with a reduction in NAA being observed (Fig. 4) along with a reduction in brain atrophy and lack of spongiform disease (Matalon et al., 2003). Huntington disease (HD) is an autosomal dominant condition involving progressive neurodegeneration, primarily in the corpus striatum and cerebral cortex. The loss of neuronal function generally leads to a reduction in NAA as observed in patients (Harms et al., 1997; Jenkins et a l , 1998), in rat (Tkac et a l , 2001), and mouse models (Ferrante et al., 2000; Jekins et al., 2000; van Dellen et al., 2000) of the disease. HD is caused by expansion of a CAG triplet repeat leading to polyglutamine expansions in huntingtin, a protein of still unknown function. Glutamate (Glu) excitotoxicity has been implicated in HD pathogenesis leading to an elevated ratio of

BA^

€li«

\

W Wldtypt

Ctamvm

Csiiiiiviiii

AAV tremteJ

^

j^p^li

2 pm

2

^m

FIGURE 4 MRS of CD in the mouse brain. The affected brain has a significant increase in N-acetylaspartate (NAA) and decrease in choline, as in children with this disease. Transfer of the asparto-acylase (ASPA) cDNA with an AAV vector resulted in partial correction. (From Matalon et al., 2003; with permission.)

Glx (combined Glu and glutamine [Gin]) to creatine (Cr) ratio in the striatum of HD patients (Taylor-Robinson et a l , 1996). In addition to excitotoxicity, impaired energy metabolism seems to play an important role in HD pathogenesis (Beal, 1992). Augmentation of intracellular energy levels by creatine supplementation leads to a delayed decrease in NAA along with increased creatine in HD mice (Ferrante et al., 2000; Jenkins et al., 2000; van Dellen et a l , 2000). In HD patients, creatine supplementation leads to a decrease in the glutamate/ glutamine to creatine ratios indicating a decrease in the glutamate excitotoxicity (Bender et a l , 2005). The mouse model of HD has been treated by gene therapy in the CNS by expressing a small interfering (si) RNA to degrade the mutant allele (Xia et al., 2004), thus the model may be useful for imaging studies.

V GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

344

25. MOLECULAR IMAGING OF GENE THERAPY FOR NEUROGENETIC DISEASES

Krabbe disease, or globoid cell leukodystrophy (GLD), is a demyelinating disorder caused by a genetic deficiency of lysosomal enzyme galactocerebrosidase (GALC), a key component in metabolic pathways of myelin turnover and breakdown (Wenger et al., 2000). Galactocerebrosides are selectively highly concentrated in myelin, almost completely lacking in the brain before the onset of myelination, and rapidly increasing during myelination accompanied by a rise of GALC activity. Deficiency of GALC results in the accumulation of cerebrosides and of psychosine galactosylsphingosine, which alters the normal myelin. Twitcher mutant mice, having the same defect as in the human disease, show initial normal development of peripheral myelin followed by a declining rate of myelination along with demyelination and the appearance of globoid cells. Oligodendrocytes vanish and are replaced by proliferated astrocytes. The pathognomonic feature of Krabbe disease is the presence of large numbers of globoid cells. Psychosine, the second substrate of GALC, known to be cytotoxic and found increased up to a 100-fold in the white matter of patients with GLD, is considered to be the key toxic agent in GLD responsible for rapid demyelination with secondary axonal degeneration and the severe encephalopathic symptoms in patients with infantile Krabbe disease. The most frequent form of Krabbe disease has an infantile onset, whereas the late-onset form is rare. The increased neuronal toxicity leads to a decrease in NAA levels, while the demylination of the white matter tracts manifests as an increase in choline and myoinositol levels in patients with Krabbe disease (Farina et a l , 2000; Brockmann et a l , 2003). In the dog model of Krabbe disease, the MRI, MTI, and MRS data are consistent with demyelination in the brain (Cozzi et al., 1998; McGowan et al., 2000, C.H. Vite and H. Poptani, unpublished). It has recently been shown that an AAV-1 vector introduced into the ventricles of Krabbe mice at birth can achieve widespread transfer of the GALC cDNA, resulting in partial rescue of the phenotype (Rafi et al., 2005), thus this model may be useful for imaging studies. Diffusion-weighted MRS (Dw-MRS) of metabolites and water may assist in understanding the cellular microenvironment of the diseased tissue. Most of the metabolites that are observed by in vivo MRS, such as NAA, Cr, Cho, are predominantly located in the intracellular compartment. Dw-MRS helps in probing the intracellular space by examining the diffusion characteristics of the metabolites (Nicolay et al., 2001). EHv-MRS has been used to estimate the dimensions of the cellular elements that restrict intracellular metabolite diffusion in muscle and nerve tissue and has provided

information on the cellular response to pathophysiological changes in brain disorders, including ischemia, epilepsy, and cancer (Wick et al., 1995; van der Toon et al., 1996; Chenevert et al., 1997; Hakumaki et al., 1998). It is hypothesized that DwMRS may aid in the measure of neuronal distension associated with many lysosomal storage diseases as well as in measures of the resolution of cell distension associated with gene therapy. VIL

OPTICAL IMAGING

Optical reporter genes have been used by molecular and cell biologists for decades, although, in general, measurement of the probe was limited to cell culture or required the sacrifice of the animal. Recent advances in optical imaging technology, using extremely sensitive cooled charge-coupled device (CCD) cameras, have expanded the role of optical imaging to non-invasive in vivo studies. Optical assays include fluorescence, using, for example, green fluorescent protein (GFP), and bioluminescence, using various forms of luciferase, such as firefly luciferase (Flue). Expression of fluorescent proteins in living cells can be imaged using various forms of microscopy, such as confocal laser microscopy, and two-photon laser microscopy. These techniques can be extended to in vivo imaging in small animals, such as mice, although the depths at which fluorescence light can be detected are very limited due to the extra path length involved for both the excitation and emitted light. Bioluminescence imaging detects light emitted from tissue when the enzyme luciferase reacts with its substrate, D-luciferin. The light emitted by Flue has a peak wavelength around 560 nm, which is strongly absorbed by tissue. However, despite this, a significant amount of the luciferase light emission occurs above 600 nm, where there is an optical window allowing light to more readily penetrate tissue. Detection of this emission with a CCD camera gives a map of the expression of the reporter gene, which can be overlaid on a normal light image of the subject (Bhaumik and Gambhir, 2002). However, in general, the problem of light penetration through tissue restricts the application of these techniques to small animals, primarily mice, where sufficient light emerges from the subject to be detected. Poor light penetration also degrades spatial resolution, makes quantification inaccurate, and prevents tomographic imaging with current technology. Probably the greatest disadvantage of optical imaging over the radionuclide techniques is the difficulty in translation to human subjects.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

345

ACKNOWLEDGMENTS

Light attenuation and scatter is so great in tissue that it is difficult if not impossible to image structures at the depth needed. Optical probes using near-infrared emissions operate in a more favorable part of the spectrum for imaging. Although this modality has not been used with a reporter gene, it has been used to study oxygenation and tumor biology in the rodent brain (Chen et al., 2003; Wang et al., 2004). Despite the obvious limitations of optical imaging, it has several attractive features. The instrumentation is relatively simple, using just a light-tight chamber and a cooled CCD camera. The substrates are readily available, and there is no requirement for radioactivity or a radiochemistry facility, with all the associated costs involved. Possibly, the most important advantage of optical imaging is the extremely high sensitivity, which exceeds even the radionuclide techniques, mainly due to the virtual absence of any background signal. In addition, unlike many PET or SPECT reporter probes, the substrates required for bioluminescence imaging cross the intact BBB (CoUaco and Geusz, 2003), and are suitable for studying gene expression in the brain. A number of applications of optical imaging of luciferase expression in the brain have been performed. Real-time tracking of neural precursor cells (NPCs) has been studied in a mouse model of glioma (Shah et al., 2005). This study used NPCs to deliver a secreted form of tumor necrosis factor-related apoptosis-inducing ligand to neoplastic cells, and tracked their migration with a luciferase reporter gene. Other applications of optical imaging have included a study of neuronal damage, in which upregulation of glial fibrillary acidic protein (GFAP) was monitored by measuring the expression of the luciferase gene under the transcriptional control of the GFAP promoter (Zhu et al., 2004). Viral infection of the brain was monitored using strains of a virus engineered to express luciferase (Cook and Griffin, 2003). An important recent development is the multimodality fusion reporter system, which can be studied using both optical and radionuclide imaging (Ray et al., 2003, 2004). This gene expresses a tri-fusion protein, which contains coding regions for red fluorescent protein, Renilla luciferase, and the HSVl-TK enzyme. Cellular fluorescence expression can be monitored using microscopic techniques, while in vivo monitoring can be performed using both optical and PET or SPECT methods. This fusion reporter offers the possibility of using the particular imaging technique that best suits the application; fluorescence for studying individual cells or for cell sorting, bioluminescence for high sensitivity in vivo imaging in small animals, and

PET or SPECT when quantitative accuracy is important, or for the translation to humans. Further developments of optical imaging techniques, such as quantum dots (Michalet et al., 2005), diffuse optical tomography (Heilscher et a l , 2002), and optical coherence tomography (Bluestone et al., 2004; Boppart et al., 2004), offer tremendous enhancements to currently available technology. Coupled with multimodality reporter systems, optical imaging has an important role to play in gene therapy.

