Lingfei Xu

CMV-β-Actin Promoter Directs Higher Expression from an Adeno-Associated Viral Vector in the Liver than the Cytomegalovirus or Elongation Factor 1α Promoter and Results in Therapeutic Levels of Human Factor X in Mice

Although AAV vectors show promise for hepatic gene therapy, the optimal transcriptional regulatory elements have not yet been identified. In this study, we show that an AAV vector with the CMV enhancer/chicken beta-actin promoter results in 9.5-fold higher expression after portal vein injection than an AAV vector with the EF1 alpha promoter, and 137-fold higher expression than an AAV vector with the CMV promoter/enhancer. Although induction of the acute-phase response with the administration of lipopolysaccharide (LPS) activated the CMV promoter/enhancer from the context of an adenoviral vector in a previous study, LPS resulted in only a modest induction of this promoter from an AAV vector in vivo. An AAV vector with the CMV-beta-actin promoter upstream of the coagulation protein human factor X (hFX) was injected intravenously into neonatal mice. This resulted in expression of hFX at 548 ng/ml (6.8% of normal) for up to 1.2 years, and 0.6 copies of AAV vector per diploid genome in the liver at the time of sacrifice. Neonatal intramuscular injection resulted in expression of hFX at 248 ng/ml (3.1% of normal), which derived from both liver and muscle. We conclude that neonatal gene therapy with an AAV vector with the CMV-beta-actin promoter might correct hemophilia due to hFX deficiency.

Therapeutic neonatal hepatic gene therapy in mucopolysaccharidosis VII dogs.

Dogs with mucopolysaccharidosis VII (MPS VII) were injected intravenously at 2–3 days of age with a retroviral vector (RV) expressing canine β-glucuronidase (cGUSB). Five animals received RV alone, and two dogs received hepatocyte growth factor (HGF) before RV in an attempt to increase transduction efficiency. Transduced hepatocytes expanded clonally during normal liver growth and secreted enzyme with mannose 6-phosphate. Serum GUSB activity was stable for up to 14 months at normal levels for the RV-treated dogs, and for 17 months at 67-fold normal for the HGF/RV-treated dog. GUSB activity in other organs was 1.5–60% of normal at 6 months for two RV-treated dogs, which was likely because of uptake of enzyme from blood by the mannose 6-phosphate receptor. The body weights of untreated MPS VII dogs are 50% of normal at 6 months. MPS VII dogs cannot walk or stand after 6 months, and progressively develop eye and heart disease. RV- and HGF/RV-treated MPS VII dogs achieved 87% and 84% of normal body weight, respectively. Treated animals could run at all times of evaluation for 6–17 months because of improvements in bone and joint abnormalities, and had little or no corneal clouding and no mitral valve thickening. Despite higher GUSB expression, the clinical improvements in the HGF/RV-treated dog were similar to those in the RV-treated animals. This is the first successful application of gene therapy in preventing the clinical manifestations of a lysosomal storage disease in a large animal.

Absence of a desmopressin response after therapeutic expression of factor VIII in hemophilia A dogs with liver-directed neonatal gene therapy

Hemophilia A (HA) is a bleeding disorder caused by factor VIII (FVIII) deficiency. FVIII replacement therapy can reduce bleeding but is expensive, inconvenient, and complicated by development of antibodies that inhibit FVIII activity in 30% of patients. Neonatal hepatic gene therapy could result in continuous secretion of FVIII into blood and might reduce immunological responses. Newborn HA mice and dogs that were injected i.v. with a retroviral vector (RV) expressing canine B domain-deleted FVIII (cFVIII) achieved plasma cFVIII activity that was 139 ± 22% and 116 ± 5% of values found in normal dogs, respectively, which was stable for 1.5 yr. Coagulation tests were normalized, no bleeding had occurred, and no inhibitors were detected. This is a demonstration of long-term fully therapeutic gene therapy for HA in a large animal model. Desmopressin (DDAVP; 1-deamino-[d-Arg8]vasopressin) is a drug that increases FVIII activity by inducing release of FVIII complexed with von Willebrand factor from endothelial cells. It has been unclear, however, if the FVIII is synthesized by endothelial cells or is taken up from blood. Because the plasma cFVIII in these RV-treated dogs derives primarily from transduced hepatocytes, they provided a unique opportunity to study the biology of the DDAVP response. Here we show that DDAVP did not increase plasma cFVIII levels in the RV-treated dogs, although von Willebrand factor was increased appropriately. This result suggests that the increase in FVIII in normal dogs after DDAVP is due to release of FVIII synthesized by endothelial cells.

