- Email: [email protected]
476. Pseudotyped Adeno-Associated Virus Gene Transfer in Wound Healing
APPLICATIONS IN DISEASE hEPO or rAd5-hEPO, respectively, via intraoral cannulation of both submandibular glands. A third group (n=3) received 108 particles of rAAV2-hEPO. Additionally, a fourth group (n=3) received 1011 genomes of naked CMV-hEPO-SV40polyA linear DNA (hEPODNA) cut from the plasmid used for the rAd5-hEPO construction. Hematocrit (Hct) and serum hEPO levels were measured over a 28week period after vector administration. Two days prior to administration, mean Hct levels were 59.8±2% (mean ± SD) and hEPO levels were undetectable in all groups. Following vector administration, mean Hct and hEPO levels in the rAd5-hEPO-treated group were maximal on week 2 (75±3% and 28±2 mU/ml, respectively) and returned to pre-administration values (60.5±3% and zero, respectively) by week 4 in all but one animal. hEPO levels in this particular animal remained >6mU/ml up until its death at week 10 without any effect on Hct levels (~61%). hEPO levels in the group infected with 2.5x109 particles of rAAV2-hEPO increased over the first 4 weeks, remained stable (~24±36 mU/ml) until week 26 and began to decline thereafter (week 28 ; 13±24 mU/ml). Mean Hct levels in this group reached 75±13% by week 4, further increased to 81±14% at week 10, and remained stable until the end of the experiment. We measured hEPO in saliva at week 5, and then values were minimal (0.6±1 mU/ml), indicating preferential secretion of hEPO into the bloodstream. hEPO and Hct levels were lower for animals in the group infected with 108 particles of rAAV2-hEPO. In the mice treated with linear hEPO-DNA, Hct levels remained normal and hEPO levels were undetectable throughout the experiment. Our results indicate that stable and long-term production of therapeutic levels of EPO in the serum can be achieved by administering low doses of rAAV2 vector to the salivary glands.
475. Delivery of α1-Antitrypsin to the Lung Via Adeno-Associated Virus Vector-Mediated Gene Transfer Is More Efficient by Intrapleural Delivery Rather Than Skeletal Muscle Administration Bishnu P. De,1 Robert Merritt,1 Michael Lam,1 Neil R. Hackett,1 Ronald G. Crystal.1 1 Weill Medical College of Cornell University, New York, NY, United States. α1-antitrypsin (α1AT), a serine proteinase inhibitor synthesized and secreted by the liver, protects the lung from degradation by neutrophil proteases. α1AT deficiency is a common autosomal recessive disorder associated with the accelerated development of emphysema if serum α1AT levels are ≤11 μM (570 μg/ml). Several studies have indicated that adeno-associated virus (AAV)-mediated delivery of α1AT is a promising candidate for a gene therapy for this disease. Because the change in scale from mice to adult humans will require higher efficiency of gene expression, we hypothesized that intrapleural administration of an AAV vector expressing α1AT may achieve higher levels of α1AT in the lung compared to administration of the vector at a distant organ. To evaluate this hypothesis, a serotype 2 AAV vector expressing the α1AT cDNA under the control of cytomegalovirus-chicken β-actin hybrid promoter (AAV2α1AT) was administered to C57Bl/6 mice by various routes. A typical batch of the purified vector had <1% empty capsids as judged by electron microscopy, and a ratio of physical particles (ELISA) to infectious units of 150 ± 50. The AAV2α1AT vector was administered at increasing doses up to 5x1011 particle units (n=5 mice/data point), via three different routes- intrapleural, intramuscular, and intravenous. Serum α1AT levels were measured by ELISA at various times post-injection. At 2 wk, α1AT was detected in serum of intrapleurally administered animals whereas the level was barely detectable in animals with other routes of delivery. At 6 wk, the level of α1AT in intramuscularly injected animals was 1.7 ± 0.5 μg/ml (mean ± SD), and in intravenously injected animals was 7.3 ± 0.7 μg/ml, whereas intrapleurally injected S186
animals had levels of 10.0 ± 3.5 μg/ml (p<0.01). To determine the relative distribution of the vector in lung following intrapleural administration, an AAV vector expressing luciferase was injected intrapleurally in C57Bl/6 mice (n=5) and after 6 wk, luciferase activity was measured in tissue homogenates. The luciferase activity in the lung was 3.7x104 RLU/mg protein whereas in the liver and skeletal muscle were 4.0x104 RLU/mg and 9.0x104 RLU/mg, respectively, indicating that lung is the major site of gene transfer. Consistent with this concept, α1AT levels in bronchoalveolar lavage fluid after 8 wk showed 30 ± 8 ng/mg protein following intramuscular administration compared to 120 ± 10 ng/mg protein following intrapleural administration (p < 0.001). These data indicate that intrapleural administration may be a more efficient strategy for delivery of α1AT to the lung than by other routes. In the context that the pleura is an easy site for administration, with its large surface area, easy infectibility of the mesothelium, and close proximity to the lung, these observations suggest that intrapleural administration may be an attractive site for gene therapy for α1AT deficiency in humans. Dr. Crystal has equity in, is a consultant to, and receives sponsored research funds from, GenVec, Inc., Gaithersburg, Maryland, a publicly-traded biotechnology company.