VIIL

SUMMARY

The ability to monitor gene expression a n d / o r changes in pathology in response to therapy will be critical for clinical gene therapy trials and applications in the CNS. Many imaging modalities are currently under development to achieve this goal and each method has distinct advantages. The spatial resolution or sensitivity is constrained by the cell biology and biochemistry of the biological processes, as well as the chemistry, physics and computational power of the instrumentation. The use of surrogate genes is preferable to developing separate methods to detect individual proteins. However, this will require that concordance of expression of the reporter gene and the therapeutic gene be demonstrated. Reporter genes ideally should be derived from mammalian systems to avoid immune responses. The genes must also be modified to avoid introducing an ectopic biological function into a target cell. Changes in neuropathology can be used as a direct measurement of the effects of gene therapy, but are only an indirect measure of gene expression; thus, they may not provide a good monitor for changes in vector function. Significant research efforts are underway to combine imaging modalities to take advantage of their most useful properties. New reagents and newer methods of detection are rapidly expanding the repertoire of methods available for analyzing and monitoring gene and cell transfer in the CNS, and the attendant therapeutic effects.

ACKNOWLEDGMENTS The work in our laboratories has been supported by the following: JHW, NIDDK (DK042707, DK046637, DK063973) and NINDS (NS029390, NS038690); PDA, NIBIB (EB001809, EB002774) and NINDS (NS048315); HP, NCI (CA102756), NICHD (HD048582), and

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

346

25. MOLECULAR IMAGING OF GENE THERAPY FOR NEUROGENETIC DISEASES

University of Pennsylvania Research Foundation; CHV, NINDS (NS002032) and NCRR (RR002512). References Acton, RD. and Kung, H.F. (2003) Small animal imaging with high resolution single photon emission tomography. Nucl. Med. Biol., 30: 889-895. AUport, J.R. and Weissleder, R. (2001) In vivo imaging of gene and cell therapies. Exp. Hematol., 29:1237-1246. Arbab, A.S., Bashaw, L.A., Miller, B.R., Jordan, E.K., Bulte, J.W. and Frank, J.A. (2003) Intracytoplasmic tagging of cells with ferumoxides and transfection agent for cellular magnetic resonance imaging after cell transplantation: methods and techniques. Transplantation, 76: 1123-1130. Auricchio, A., Acton, P.D., Hildinger, M., Plossl, K., Kung, H.F. and Wilson, J.M. (2003) In vivo quantitative non-invasive imaging of gene transfer with single-photon emission computerized tomography. Hum. Gene Ther., 14: 255-261. Aydinli, N., Caliskan, M., Calay, M. and Ozmen, M. (1998) Use of localized proton nuclear magnetic resonance spectroscopy in Canavan's disease. Turkish J. Pediatr., 40: 549-557. Barker, RB., Gillard, J.H., van Zijl, P C , Sober, B.J., Hartley, D.R, Agildere, A.M., Oppenheimer, S.M. and Bryan, R.N. (1994) Acute stroke: evaluation with serial proton MR spectroscopic imaging. Radiology, 192: 723-732. Barone, R., Parano, E., Trifiletti, R.R., Fiumara, A. and Pavone, P. (2002) White matter changes mimicking a leukodystrophy in a patient with mucopolysaccharidosis: characterization by MRI. J. Neurol. Sci., 195: 171-175. Basser, P.J. and Pierpaoh, C. (1996) Microstructural and physiological features of tissues elucidated by quantitative-diffusiontensor MRI. J. Magn. Reson. B, 111: 209-219. Beal, M.F. (1992) Does impairment of energy metabolism result in excitotoxic neuronal death in neurodegenerative illnesses? Ann. Neurol., 31: 119-130. Beaulieu, C. (2002) The basis of anisotropic water diffusion in the nervous system — a technical review. NMR Biomed., 15: 435-455. Bender, A., Auer, D.P., Merl, T., Reilmann, R., Saemann, P., Yassouridis, A., Bender, J., Weindl, A., Dose, M., Gasser, T. and Klopstock, T. (2005) Creatine supplementation lowers brain glutamate levels in Huntington's disease. J. Neurol., 252: 36-41. Bhaumik, S. and Gambhir, S.S. (2002) Optical imaging of Renilla luciferase reporter gene expression in living mice. Proc. Natl. Acad. Sci. USA, 99: 377-382. Bluestone, A.Y., Stewart, M., Lasker, J., Abdoulaev, G.S. and Hielscher, A.H. (2004) Three-dimensional optical tomographic brain imaging in small animals, part 1: hypercapnia. J. Biomed. Opt., 9: 1046-1062. Bluml, S. (1999) In vivo quantitation of cerebral metabolite concentrations using natural abundance 13C MRS at 1.5 T. J. Magn. Reson., 136: 219-225. Boppart, S.A., Luo, W., Marks, D.L. and Singletary, K.W (2004) Optical coherence tomography: feasibility for basic research and image-guidedsurgery of breast cancer. Breast Cancer Res. Treat., 84: 85-97. Brockmann, K., Dechent, P., Wilken, B., Rusch, O., Frahm, J. and Hanefeld, R (2003) Proton MRS profile of cerebral metabolic abnormalities in Krabbe disease. Neurology, 60: 819-825. Bulte, J.W, Douglas, T., Witwer, B., Zhang, S.C, Strable, E., Lewis, B.K., Zywicke, H., Miller, B., van Gelderen, P., Moskowitz, B.M.,

Duncan, I.D. and Frank, J.A. (2001) Magnetodendrimers allow endosomal magnetic labeling and in vivo tracking of stem cells. Nat. Biotechnol., 19: 1141-1147. Bulte, J.W, Zhang, S., van Gelderen, P., Herynek, V., Jordan, E.K., Duncan, I.D. and Frank, J.A. (1999) Neurotransplantation of magnetically labeled oligodendrocyte progenitors: magnetic resonance tracking of cell migration and myelination. Proc. Natl. Acad. Sci. USA, 96: 15256-15261. Castillo, M., Kwock, L. and Mukherji, S.K. (1996) Clinical apphcations of proton MR spectroscopy. Am. J. Neuroradiol., 17:1-15. Chatziioannou, A.F., Cherry, S.R., Shao, Y, Silverman, R.W, Meadors, K., Farquhar, T.H., Pedarsani, M. and Phelps, M.E. (1999) Performance evaluation of microPET: a high resolution lutetium oxyorthosilicate PET scanner for animal imaging. J. Nucl. Med., 40:1164-1175. Chen, Y, Intes, X., Tailor, D.R., Regatte, R.R., Ma, H., Ntziachristos, V., Leigh, J.S., Reddy, R. and Chance, B. (2003) Probing rat brain oxygenation with near-infrared spectroscopy (NIRS) and magnetic resonance imaging (MRI). Adv. Exp. Med. Biol., 510: 199-204. Chenevert, T.L., Brunberg, J.A. and Pipe, J.G. (1990) Anisotropic diffusion in human white matter: demonstration with MR techniques in vivo. Radiology, 177: 401^05. Chenevert, T.L., McKeever, PE. and Ross, B.D. (1997) Monitoring early response of experimental brain tumors to therapy using diffusion magnetic resonance imaging. Clin. Cancer Res., 3: 1457-1466. Cherry, S.R., Shao, Y, Silverman, R.W, Meadors, K., Siegel, S., Chatziioannou, A., Young, J.W., Jones, W., Moyers, J.C, Newport, D., Boutefnouchet, A., Farquhar, T.H., Andreaco, M., Paulus, M.J., Binkley, D.M., Nutt, R. and Phelps, M.E. (1997) MicroPET: a high resolution PET scanner for imaging small animals. IEEE Trans. Nucl. Sci., 44:1161-1166. Cho, J.Y, Shen, D.H., Yang, W, WiUiams, B., Buckwalter, T.L., La Perle, K.M., Hinkle, G., Pozderac, R., Kloos, R., Nagaraja, H.N., Barth, R.F. and Jhiang, S.M. (2002) In vivo imaging and radioiodine therapy following sodium iodide symporter gene transfer in animal model of intracerebral gliomas. Gene Ther., 9: 1139-1145. Choi, S.R., Zhuang, Z.P, Chacko, A.M., Acton, RD., Gelovani, J., Doubrovin, M., Chu, D.C.K. and Kung, H.R (2005) SPECT imaging of herpes simplex virus type 1 thymidine kinase gene expression by [i23l]FIAU. Acad. Radiol., 12: 795-805. Chung, J.K. (2002) Sodium iodide symporter: its role in nuclear medicine. J. Nucl. Med., 43: 1188-1200. CoUaco, A.M. and Geusz, M.E. (2003) Monitoring immediate-early gene expression through firefly luciferase imaging of HRS/J hairless mice. BMC Physiol., 3: 8. Cook, S.H. and Griffin, D.E. (2003) Luciferase imaging of a neurotropic viral infection in intact animals. J. Virol., 71: 5333-5338. Cozzi, R, Vite, C.H., Wenger, D.A., Victoria, T. and Haskii\s, M.E. (1998) MRI and electrophysiological abnormalities in a case of carune globoid cell leucodystrophy. J. Small Anim. Pract., 39: 401^05. Delikatny, E.J. and Poptani, H. (2005) MR techniques for in vivo molecular and cellular imaging. Radiol. Clin. North Am., 43: 205-220. Dodd, C.H., Hsu, H.C., Chu, WJ., Yang, R, Zhang, H.G., Mountz, J.D., Jr., Zinn, K., Forder, J., Josephson, L., Weissleder, R., Mountz, J.M. and Mountz, J.D. (2001) Normal T-cell response and in vivo magnetic resonance imaging of T ceUs loaded with HIV transactivator-peptide-derived superparamagnetic nanoparticles. J. Immunol. Meth., 256: 89-105.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