Gene Therapy Ameliorates Cardiovascular Disease in Dogs With Mucopolysaccharidosis VII

Background—Mucopolysaccharidosis VII (MPS VII) is a lysosomal storage disease caused by deficient &bgr;-glucuronidase (GUSB) activity resulting in defective catabolism of glycosaminoglycans (GAGs). Cardiac disease is a major cause of death in MPS VII because of accumulation of GAGs in cardiovascular cells. Manifestations include cardiomyopathy, mitral and aortic valve thickening, and aortic root dilation and may cause death in the early months of life or may be compatible with a fairly normal lifespan. We previously reported that neonatal administration of a retroviral vector (RV) resulted in transduction of hepatocytes, which secreted GUSB into the blood and could be taken up by cells throughout the body. The goal of this study was to evaluate the effect on cardiac disease. Methods and Results—Six MPS VII dogs were treated intravenously with an RV-expressing canine GUSB. Echocardiographic parameters, cardiovascular lesions, and biochemical parameters of these dogs were compared with those of normal and untreated MPS VII dogs. Conclusions—RV-treated dogs were markedly improved compared with untreated MPS VII dogs. Most RV-treated MPS VII dogs had mild or moderate mitral regurgitation at 4 to 5 months after birth, which improved or disappeared when evaluated at 9 to 11 and at 24 months. Similarly, mitral valve thickening present early in some animals disappeared over time, whereas aortic dilation and aortic valve thickening were absent at all times. Both myocardium and aorta had significant levels of GUSB and reduction in GAGs.

High expression reduces an antibody response after neonatal gene therapy with B domain-deleted human factor VIII in mice.

Summary.  Background: Gene therapy could prevent bleeding in patients with hemophilia A, but might induce antibodies that block factor VIII (FVIII) function. Objectives: To test the efficacy of gene therapy in the newborn period for preventing a response to human FVIII (hFVIII) because of immaturity of the immune system. Methods: Varying doses of a retroviral vector (RV) expressing a B domain‐deleted hFVIII cDNA were injected i.v. into newborn hemophilia A C57BL/6 or normal C3H mice. Mice were evaluated for hFVIII expression, hemostasis, and development of anti‐hFVIII antibodies with inhibitory activity. Results and conclusions: Injection of a high RV dose [1010 transducing units (TU) kg−1] into newborn hemophilia A or C3H mice resulted in 61% and 13% of normal hFVIII antigen in plasma, respectively; most mice did not produce anti‐hFVIII antibodies, and hemophilia A mice did not bleed. Furthermore, most mice with >20 ng mL−1 of hFVIII in plasma (10% normal, 1 × 10−10 m) were tolerant to hFVIII, as an antibody response was markedly reduced after challenge with hFVIII with or without adjuvant. However, most RV‐treated animals with lower antigen levels developed antibodies before or after challenge. Thus, initiation of a gene therapy trial with low RV doses might increase inhibitor formation. Furthermore, frequent hFVIII infusions in newborns with hemophilia A might reduce inhibitor formation. Finally, difficulties in achieving tolerance after gene therapy for hemophilia A as compared to hemophilia B may relate to lower expression of FVIII than FIX, as high antigen levels are most effective at inducing tolerance.

Cutting Edge: Treatment of Complement Regulatory Protein Deficiency by Retroviral In Vivo Gene Therapy

Gene therapy is an attractive means to replace a deficient or defective protein. Using a murine retroviral vector, we provide an example of reconstituting a C regulator by neonatal in vivo gene transfer. The fusion gene containing the mouse C receptor 1-related gene/protein y (Crry) and a single chain Ab fragment with specificity for mouse glycophorin A was placed under transcriptional control of a liver-specific promoter. Shortly after birth, Crry KO mice were injected with the retroviral vectors. Protein expression progressively increased over the next 6–8 wk after which an equilibrium was established. Coating levels on RBCs were obtained that inhibited C activation similar to wild-type cells and remained constant for >1 year. Thus, gene therapy with targeted regulators represents a treatment option to provide a long-term and sustained protein supply for the site-specific blockade of undesirable complement activation.

