RESEARCH ARTICLE Delivery of glucose-6-phosphatase in a canine model for glycogen storage disease, type Ia, with adenoassociated

Transcription

1 (2002) 9, Nature Publishing Group All rights reserved /02 $ RESEARCH ARTICLE Delivery of glucose-6-phosphatase in a canine model for glycogen storage disease, type Ia, with adenoassociated virus (AAV) vectors RM Beaty 1, M Jackson 2, D Peterson 1, A Bird 1, T Brown 2, DK Benjamin Jr 3, T Juopperi 2, P Kishnani 1, A Boney 1, YT Chen 1 and DD Koeberl 1 1 Division of Medical Genetics, Duke University Medical Center, Durham, NC, USA; 2 College of Veterinary Medicine, North Carolina State University, Raleigh, NC, USA; and 3 Division of Infectious Diseases, Department of Pediatrics, Duke University Medical Center, Durham, NC, USA Therapy in glycogen storage disease type Ia (GSD Ia), an inherited disorder of carbohydrate metabolism, relies on nutritional support that postpones but fails to prevent longterm complications of GSD Ia. In the canine model for GSD Ia, we evaluated the potential of intravenously delivered adeno-associated virus (AAV) vectors for gene therapy. In three affected canines, liver glycogen was reduced following hepatic expression of canine glucose-6-phosphatase (G6Pase). Two months after AAV vector administration, one affected dog had normalization of fasting glucose, cholesterol, triglycerides, and lactic acid. Concatamerized AAV vector DNA was confirmed by Southern blot analysis of liver DNA isolated from treated dogs, as head-to-tail, head-tohead, and tail-to-tail concatamers. Six weeks after vector administration, the level of vector DNA signal in each dog varied from one to five copies per cell, consistent with variation in the efficiency of transduction within the liver. AAV vector administration in the canine model for GSD Ia resulted in sustained G6Pase expression and improvement in liver histology and in biochemical parameters. (2002) 9, doi: /sj.gt Keywords: gene therapy; adeno-associated virus vectors; glycogen storage disease; disorders of metabolism; genetic disease; animal model Introduction AAV vectors have been shown to transduce liver and skeletal muscle, and to provide long-term expression of secreted proteins. AAV vectors preferentially transduce hepatocytes in liver following intravenous administration, 1 7 and transduce myofibers in skeletal muscle after local injection. 6,8 10 Indeed, a sustained, therapeutic response to factor IX expressed from an AAV vector was observed in hemophilic mice and dogs after liver-targeted delivery. 2,5 In the canine model for hemophilia B, therapeutically relevant levels of canine factor IX were produced with an AAV vector in liver 2 or muscle, 10 and a partial correction of hemophilia was observed for over 5 months. These preliminary results justified a phase I clinical trial for AAV vector-mediated gene therapy in hemophilia B, which showed evidence of factor IX expression and decreased bleeding following muscle-targeted vector administration in hemophiliacs. 11 GSD Ia is a classic inherited disorder of metabolism with life-threatening complications, first described by von Gierke in 1929 as hepato-nephromegalia glycogenica. 12 The cause of GSD Ia is deficiency of glucose-6- Correspondence: DD Koeberl, Division of Medical Genetics, Duke University Medical Center, Durham, NC, USA Received 25 September 2001; accepted 11 February 2002 phosphatase (G6Pase), a key regulatory enzyme in glucose production from glycogen. This defect prevents the conversion of glucose-6-phosphate to glucose, with secondary accumulation of glycogen and fat in tissues. 13 The metabolic consequences of G6Pase deficiency are hypoglycemia upon fasting and excess production of lactic acid from increased glycolysis. Patients with GSD Ia typically present with hypoglycemia and seizures at a few months of age, and develop long-term complications of hypertriglyceridemia, growth failure, renal dysfunction, gout, osteoporosis and hepatic adenomas that can degenerate to hepatocellular carcinoma. 14 Pulmonary hypertension and early-onset atherosclerosis have been reported Thus, GSD Ia is a severe, multisystem disorder that results in significant morbidity and mortality. Prior to the development of current nutritional treatment, many patients with GSD Ia died, and the prognosis was guarded in those who survived. 14 Long-term complications of treated GSD Ia include short stature, gout, osteoporosis, hepatic adenomas and renal dysfunction. 14,18 22 Current nutritional therapy for GSD Ia consists of sufficient carbohydrate supplementation, either by uncooked cornstarch or continuous glucose infusion, to prevent hypoglycemia. Uncooked cornstarch, dosed 1.5 to 2.5 gm/kg every 4 to 6 h, provides a slow-release form of glucose supplementation. However, treatment in infancy can require continuous nasogastric tube feeding of glucose due to starch intolerance. These treatments