476. Pseudotyped Adeno-Associated Virus Gene Transfer in Wound Healing Sundeep G. Keswani,1 Anna B. Katz,1 Foong-Yen Lim,1 Philip Zoltick,1 Gary P. Kobinger,2 James M. Wilson,2 Daniel J. Weiner,3 Timothy M. Crombleholme.1,2 1 Division of General, Thoracic, and Fetal Surgery, The Children’s Hospital of Philadelphia, Philadelphia, PA; 2Department of Medical Genetics, University of Pennsylvania School of Medicine, Philadelphia, PA; 3Division of Pulmonary Medicine, The Children’s Hospital of Philadelphia, Philadelphia, PA. Sustained transgene expression and cell specific gene transfer are goals of cutaneous gene therapy for tissue repair and regeneration. Adeno-associated virus serotype 2 (AAV2) mediated gene transfer to the skin produces stable transgene expression limited to the parniculus carnosus (PC) in rodents and the epidermis in porcine skin. This restricted distribution limits the clinical applicability of AAV2 vectors in human chronic wounds that frequently lack an epidermis and do not have a PC. The pseudotyping of AAV vectors using cap genes of recently identified AAV serotypes has been shown to alter transduction efficiency and tropism profiles in non cutaneous applications. This led us to hypothesize that pseudotyped AAV vectors may have an enhanced tropism and transduction efficiency in cells specific to cutaneous wounds. AAV2 genomes were packaged in capsids of AAV serotypes (AAV5, AAV7, AAV8), producing corresponding pseudotyped AAV vectors 2/5, 2/7 and 2/8. Each vector has an E. coli β-galactosidase gene inserted as a reporter, and was driven by a CMV promoter. 1x1011 genome copies of each AAV vector (AAV2, 2/5, 2/7, 2/8) were injected on to 8 mm full thickness skin wounds in C57 mice and labeled with india ink (n=3/ each vector). Wounds harvested at 4 weeks post transduction were processed for transgene expression. Histologic analysis was performed and transduction efficiencies were determined for gene transfer to the epidermis, the wound matrix and the PC. Data for epidermis and PC expressed as mean percentage of positive cells ± SEM. Data for wound matrix expressed as positive cells per mm² ± SEM. P values determined by ANOVA. At 28 days there was no evidence of vector toxicity and all wounds were in the remodeling phase of wound healing. AAV2/7 (83.6%±10.6, p<.001) and 2/8 (91.2 %±4.4, p<.0001) produced significantly greater gene transfer to the PC than AAV2 (33.1%±2.1). AAV2/5 resulted in limited gene transfer (3.6%±1.3) to the PC. The few cells that were transduced by AAV2/5 consisted of newly formed muscle Molecular Therapy Vol. 7, No. 5, May 2003, Part 2 of 2 Parts
Copyright © The American Society of Gene Therapy
APPLICATIONS IN DISEASE fibers at the regenerating edge of the PC. AAV2/5(80.3%±11.6) followed by AAV2/8 (59.5%±30.3) produced highly efficient gene transfer to the epidermis and both were significantly (p<.05) greater than AAV2 (11.7%±2.9) and AAV2/7 (5.5%±5.4). AAV2/5 expression was in both basilar and suprabasilar keratinocytes, while AAV 2/8 transduction was mainly in suprabasilar keratinocytes. AAV2/5 (3686±424cells/mm²) was significantly more efficient in transducing cells in the remodeling wound matrix than, AAV2 (553±94cells/mm², p<.0001), AAV2/7 (520±128cells/mm², p<.0001), and AAV2/8 (1440±276cells/mm², p<.0001). The change in capsid of the different pseudotyped AAV vectors produces distinct tropism and efficiency profiles in the wound healing model. AAV2/5 and AAV 2/8 produced highly efficient gene transfer compared to AAV2. AAV2/5 was the only vector to efficiently transduce proliferating cells of the epidermis, regenerating myocytes and the spectrum of cells of the wound matrix, suggesting that 2/5 may be the best pseudotyped AAV vector for gene therapy applications in wound repair and regeneration.
