Gene therapy, also known by the compounds and materials deposited in the brain in search of a therapeutic response shows much promise. This paper = is by Dr. K.S. Bankiewicz, a scientist with Somatix Therapy Corp., Alameda, = CA. The Neurosciences Institute, which is associated with Good Samaritan=20 Hospital in Los Angeles, sponsored a symposium on Nov. 2-3, 1996 in Rancho Mirage, CA. This review was included in their syllabus. I wish to thank Carole Hilton for making this available to the archive. Application of Gene Therapy for Parkinson's Disease;=20 Non-Human Primate Experience K.S. Bankiewicz Somatix Therapy Corporation, 850 Marina Village Parkway,=20 Alameda, CA 94501 Introduction Parkinson's disease (PD) is characterized by the=20 progressive loss of dopaminergic cells and a deficiency of=20 tyrosine hydroxylase (TH) which is necessary for the synthesis=20 of dopamine (DA). As the degeneration of nigral dopamine neurons=20 continues, intracellular biochemical responses occur in the=20 presynaptic dopamine neurons and the postsynaptic striatal=20 neurons that enable the nigrostriatal dopamine system to=20 compensate for great depletions of DA prior to the onset of clinical symptoms. As the disease progresses, and up to 85%=20 of nigral dopaminergic neurons have degenerated, the synthesis=20 of TH and the release of DA from the presynaptic terminals is increased in a compensatory manner. Early in the disease, this TH deficiency can be offset=20 by oral administration of L-dopa. The exogenous L-dopa is taken=20 into the cell via active amino acid transport, where it is acted=20 upon by L-AADC - thus circumventing the rate limiting TH. Currently=20 it is thought that the exogenous L-dopa exerts its therapeutic effects=20 by converting to dopamine in the remaining striatonigral DA cells=20 which then release DA in a physiologically relevant manner. Oral administration of L-dopa continues to be the mainstay=20 of current therapy for Parkinson's disease but is limited by:=20 (i) the inability to achieve site specific delivery, which results in unwanted side-effects and limits the amount of drug which can=20 be given, (ii) the inability to maintain a sustained drug level within=20 the central nervous system (CNS), which is thought to contribute=20 to unpredictable "on-off" effects in the treatment, and (iii) the=20 inability to prevent the progressive degeneration of dopamine- secreting nerve cells (1). Development of new therapeutic approaches to Parkinson's disease must address these issues. Local Delivery of Therapeutic Agents into CNS Delivery of macromolecular therapeutic agents into the=20 CNS presents unique challenges. Systematically-administered=20 proteins do not enter the brain tissue due to exclusion by the=20 brain capillary endothelial cells. Since the blood brain barrier=20 (BBB) is permeable to lipid-soluble compounds, enhancement of=20 the lipid solubility of polypeptides has been attempted, but with limited success. The permeability through brain capillaries can=20 also be increased with chemical modifications of proteins such as=20 cationization, Conjugation of peptides and proteins to anti- transferin receptor antibodies also increases uptake into the brain,=20 but this has resulted in systemic toxicity. Drug-impregnated=20 polymer devices can provide sustained and local release into the CNS. However, such delivery is only possible for a limited period of time,=20 and "empty" polymer capsules that have limited biodegradability=20 remain in the CNS after therapy. Regardless of the method used for=20 direct drug delivery, the ability of agents to distribute within the=20 brain following introduction at a local site is a critical issue in=20 determining drug efficacy and toxicity, Direct administration of therapeutic agents into the=20 CNS eliminates the need to bypass the (BBB) and thereby, reducing=20 systemic side effects. Delivery of proteins into the cerebral=20 ventricles results in high protein concentration in the tissue=20 adjacent to the ventricle, but there is limited penetration into=20 surrounding regions and clearance into the peripheral circulation=20 is rapid, Direct injection into the brain parenchyma at the target=20 site has been shown to be a more effective means of achieving high=20 drug concentration at specific sites within the brain,=20 Intraparenchymal cannulae and infusion pumps are required for=20 chronic delivery, making this approach less desirable, The short penetration distances of proteins pose an even=20 greater problem in the human brain than in the relatively small=20 rat brain, The treatment of many neurodegenerative disorders will require protocols which allow for delivery of sufficient levels of=20 therapeutic agents over significant distances in brain parenchyma. Cell-based delivery into CNS Cellular implants can overcome the need for persistent=20 and local physical delivery of therapeutic agents into the CNS=20 Limited but striking successes have been achieved through the implantation of DA-producing fetal cells in Parkinsonian patients,=20 The requirement of many embryos per patient is a major factor=20 restricting the wide scale adoption of this approach as a possible treatment for PD. Several alternative approaches utilizing genetic=20 engineering have been proposed by many investigators (2).