205. Intrathecal Administration of AAV Vector for the Treatment of Lysosomal Storage Disease in the Brains of MPS I Mice

205. Intrathecal Administration of AAV Vector for the Treatment of Lysosomal Storage Disease in the Brains of MPS I Mice

GENETIC AND METABOLIC DISEASES: PART ONE deficient insulin secretion and hyperglycemia. Despite intensive insulin therapy, properly controlled glucose...

77KB Sizes 6 Downloads 24 Views

GENETIC AND METABOLIC DISEASES: PART ONE deficient insulin secretion and hyperglycemia. Despite intensive insulin therapy, properly controlled glucose values are hardly achieved, therefore degenerative chronic complications may arise. In search for alternative therapeutic options, this project aims at engineering non-beta cells for the production of insulin in a glucoseregulated manner. The liver is the candidate target organ for this purpose, as hepatocytes are able to ‘sense’ extracellular glucose variations and modulate consequentially gene expression. To obtain efficient insulin expression in the liver, we are exploring HelperDependent Adenoviral Vectors (HD-AdV), which should also provide a better safety profile. We generated a vector coding for Furine cleavable-Human Proinsulin (FurHPI) under the control of the glucose-responsive liver-specific Pyruvate Kinase gene promoter (L-PKp) coupled to the SV40 enhancer. The vector was injected into STZ-induced diabetic immunodeficient’nude’ mice (n=4/group) at two different doses: 1.4x10e11 (low dose) or 2,8x10e11 (high dose) vp/mouse. Human insulin was detectable as early as the first week after virus injection in both groups, reaching a maximum between the 2nd and 3rd week of 1.1-3.6 μU/ml in the low-dose group and of 129-192 μU/ml in the high-dose group, and remaining at similar levels thereafter. Mice injected with the low dose remained hyperglycemic whereas with the high dose blood glucose returned to normal value, starting from the 2nd week. At 4 weeks, the high dose produced exceedingly high insulin levels (normal insulin values in fed nude mice =49.3+8.1 μU/ml). Consequently, blood glucose level was reduced to moderate-severe hypoglycemic value (53±8 mg/dl). Glucose loads and fasting tests demonstrated that L-PKp displays a slow and weak (3.6-fold over 4 h) induction and 6-fold repression of insulin synthesis upon glucose increase/decrease. We then generated a second vector containing a Hybrid(L-PK/Spot14) promoter, which displayed a stronger response to glucose than the L-PKp once transfected in rat hepatocytes. We linked this promoter to the nuclear LacZ reporter gene in order to quantify liver expression and tissue specificity. This vector was injected into ‘nude’ mice (n=2) at the dose of 2,8x10e11 particles/mouse. Animals were sacrificed and tissues collected 4 weeks post-gene transfer and analyzed for histochemistry and β-Galactosidase activity. Liver sections showed ∼15% transduced hepatocytes. β-Galactosidase activity was detected in liver extracts only, indicating that this promoter maintains liver-restricted expression. Hepatocytes toxicity was determined by liver enzyme function test. AST and ALT peaked (x5) at day 1 after injection, then returned to normal level at day 7. The reporter gene will be replaced with FurHPI to obtain faster glucose-responsive induction of insulin synthesis in liver cells. These results indicate that HD-AdVv are suitable vectors to pursue a finetuned glucose-responsive insulin expression in the liver for gene therapy of type 1 diabetes.

