Adam G. Campbell

Adeno-associated viral delivery of a metabolically regulated insulin transgene to hepatocytes

Transduction with a liver specific, metabolically responsive insulin transgene produces near-normal blood sugars in STZ-diabetic rats. To overcome the limited duration of hepatic transgene expression induced by E1A-deleted adenoviral vectors, we evaluated recombinant adeno-associated virus (rAAV2) for cell type specificity and glucose responsiveness in vitro. Co-infection of AAV2 containing the glucose responsive, liver-specific (GlRE)(3)BP-1 promoter with an empty adenovirus enhanced transduction efficiency, and shortened the duration of transgene expression in HepG2 hepatoma cells, but not primary hepatocytes. However, in the context of rAAV2, (GlRE)(3)BP-1 promoter activity remained confined to cells of hepatocyte lineage, and retained glucose responsiveness. While isolated infection with an insulin expressing rAAV2 failed to attenuate blood sugars in diabetic mice, adenoviral co-administration with the same rAAV2 induced transient, near-normal random blood sugars in a diabetic animal. We conclude that rAAV2 can induce metabolically responsive insulin secretion from hepatocytes in vitro and in vivo. However, alternative AAV serotypes will likely be required to efficiently deliver therapeutic genes to the liver for the treatment of diabetes mellitus.

Hepatic Insulin Gene Therapy Normalizes Diurnal Fluctuation of Oxidative Metabolism in Diabetic BB/Wor Rats

Previous studies of hepatic insulin gene therapy (HIGT) focused on glycemic effects of insulin produced from hepatocytes. In this study, we extend the observations of glycemic control with metabolically regulated HIGT to include systemic responses and whole-body metabolism. An insulin transgene was administered with an adenoviral vector [Ad/(GlRE)(3)BP1-2xfur] to livers of BB/Wor rats made diabetic with polyinosinic polycytidilic acid (poly-I:C) (HIGT group), and results compared with nondiabetic controls (non-DM), and diabetic rats receiving different doses of continuous-release insulin implants (DM-low BG and DM-high BG). Blood glucose and growth normalized in HIGT, with lower systemic insulin levels, elevated glucagon, and increased heat production compared with non-DM. Minimal regulation of systemic insulin levels were observed with HIGT, yet the animals maintained normal switching from carbohydrate to lipid metabolism determined by respiratory quotients (RQs), and tolerated 24-hour fasts without severe hypoglycemia. HIGT did not restore serum lipids as we observed increased triglycerides (TGs) and increased free fatty acids, but reduced weight of visceral fat pads despite normal total body fat content and retroperitoneal fat depots. HIGT favorably affects blood glucose, normalizes metabolic switching in diabetic rats, and reduces intra-abdominal fat deposition.

Host Cells Reduce Glucose Uptake and Glycogen Deposition in Response to Hepatic Insulin Gene Therapy

Background Hepatic insulin gene therapy (HIGT) restores weight gain and near-normal glycemia in rodent models of insulin- deficient diabetes mellitus. However, the effect of transgenic insulin on endogenous genes and recipient cell function is relatively unexplored. To investigate hepatocellular effects of transgenic insulin expression, we evaluated intermediary glucose metabolism in primary cultured hepatocytes treated with HIGT. Methods Rat hepatocytes were transduced with adenovirus expressing a glucose-responsive human insulin transgene and cultured in high-glucose and high-insulin conditions. We determined glycogen content in cell cultures and intact liver directly. Glycogenolysis was compared using glucose production of cultured cells. Glucose uptake, oxidative, and glycolytic processing were determined by radiotracer analysis or direct end-product assessment. Quantitative real-time reverse transcriptase polymerase chain reaction was used to determine expression of glucose transporter 2 (GLUT2) and glucokinase genes. GLUT2 protein abundance was determined by Western blot analysis. Results HIGT-treated hepatocytes contained significantly less glycogen than either untreated hepatocytes or those treated with an empty virus. Glucose release owing to glycogenolysis remained normal. However, HIGT treatment significantly impaired glucose uptake and processing. Metabolic synthetic processes were not generally inhibited, as indicated by enhanced β-hydroxybutyrate secretion. While preserving cell viability, HIGT treatment diminished expression of both glucokinase and GLUT2. In HIGT-treated streptozocin-treated diabetic rats, total liver glycogen was intermediate between diabetic animals and normal controls. Conclusions These results suggest gene-specific effects in recipient hepatocytes following HIGT treatment and underscore the need for expanded studies examining host cell responses to the transfer of metabolically active transgenes.