=20 Primary or immortalized cells have been engineered to produce=20 a specific protein in culture and then implanted into the host=20 CNS either by direct cell transplantation or by encapsulating=20 cells into semipermeable membranes. Other approaches consist of=20 in vivo gene transfer based on direct introduction of genetic=20 material into the CNS using viral or synthetic vectors. Ex vivo=20 methods using primary fibroblasts transduced with retroviral=20 vectors and the in vivo method utilizing adeno-associated virus=20 (AAV) are described here as possible means of chronic delivery of=20 therapeutic agents into CNS. In order to develop a viable gene=20 therapy for any neurological disorder, four requirements must be=20 met: (i) demonstration of long-term expression; (ii) supply=20 of local and therapeutic levels of the desired protein (iii) lack=20 of toxicity of the procedure; and (vi) the ability to reverse the=20 treatment if adverse events occur. Ex-vivo Gene Therapy using Autologous Cells One method for delivery of L-DOPA locally into the CNS=20 is to transplant cells that synthesize and secrete L-DOPA after=20 ex vivo genetic modification. In this scenario, a PD patient's=20 own cells would be genetically modified to express TH. Depending=20 on the cell type being used as a vehicle, it may be necessary to=20 express a second enzyme, GTP cyclohydrolase 1, to facilitate the=20 production of the cofactor for TH activity, tetrahydrobiopterin=20 (BH4). These cells then would be grafted into appropriate sites=20 in the striatum to provide a local supply of L-DOPA at the sites=20 in the brain that are normally innervated by dopaminergic neurons.=20 The L-DOPA secreted by these cells may then be taken up by=20 remaining neuronal and non-neuronal cells, converted to DA and=20 released either in a normally regulated or in a unregulated fashion to ameliorate symptoms of the underlying Parkinsonism in these patients. In-vivo Gene Transfer Significant advances have recently been made regarding=20 the introduction of genetic material into the mammalian cells in vivo. Several vehicles have been used for in vivo transfer of cDNA, including herpes simplex viral vectors, adeno viral vectors,=20 direct plasmid DNA transfer and, most recently, retroviral=20 vectors (lentivirus) (2). An additional promising vector system utilizes AAV, a nonpathogenic in human single stranded DNA=20 parvovirus (3). In addition to having no known pathology in man=20 the wild-type virus is incapable of replication without helper functions provided by an adenovirus. Wild type AAV has been shown=20 to infect post-mitotic mammalian cells and to insert its DNA=20 into the host genome with some specificity for chromosome 19.=20 It should be stressed, however, that at the present time it is=20 difficult to predict which of these in vivo delivery systems will=20 prove to be the most optimal and clinically applicable. Delivery of TH-AAV vector into monkey striatum In addition to the difficulty of introducing therapeutic=20 agents into the CNS another significant challenge is presented=20 by PD. Since the disease affects mostly nigrostriatal and partially mesolimbic systems, therapeutic agents must be spread over a=20 considerable area of the brain. This can be accomplished by=20 utilizing multiple delivery sites. However, with the increasing=20 number of penetration sites into the CNS, the risk of complication=20 increases as well. At the present time direct gene delivery into=20 the CNS can be accomplished only by introduction of viral vectors=20 into the brain parenchyma. Injection of the AAV vector results in=20 a limited and highly concentrated infection of the brain located=20 within close proximity to the injection cannulae. Such a limited=20 area of coverage is not desirable when large regions of the human=20 brain are to be targeted. In addition, due to the limited=20 diffusion of therapeutic agents in the brain, infections are=20 highly localized. An overproduction of a gene product in a limited=20 area can result in local toxicity, which give