347

REFERENCES Dodd, S.J., Williams, M., Suhan, J.R, Williams, D.S., Koretsky, A.R and Ho, C. (1999) Detection of single mammalian cells by highresolution magnetic resonance imaging. Biophys. J., 76:103-109. Dousset, V. (1993) Magnetization transfer imaging in vivo study of normal brain tissues and characterization of multiple sclerosis and experimental allergic encephalomyelitis lesions. J. NeuroradioL, 20: 297. EUinwood, N.M., Vite, C.H. and Raskins, M.E. (2004) Gene therapy for lysosomal storage diseases: the lessons and promise of animal models. J. Gene Med., 6: 481-506. Engelbrecht, V., Rassek, M., Preiss, S., Wald, C. and Modder, U. (1998) Age-dependent changes in magnetization transfer contrast of white matter in the pediatric brain. Am. J. Neuroradiol., 19:1923-1929. Engelbrecht, V., Scherer, A., Rassek, M., Witsack, H.J. and Modder, U. (2002) Diffusion-weighted MR imaging in the brain in children: findings in the normal brain and in the brain with white matter diseases. Radiology, 222: 410-418. Farina, L., Bizzi, A., Finocchiaro, G., Pareyson, D., Sghirlanzoni, A., Bertagnolio, B., Savoiardo, M., Naidu, S., Singhal, B.S. and Wenger, D.A. (2000) MR imaging and proton MR spectroscopy in adult Krabbe disease. Am. J. Neuroradiol, 21:1478-1482. Fecci, P.E., Gromeier, M. and Sampson, J.H. (2002) Viruses in the treatment of brain tumors. Neuroimag. Clin. N. Am., 12: 553-570. Ferrante, R.J., Andreassen, O.A., Jenkins, B.G., Dedeoglu, A., Kuemmerle, S., Kubilus, J.K., Kaddurah-Daouk, R., Hersch, S.M. and Beal, M.F. (2000) Neuroprotective effects of creatine in a transgenic mouse model of Huntington's disease. J. Neurosci., 20: 4389-4397. Filippi, M., Campi, A., Dousset, V., Baratti, C , Martinelli, V., Canal, N., Scotti, G. and Comi, G. (1995) A magnetization transfer imaging study of normal-appearing white matter in multiple sclerosis. Neurology, 45: 478-482. Fleige, G., Seeberger, R, Laux, D., Kresse, M., Taupitz, M., Pilgrimm, H. and Zimmer, C. (2002) In vitro characterization of two different ultrasmall iron oxide particles for magnetic resonance cell tracking. Invest. Radiol., 37: 482-488. Frank, J.A., Miller, B.R., Arbab, A.S., Zywicke, H.A., Jordan, E.K., Lewis, B.K., Bryant, L.H., Jr. and Bulte, J.W (2003) Clinically applicable labeling of mammalian and stem cells by combining superparamagnetic iron oxides and transfection agents. Radiology, 228: 480^87. Fulham, M.J., Bizzi, A., Dietz, M.J., Shih, H.H., Raman, R., Sobering, G.S., Frank, J.A., Dwyer, A.J., Alger, J.R. and Di Chiro, G. (1992) Mapping of brain tumor metabolites with proton MR spectroscopic imaging: clinical relevance. Radiology, 185: 675-686. Gambhir, S.S., Barrio, J.R., Phelps, M.E., Iyer, M., Namavari, M., Satyamurthy, N., Wu, L., Green, L.A., Bauer, E., MacLaren, D.C., Nguyen, K., Berk, A.J., Cherry, S.R. and Herschman, H.R. (1999) Imaging adenoviral-directed reporter gene expression in living animals with positron emission tomography. Proc. Natl. Acad. Sci. USA, 96: 2333-2338. Gambhir, S.S., Barrio, J.R., Wu, L., Iyer, M., Namavari, M., Satyamurthy, N., Bauer, E., Parrish, C , MacLaren, D.C., Borghei, A.R., Green, L.A., Sharfstein, S., Berk, A.J., Cherry, S.R., Phelps, M.E. and Herschman, H.R. (1998) Imaging of adenoviraldirected herpes simplex virus type 1 thymidine kinase reporter gene expression in mice with radiolabeled ganciclovir. J. Nucl. Med., 39: 2003-2011. Gambhir, S.S., Bauer, E., Black, M.E., Liang, Q., Kokoris, M.S., Barrio, J.R., Iyer, M., Namavari, M., Phelps, M.E. and Herschman, H.R. (2000) A mutant herpes simplex virus type 1 thymidine

kinase reporter gene shows improved sensitivity for imaging reporter gene expression with positron emission tomography. Proc. Natl. Acad. Sci. USA, 97: 2785-2790. Gelovani Tjuvajev, J. and Blasberg, R.G. (2003) In vivo imaging of molecular-genetic targets for cancer therapy. Cancer Cell, 3: 327-332. Genove, G., DeMarco, U., Xu, H., Goins, WR and Ahrens, E.T. (2005) A new transgene reporter for in vivo magnetic resonance imaging. Nat. Med., 11: 450-454. Green, L.A., Nguyen, K., Berenji, B., Iyer, M., Bauer, E., Barrio, J.R., Namavari, M., Satyamurthy, N. and Gambhir, S.S. (2004) A tracer kinetic model for 18F-FHBG for quantitating herpes simplex virus type 1 thymidine kinase reporter gene expression in Iving animals using PET. J. Nucl. Med., 45:1560-1570. Green, M.V., Seidel, J., Vaquero, J.J., Jagoda, E., Lee, I. and Eckelman, WC. (2001) High resolution PET, SPECT and projection imaging in small animals. Comput. Med. Imag. Graph., 25: 79-86. Grodd, W , Krageloh-Mann, I., Klose, U. and Sauter, R. (1991) Metabolic and destructive brain disorders in children: findings with localized proton MR spectroscopy. Radiology, 181:173-181. Grossman, R.I. (1999) Application of magnetization transfer imaging to multiple sclerosis. Neurology, 53: S8-S11. Hakumaki, J.M., Poptani, H., Puumalainen, A.M., Loimas, S., Paljarvi, L.A., Yla-Herttuala, S. and Kauppinen, R.A. (1998) Quantitative I H nuclear magnetic resonance diffusion spectroscopy of BT4C rat glioma during thymidine kinase-mediated gene therapy in vivo: identification of apoptotic response. Cancer Res., 58: 3791-3799. Harms, L., Meierkord, H., Timm, G., Pfeiffer, L. and Ludolph, A.C. (1997) Decreased N-acetyl-aspartate/choline ratio and increased lactate in the frontal lobe of patients with Huntington's disease: a proton magnetic resonance spectroscopy study. J. Neurol. Neurosurg. Psychiatry 62: 27-30. Haskins, M., Abkowitz, J., Aguirre, G., Evans, S., Hasson, C , Just, C , Lexa, F , Miranda, S., Ray, J., Schuchman, E., Simonaro, C , Thrall, M., Wang, P , Weil, M., Weimelt, S., Wolfe, J. and Patterson, D. (1997) Bone marrow transplantation in animal models of lysosomal storage diseases. In: Rigden, O., Hobbs, J.R. and Stewart, C.G. (Eds.), Correction of Genetic Diseases by Transplantation. COGENT, London, pp. 1-11. Haubner, R., Avril, N., Hantzopoulos, P.A., Gansbacher, B. and Schwaiger, M. (2000) In vivo imaging of herpes simplex virus type 1 thymidine kinase gene expression: early kinetics of radiolabelled FIAU. Eur. J. Nucl. Med., 27: 283-291. Heilscher, A.H., Bluestone, A.Y., Abdoulaev, G.S., Klose, A.D., Lasker, J., Stewart, M., Netz, U. and Beuthan, J. (2002) Near-infrared diffuse optical tomography. Dis. Markers, 18: 313-337. Hennig, A.K., Levy, B., Ogilvie, J.M., Vogler, C.A., Galvin, N., Bassnett, S. and Sands, M.S. (2003) Intravitreal gene therapy reduces lysosomal storage in specific areas of the CNS in mucopolysaccharidosis VII mice. J. Neurosci., 23: 3302-3307. Herschman, H.R. (2004). Noninvasive imaging of reporter gene expression in living subjects. Adv. Cancer Res., 92: 29-80. Herschman, H.R., MacLaren, D.C., Iyer, M., Namavari, M., Bobinski, K., Green, L.A., Wu, L., Berk, A.J., Toyokuni, T, Barrio, J.R., Cherry, S.R., Phelps, M.E., Sandgren, E.R and Gambhir, S.S. (2000) Seeing is believing: non-invasive, quantitative and repetitive imaging of reporter gene expression in living animals, using positron emission tomography. J. Neurosci. Res., 59: 699-705. Hinds, K.A., Hill, J.M., Shapiro, E.M., Laukkanen, M.O., Silva, A.C, Combs, C.A., Vamey, T. R., Balaban, R.S., Koretsky, A.R and Dunbar, C.E. (2003) Highly efficient endosomal labeling of progenitor