Hepatic insulin gene therapy diminishes liver glycogen despite insulin responsive transcriptional effects in diabetic CD-1 mice.

Hepatic insulin gene therapy (HIGT) produces near‐normal glycemia in diabetic rats. Hepatic insulin production is expected to stimulate glycogen storage. However, the effect of HIGT on hepatic glycogen metabolism in vivo is unknown.

Long-term glycemic control with hepatic insulin gene therapy in streptozotocin-diabetic mice.

Insulin self‐administration is burdensome and can produce dangerous hypoglycemia. Insulin gene therapy may improve and simplify the treatment of diabetes mellitus. In rats, metabolically responsive hepatic insulin gene therapy (HIGT) delivered by adenovirus normalizes random blood sugars but with a limited duration. To prolong glycemic control, we delivered a metabolically regulated insulin transgene by adeno‐associated virus (AAV).

2 PEROXISOME PROLIFERATOR-ACTIVATED RECEPTOR GAMMA LIGAND, ROSIGLITAZONE, ATTENUATES VASCULAR OXIDATIVE STRESS IN A MOUSE MODEL OF TYPE 2 DIABETES.

Purpose We have previously shown that peroxisome proliferator-activated gamma (PPARg) ligands reduce superoxide anion (O2 2 ×) generation in vascular endothelial cells in vitro by suppressing expression of selected subunits of NADPH oxidase and by increasing the expression and activity of Cu/Zn superoxide dismutase (SOD). The current study was designed to determine if PPARg ligands modulate vascular endothelial O2 2 × generation in vivo through these same mechanisms. Methods Lean control (db +/db 2) and obese, leptin receptor-deficient (db 2 /db 2) mice were treated with either vehicle or rosiglitazone (3 mg/kg/day) by gavage for 7 days. Aortas were prepared for analysis of O2 2 × production using ESR spectroscopy and for RNA analysis, and serum was collected for analysis of metabolic parameters. Results Compared to db +/db 2 mice, obese, db 2 /db 2 mice had higher serum glucose, insulin, leptin, triglyceride, and fatty acid levels and lower adiponectin levels. Rosiglitazone had no effect on these metabolic derangements. Aortic O2 2 × generation measured with ESR spectroscopy was significantly increased in db 2 /db 2 mice. Aortic tissue from these mice also demonstrated higher relative mRNA levels of the NADPH oxidase subunits, Nox-1 and Nox-4, as measured by real-time PCR analysis and lower mRNA levels of PPARg. Rosiglitazone treatment decreased O2 2 × generation and mRNA levels of Nox homologues in db 2 /db 2 mice. Conclusions These data indicate that short-term treatment with the PPARg agonist rosiglitazone suppressed vascular NADPH oxidase expression and O2 2 × production in an animal model of vascular oxidative stress. Because these findings occurred in the absence of significant metabolic effects, these results indicate that rosiglitazone and other PPARg ligands may exhibit direct vascular protective effects.