247. Gene Therapy with a Retroviral Vector Expressing Canine |[beta]|-Glucuronidase to Juvenile (7 Week Old ) Dogs Improves the Biochemical Manifestations of MPS VII

Top of pageAbstract Mucopolysaccharidosis VII (MPS VII) is a lysosomal storage disease due to deficiency of |[beta]|-glucuronidase (GUSB). Manifestations include bone and joint disease, heart disease, and neurological dysfunction. The MPS VII dog has a missense mutation in GUSB (R166H) that results in clinical manifestations that resemble those in humans. We have previously demonstrated that neonatal gene therapy with 3|[times]|109 TU/kg of a retroviral vector (RV) expressing canine GUSB from the human |[alpha]|1-antitrypsin promoter (hAAT-cGUSB-WPRE) resulted in transduction of hepatocytes and 195+/-36 U/ml of GUSB in serum. This resulted in a marked improvement in cardiac, bone and joint, and other manifestations. However, most patients with MPS VII are not identified at birth, and it will be necessary to determine the effect of transfer into older animals. Seven week-old MPS VII dogs were injected IV with 1|[times]|1010 transducing units (TU)/kg of hAAT-cGUSB-WPRE. Some received a cumulative dose of hepatocyte growth factor (HGF) over 24 hours of 2.5 mg/kg, then were injected with RV at 24, 48, 72, and 96 hours after the first dose of HGF (HGF/RV). Others received RV once a day for 4 days without preceding HGF (RV alone). Dogs that received HGF/RV or RV alone at 7 weeks had transduced hepatocytes and achieved stable expression of GUSB in serum at 185+/-54 U/ml or 51+/-5 U/ml for up to 1 year, respectively, although differences were not statistically significant. The serum GUSB achieved per TU/kg given was 28% and 8%, respectively, of that achieved after neonatal gene transfer, which likely reflects lower levels of hepatocyte replication in juveniles. Some animals were sacrificed at 6 months after gene transfer, and organs analyzed for biochemical correction of disease. The results in both juvenile transfer groups were similar and were pooled for statistical analyses. MPS VII results in elevation of the secondary lysosomal enzyme |[beta]|-hexosaminodase (|[beta]|-hex) and glycosaminoglycan (GAG) levels, and normalization of these occurs with successful therapies. In liver, spleen, jejunum, and lung, GUSB activity was 1% to 15% of normal levels and there was a marked reduction in |[beta]|-hex and GAG levels. In thymus, pancreas, kidney, and muscle, GUSB activity was 0.25% to 1% of normal and there was a marked or partial reduction in |[beta]|-hex and GAG levels. However, brain GUSB activity was only 0.2% of normal and there was little reduction in |[beta]|-hex (GAG was not evaluable as untreated MPS VII dogs do not have elevated brain GAG levels). Despite these biochemical improvements, the mobility of the dogs was only slightly improved and all were unable to walk by 12 months or earlier. We conclude that gene therapy into juvenile dogs results in expression similar to that observed after neonatal gene transfer, and marked biochemical improvements in somatic organs. This is the first demonstration of using juvenile gene therapy to improve disease in a large animal model for LSD. However, there was little biochemical improvement in brain, and the bone disease remained severe.

602. Induction of Tolerance to Human Factor VIII after Neonatal Gene Transfer of a Retroviral Vector Is Dose-Dependent in Mice and Is Not Effective in Cats