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

348

25. MOLECULAR IMAGING OF GENE THERAPY FOR NEUROGENETIC DISEASES

and stem cells with large magnetic particles allows magnetic resonance imaging of single cells. Blood, 102: 867-872. Hoehn, M., Kustermann, E., Blunk, J., Wiedermann, D., Trapp, T., Wecker, S., Focking, M., Arnold, H., Fleischmann, B.K., Schwindt, W. and Buhrle, C. (2002) Monitoring of implanted stem cell migration in vivo: a highly resolved in vivo magnetic resonance imaging investigation of experimental stroke in rat. Proc. Natl. Acad. Sci. USA, 99: 16267-16272. Ishizu, K., Mukai, T, Yonekura, Y., Pagani, M., Fujita, T., Magata, Y, Nishizawa, S., Tamaki, N., Shibasaki, H. and Konishi, J. (1995) Ultra-high resolution SPECT system using four pinhole collimators for small animal studies. J. Nucl. Med., 36: 2282-2287. Iyer, M., Barrio, J.R., Namavari, M., Bauer, E., Satyamurthy, N., Nguyen, K., Toyokuni, T., Phelps, M.E., Herschman, H.R. and Gambhir, S.S. (2001) 8-[18F]fluoropenciclovir: an improved reporter probe for imaging HSVl-tk reporter gene expression in vivo using PET. J. Nucl. Med., 42: 96-105. Jendelova, P., Herynek, V., DeCroos, J., Glogarova, K., Andersson, B., Hajek, M. and Sykova, E. (2003) Imaging the fate of implanted bone marrow stromal cells labeled with superparamagnetic nanoparticles. Magn. Reson. Med., 50: 767-776. Jenkins, B.C., Klivenyi, P., Kustermann, E., Andreassen, O.A., Ferrante, R.J., Rosen, B.R. and Beal, M.F. (2000) Nonlinear decrease over time in N-acetyl aspartate levels in the absence of neuronal loss and increases in glutamine and glucose in transgenic Huntington's disease mice. J. Neurochem., 74: 2108-2119. Jenkins, B.C., Rosas, H.D., Chen, Y.C., Makabe, T., Myers, R., MacDonald, M., Rosen, B.R., Beal, M.R and Koroshetz, W.J. (1998) I H NMR spectroscopy studies of Huntington's disease: correlations with CAG repeat numbers. Neurology, 50: 1357-1365. Josephson, L., Tung, C.H., Moore, A. and Weissleder, R. (1999) High-efficiency intracellular magnetic labeling with novel superparamagnetic-Tat peptide conjugates. Bioconjugate Chem., 10: 186-191. Kim, YH., Lee, D.S., Kang, J.H., Lee, Y.J., Chung, J.K., Roh, J.K., Kim, S.U. and Lee, M.C. (2005) Reversing the silencing of reporter sodium/iodide symporter transgene for stem cell tracking. J. Nucl. Med., 46: 305-311. Kinoshita, Y, Kajiwara, H., Yokota, A. and Koga, Y (1994) Proton magnetic resonance spectroscopy of brain tumors: an in vitro study. Neurosurgery, 35: 606-614. Le Bihan, D. (1992) Theoretical principles of perfusion imaging. Application to magnetic resonance imaging. Invest. Radiol., 27(Suppl2):S6-Sll. Lenkinski, R. and Schnall, M.D. (1996) MR spectroscopy and the biochemical basis of neurological disease. In: Atlas, S. (Ed.), Magnetic Resonance Imaging of the Brain and Spine. LippencottRaven, Philadelphia, pp. 1619-1653. Leone, P., Janson, C.G., Bilaniuk, L., Wang, Z., Sorgi, R, Huang, L., Matalon, R., Kaul, R., Zeng, Z., Freese, A., McPhee, S.W, Mee, E. and During, M.J. (2000) Aspartoacylase gene transfer to the mammalian central nervous system with therapeutic implications for Canavan disease. Ann. Neurol., 48: 27-38. Lewin, M., Carlesso, N., Tung, C.H., Tang, X.W., Cory, D., Scadden, D.T. and Weissleder, R. (2000) Tat peptide-derivatized magnetic nanoparticles allow in vivo tracking and recovery of progenitor cells. Nat. BiotechnoL, 18: 410^14. Lexa, F.J. and Grossman, R.I. (1994) MR of sarcoidosis in the head and spine: spectrum of manifestations and radiographic response to steroid therapy Am. J. NeuroradioL, 15: 973-982.

Liang, Q., Satyamurthy, N., Barrio, J.R., Toyokuni, T., Phelps, M.E., Gambhir, S.S. and Herschman, H.R. (2001) Noninvasive, quantitative imaging in living animals of a mutant dopanrdne D2 receptor reporter gene in which ligand binding is uncoupled from signal tiransduction. Gene Ther., 8: 1490-1498. Loevner, L.A., Grossman, R.L, Cohen, J.A., Lexa, F.J., Kessler, D. and Kolson, D.L. (1995) Microscopic disease in normal-appearing white matter on conventional MR images in patients with multiple sclerosis: assessment with magnetization-transfer measurements. Radiology, 196: 511-515. MacDonald, L.R., Patt, B.E., Iwanczyk, J.S., Tsui, B.M.W., Wang, Y, Frey, E.G., Wessell, D.E., Acton, PD. and Kung, H.F. (2001) Pinhole SPECT of mice using the LumaGEM gamma camera. IEEE Trans. Nucl. Sci., 48: 830-836. MacLaren, D.C., Gambhir, S.S., Satyamurthy, N., Barrio, J.R., Sharfstein, S., Toyokuni, T., Wu, L., Berk, A.J., Cherry, S.R., Phelps, M.E. and Herschman, H.R. (1999) Repetitive, non-invasive imaging of the dopamine D2 receptor as a reporter gene in living animals. Gene Ther., 6: 785-791. MacLaren, D.C., Toyokuni, T., Cherry, S.R., Barrio, J.R., Phelps, M.E., Herschman, H.R. and Gambhir, S.S. (2000) PET imaging of transgene expression. Biol. Psychiatry, 48: 337-348. Magnitsky, S., Watson, D.J., Walton, R.M., Pickup, S., Bulte, J.W.M., Wolfe, J.H. and Poptani, H. (2005). In vivo and ex vivo MRI detection of localized and disseminated neural stem cell grafts in the mouse brain. Neuroimage, 26: 744-754. Mascalchi, M., LoUi, R, Delia Nave, R., Tessa, C , Petralli, R., Gavazzi, C , Politi, L.S., Macucci, M., Filippi, M. and Piacentini, S. (2004) Huntington disease: volumetric, diffusion-weighted, and magnetization transfer MR imaging of brain. Radiology, 232: 867-873. Massoud, T.R and Gambhir, S.S. (2003) Molecular imaging in living subjects: seeing fundamental biological processes in a new light. Gene. Dev, 17: 545-580. Matalon, R., Rady, PL., Piatt, K.A., Skinner, H.B., Quast, M.J., Campbell, G.A., Matalon, K., Ceci, J.D., Tyring, S.K., Nehls, M., Surendran, S., Wei, J., Ezell, E.L. and Szucs, S. (2000) Knock-out mouse for Canavan disease: a model for gene transfer to the central nervous system. J. Gene Med., 2:165-175. Matalon, R., Surendran, S., Rady, PL., Quast, M.J., Campbell, G.A., Matalon, K.M., Tyring, S.K., Wei, J., Peden, C.S., Ezell, E.L., Muzyczka, N. and Mandel, R.J. (2003) Adeno-associated virusmediated aspartoacylase gene transfer to the brain of knockout mouse for canavan disease. Mol. Ther., 7: 580-587. Matheus, M.G., Castillo, M., Smith, J.K., Armao, D., Towle, D. and Muenzer, J. (2004) Brain MRI findings in patients with mucopolysaccharidosis types I and II and mild clinical presentation. Neuroradiology, 46: 666-672. McElroy, D.P, MacDonald, L.R., Beekman, RJ., Wang, Y, Patt, B.E., Iwanczyk, J.S., Tsui, B.M.W. and Hoffman, E. (2002) Performance evaluation of A-SPECT: a high resolution desktop pinhole SPECT system for imaging small animals. IEEE Trans. Nucl. Sci., 49: 2139-2147. McGowan, J.C, Haskins, M., Wenger, D.A. and Vite, C. (2000) Investigating demyelination in the brain in a canine model of globoid cell leukodystrophy (Krabbe disease) using magnetization transfer contrast: preliminary results. J. Comput. Assist. Tomogr., 24: 316-321. Michalet, X., Pinaud, RR, Bentolila, L.A., Tsay, J.M., Doose, S., Li, J.J., Sundaresan, G., Wu, A.M., Gambhir, S.S. and Weiss, S. (2005) Quantum dots for live cells, in vivo imaging, and diagnostics. Science, 307: 538-544.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