administration group. Conversely, we found significant elevations in Hct and serum Epo levels in mice previously treated with saline. We conclude that after a single administration of rAAV2 to murine SGs there is no significant acute immune response. However, rAAV2 administration to murine SGs results in both cellular and humoral immune responses. The latter may interfere with the efficacy of rAAV2 vectors after readministration.
477. Immune Responses Following Salivary Gland Administration of Recombinant Adenoassociated Virus Serotype 2 Vectors
AAV offers multiple advantages over alternative viral vectors, including minimal induction of secondary immune responses, broad tissue tropism and prolonged transgene expression. However, inherent disadvantages have limited the application of AAV. Choice of transgene is limited by a packaging constraint of ~5kb of DNA. It remains costly and technically demanding to produce large quantities of AAV at high titers. Finally, transgene expression is delayed compared to other commonly used viral vectors, a factor that can complicate experiments in vivo. Recently, reports of selfcomplementary (SC) AAV utilizing modified AAV plasmid constructs that permit formation and packaging of double-stranded recombinant AAV genomes which limit the need for second-strand synthesis, have demonstrated more rapid transgene expression and enhanced transduction efficiency. Here we describe the production and function of a SC-AAV encoding the glucose regulated insulin transgene. The SC-AAV backbone was created by removing the Rep Binding Site sequence from one of the viral ITRs and further incorporating a previously described (GlRE)3BP1.INS glucose sensitive insulin transgene or a CMV.EGFP. The population of viruses produced from such constructs are heterogeneous containing ~30% doublelength, double-stranded genomes of the SC-AAV. The remaining viruses contain one or two copies of a single length genome. Standard (SD)-AAV containing the identical but only single stranded sequences were used as controls. To compare requirements for stimulators of second strand synthesis HepG2 cells were transduced with or without helper adenovirus (Ad dl312, MOI - 10), and equal amounts either SC-AAV (GlRE)3BP1.INS or SD-AAV (GlRE)3BP1.INS. Secreted insulin in conditioned medium was detectable following SD-AAV (GlRE)3BP-1INS infection after 4 days. As expected, coinfection with Ad advanced insulin production to 2 days. Most importantly, however, SC-AAV (GlRE)3BP1.INS infection in absence of Ad. produced insulin levels that were greater than SDAAV (GlRE)3BP-1INS infection at the both 2 and 4 days. On day 4 insulin secretion from cultures infected with the self-complementary virus, SC-AAV (GlRE)3BP.1INS, was 7-fold greater than in those infected with the standard AAV. Similar results were obtained with self-complementary viruses expressing EGFP: photomicrographs show earlier and greater GFP expression with SC-AAV CMV.EGFP compared to SD-AAV CMV.EGFP infected cells. Preliminary data indicate the same phenomenon in primary cultures of hepatocytes. In sum, these data indicate that transduction with SC-AAV causes more rapid and enhanced transgene expression compared to that of SD-AAV, and underscore potential benefits of SC-AAV in gene transfer in vivo.