177 COUNTERREGULATORY HORMONES IN RATS TREATED WITH HEPATIC INSULIN GENE THERAPY FOR DIABETES.

Hepatic insulin gene therapy (HIGT) using a metabolically regulated insulin transgene expressed in the liver is an effective way to supply insulin to diabetic rats to maintain growth and normalize random daily blood glucose. Previous results in rodent models of diabetes support the role of transcriptional regulation of insulin secretion as part but not all of the mechanism that allows normal random daily blood glucose, improved glucose tolerance compared to diabetic rats, and the ability to tolerate extended fasts with blood glucose similar to nondiabetic rats. Glucagon was previously found to be uniquely elevated in the BB/Wor model of spontaneous autoimmune diabetes. We further explored the roles of counterregulatory hormones in the mechanism of glycemic control with HIGT. Sprague-Dawley rats were made diabetic with streptozotocin and treated with an adenovirus carrying a metabolically regulated insulin transgene that is expressed in the liver (Ad/(GlRE)3BP1-2xFur, 2 3 1010 PFU/kg). Glucagon, growth hormone, and cortisol were measured in HIGT-treated diabetic rats (DM + HIGT), diabetic rats treated with subcutaneous insulin (DM), nondiabetic rats treated with HIGT (non-DM + HIGT), and nondiabetic control rats (non-DM). HIGT-treated rats were also given long-acting octreotide analogues to block the release of glucagon and growth hormone while monitoring blood glucose, insulin, and counterregulatory hormones. Glucagon was found to be consistently and uniquely elevated in DM + HIGT-treated diabetic rats. Insulin levels were low in HIGT-treated rats. Growth hormone was elevated in all diabetic rats and cortisol was elevated in all rats tested. Blocking glucagon secretion with octreotide moderately decreased blood glucose levels in HIGT-treated rats examined during 12 hours of random feeding and during a fast. Novel methods of insulin administration may have important interactions with endogenous systems of glycemic control. Unique elevations in glucagon appear to have moderate effects to enhance the safety and efficacy of HIGT in diabetic rats by allowing physiologic controls in DM + HIGT and by interaction with the transgene promoter. Further evaluation of hepatic glucose handling as a result of HIGT and elevated levels of glucagon are expected to identify the mechanism of glycemic regulation in diabetic rats treated with metabolically regulated insulin transgenes expressed in the liver.

893. Effect of hepatic insulin gene therapy on intrahepatic carbohydrate metabolism in vivo

Metabolically responsive insulin transgenes can be transduced into the livers of diabetic rodents and lead to normalization of randomly sampled blood glucose. Hepatocytes can only produce insulin through a constitutive pathway. Therefore, rapidly changing serum levels of insulin are not possible with this form of gene therapy for diabetes. However, blood glucose remains under better control than can be accounted for by the systemic levels of insulin produced by hepatocytes transduced by an adenoviral vector to express a metabolically sensitive transgene, Ad/(GlRE)3BP-1 2xFur. We hypothesize that intrahepatic carbohydrate metabolism is changed during hepatic insulin gene therapy to contribute to the mechanism of glucose regulation in rats treated with this form of gene therapy. We previously measured in vitro carbohydrate metabolism in primary hepatocytes transduced with Ad/(GlRE)3BP-1 2xFur showing decreased glycogen, decreased gluconeogenesis, and decreased glucose uptake. We now expand the observations to in vivo carbohydrate metabolism. Sprague-Dawley rats were made diabetic with streptozotocin then received Ad/(GlRE)3BP-1 2xFur (2 x 1010 PFU/kg), and compared to non-diabetic rats, diabetic controls, and diabetic rats treated with peripherally administered insulin. Blood glucose and carbohydrate-regulatory hormones were monitored. Glycogen was measured in a group of rats sacrificed during unrestrained feeding. Another group of rats underwent 24 hour fasts with administration of deuterium oxide near the end of the fast to measure pathways of hepatic glucose production by NMR spectroscopy to demonstrate different pathways of gluconeogenesis and glycogenolysis that are active depending on the sites of incorporation of deuterium onto carbon molecules of glucose produced in the liver. Blood glucose was similar in rats producing hepatic insulin and normal rats, with higher levels of glucagon and growth hormone in rats treated with gene therapy. Random levels of hepatic glycogen were highest in normal rats (34.2 + 6.4 mg/g of liver tissue), with less in rats producing insulin within the liver (26.0 + 9.0 mg/g), and lowest in diabetic rats (15.1 + 4.4 mg/g). Normal and gene therapy treated rats had similar blood glucose during the fast, with higher glucose in the diabetic controls. Plasma glucose was extracted and converted to mono-acetylated glucose for evaluation by NMR spectroscopy. The NMR spectra show different relative activity in pathways of hepatic glucose metabolism. Hepatic insulin gene therapy is associated with normalization of random and fasting blood glucose levels, but the mechanism of glucose regulation is different than normal animals due to changes in intrahepatic carbohydrate metabolism and counterregulatory hormones. Potential mechanisms are: (1) autocrine or paracrine activity of hepatic insulin, (2) unique counterregulatory hormone responses due to the anatomical site of hepatic insulin production, (3) changes in hepatic carbohydrate metabolism induced by intracellular insulin, and (4) changes in metabolic enzymes due to transcription factors used by the metabolically responsive transgene.