349

REFERENCES Min, J.J. and Gambhir, S.S. (2004) Gene therapy progress and prospects: noninvasive imaging of gene therapy in living subjects. Gene Ther., 11:115-125. Modo, M., Cash, D., Mellodew, K., Williams, S.C., Fraser, S.E., Meade, T.J., Price, J. and Hodges, H. (2002) Tracking transplanted stem cell migration using bifunctional, contrast agent-enhanced, magnetic resonance imaging. Neuroimage, 17: 803-811. Moreno, A., Ross, B.D. and Bluml, S. (2001) Direct determination of the N-acetyl-L-aspartate synthesis rate in the human brain by (13)C MRS and [l-(13)C]glucose infusion. J. Neurochem., 77'. 347-350. Nanda, D., de Jong, M., Vogels, R., Havenga, M., Driesse, M., Bakker, W., Bijster, M., Avezaat, C., Cox, R, Morin, K., Naimi, E., Knaus, E., Wiebe, L. and Smitt, RS. (2002) Imaging expression of adenoviral HSVl-tk suicide gene transfer using the nucleoside analogue FIRU. Eur. J. Nucl. Med., 29: 939-947. Nanda, D., Vogels, R., Havenga, M., Avezaat, C , Bout, A. and Smitt, RS. (2001) Treatment of malignant gliomas with a replicating adenoviral vector expressing herpes simplex virus-thymidine kinase. Cancer Res., 61: 8743-8750. NegendarO<, W.G., Sauter, R., Brown, T.R., Evelhoch, J.L., Falini, A., Gotsis, E.D., Heerschap, A., Kamada, K., Lee, B.C., Mengeot, M.M., Moser, E., Padavic-Shaller, K.A., Sanders, J.A., Spraggins, T.A., Stillman, A.E., Terwey, B., Vogl, T.J., Wicklow, K. and Zimmerman, R.A. (1996) Proton magnetic resonance spectroscopy in patients with glial tumors: a multicenter study. J. Neurosurg., 84: 449^58. Ng, T.C., Comair, Y.G., Xue, M., So, N., Majors, A., Kolem, H., Luders, H. and Modic, M. (1994) Temporal lobe epilepsy: presurgical localization with proton chemical shift imaging. Radiology, 193: 465-472. Nicolay, K., Braun, K.P., Graaf, R.A., Dijkhuizen, R.M. and Kruiskamp, M.J. (2001) Diffusion NMR spectroscopy. NMR Biomed., 14: 94-111. Passini, M.A., Lee, E.B., Heuer, G.G. and Wolfe, J.H. (2002) Distribution of a lysosomal enzyme in the adult brain by axonal transport and by cells of the rostral migratory stream. J. Neurosci., 22: 6437-6446. Patterson, D.R, Haskins, M.E., Jezyk, PR, Giger, U., Meyers-Wallen, V.N., Aguirre, G., Fyfe, J.C. and Wolfe, J.H. (1988) Research on genetic diseases: reciprocal benefits to animals and man. J. Am. Vet. Med. Assoc, 193:1131-1144. Poptani, H., Gupta, R.K., Roy, R., Pandey R., Jain, VK. and Chhabra, D.K. (1995) Characterization of intracranial mass lesions with in vivo proton MR spectroscopy Am. J. Neuroradiol., 16: 1593-1603. Rademacher, J., Engelbrecht, V, Burgel, U., Freund, H. and Zilles, K. (1999) Measuring in vivo myelination of human white matter fiber tracts with magnetization transfer MR. Neuroimage, 9: 393-406. Rafi, M.A., Zhi Rao, H., Passini, M.A., Curtis, M., Vanier, M.T., Zaka, M., Luzi, P, Wolfe, J.H. and Wenger, D.A. (2005) AAVmediated expression of galactocerebrosidase in brain results in attenuated symptoms and extended life span in murine models of globoid cell leukodystrophy. Mol. Ther., 11: 734-744. Ray, P, De, A., Min, J.J., Tsien, R.Y. and Gambhir, S.S. (2004) Imaging tri-fusion multimodality reporter gene expression in living subjects. Cancer Res., 64:1323-1330. Ray, P , Wu, A.M. and Gambhir, S.S. (2003) Optical bioluminescence and positron emission tomography imaging of a novel fusion reporter gene in tumor xenografts of living mice. Cancer Res., 63:1160-1165.

Ross, B.D., Jacobson, S., Villamil, R, Korula, J., Kreis, R., Ernst, T., Shonk, T. and Moats, R.A. (1994) Subclinical hepatic encephalopathy: proton MR spectroscopic abnormalities. Radiology, 193: 457-463. Safarik, I. and Safarikova, M. (1999) Use of magnetic techniques for the isolation of cells. J. Chromatogr. B. Biomed. Sci. AppL, 722: 33-53. Scriver, C.R.A., Beaudet, A.L.A., Sly, W.S.A., Valle, D.A., Childs, B.A., Kinzler, K.W.A. and Vogelstein, B.A. (2001) The Metabolic and Molecular Bases of Inherited Disease. McGraw-Hill Professional Publishing; Blacklick: McGraw-Hill Companies, New York. Sener, R.N. (2003) Canavan disease: diffusion magnetic resonance imaging findings. J. Comput. Assist. Tomogr., 27: 30-33. Sener, R.N. (2004) Diffusion magnetic resonance imaging patterns in metabolic and toxic brain disorders. Acta Radiol., 45: 561-570. Shah, K., Bureau, E., Kim, D.E., Yang, K., Tang, Y, Weissleder, R. and Breakefield, X.O. (2005) Glioma therapy and real-time imaging of neural precursor cell migration and tumor regression. Ann. Neurol, 57\ 34-41. Shapiro, E.M., Skrtic, S., Sharer, K., Hill, J.M., Dunbar, C.E. and Koretsky, A.P (2004) MRI detection of single particles for cellular imaging. Proc. Natl. Acad. Sci. USA, 101:10901-10906. Shin, J.H., Chung, J.K., Kang, J.H., Lee, Y.J., Kim, K.I., Kim, C.W., Jeong, J.M., Lee, D.S. and Lee, M.C. (2004) Feasibility of sodium/ iodide symporter gene as a new imaging reporter gene: comparison with HSVl-tk. Eur. J. Nucl. Med., 31: 425-432. Sipe, J.C, Filippi, M., Martino, G., Furlan, R., Rocca, M.A., Rovaris, M., Bergami, A., Zyroff, J., Scotti, G. and Comi, G. (1999) Method for intracellular magnetic labeling of human mononuclear cells using approved iron contrast agents. Magn. Reson. Imaging, 17: 1521-1523. Surendran, S., Rady, PL., Michals-Matalon, K., Quast, M.J., Rassin, D.K., Campbell, G.A., Ezell, E.L., Wei, J., Tyring, S.K., Szucs, S. and Matalon, R. (2003) Expression of glutamate transporter, GABRA6, serine proteinase inhibitor 2 and low levels of glutamate and GABA in the brain of knock-out mouse for Canavan disease. Brain Res. Bull., 61: 427-435. Surti, S., Karp, J.S., Freifelder, R., Perkins, A.E. and Muehllenher, G. (2003) Design evaluation of A-PET: a high sensitivity animal PET camera. IEEE Trans. Nucl. Sci., 50: 1357-1364. Taylor, R.M. and Wolfe, J.H. (1997) Decreased lysosomal storage in the adult MPS VII mouse brain in the vicinity of grafts of retroviral vector-corrected fibroblasts secreting high levels of betaglucuronidase. Nat. Med., 3: 771-774. Taylor-Robinson, S.D., Weeks, R.A., Bryant, D.J., Sargentoni, J., Marcus, CD., Harding, A.E. and Brooks, D.J. (1996) Proton magnetic resonance spectroscopy in Huntington's disease: evidence in favour of the glutamate excitotoxic theory. Movement Disord., 11: 167-173. Tjuvajev, J.G., Avril, N., Oku, T., Sasajima, T., Miyagawa, T., Joshi, R., Safer, M., Beattie, B., DiResta, G., Daghighian, R, Augensen, R, Koutcher, J., Zweit, J., Humm, J., Larson, S.M., Finn, R. and Blasberg, R.G. (1998) Imaging herpes virus thymidine kinase gene transfer and expression by positron emission tomography. Cancer Res., 58: 4333-4341. Tjuvajev, J.G., Chen, S.H., Joshi, A., Joshi, R., Guo, Z.S., Balatoni, J., Ballon, D., Koutcher, J., Finn, R., Woo, S.L. and Blasberg, R.G. (1999) Imaging adenoviral-mediated herpes virus thymidine kinase gene transfer and expression in vivo. Cancer Res., 59: 5186-5193. Tjuvajev, J.G., Finn, R., Watanabe, K., Joshi, R., Oku, T., Kennedy, J., Beattie, B., Koutcher, J., Larson, S. and Blasberg, R.G. (1996) Noninvasive imaging of herpes virus thymidine kinase gene transfer