Marc R. Kok,1,2 Antonis Voutetakis,1 Seiichi Yamano,1 Jianghua Wang,1 Hisako Katano,1 Ioannis Bossis,1 Sandra Afione,1 Micheal Schmidt,1 John A. Chiorini,1 Paul-Peter Tak,2 Bruce J. Baum.1 1 Gene Therapy and Therapeutics Branch, NIH, Bethesda, MD; 2 Division of Clinical Immunology and Rheumatology, AMC, University of Amsterdam, Amsterdam, Netherlands. Salivary glands (SGs) provide a novel target site for several potentially useful clinical gene transfer applications. The SGs are capable of producing large amounts of proteins, and are a site where gene transfer can be readily accomplished in a minimally invasive manner (intraductal cannulation). Human SGs are well encapsulated, a circumstance likely to minimize the undesirable access of administered vectors and transgenes to other tissues (Baum et al, Int Rev Cytol, 2002). Numerous studies have shown that recombinant adenoassociated virus serotype 2 (rAAV2) vectors are useful for long term gene transfer applications to many tissues, including SGs (e.g., Yamano et al, Hum Gene Ther, 2002). Previous studies have indicated that intravenous, intramuscular and intranasal administration of rAAV2 vectors induces host immune responses. SGs are part of the mucosal immune system, and as yet there are no reported studies on the effects of administration of rAAV2 vectors into SGs on immune responsiveness. To examine this issue, the main excretory ducts of the submandibular glands of Balb/c mice were cannulated and vector administered by retrograde infusion (Baum et al, ibid). On day zero, we delivered a rAAV2 encoding βgalactosidase (rAAV2LacZ; 2.5x109 particles/gland) or saline (control) to adult mice (n=15/group). On day 28, we administered a second rAAV2 vector encoding human erythropoeitin (Epo; rAAV2Epo; 2.5x109 particles/gland) to 5 mice of each group. Immune activities were evaluated in saliva, serum, SGs and spleens collected on days 1, 28, and 56. There were no differences in salivary flow rates at all time points among the control, vector, and vector readministered groups. Histological examination of SGs on day 1 did not indicate any significant mononuclear cell infiltration in any of the treatment groups. Ex-vivo stimulation, with rAAV2, of splenocytes collected from mice sacrificed on days 28 and 56, resulted in elevated interferon γ levels in culture media from cells of mice administered rAAV2 in vivo but not from cells of control mice. In addition, significant titers of neutralizing antibodies to rAAV2 were detected in serum (from 1:200 - 1:6400). Also, we were unable to observe transduction of SG cells by rAAV2Epo in mice previously infected with rAAV2LacZ. No elevations in hematocrit (Hct) and serum Epo levels were seen at day 56 in mice from the virus reMolecular Therapy Vol. 7, No. 5, May 2003, Part 2 of 2 Parts Copyright © The American Society of Gene Therapy
478. Self-Complementary AAV (SC-AAV) Delivery Accelerates and Enhances Transgenic Insulin Production in Hepatocytes Miroslaw Kozlowski,1,3 Dorota Lyszkowicz,1 Sara A. Paveglio,2 Adam G. Campbell,2 Darrin E. Olson,2 Janet Rubin,1,2 Peter M. Thule.1,2 1 Atlanta VA Medical Center, Atlanta, GA; 2Div. of Endocrinology and Metabolism, EMORY University School of Medicine, Atlanta, GA; 3Dep. of Orthopedics, EMORY University School of Medicine, Atlanta, GA.