204. Systemic Gene Therapy with Interleukin10 (IL-10) for Prevention of Type 1 Diabetes in NOD Mice: Comparison of Mutant (I87A Substitution) to Recombinant Murine IL-10 Matthias H. Kapturczak,1 Clive H. Wasserfall,2 Scott Loiler,3 Martha Campbell-Thompson,2 Marda Scott-Jorgensen,2 Jeff Cross,2 James M. Crawford,2 Tamir M. Ellis,2 Terence Flotte,3 Mark A. Atkinson.2 1 Medicine, University of Florida, Gainesville, FL; 2Pathology, Immunology and Laboratory Medicine, University of Florida, Gainesville, FL; 3Pediatrics, University of Florida, Gainesville, FL. Type 1 diabetes results from the autoimmne destruction of the insulin producing β cells. While the immunological mechanisms underlying this process are unclear, immunoregulatory defects appear to be associated with genetic susceptibility and disease progression. Previous studies have demonstrated that a single intramuscular (I.M). S80

injection of recombinant adeno-associated virus (rAAV) vector (serotype 1) containing the murine IL-10 gene, in a dose and timedependent fashion, consistently and dependably prevents type 1 diabetes and insulitis development in female non-obese diabetic (NOD) mice. However, histological evaluation of the muscle injection sites receiving rAAV-IL-10 revealed long term evidence of myositis and atrophic muscular changes. Additional studies have suggested that a substitution of isoleucine at position 87 of IL-10 with alanine inhibits some of the immunostimulatory functions of IL-10, rendering the mutated molecule a potentially more safe agent for studies aimed at disease prevention. To compare IL-10 with I87A-IL10 in terms of diabetes prevention and side effect profiles, groups of 4 week-old female NOD mice (n=13 animals per group) were I.M. injected with 109 infectious units (IU) of rAAV-IL-10, 109 IU of rAAV-I87A-IL10, 107 IU of rAAV-I87A-IL-10, 109 IU of rAAV-delta-IL-10 (encoding the first 33 amino acids of IL-10 and not biologically active) and saline as controls. Urine glucose levels were monitored weekly for disease development, with disease rates assessed by life-table (Kaplan-Meier) analysis. At 14 weeks post injection, 3 animals from each group were sacrificed for histological and immunological assesments. At 38 weeks post injection, diabetes frequency was as follows: IL-10 0% p=0.0016 vs. saline, 0.0671 vs. delta-IL-10), 109 IU I87A 20% NS vs. controls), 107 IU I87A 60%, delta-IL-10 30% and saline 50%. In terms of the prevention of insulitis, the degree of infiltration essentially mirrored that of diabetes development with high-dose IL-10 proving most effective.Unlike the saline and delta-IL-10 groups, the IL-10 group and both I87AIL-10 groups showed development of significant infiltrate of the muscle at the injection site. Phenotyic analysis (CD3, CD4, CD8, macrophage, B-lymphocyte) revealed this inflitrate to be predominantly B-lymphocyte in origin. Significant muscle fiber atrophy, however, was seen only in the IL-10 group. Our study demonstrates that I87A-IL-10, was less efficient than IL-10 in terms of diabetes prevention. Furthermore, such substitution did not abrogate the chemotactic properties of IL-10. Studies of other mutations in the IL-10 structure will be necessary to select for only the desirable parts of the IL-10 function repertoire; allowing for effective disease prevention and devoid of sequence allowing for potentially deleterious inflammation.

205. Intrathecal Administration of AAV Vector for the Treatment of Lysosomal Storage Disease in the Brains of MPS I Mice Gordon Watson,1 Jacob Bastacky,1 Steve Jungles,2 Michael Vellard,2 Pavel Belichenko,3 Emil Kakkis.2 1 Children’s Hospital Oakland Research Institute, Oakland, CA; 2 BioMarin Pharmaceutical, Novato, CA; 3Stanford University, Stanford, CA. Mucopolysaccharidosis Type I (MPS I) is caused by an inherited deficiency of α-l-iduronidase (IDU). The result is a progressive, lysosomal storage disease that includes CNS as well as systemic involvement. Enzyme replacement therapy by periodic intravenous infusion of purified IDU has been shown to be effective in treating the non-CNS manifestations of the disease; however, the blood brain barrier excludes therapeutic levels of IDU in the brain. Thus, our goal was to develop gene therapy for MPS I specifically targeting the brain. Although the blood brain barrier also excludes viral vectors that are administered intravenously, initial studies with a similar storage disease, MPS VII, indicated that intrathecal administration of an AAV vector could circumvent this problem. For MPS VII experiments, which eliminated storage vacuoles throughout the brain, a relatively high dose of 5 x 1011 vector particles per adult mouse was used. To treat MPS I mice, AAV-IDU vectors with and without the woodchuck hepatitis virus posttranslational regulatory element (WPRE) were constructed. These vectors contained human IDU Molecular Therapy Vol. 7, No. 5, May 2003, Part 2 of 2 Parts