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

350

25. MOLECULAR IMAGING OF GENE THERAPY FOR NEUROGENETIC DISEASES

and expression: a potential method for monitoring clinical gene therapy. Cancer Res., 56: 4087-4095. Tkac, I., Keene, CD., Pfeuffer, J., Low, W.C. and Gruetter, R. (2001) MetaboUc changes in quinolinic acid-lesioned rat striatum detected non-invasively by in vivo (1)H NMR spectroscopy. J. Neurosci. Res., 66 891-898. Tsukada, H., Maramatsu, S., Nishiyama, S., Harada, N., Fukumoto, D., Kakiuchi, T., Nakano, I. and Ozawa, K. (2005) PET imaging of therapeutic gene expression in primate model of Parkinson's disease (abstract). Mol. Imag. Biol, 7: 180. Tzika, A., Ball, W.S., Vigneron, D.B., Dunn, R.S., Kirks, D.R. (1993) Clinical proton MR spectroscopy of neurodegenerative disease in childhood. Am. J. Neuroradiol., 14: 1267-1281. van Dellen, A., Welch, J., Dixon, R.M., Cordery, P, York, D., Styles, P., Blakemore, C. and Hannan, A.J. (2000) N-Acetylaspartate and DARPP-32 levels decrease in the corpus striatum of Huntington's disease mice. Neuroreport, 11: 3751-3757. van der Knapp, M., Ross, B., Valk, J. (1994) Uses of MR in inborn errors of metabolism. In: Kucharczyk, J., Moseley, M. and Barkovich, A.J. (Eds.), Magnetic Resonance Neuroimaging. CRC Press, Boca Raton, PL, pp. 245-318. van der Toon, A., Dijkhuizen, R., Tulleken, C. and Nicolay, K. (1996) Diffusion of metabolites in normal and ischemic rat brain measured by localized IH-MRS. Magn. Reson. Med., 36: 914-922. Vite, C.H., McGowan, J.C, Braund, K.G., Drobatz, K.J., Glickson, J.D., Wolfe, J.H. and Raskins, M.E. (2001) Histopathology, electrodiagnostic testing, and magnetic resonance imaging show significant peripheral and central nervous system myelin abnormalities in the cat model of alpha-mannosidosis. J. Neuropathol. Exp. Neurol., 60: 817-828. Vite, C.H., McGowan, J.C, Niogi, S.N., Passini, M.A., Drobatz, K.J., Haskins, M.E. and Wolfe, J.H. (2005) Effective gene therapy for an inherited CNS disease in a large animal model. Ann. Neurol., 57'. 355-364. Wang, W., Ke, S., Wu, Q., Chamsangavej, C , Gurfinkel, M., Gelovani, J.G., Abbruzzese, J.L., Sevick-Muraca, E.M. and Li, C. (2004) Near-infrared optical imaging of integrin alphavbeta3 in human tumor xenografts. Mol. Imaging, 3: 343-351. Watson, D.J. and Wolfe, J.H. (2003) Lentiviral vectors for gene transfer to the central nervous system. Applications in lysosomal storage disease animal models. Meth. Mol. Med., 76: 383-403. Weber, D.A. and Ivanovich, M. (1995) Pinhole SPECT: ultra-high resolution imaging for small animal studies. J. Nucl. Med., 36: 2287-2289.

Weissleder, R., Moore, A., Mahmood, U., Bhorade, R., Benveniste, H., Chiocca, E.A. and Basilion, J.P. (2000) In vivo magnetic resonance imaging of transgene expression. Nat. Med., 6: 351-355. Wenger, D., Suzuki, K., Suzuki, Y. and Suzuki, K. (2000) Galactosyleer amide lipidosis: globoid cell leukodystrophy (Krabbe disease). In: Scriver, C., Beaudet, A., Valle, D. and Sly, W. (Eds,), The Metabolic and Molecular Bases of Inherited Disease, Vol. III. McGraw-Hill, New York, pp. 3669-3694. Wick, M., Nagatomo, Y, Prielmeier, F. and Frahm, J. (1995) Alteration of intracellular metabolite diffusion in rat brain in vivo during ischemia and reperfusion. Stroke, 26:1930-1933; discussion 1934. Wolfe, J.H. (1994) Recent progress in gene therapy for inherited diseases. Curr. Opin. Pediatr., 6: 213-218. Wolff, S.D. and Balaban, R.S. (1989) Magnetization transfer contrast (MTC) and tissue water proton relaxation in vivo. Magn. Reson. Med., 10:135-144. Wolff, S.D., Eng, J. and Balaban, R.S. (1991) Magnetization transfer contrast: method for improving contrast in gradient-recalledecho images [see comments]. Radiology, 179:133-137. Xia, H., Mao, Q., EUason, S.L., Harper, S.Q., Martins, I.H., Orr, H.T., Paulson, H.L., Yang, L., Kotin, R.M. and Davidson, B.L. (2004) RNAi suppresses polyglutamine-induced neurodegeneration in a model of spinocerebellar ataxia. Nat. Med., 10: 816-820. Yaghoubi, S., Barrio, J.R., Dahlbom, M., Iyer, M., Namavari, M., Satyamurthy, N., Goldman, R., Herschman, H.R., Phelps, M.E. and Gambhir, S.S. (2001) Human pharmacokinetic and dosimetry studies of [18F]FHBG: a reporter probe for imaging herpes simplex virus type 1 thymidine kinase reporter gene expression. J. Nucl. Med., 42:1225-1234. Yeh, T.C., Zhang, W , Ildstad, S.T. and Ho, C. (1995) In vivo dynamic MRI tracking of rat T-cells labeled with superparamagnetic ironoxide particles. Magn. Reson. Med., 33: 200-208. Zafeiriou, D.I., Savvopoulou-Augoustidou, P.A., Sewell, A., Papadopoulou, R, Badouraki, M., Vargiami, E., Gombakis, N.P and Katzos, G.S. (2001) Serial magnetic resonance imaging findings in mucopolysaccharidosis IIIB (Sanfilippo's syndrome B). Brain Dev, 23: 385-389. Zelivyanskaya, M.L., Nelson, J.A., Poluektova, L., Uberti, M., Mellon, M., Gendelman, H.E. and Boska, M.D. (2003) Tracking superparamagnetic iron oxide labeled monocytes in brain by highfield magnetic resonance imaging. J. Neurosci. Res. 73: 284-295. Zhu, L., Ramboz, S., Hewitt, D., Boring, L., Grass, D.S. and Purchio, A.R (2004) Non-invasive imaging of GFAP expression after neuronal damage in mice. Neurosci. Lett., 367: 210-212.

V. GENE THERAPY FOR BRAIN TUMORS AND NEUROGENETIC DISEASES

Subject Index

A a-synuclein, 8, 92,100-101,114,139,140, 142 Acid sphingomyelinase (ASM), 50 Adaptive immunity and viral vectors AAV, 47-48 adenovirus vectors, 48 HSV vectors, 48 lentivirus vectors, 4 8 ^ 9 Adeno-associated virus (AAV), 47, 111, 154, 183-184,186, 244-245 AAV2-mediated gene transfer, 322 AAV2cu HCLN2 vector, manufacturing, 326-328 AAVSl, 183 biology, 18-19 as gene transfer vectors, 22 hunt for, 19 molecular identification and rescue, 19-20 novel primate clades, 21-22 confounding factor for serotyping, 20-21 cross-species transmission, 21-22 evolution, 21-22 identification and isolation, 19-20 serotypes, 20 pseudotyping, 260-261 recombinant AAV (rAAV) cassette, design, 3-4 cell-type-specific tropism, 4-5 expression cassettes, 7 vectors, 18, 94, 96,154, 244, 262 estrogen receptor expression, 196-204 as virus, 17-18 Adenosine, 156 Adenoviral (Ad) vectors. 111, 277-278 Adenovirus, 289, 292, 293, 309, 313 Allogenic stem theory, 320 Alzheimer's disease, 110,112,137-138, 184-186 Amplicon-based HSV vectors, 183,184-186 Amygdala, 172, 231, 235 Amyloid structures, targeting, 141,142, 242

Amyotrophic lateral sclerosis (ALS), 113 gene therapy approaches adeno-associated viruses, 244-245 herpes simplex virus vectors, 246-247 lentiviral vectors, 245-246 motoneurons targeting, 243-244 intrabody therapeutics, 142 treatment, 243 Animal models Canavan disease ASPA knockout mouse, 307-308 ASPA null carriers, 309 tremor rat, 308-309 Antigen-presenting cells (APC), 50 Apoptosis prevention, 100 Apparent diffusion coefficient (ADC), 340 Aromatic amino acid decarboxylase (AADC), 95, 339 Aspartoacylase (ASPA), 160-161 deficient mice, 307, 308, 309 gene therapy, 309-313 chronicles, 310 clinical trials, 310-311 preclinical trials, 311-313 knockout mouse, 307-308 null carriers, 309 Ataxin-1, intrabody targeting, 142 Axonal transport, 126-127, 253-254

B Baculoviral vector, 258 Basal ganglia circuits, 127 Batten disease, late infantile form AAV2cu HCLN2 vector, manufacturing, 326-328 CL2 gene and protein, 319-320 clinical protocol design, 328-330 LINCL, 317-319 CNS manifestations, 320-321 gene therapy challenges, 321-322 pre-clinical efficacy studies, 322-324 toxicology studies, 324-326

351

Behavioral neuroscience viral-mediated RNA interference, 196 Biologies development, 81 Bioluminescence imaging, 344 Blood-brain barrier (BBB), 45,121-123, 339 Botulinum neurotoxin, 256 Brain pathology Canavan diseased patients, 304 Brain tumors gene therapy, 35-36 treatment local vector delivery, 128-129 systemic vector delivery, 128 Bystander effect, 292, 296

C proteins synthesis, 46 CAG hybrid promoter, 323 Calcineurin, 170-171 CaMKII, 169-170 Canavan disease animal models, 307-309 ASPA, in gene therapy, 309-313 brain pathology, 304 clinical trials, 303 NAA metabolic cycle, 304-307 nosology, 303-304 Canine adenovirus serotype 2 (CAV-2), 256-257 Carboxylesterase (CE), 297 CB1954, 296-297 Cell genesis, adult CNS, 55 Central sensitization, 223-224 Chronic pain, 209, 210, 214-218, 231-232 Clinical gene transfer trial, implementation IBC approval, 83 IND submission, 85-87 IRB approval, 83-85 pre-IND meeting, 82 RAC meeting, 82-83 Clinical protocol design, 328-330 CLN2 gene and protein, 319-320 Closed injuries, 275-276