S187
475. Delivery of α1-Antitrypsin to the Lung Via Adeno-Associated Virus Vector-Mediated Gene Transfer Is More Efficient by Intrapleural Delivery Rather Than Skeletal Muscle Administration Bishnu P. De,1 Robert Merritt,1 Michael Lam,1 Neil R. Hackett,1 Ronald G. Crystal.1 1 Weill Medical College of Cornell University, New York, NY, United States. α1-antitrypsin (α1AT), a serine proteinase inhibitor synthesized and secreted by the liver, protects the lung from degradation by neutrophil proteases. α1AT deficiency is a common autosomal recessive disorder associated with the accelerated development of emphysema if serum α1AT levels are ≤11 μM (570 μg/ml). Several studies have indicated that adeno-associated virus (AAV)-mediated delivery of α1AT is a promising candidate for a gene therapy for this disease. Because the change in scale from mice to adult humans will require higher efficiency of gene expression, we hypothesized that intrapleural administration of an AAV vector expressing α1AT may achieve higher levels of α1AT in the lung compared to administration of the vector at a distant organ. To evaluate this hypothesis, a serotype 2 AAV vector expressing the α1AT cDNA under the control of cytomegalovirus-chicken β-actin hybrid promoter (AAV2α1AT) was administered to C57Bl/6 mice by various routes. A typical batch of the purified vector had <1% empty capsids as judged by electron microscopy, and a ratio of physical particles (ELISA) to infectious units of 150 ± 50. The AAV2α1AT vector was administered at increasing doses up to 5x1011 particle units (n=5 mice/data point), via three different routes- intrapleural, intramuscular, and intravenous. Serum α1AT levels were measured by ELISA at various times post-injection. At 2 wk, α1AT was detected in serum of intrapleurally administered animals whereas the level was barely detectable in animals with other routes of delivery. At 6 wk, the level of α1AT in intramuscularly injected animals was 1.7 ± 0.5 μg/ml (mean ± SD), and in intravenously injected animals was 7.3 ± 0.7 μg/ml, whereas intrapleurally injected S186
animals had levels of 10.0 ± 3.5 μg/ml (p<0.01). To determine the relative distribution of the vector in lung following intrapleural administration, an AAV vector expressing luciferase was injected intrapleurally in C57Bl/6 mice (n=5) and after 6 wk, luciferase activity was measured in tissue homogenates. The luciferase activity in the lung was 3.7x104 RLU/mg protein whereas in the liver and skeletal muscle were 4.0x104 RLU/mg and 9.0x104 RLU/mg, respectively, indicating that lung is the major site of gene transfer. Consistent with this concept, α1AT levels in bronchoalveolar lavage fluid after 8 wk showed 30 ± 8 ng/mg protein following intramuscular administration compared to 120 ± 10 ng/mg protein following intrapleural administration (p < 0.001). These data indicate that intrapleural administration may be a more efficient strategy for delivery of α1AT to the lung than by other routes. In the context that the pleura is an easy site for administration, with its large surface area, easy infectibility of the mesothelium, and close proximity to the lung, these observations suggest that intrapleural administration may be an attractive site for gene therapy for α1AT deficiency in humans. Dr. Crystal has equity in, is a consultant to, and receives sponsored research funds from, GenVec, Inc., Gaithersburg, Maryland, a publicly-traded biotechnology company.
476. Pseudotyped Adeno-Associated Virus Gene Transfer in Wound Healing Sundeep G. Keswani,1 Anna B. Katz,1 Foong-Yen Lim,1 Philip Zoltick,1 Gary P. Kobinger,2 James M. Wilson,2 Daniel J. Weiner,3 Timothy M. Crombleholme.1,2 1 Division of General, Thoracic, and Fetal Surgery, The Children’s Hospital of Philadelphia, Philadelphia, PA; 2Department of Medical Genetics, University of Pennsylvania School of Medicine, Philadelphia, PA; 3Division of Pulmonary Medicine, The Children’s Hospital of Philadelphia, Philadelphia, PA. Sustained transgene expression and cell specific gene transfer are goals of cutaneous gene therapy for tissue repair and regeneration. Adeno-associated virus serotype 2 (AAV2) mediated gene transfer to the skin produces stable transgene expression limited to the parniculus carnosus (PC) in rodents and the epidermis in porcine skin. This restricted distribution limits the clinical applicability of AAV2 vectors in human chronic wounds that frequently lack an epidermis and do not have a PC. The pseudotyping of AAV vectors using cap genes of recently identified AAV serotypes has been shown to alter transduction efficiency and tropism profiles in non cutaneous applications. This led us to hypothesize that pseudotyped AAV vectors may have an enhanced tropism and transduction efficiency in cells specific to cutaneous wounds. AAV2 genomes were packaged in capsids of AAV serotypes (AAV5, AAV7, AAV8), producing corresponding pseudotyped AAV vectors 2/5, 2/7 and 2/8. Each vector has an E. coli β-galactosidase gene inserted as a reporter, and was driven by a CMV promoter. 