Copyright © The American Society of Gene Therapy

GENETIC AND METABOLIC DISEASES: PART ONE cDNA driven by a CMV promoter/enhancer. The vector containing the WPRE averaged higher expression levels and was chosen for further studies. Different doses of vector were administered by subdural injection into the thoracic spines of adult MPS I mice. Enzyme activities in brain were then measured several weeks later. Lower doses were explored particularly to establish the feasibility of future scale-up experiments in larger animals. Doses of 0.2, 0.7, 2 and 6 x 109 particles of AAV-IDU provided 3, 3, 8 and 48% of normal IDU activity in brain (mean activities from coronal sections in mid brain using 2 or 3 MPS I animals at each dose). With the exception of higher activity in the olfactory bulbs, IDU activity in brain coronal sections tended to increase rostral to caudal. This distribution of IDU positive cells was confirmed in sagittal sections using immunofluorescence staining with anti-IDU antibody and confocal microscopy. Furthermore, it was noted that neurons in particular were expressing IDU. Since only a small fraction of normal activity should be sufficient to reduce storage, histopathology of mice treated with 2 x 109 particles was investigated in detail. Storage in the MPS I mice was particularly prominent in perivascular cells. To estimate the frequency of brain cells that had storage material, 10 high-power fields per section were scanned using light microscopy, and all cells having apparent storage vacuoles were scored. The fraction of capillaries showing storage in perivascular cells was 46% in untreated mice and 19% in treated mice. Furthermore, among the remaining cells with any storage vacuoles at all, the amount of storage per cell was much reduced after the AAV-IDU treatment. In the same sets of 10 fields, there were 5 parenchymal cells with storage from untreated and none from treated mice. We conclude that the intrathecal route of administration is effective for delivering vector to brain cells, and that relatively low doses of AAV-IDU are sufficient to substantially reduce, but not completely eliminate storage vacuoles. Co-authors who are employed by BioMarin Pharmaceutical have a vested interest in results that may contribute to the financial success of the company. The presenter has no financial interest in the company.

206. Muscle-Mediated Human Factor IX Expression in Cynomolgus Monkey Using AAV Serotype-Derived Vectors Hiroaki Mizukami,1 Jun Mimuro,2 Tsuyoshi Ogura,1 Fumiko Ono,3 Eiji Kobayashi,4 Shin-ichi Muramatsu,5 Seiji Madoiwa,2 Takashi Matsushita,1 Takashi Okada,1 Yutaka Hanazono,1 Akihiro Kume,1 Keiji Terao,3 Keiya Ozawa,1 Yoichi Sakata.2 1 Division of Genetic Therapeutics, Center for Molecular Medicine, Jichi Medical School, Minamikawachi-machi, Tochigi, Japan; 2 Division of Cell and Molecular Medicine, Center for Molecular Medicine, Jichi Medical School, Minamikawachi-machi, Tochigi, Japan; 3Tsukuba Primate Center, National Institute of Infections Diseases, Tsukuba, Ibaraki, Japan; 4Division of Organ Replacement Research, Center for Molecular Medicine, Jichi Medical School, Minamikawachi-machi, Tochigi, Japan; 5 Department of Neurology, Jichi Medical School, Minamikawachimachi, Tochigi, Japan. Hemophilia B suits with gene therapy utilizing AAV vectors. For the muscle-directed approach, AAV 1-derived vectors were most advantageous among the serotypes, at least in small animals. In order to estimate the utility of AAV serotype-derived vectors in humans, we compared the efficacy and the consequences of these vectors in nonhuman primate model. A total of 10 male cynomolgus monkeys, aged 4-12, were included in this study. All of the animal experiments were performed under the institutional guidelines. The monkeys were divided into three groups and received intramuscular injection of human factor IX-encoding AAV vectors based on serotype 1, 2 and 5. Vector dose at 1012 vector genomes/kg body Molecular Therapy Vol. 7, No. 5, May 2003, Part 2 of 2 Parts Copyright © The American Society of Gene Therapy