Hemophilia is a bleeding disorder due to deficiency of Factor VIII (FVIII) or Factor IX (FIX). Approximately 30% and 5% of patients with hemophilia A and B develop antibodies that inhibit the coagulation function of the factor (inhibitors). Neonatal gene transfer induces tolerance to human FIX in mice, dogs, and cats. However, it was unclear if this approach induces tolerance for the more immunogenic human FVIII (hFVIII) protein. Indeed, some mice that received a VSV-G pseudotyped retroviral vector (RV) expressing hFVIII as newborns developed inhibitors. In this study, an amphotropic gamma RV expressing human B domain-deleted FVIII (hFVIII) from the human |[alpha]|1-antitrypsin promoter was used to define the level of hFVIII necessary to achieve tolerance after neonatal gene transfer in mice. Hemophilia A mice were injected IV with 1010 transducing units (TU)/kg of RV at 2 to 3 days after birth. Plasma hFVIII antigen levels were 204|[plusmn]|33% of normal, and activity by COATEST assay was 326% of normal. None of 8 mice developed anti-hFVIII antibodies. To evaluate the level of hFVIII necessary to induce tolerance, a dose response study was done in normal C3H mice, which produced anti-hFVIII antibodies of 13|[plusmn]|4 mg/ml and inhibitor titers of 170 Bethesda units (BU)/ml after gene transfer into adults. Neonatal C3H mice were injected IV with 1010 (high), 109 (medium) or 108 (low) TU/kg of RV. The high dose group achieved 278|[plusmn]|45% of normal hFVIII, and 0 of 8 mice developed anti-hFVIII antibodies. The medium dose group achieved 19|[plusmn]|6% of normal hFVIII, and 8 of 11 mice developed anti-hFVIII antibodies. Average antibody levels for the group were 0.09|[plusmn]|0.03 mg/ml and 10|[plusmn]|4 BU/ml. These data suggest that 2|[times]|10|[ndash]|9 M of hFVIII in plasma induces tolerance, but 2|[times]|10|[ndash]|10 M is not sufficient. Nevertheless, neonatal transfer of the medium dose of RV resulted in antibody levels that were only 0.7% and Bethesda titers that were only 6% of those achieved after transfer into adults. Animals will be followed long-term to determine if antibody levels decline over time, which can occur after immune tolerance induction. Mice treated with low dose RV had undetectable levels of hFVIII (<3% of normal), and 0 of 13 developed anti-hFVIII antibodies, which was probably due to insufficient levels of hFVIII to stimulate an immune response. Mice will be challenged with hFVIII protein to determine if they are truly tolerant. The immune response to hFVIII after neonatal gene transfer was also evaluated in normal cats, which appear to have a more mature immune system. Four neonatal cats were injected IV with 9|[times]|109 TU/kg of the RV expressing hFVIII at day 5 after birth. Plasma hFVIII levels were undetectable, but 2 out of 4 cats developed anti-hFVIII IgG antibodies at 0.2 mg/ml within 4 weeks after gene transfer. We conclude that a high level of hFVIII is necessary to achieve tolerance after neonatal gene transfer in mice, and neonatal gene transfer is not effective at inducing tolerance in cats.

856. Neonatal Liver-Directed Gene Therapy Markedly Reduces Bone Disease in MPS I

Mucopolysaccharidosis (MPS) is due to deficiencies in enzymes that degrade glycosaminoglycans (GAGs). MPS I is due to deficient α-L-iduronidase activity (IDUA) and results in the accumulation of heparin sulfate (HS) and dermatan sulfate (DS). MPS VII is due to deficient β-glucuronidase activity and results in the accumulation of chondroitin sulfate (CS) in addition to HS and DS. Both disorders result in dysostosis multiplex in patients, which is characterized by short and thick bones. MPS can be treated with liver-directed gene therapy in which hepatocytes continuously secrete enzyme with mannose 6-phosphate, which can be taken up from blood. Although hepatic gene therapy reduced many manifestations of MPS VII, bone disease has not been completely corrected. The purpose of this study was to evaluate the effect of neonatal retroviral vector (RV)-mediated gene therapy on bone disease in MPS I mice. In contrast to untreated MPS VII mice in which long bones have only 81% to 87% of normal length, untreated MPS I mice had normal bone lengths at 6 weeks and 8 months after birth. Normal lengths were likely due to the fact that untreated MPS I mice had much less lysosomal storage material than did MPS VII mice in cartilage cells of the growth plate, which plays a pivotal role in bone elongation. This is likely due to the fact that CS is the major GAG in chondrocytes, and CS only accumulates in MPS VII. For untreated MPS I mice, bone mineral density (BMD) was normal at 6 weeks, but markedly increased at 0.069 gm/cm2 at 8 months [normal is 0.054 (p < 0.0001)]. The long bones of untreated MPS I mice had an increased diameter and were sclerotic, and there was substantial storage in osteocytes. The canine IDUA cDNA was cloned into a retroviral vector (RV) with a strong liver promoter, and the RV was injected into newborn MPS I mice at a high [1 × 109 transducing units (TU)/kg] or a low (1 × 108 TU/kg) dose. This resulted in transduction of liver cells and stable serum IDUA activity in most animals for 8 months at 1241 ± 147 and 110 ± 31 units (U)/ml for the high and low dose animals, respectively. At 8 months, mice that received a high dose of RV had normal long bone diameters without sclerosis, normal BMD [0.054 gm/cm2 (p < 0.0001 vs. values in untreated MPS I)], and marked reduction in lysosomal storage in the osteocytes. Mice that received a low dose of RV had some increase in diameter and sclerosis of the long bones, although they were not as severe as untreated MPS I mice. Similarly, the BMD (0.060 gm/cm2) was partially increased. We conclude that bone disease is much less severe in untreated MPS I mice than in untreated MPS VII mice. This is likely due to the fact that CS does not accumulate in MPS I, and CS is high in cartilage and accumulates in MPS VII. However, MPS I mice still had marked thickening and sclerosis of the bone, which were prevented with a high dose of RV at birth. However, the low dose of RV was less effective. These data help to define the serum activity necessary to prevent bone disease in MPS I.