352

SUBJECT INDEX

CNS drug delivery convection-enhanced (CED), 123-124 intranasal, 126 macromolecule distribution enhancements within brain, 124-126 adaptive immunity and viral vectors, 47-49 immune system influence, 45 transgenes immunology, 50 Colchicine, 256 Collimator, 338 Concatemer, 182 Convection-enhanced delivery (CED), 72, 123-124 Cooled charge-coupled device (CCD) camera, 344 CPT-11, 297 Cross-species transmission, 21-22 CycHc AMP response element binding protein (CREB), 34,171-172,186,187 binding protein, 172 targets, 172 Cyclophosphamide (CPA), 294-295 Cytochrome P450 (CYP), 294-295 Cytomegalovirus (CMV), 4, 6, 8, 9, 31, 211, 220, 319 Cytosine deaminase, 294

D Dependovirus, 183 Depression, 187 Dopamine D, receptor (D2R), 339 Dopamine synthesis pathway, 94 Dorsal root ganglia (DRG), 209-210

Elucidating neuronal functions, 34-35 Encephalomyocarditis virus (EMCV), 11,12 Endogenous induction glial progenitors as targets, 60-61 Enkephalin, 217, 218 Enzyme replacement therapy, 320 Epilepsy delivery methods, 153-154 AAV vectors, 154 seizures, experimental models, 155-161 therapeutic implications, 161-162 therapeutic targets anticonvulsant activity, 152 neuroprotection, 152 Estrogen receptor expression mouse hypothalamus, 196-201 transgenic knockouts, 201-204 Ex vivo gene therapy, 95,110, 111, 313 Expression cassettes design and optimization, 3 glial promoters, 9-10 hybrid promoters, 8-9

neuron-specific promoters, 6-8 promoter, choice, 5-10 rAAV cassette design, 3-4 cell-type-specific tropism, 4-5 mediated expression, 10 regulatory elements, 10-12 targeting glia, 9 viral promoters, 6

5-Fluorocytosine (5-FC), 294 5-Fluorouracil (5-FU), 294 Ferritin, 341 Fibroblasts, 279-280 Forebrain targets, 235

GABA-based strategies, 235 Gamma-aminobutyric acid (GABA), 156-157, 218, 235 Ganciclovir (GCV), 128, 291-294 Gene combinations, 95-96 Gene delivery, 32-34, 61, 67 Genetic diseases, 335 Gene expression spatial restriction, 168 temporal restriction, 168 Gene regulation, 31-32 Gene targeting, spatial-temporal pain receptors characterization, 223-228 Gene therapy approaches for ALS, 243-247 neurodegenerative disorders, 114,122, 246 Batten disease late infantile form, 317 Canavan disease approaches, 313 clinical trials, 303 rAAV-mediated, 311 CNS diseases, using intrabodies, 133 delivery methods, 153-155 direct implantation, 71 epilepsy, 151,152 ex vivo, 95, 111, 313 in vivo, 110, 111, 297, 298-299, 342 intraparenchymal infusion, 71, 72 intraventricular injection, 71 landscape gene transfer, clinical approaches, 77-78 serious adverse events, impact, 78 neurodegenerative disorders, 109 neurogenetic diseases, molecular imaging, 335 neurologic disease, 53 and brain tumors, 35-36 neuronal circuitry reset, 99

Parkinson's disease, 59, 91, 92,113-116 hurdles in therapy, 101 therapy intervention strategies, 93-94 UPS and therapy 100-101 prodrug-activation, 291 protocols, oversight, 78 federal agencies, 79-80 good practice requirements, 81 local agencies, 80-81 spinal cord injury, 273 studies in monkeys, 116-117 in nonhuman primate models, 115 successful strategy considerations. 111 testing neurodegenerative disorders, 109 viral vector mediated, for pain forebrain targets, 235 GABA-based sti-ategies, 235 growth factor-based strategy, 233-235 non-pain directed strategies, 235-236 opiate-based strategy, 232-233 Gene therapy agents, 67 intraventricular injection, 71 methods central nervous system accessing, 70-72 intracranial gene transfer enhancing, 72-73 stereotactic neurosurgery anatomical targeting, 68-70 Gene transfer, 337, 338 dopamine synthesis, 94-97 GDNF, 94, 98 growth factors, 97-99 Genetic manipulation learning and memory, 167 techniques, 167-168 memory formation, aspects, 168-176 Glial cell line-derived neurotrophic factor (GDNF), 94 effects on nigrostriatal system, 97-98 gene transfer, 98 use in humans, 98-99 Glial progenitors, parenchymal targets for exogenous mobilization, 60—61 gene delivery, 61 Glial promoters F4/80 microglial, 10 GFAP, 9-10 MBP, 10 Glial suppressive strategies heterotopic neurogenesis, 58-59 Globoid cell leukodystrophy, see Krabbe disease Glutamic acid decarboxylase, 99,114-116, 235 Green fluorescent protein (GFP), 29,126-127,154,155,186, 224, 225 Growth factor-based strategy, 233-235

353

SUBJECT INDEX GTP-cyclohydrolase I, 95 Guanine phosphoribosyl-transferase, 296

H Herpes simplex virus (HSV), 25, 48, 184-186, 276-277 biology, 210-211 gene transfer chronic pain, 214-218 polyneuropathy, 212-214 type 1,182-183, 246-247 thymidine kinase (HSVtk), 291-294 vector, 48,182-183, 246-247, 276-277 vector construction, propagation, 211-212 Homozygous tm/tm rats, 308, 309 HSV amplicon vectors, gene delivery HSV-1 amplicon advantages, 28-32 basics, 25-28 nervous system uses in, 32-36 perspectives, 36-38 HSV-1 amplicon, 25-28 advantages gene regulation, 31-32 hybrid amplicons, 29-31 transgene capacity, 29 virions, modification, 28-29 characteristics, 25-26 vector, 26-28 HSVl-tk, 338 HSV-mediated neurotrophin gene transfer, 212-214 Human a-synapsin-1 (hSYN) promoter, 8 Huntington disease (HD), 59,112-113, 136-139, 343 Hybrid amplicons, 29-31 Hybrid promoters CBA,9 CMV/E-PDGF, 8

Institutional Biosafety Committee (IBC), 79, 80-81 approval, 83 Institutional Review Board (IRB), 80, 83-85 Internal ribosome entry site (IRES) EMCV, 11 hepatitis C, 12 poliovirus, 11 Intrabodies expression, 134,143 function, 136,143 gene delivery, 143,144 generation and selection, 133-135 neurodegeneration, approaches to amyloid structures, 141 pathogenic proteolysis, prevention, 142-143 post-translational modifications, 141-142 protein folding and interactions, 136-141 perspective, 145 potential toxicity, 143-144 power, 136 targeting, 141,142 therapeutics, 142 Intracranial gene transfer, 72-73 Intranasal delivery, 126 Intraplantar formalin, 223, 226 Intravascular injection accessing CNS, 70-71 Inverted terminal repeats (ITRs), 5,10,17, 31, 37,190 Investigational New Drug (IND) submission, 85-87

K Kindling, 153,157,160 Kit style system, 184 Krabbe disease, 344

L I Ifosfamide (IFA), 294-295 Immune system, on CNS gene transfer, 45 adaptive immunity and viral vectors AAV, 47-48 adenovirus vectors, 48 HSV vectors, 48 lentivirus vectors, 48-49, 111, 154,190 transgenes, immunology, 50 In vivo gene therapy, 110, 111 In vivo magnetic resonance spectroscopy, 311, 342, 344 Induced neurogenesis in adult neostriatum, 58 hippocampal atrophies, 59 as therapeutic strategy in Huntington's disease, 59 Innate immune system, 45, 46, 50

Late infantile form of neuronal ceroid lipofuscinosis (LINCL), 317-319 CNS manifestations therapeutic options, 320-321 effective gene therapy challenges, 321-322 Lentipseudotyping, 261 Lentiviral vectors, 48-49, 61,184, 244, 245-246, 259, 261, 263, 278-279 Lentivirus, 48-49, 94, 98, 111, 154,184, 261 Lewy bodies, 91, 92,113,139 Lipid-entrapped, polycation-condensed delivery system (LPD), 310-311 Local vector delivery, 128-129 Long-term depression (LTD), 170,171 Long-term potentiation (LTP), 167,169,170, 171,173,175-176 Lysosomal storage diseases, 320, 325, 342

M Macromolecule distribution enhancements within brain, 124-126 Magnetic resonance methods CNS pathologic changes, 342-344 stem cell tracking and reporter gene expression, 340-342 Magnetic resonance spectroscopy (MRS), 304, 330, 336, 342, 343, 344 Magnetic susceptibility iron-oxide particles, 341 Melanin-concentrating hormone (MCH), 8 Memory formation, aspects extrasynaptic modulation, 175-176 genes and environment, interaction, 174-175 memory acquisition, 168-169 memory consolidation, 174 memory molecules calcineurin, 170-171 CaMKII, 169-170 CREB, 171-172 protein kinase A, 171 postS5maptic receptor clustering, 169 synaptic tagging, 172-174 Mitogen-activated protein kinase (MARK), 173, 203 Molecular therapeutics blood-brain barrier (BBB) bypassing, 121-123,126, 261 brain tumors local vector delivery, 128-129 systemic vector delivery, 128 distribution, within brain, 121 axonal transport, 126-127 perivascular pump, 127 drug delivery to CNS local methods, 123-126 Motoneurons targeting, 243-244 Motor neuron diseases, 262-263 MPSVII mice, 9, 322 MPTP, 97,114,115 MRI-detectable reporter gene, 341-342