weight was chosen throughout the study. The assay system for quantifying human factor IX apart from monkey counterpart was already established and utilized in this study. One week following injection, the expression of human factor IX was recognized and surged thereafter. Two of the monkeys received AAV1-based vectors showed high levels of human factor IX, reaching to the therapeutic level (86 and 138 ng/ml). After four weeks, serum levels of human factor IX went down and became undetectable in most cases. Generation of anti-human FIX antibody of IgG class was demonstrated, despite the high degree of homology of these molecules in amino acid sequences (greater than 97%). Antibody against vector capsid corresponding to the injected serotype was also demonstrated in 9 out of 10 subjects. The third monkey with AAV-1 injection did not exhibit human factor IX and showed antibody against AAV-1 capsid, which may affect gene transfer efficiencies. These results imply that 1) serotype 1-based vectors achieved much higher expression of factor IX in monkey muscle than those with serotype 2 or 5, suggesting the advantage of this vector in humans, 2) human factor IX showed immunogenicity in monkeys, despite the extreme homology, 3) antibody against vector capsid with neutralizing activity was developed following 2-4 weeks of injection. These results will give insight for designing therapeutic approaches into humans.

207. In Vivo Assessment of the Efficacy of Gene Therapy Targeted to Diseased Human Hematopoietic Progenitor Cells in a Xenotransplantable Murine Model of a Lysosomal Storage Disorder Alex Hofling,1 Steven Devine,1 Carole Vogler,2 Mark Sands.1 Internal Medicine, Washington University School of Medicine, St. Louis, MO, United States; 2Pathology, Saint Louis University School of Medicine, St. Louis, MO, United States. 1

Lysosomal storage diseases (LSDs) are inherited disorders that typically result in a systemic accumulation of undegraded catabolites. The strong therapeutic response of many of these disorders to bone marrow transplantation suggests that hematopoietic progenitor cells (HPCs) would be an effective target for gene therapy. To evaluate the efficacy of such an approach directly on diseased human HPCs in vivo, a mouse model of LSD that was also capable of engrafting human cells was needed. By moving the murine mucopolysaccharidosis type VII (MPS VII) mutation onto the widely used xenotransplantation strain, non-obese diabetic/ severe combined immunodeficient (NOD/SCID), such a model was generated (NOD/SCID/MPSVII). MPS VII is an LSD caused by a deficiency in β-glucuronidase (GUSB) activity and results in a broad range of clinical manifestations in both human patients and NOD/ SCID/MPSVII mice. A therapeutic baseline was established in these mice through the transplantation of human CD34+ hematopoietic cells from healthy GUSB+ donors. NOD/SCID/MPSVII mice engrafted highly with human cells of both lymphoid and myeloid lineages in the bone marrow, spleen, and blood 6-12 weeks following transplantation. In situ histochemical staining for GUSB activity identified large numbers of human cells in other organs such as the liver, kidney, lung, heart, meninges, and retina. The widespread distribution of GUSB+ human cells resulted in a reduction of lysosomal storage material in many host tissues including the liver, spleen, and bone. To determine the relative efficacy of a gene therapy protocol directed at diseased human HPCs, CD34+ cells were purified from the mobilized peripheral blood apheresis product of a human MPS VII patient. The peripheral blood leukocytes of this patient had only 1.3% of the normal level of GUSB specific activity. An HIV-based lentiviral vector with a poly-purine tract was used to transfer a functional GUSB cassette to the patient’s CD34+ cells. A 16 hour transduction protocol in the presence of IL-3, IL-6, SCF, S81