162. Neonatal gene therapy induces tolerance in dogs but not cats

Immune responses after gene therapy could include an antibody response that blocks the activity of a blood protein, or a cytotoxic T lymphocyte (CTL) response that destroys transduced cells. Either could reduce the efficacy of gene therapy. We have previously shown that high dose neonatal gene therapy resulted in tolerance to canine Factor IX (cFIX), human FIX (hFIX), and canine β-glucuronidase in mice. However, since large animals may have a more mature immune system at birth, we evaluated immune responses after neonatal gene therapy in dogs and cats. Neonatal normal dogs that were transduced with a medium (3 × 109 transducing units (TU)/kg) or a low (8 × 107 TU/kg) dose of an RV expressing hFIX achieved stable expression of hFIX for over 6 months at 494 +/− 132 ng/ml (10% of normal) and 26 +/− 12 ng/ml (0.5% of normal), respectively. None of the neonatal RV-treated dogs developed anti-hFIX IgG. Further, the low dose group did not develop antibodies after infusion of 10 doses of hFIX (30 IU/kg/dose, given once per week starting at 2 months after birth), and thus were truly tolerant. Similar results were obtained in one hemophilia B dog that was transduced with 3 × 109 TU/kg of RV at birth and achieved 220 ng/ml of hFIX, as he did not develop anti-hFIX antibodies either before or after stimulation. In contrast, normal dogs that did not receive neonatal gene therapy developed high levels of anti-hFIX antibodies in response to an identical regimen of hFIX protein infusion. We conclude that neonatal gene therapy results in tolerance to hFIX in dogs. A similar gene therapy approach was tested in cats with mucopolysaccharidosis I (MPS I). Six MPS I cats received 1 × 109 TU/kg of an RV expressing canine α-L-iduronidase (IDUA) shortly after birth. All achieved detectable IDUA activity in blood within 2 weeks after gene transfer, with average serum levels of 25.2 units (U)/ml. However, serum activity declined to undetectable levels (<0.2 U/ml) by 2 months. This decline was not associated with anti-cIDUA antibodies. None of 5 cats analyzed at 2 months had detectable IDUA activity or RV DNA sequences in the liver, although these were high at 10 days in 1 cat (DNA copy number 0.25 copies/cell). Cats are capable of expressing this RV long-term, as MPS VI cats that received neonatal injection of an otherwise-identical vector expressing the feline N-acetylgalactosamine 4-sulfatase have maintained expression for 4 months. We infer that the cats most likely developed a CTL response to cIDUA after neonatal gene therapy. Interestingly, 3 of 18 MPS I mice that received 1 × 108 TU/kg of the cIDUA-expressing RV at birth probably developed a CTL response, although 0 of 25 MPS I mice that received 1 × 109 TU/kg did so. We conclude that cats develop a potent immune response to cIDUA but dogs do not develop an immune response to hFIX after neonatal gene therapy. It is possible that the intracellular cIDUA is more potent at inducing a CTL response than is the secreted hFIX. Alternatively, cats may have more mature immune systems at birth than do dogs. Further experiments will determine if cats produce immune responses to hFIX after neonatal gene therapy, and will further evaluate the CTL response to cIDUA in cats.