N N-acetyl-L-aspartate (NAA), 160, 307, 308, 309, 343, 344 metabolic cycle, 300-303 degradation, 306-307 production, 304-306 shuttling, 306 Af-methyl-D-aspartate (NMDA) receptor, 157-158,169, 223, 224 NAAG, 305, 307, 308 National Institute of Health (NIH), 79-80, 83, 218 Neural progenitor cells adult human brain distribution, 55-56 human brain, 53-54

354 Neural stem cells (NSCs), 55, 56, 58, 241, 282, 340, 341 endogenous, 53 human brain, 53-54 Neurodegeneration, intrabody approaches, 136-143 Neurodegenerative disorders, gene therapy, 109,110 nonhuman primate models, 109 Alzheimer's disease, 112 amyotrophic lateral sclerosis, 113 Huntington's disease, 112-113 need for, 112 Parkinson's disease, 113-116 strategy considerations. 111 study relevance in monkeys, 116-117 therapeutic genes delivery methods, 110-111 usage in, 110 Neurogenesis, 58, 59, 60,102,158,175 Neurogenetic diseases, molecular imaging, 335 animal models, 336-337 gene transfer, 337-338 imaging modalities, 336 magnetic resonance methods pathological changes, 342-344 reporter gene expression, 341-342 stem cell tracking, 340-341 optical imaging, 337, 344-345 radioisotope methods, 338-340 Neurologic disease gene therapy, 35-36 gene transfer trial, imple mentation, 82-87 landscape, 77-78 gene therapy protocols good practice requirements, 81 oversight, 78-81 gene transfer, 77 biologies development, clinical phases, 81 Neurogenic niches, 53 Neuron-specific promoters, 6-8 enolase (NSE) promoter, 4, 6-8,154,160 hSYN, 7, 8,13 MCH peptide, 8 PDGF, 4, 8 Neuronal axonal transport, 253-254 Neuronal ceroid lipofuscinoses, 317 Neuronal circuitry gene therapy for resetting, 99-100 Neuropathic pain, 216, 231 Neuropathy, 212, 240, 263 Neuropeptides enzyme replacement, 160-161 galanin, 158-159 neuropeptide Y (NPY), 159-160 Neurotropic viruses as gene transfer vectors, 258-259 uptake and retrograde transport mechanisms, 254-256

SUBJECT INDEX Nitroreductase, 296-297 Nonhuman primate models need for, 112 neurodegenerative disorders for testing gene therapy, 109,112-116 Non-pain directed strategies, 235-236 Nonviral vectors, 110 Nuclear magnetic resonance (NMR), 340, 336

O Office for Human Research Protections (OHRP), 79, 80 Office of Biotechnology Activities (OBA), 79,83 Olfactory ensheathing glia cells, 281-282 Olfactory route CNS gene delivery, 70 Oncolytic adenovirus, 297 Oncolytic virus (OV), 212, 295 Opiate-based strategy, 232-233

Pain, 210, 214-218, 231, 232, 234, 263-264 Parenchymal volume loss, 318 Parkinson's disease, 59, 91, 92,113-116,139, 264^266 animal models, 93 apoptosis, prevention, 100 dopamine synthesis biochemical augmentation, 94-97 gene therapy hurdles, 101 intervention strategies, 93-94 to reset neuronal circuitry, 99-100 UPS, 100-101 gene transfer, growth factors, 97-99 neuroprotective strategies, 116 therapy comparison pharmacological vs. gene, 92-93 Passaging, 183 Pathogenic proteolysis, prevention, 142-143 Perikaryon, see Soma Peripheral nerve bridges, 281 Peripheral nervous system, 209, 212, 216, 276 Perivascular pump, 127 Platelet derived growth factor (PDGF) p-chain promoter, 8 Polyneuropathy, 209, 210, 212-214 Porcine pseudorabies virus (PrV), 255 Positron emission tomography (PET), 101, 338-339, 345 tracer, 336 Post-translational modifications intrabody approaches, 141-142 PP2B, see Calcineurin Pre-IND meeting, 82 Progenitor cells, human brain, 53-54 mobility, 56-58

neural, 55-58 in storage diseases, 61-62 in Parkinson's disease, 59 substrates for repair, 56 Promoter, choice, 5-10 Protein folding and interactions Alzheimer's disease, 139-140 Huntington's disease, 136-139 Parkinson's disease, 139 prion, 140-141 synucleinopathies, 139 tauopathies, 139-140 Protein kinase A, 171 Proton MRS, 342 Pseudotyping, 18 vector, 245 AAV, 260-261 lentiviral, 259, 261 viral, 260 Psychiatric disorders, clinically viable gene therapy better vectors, designing, 189 controlling expression, 189-190 good targets, finding, 188-189 safety issues, 190-191 Psychiatry, viral vectors utility, 181,184, 186-191 Purine analogs, 297-299 Purine nucleoside phosphorylase (PNP), 297-299

R Rabies virus, 254-256, 258 RAC meeting, 79, 82-83 Radioisotope, 336, 338-340 Rat models spinal cord injury, 275-276 Recombinant vectors viral vectors AAV, 183-184 HSV, 182-183 lentivirus, 184 sindbis virus, 184 Regulatory elements bi-directional promoters, 12 expression, two or more genes, 11 IRES, 11-12 polyproteins cleavage, 12 WPRE, 10-11 Reporter genes, 127, 280, 336, 337-338 Retrovirus, 31, 48, 111, 184, 292, 293, 321 Rostral agranular insular cortex (RAIC), 235 Rous sarcoma virus (RSV), 5, 6, 7

Schwann cells, 241, 279, 280-281 Seizures, experimental models genetically modified cells, transplantation, 155-157

SUBJECT INDEX nonviral delivery systems, 161 viral vector, gene transfer, 157-161 Setting back the clock concept, 321-322 Sindbis virus, 184 Single positron emission computed tomography (SPECT), 336, 338-340, 345 Sodium-iodide symporter (MS), 339 Soma, 48, 253-254 Spasticity, 256, 266-267 Spatial-temporal gene targeting, 223 Spinal cord dorsal horn (SCDH), 223, 224, 225 Spinal cord injury future perspective, 282-283 pathophysiology acute phase, 274 chronic phase, 275 subacute phase, 274-275 rat models, 275-276 for successful regeneration, 276 transplantation and gene therapy, combination fibroblasts, 279-280 neural stem cells, 282 olfactory ensheathing glia cells, 281-282 peripheral nerve bridges, 281 Schwann cells, 280-281 viral vector-mediated gene transfer adeno-associated vectors, 278-279 adenoviral vectors, 277-278 herpes simplex viral vector, 276-277 lentiviral vectors, 278-279 Spinal muscular atrophy (SMA), 246, 262 Status epilepticus, 157,159,160 Stereotactic neurosurgery, 68-70 Storage disease, 61-62 Suicide gene therapy, 292, 294, 296-297 Superparamagnetic iron oxide (SPIO), 340-341 Synaptic plasticity, 173,175 Synucleinopathies, 139,142 Systemic vector delivery, 128

6-Thioxanthine (6-TX), 296 Targeting glia, 9 Tau, intrabody targeting, 141 Tauopathies, 139-140 Tegument, 25, 28, 29, 32, 37, 38, 210, 255 Tetanus toxin, 261 heavy chain (HC), 255-256 as protein carrier targeting CNS, 259-260 light chain (LC), 255, 259 Tetracycline-responsive system, 32,154 Tetracycline transactivator, 168 Therapeutic genes, 128,157, 259 methods of delivery, 110-111 Transgene capacity, 29, 31, 36, 38 immunology, 50 Transplantation and gene therapy, combination, 279-282 Tremor rat, 308-309 Tyrosine hydroxylase, 94-95, 98,183, 339

U Ubiquitin-proteasome system (UPS), 92, 93, 100-101 US Food and Drug Administration (FDA), 79

V Van Bogaert-Bertrand syndrome, 303 Vascular endothelial growth factor (VEGF), 172-173 ALS, 242-243 gene therapy approaches, 243-247 therapy, 243 angiogenic player, 239-240 delivery with EIAV lentiviral vector, 247

355 in nervous system, 241 in neural development and disease angiogenic role, 240 neural role, 240-242 protein therapy, 247 therapeutic mechanisms, 248 therapeutic potential, for ALS, 247-248 VEGF therapy, perspectives, 248 Ventromedial nucleus (VMN), 197-201 Vesicular monoamine transporter-2 (VMAT2), 96 Viral gene transfer technology, 187 Viral promoters CMV, 4, 6, 8, 9, 31, 211, 220, 319 RSV, 5, 6, 7 Viral vector, 110, 111 current data, 256-258 psychiatric applications, 181 recombinant vectors, 181-184 study of behavior, application, 184-186 utility, in psychiatry, 186 uses, 188-191,195 AAV vectors, 196, 201 behavioral neuroscience, 196 Viral vector, axonal uptake and retrograde transport current data, 256-258 neuronal axonal transport, 253-254 neurotropic viruses as gene transfer vectors, 258-259 mechanisms, 254-256 retrograde transport, application Alzheimer's disease, 264-266 motor neuron diseases, 262-263 neuropathy, 263 pain, 263-264 Parkinson's disease, 264-266 spasticity, 266-267 tetanus toxin HC, as protein carrier targeting CNS, 259 vector pseudotyping, 260-261 Virions, modification, 28-29