1x1011 genome copies of each AAV vector (AAV2, 2/5, 2/7, 2/8) were injected on to 8 mm full thickness skin wounds in C57 mice and labeled with india ink (n=3/ each vector). Wounds harvested at 4 weeks post transduction were processed for transgene expression. Histologic analysis was performed and transduction efficiencies were determined for gene transfer to the epidermis, the wound matrix and the PC. Data for epidermis and PC expressed as mean percentage of positive cells ± SEM. Data for wound matrix expressed as positive cells per mm² ± SEM. P values determined by ANOVA. At 28 days there was no evidence of vector toxicity and all wounds were in the remodeling phase of wound healing. AAV2/7 (83.6%±10.6, p<.001) and 2/8 (91.2 %±4.4, p<.0001) produced significantly greater gene transfer to the PC than AAV2 (33.1%±2.1). AAV2/5 resulted in limited gene transfer (3.6%±1.3) to the PC. The few cells that were transduced by AAV2/5 consisted of newly formed muscle Molecular Therapy Vol. 7, No. 5, May 2003, Part 2 of 2 Parts
Copyright © The American Society of Gene Therapy
APPLICATIONS IN DISEASE fibers at the regenerating edge of the PC. AAV2/5(80.3%±11.6) followed by AAV2/8 (59.5%±30.3) produced highly efficient gene transfer to the epidermis and both were significantly (p<.05) greater than AAV2 (11.7%±2.9) and AAV2/7 (5.5%±5.4). AAV2/5 expression was in both basilar and suprabasilar keratinocytes, while AAV 2/8 transduction was mainly in suprabasilar keratinocytes. AAV2/5 (3686±424cells/mm²) was significantly more efficient in transducing cells in the remodeling wound matrix than, AAV2 (553±94cells/mm², p<.0001), AAV2/7 (520±128cells/mm², p<.0001), and AAV2/8 (1440±276cells/mm², p<.0001). The change in capsid of the different pseudotyped AAV vectors produces distinct tropism and efficiency profiles in the wound healing model. AAV2/5 and AAV 2/8 produced highly efficient gene transfer compared to AAV2. AAV2/5 was the only vector to efficiently transduce proliferating cells of the epidermis, regenerating myocytes and the spectrum of cells of the wound matrix, suggesting that 2/5 may be the best pseudotyped AAV vector for gene therapy applications in wound repair and regeneration.
administration group. Conversely, we found significant elevations in Hct and serum Epo levels in mice previously treated with saline. We conclude that after a single administration of rAAV2 to murine SGs there is no significant acute immune response. However, rAAV2 administration to murine SGs results in both cellular and humoral immune responses. The latter may interfere with the efficacy of rAAV2 vectors after readministration.
477. Immune Responses Following Salivary Gland Administration of Recombinant Adenoassociated Virus Serotype 2 Vectors
AAV offers multiple advantages over alternative viral vectors, including minimal induction of secondary immune responses, broad tissue tropism and prolonged transgene expression. However, inherent disadvantages have limited the application of AAV. Choice of transgene is limited by a packaging constraint of ~5kb of DNA. It remains costly and technically demanding to produce large quantities of AAV at high titers. Finally, transgene expression is delayed compared to other commonly used viral vectors, a factor that can complicate experiments in vivo. Recently, reports of selfcomplementary (SC) AAV utilizing modified AAV plasmid constructs that permit formation and packaging of double-stranded recombinant AAV genomes which limit the need for second-strand synthesis, have demonstrated more rapid transgene expression and enhanced transduction efficiency. Here we describe the production and function of a SC-AAV encoding the glucose regulated insulin transgene. The SC-AAV backbone was created by removing the Rep Binding Site sequence from one of the viral ITRs and further incorporating a previously described (GlRE)3BP1.INS glucose sensitive insulin transgene or a CMV.EGFP. The population of viruses produced from such constructs are heterogeneous containing ~30% doublelength, double-stranded genomes of the SC-AAV. The remaining viruses contain one or two copies of a single length genome. Standard (SD)-AAV containing the identical but only single stranded sequences were used as controls. To compare requirements for stimulators of second strand synthesis HepG2 cells were transduced with or without helper adenovirus (Ad dl312, MOI - 10), and equal amounts either SC-AAV (GlRE)3BP1.INS or SD-AAV (GlRE)3BP1.INS. Secreted insulin in conditioned medium was detectable following SD-AAV (GlRE)3BP-1INS infection after 4 days. As expected, coinfection with Ad advanced insulin production to 2 days. Most importantly, however, SC-AAV (GlRE)3BP1.INS infection in absence of Ad. produced insulin levels that were greater than SDAAV (GlRE)3BP-1INS infection at the both 2 and 4 days. On day 4 insulin secretion from cultures infected with the self-complementary virus, SC-AAV (GlRE)3BP.1INS, was 7-fold greater than in those infected with the standard AAV. Similar results were obtained with self-complementary viruses expressing EGFP: photomicrographs show earlier and greater GFP expression with SC-AAV CMV.EGFP compared to SD-AAV CMV.EGFP infected cells. Preliminary data indicate the same phenomenon in primary cultures of hepatocytes. In sum, these data indicate that transduction with SC-AAV causes more rapid and enhanced transgene expression compared to that of SD-AAV, and underscore potential benefits of SC-AAV in gene transfer in vivo.