376. A Metabolically Responsive Insulin Transgene Delivered by AAV2/8 Pseudotyped, Self-Complementary AAV (SC-AAV2/8) Controls Glycemia in STZ-Diabetic Mice

Metabolically responsive insulin gene therapy promises advances in the treatment of type 1 diabetes mellitus. We have shown that hepatic expression of a transcriptionally regulated human insulin transgene successfully treats diabetes in rodents. However, consistent with first generation adenoviral delivery transgenic insulin production is transient in vivo. In contrast, transgenes delivered by recombinant adeno-associated virus (rAAV) express in murine liver for more than a year. Despite their advantages application of AAV has been limited. Transgene expression is delayed, a factor that can complicate experiments in vivo. Moreover, producing large quantities of AAV at high titers remains costly and technically demanding. Finally, common vectors infect some tissues poorly, or express weakly due to limited second-strand synthesis. Two recent developments in rAAV technology address these issues. A novel serotype, rAAV8 was shown to infect hepatocytes 10-100 fold more efficiently than rAAV2. Separately, self-complimentary (SC)-AAV particles produced by eliminating the terminal resolution sequence (TRS) from one inverted terminal repeat (ITR) were shown to accelerate and enhance transgene expression in vivo. Here we describe the combination of SC and pseudotyping strategies to obtain long-term glycemic control in diabetic mice. We created pSC-AAV2(GlRE)3BP-1 2fur by inserting a previously described glucose and insulin responsive insulin transgene into an AAV2 transfer plasmid from which the TRS had been removed from the left-ITR. Compared to vector derived from a longer, AAV2(GlRE)3BP-1 2furGFP, or similarly short sequence, AAV2(GlRE)3BP-1 2fur, infection with SC-AAV2(GlRE)3BP-1 2fur accelerated, and more than tripled secretion of human insulin from primary hepatocytes. Co-transfection of 293 cells with pSC-AAV2(GlRE)3BP-1 2fur, and a REP2/CAP8 expressing plasmid produced pseudotyped SC-AAV2/8 vector. Twenty-three male, STZ-diabetic CD-1 (BG>200 mg/dl) mice received a portal vein injection of either SC-AAV2(GlRE)3BP-1 2fur, or increasing doses of SC-AAV2/8(GlRE)3BP-1 2fur. Despite enhanced transgene expression in vitro, SC-AAV2(GlRE)3BP-1 2fur administration (3.4-4.8×1010vg, n=4) failed to reduce hyperglycemic by 18 days. In contrast, SC-AAV2/8(GlRE)3BP-1 2fur administration lowered blood glucose in 18 of 19 diabetic mice by 7 days, beginning with doses as low as 1.3×1010vg. Six of seven mice receiving the three highest doses (5.2×1010vg n=3, 7.8×1010vg n=1, 1.3×1011vg n=3) succumbed to hypoglycemia within 4 days. Two animals receiving 2.6×1010vg (n=6) died of hypoglycemia at day 34. However, 9 of the remaining 10 mice receiving 1.3-2.6×1010vg continue to grow normally and maintain near euglycemia (105±6 mg/dl). In conclusion, we have combined SC-AAV, AAV2/8 pseudotyping techniques, and metabolically responsive hepatic insulin gene therapy to treat STZ-diabetic mice in vivo. Effective viral doses may be 10-fold less than standard, non-pseudotyped vector delivering the same transgene. On-going investigations will determine the long-stability of glycemic control in this model.