rise to unwanted=20 side effects. Attempts to spread the infection into larger areas of=20 the brain using a slow infusion technique and fused silica=20 cannulae resulted in a higher number of infected cells and a less concentrated area of infection. No signs of toxicity in the=20 striatum related to the volumes or techniques used were observed (4). Gene expression at 4 months in vivo approach Gene expression was detected by immunostaining in monkey=20 striatum after l4-21 days and at 3 months survival for both=20 =DF-galactosidase (=DF-gal) and TH-AAV vectors. Gene expression after 3 months appeared less robust than at the earlier time points,=20 however, the level of gene expression over time is now under=20 investigation. Based on double labeling immunocytochemistry using antibodies against the transgene and neurofilament or glial=20 fibrillary acidic protein (GFAP), it appears that the AAV vector=20 used infects almost exclusively neurons in the monkey brain. Some glia-like cells were detected expressing =DF-gal. =20 However, they constituted a minority of AAV-infected cells as=20 they were estimated to be less than 5% of all the cells detected to express =DF-gal. Furthermore, no GFAP-positive cell was observed to=20 express also =DF-gal. The number of =DF-gal positive cells was the greatest when=20 the slow infusion methods were applied. In this case there was=20 estimated to be between 31,000 to 14,000 cells per single delivery site (4). ex-vivo approach Expression of the TH gene was examined at four months=20 after the fibroblast implantation by in situ, hybridization,=20 immunostaining with a TH antibody and Southern PCR. The application of the photographic emulsion combined with H&E counter staining=20 allows the cellular resolution of in-situ hybridization. The TH=20 mRNA-expressing cells were recognized by the accumulation of silver grains over the cell bodies. The grains were highly=20 concentrated over the groups of elongated fibroblast inside the=20 implants. Positive TH staining was observed in the fibroblast implants as well. So far, we have examined the gene expression in=20 genetically modified fibroblasts for up to four months. We have=20 demonstrated persistent expression in the grafted cells for this=20 period. The addition of GTP-cyclohydrolase-l vectors, and therefore the production of BH4, appears to facilitate TH immunostaining,=20 perhaps by lengthening the half-life of the TH enzyme in co- infected fibroblasts. As has been shown in rat experiments, the=20 combination of expressing TH and GTP-cyclohydrolase- 1 cDNAs in=20 grafted cells also makes it possible to detect L-DOPA by microdialysis. The application of AAV-based vectors into the monkey=20 striatum results in robust gene expression which is limited mostly=20 to neuronal populations of cells in the stratum. The mechanism for specific targeting of the AAV vector expression to neurons is=20 not clear at this time. However, we have observed this pattern=20 using both the cytomegalovirus (CMV) immediate-early promoter and a Moloney Murine leukemia virus (MoMuLV) promoter driving=20 expression in the AAV vector. A critical question still remains=20 regarding the regulation of gene expression. The development=20 of regulatable promoters might help to overcome this in the=20 near future. Further technological advances are required to=20 optimize gene delivery, regulation of gene expression and testing in appropriate functional models before gene therapy=20 can be used extensively. References 1 Chase TN, Juncos J, Serrati C, Fabbrini G, Bruno G.=20 Fluctuations in response to chronic levodopa treatment therapy:=20 pathogenic and therapeutic considerations. In: Yahr MD, Bergmann KJ (eds) Advances in Neurology: Parkinson's disease. New York:=20 Raven Press, 1986, 477-480. 2 Freese A, Stern M, Kaplitt MG, O'Connor W, Abbey M,=20 O'Connor MJ, During MJ, Prospects for gene therapy in=20 Parkinson's disease. Mov Dis 1996; 5:469-488 3 Muzyczka N, Use of adeno-associated virus as general=20 transduction vector for mammalian cells. Curr Top Microbiol=20 Imunol 1992; 158: 97-l29. 4 Bankiewicz KS, Snyder R, Zhou SZ, Morton M, Conway J,=20 Nagy D. Adeno-associated (AAV) viral vector-mediated gene=20 delivery in non-human primates. Soc. Neurosci Abst 1996:22:768 John Cottingham To search the Parkinsn archive, send search requests to [log in to unmask] with Archive Search as the subject. LibraryH Searches of the Subject: line, From: line and Body are possible. Look for "Revised Current Topics...." messag= e HomeBoy for Articles and Studies available by e-mail. PARKINSONIANS WORLD-WIDE GIVE THANKS IN MEMORY OF ALAN= =20 [log in to unmask] BONANDER.....WHERE EVER YOU ARE!