Marc R. Kok,1,2 Antonis Voutetakis,1 Seiichi Yamano,1 Jianghua Wang,1 Hisako Katano,1 Ioannis Bossis,1 Sandra Afione,1 Micheal Schmidt,1 John A. Chiorini,1 Paul-Peter Tak,2 Bruce J. Baum.1 1 Gene Therapy and Therapeutics Branch, NIH, Bethesda, MD; 2 Division of Clinical Immunology and Rheumatology, AMC, University of Amsterdam, Amsterdam, Netherlands. Salivary glands (SGs) provide a novel target site for several potentially useful clinical gene transfer applications. The SGs are capable of producing large amounts of proteins, and are a site where gene transfer can be readily accomplished in a minimally invasive manner (intraductal cannulation). Human SGs are well encapsulated, a circumstance likely to minimize the undesirable access of administered vectors and transgenes to other tissues (Baum et al, Int Rev Cytol, 2002). Numerous studies have shown that recombinant adenoassociated virus serotype 2 (rAAV2) vectors are useful for long term gene transfer applications to many tissues, including SGs (e.g., Yamano et al, Hum Gene Ther, 2002). Previous studies have indicated that intravenous, intramuscular and intranasal administration of rAAV2 vectors induces host immune responses. SGs are part of the mucosal immune system, and as yet there are no reported studies on the effects of administration of rAAV2 vectors into SGs on immune responsiveness. To examine this issue, the main excretory ducts of the submandibular glands of Balb/c mice were cannulated and vector administered by retrograde infusion (Baum et al, ibid). On day zero, we delivered a rAAV2 encoding βgalactosidase (rAAV2LacZ; 2.5x109 particles/gland) or saline (control) to adult mice (n=15/group). On day 28, we administered a second rAAV2 vector encoding human erythropoeitin (Epo; rAAV2Epo; 2.5x109 particles/gland) to 5 mice of each group. Immune activities were evaluated in saliva, serum, SGs and spleens collected on days 1, 28, and 56. There were no differences in salivary flow rates at all time points among the control, vector, and vector readministered groups. Histological examination of SGs on day 1 did not indicate any significant mononuclear cell infiltration in any of the treatment groups. Ex-vivo stimulation, with rAAV2, of splenocytes collected from mice sacrificed on days 28 and 56, resulted in elevated interferon γ levels in culture media from cells of mice administered rAAV2 in vivo but not from cells of control mice. In addition, significant titers of neutralizing antibodies to rAAV2 were detected in serum (from 1:200 - 1:6400). Also, we were unable to observe transduction of SG cells by rAAV2Epo in mice previously infected with rAAV2LacZ. No elevations in hematocrit (Hct) and serum Epo levels were seen at day 56 in mice from the virus reMolecular Therapy Vol. 7, No. 5, May 2003, Part 2 of 2 Parts Copyright © The American Society of Gene Therapy
478. Self-Complementary AAV (SC-AAV) Delivery Accelerates and Enhances Transgenic Insulin Production in Hepatocytes Miroslaw Kozlowski,1,3 Dorota Lyszkowicz,1 Sara A. Paveglio,2 Adam G. Campbell,2 Darrin E. Olson,2 Janet Rubin,1,2 Peter M. Thule.1,2 1 Atlanta VA Medical Center, Atlanta, GA; 2Div. of Endocrinology and Metabolism, EMORY University School of Medicine, Atlanta, GA; 3Dep. of Orthopedics, EMORY University School of Medicine, Atlanta, GA.
S187