2 1016 Gene therapy in canine GSD Ia with AAV vectors have improved the outcome of GSD Ia. However, neither treatment prevents hypoglycemia and lactic acidosis entirely, and nasogastric feeding has the accompanying risk of life-threatening equipment failure. 14 Biochemical monitoring of 17 children and young adults who were treated with nutritional therapy for GSD I demonstrated early morning hypoglycemia (24%) and elevated lactic acid (88%), despite compliance with cornstarch treatment. 18 Hence, current nutritional therapy for GSD Ia is inadequate to prevent immediate and long-term complications of this disorder. A Maltese canine model for glycogen storage disease type Ia has been described by our group, and its pathologic manifestations are equivalent to severe GSD Ia, in humans. The clinical presentation of GSD Ia in Maltese puppies was typical for untreated GSD Ia, including lethargy, coma, and early death. 23 Postmortem examination revealed hepatomegaly, nephromegaly, and hepatic glycogen accumulation. The diagnosis of GSD Ia was confirmed biochemically on the basis of deficient G6Pase in liver. 23 The causative mutation was a G to C transversion at bp 450 in the G6Pase cdna. 24 The missense mutation changed methionine to isoleucine at codon 121, a position that was conserved across 4 species, and decreased G6Pase activity over 10-fold in transfected COS-7 cells. 24 Initial biochemical and histopathological characterization of the canine model for GSD Ia showed a close resemblance to the human disorder, and further clinical evaluation in canine GSD Ia confirmed its similarity to human GSD Ia. 25 We report the administration of AAV vectors encoding G6Pase to three affected dogs early in life. Delivery of G6Pase by transduction of cells in liver with AAV vectors was a logical choice in GSD Ia, where liver expression would be therapeutic based on the biochemical effect of liver transplantation in GSD Ia patients. 14,26 Fasting glucose was measured on the day of vector administration, and every 2 weeks subsequently. Other biochemical parameters were monitored in parallel, including blood cholesterol, triglycerides and lactic acid. Liver G6Pase and glycogen content were analyzed, as was vector DNA in the liver. The results of this study demonstrated a beneficial effect of AAV vector administration on blood glucose and other biochemical abnormalities by liver-targeted G6Pase expression in canine GSD Ia. Results An AAV vector encoding canine G6Pase was administered intravenously to three affected puppies on day 3 or 4 of life (Table 1; Figure 1). Fasting glucose levels nor- Figure 1 AAV vectors encoding canine G6Pase. TR is the AAV2 terminal repeat sequence, which flanks the transgene in each vector plasmid. Albumin is the mouse albumin promoter/enhancer, cg6pase is the canine G6Pase cdna, the intron is the human growth hormone intron 4, and pa is the human growth hormone gene polyadenylation signal. malized for two puppies (V4 and P3) (Figure 2). Fasting studies for one of these animals (P3) also revealed correction of fasting cholesterol, triglycerides, and lactic acid after 8 weeks of age. The observation of normal values for these fasting blood measurements was dramatic, compared with untreated, affected puppies. Liver G6Pase and glycogen content were analyzed for affected puppies following AAV vector administration (Table 1), and compared with values for the untreated affected control (D3), as well as affected Maltese puppies previously reported (Figure 3a and b). 23 The liver G6Pase activity was significantly increased (P 0.05 by the two sample t test) and liver glycogen content was significantly reduced (P 0.02 by the two sample t test) for affected puppies following AAV vector administration, compared with untreated, affected puppies (Figure 3a and b). Normalized liver G6Pase was increased for affected puppies following AAV vector administration (P = 0.038) and for unaffected carriers (P = 0.043) compared with the untreated, affected control (Tables 2 and 3, Figure 3c). Vector DNA was analyzed in liver DNA following AAV vector administration by Southern blot analysis following BamHI digest of total liver DNA (Figure 4a). BamHI cuts once in the AAV vector, and the size of the detected fragment reflects the orientation of concatamerized AAV vector DNA (Figure 4b). At 11 weeks following vector administration (P3), the Southern blot reflected a high level of head-to head concatamers (5.0 kbp), while lower strength signals indicated the presence of head-totail and tail-to-tail concatamers (4.3 and 3.6 kbp, respectively, Figure 4a and b). The signals for vector DNA in two other affected livers at 5 weeks (V4) and 7 weeks (J5) Table 1 AAV vector administration to affected GSD Ia puppies Puppy/day of life vector Vector Vector particles Liver G6Pase a Liver glycogen content b administered V4/day 4 AAV-AlbG6PGH ± P2/day 3 AAV-AlbG6PGH( ) ± P3/day 3 AAV-AlbG6PGH( ) ± D3 None ± Normals (n = 8) c None ± ± 1.4 a G6Pase (glucose-6-phosphate/g liver per min), determined four (V4) or five times (P2, P3 and D3) independently. b Percent glycogen (g glycogen/100 g liver). c From Ref. 23.

3 Gene therapy in canine GSD Ia with AAV vectors 1017 Figure 2 Fasting blood levels following AAV vector administration. Blood sampling occurred 3 h following a standard meal. Three affected puppies received AAV vectors as detailed in Table 1 (P2, P3 and V4). Two affected puppies received no vector (D3 and G1). Three carriers served as normal controls (T2, T4 and P1). Normal values were provided by the North Carolina State University Veterinary Teaching Hospital, except for lactic acid which was collected for 16 normal and carrier dogs from 1 to 6 months old. reflected a high level of tail-to-tail concatamers (4.4 kbp fragment, Figure 4a and b). Uncut liver DNA showed higher molecular weight vector DNA in V4 and P3, consistent with concatamerization of AAV vector DNA. The signal for vector DNA was absent for the affected, untreated control (D3) in BamHI, EcoRI, and uncut DNA samples (Figure 4a and c). Vector DNA analysis unexpectedly showed intraindividual variation in the level of vector DNA within liver DNA (Figure 4c). The signal for vector DNA digested with EcoRI (1.7 kbp) was normalized to a genomic signal (4 kbp), and quantitated versus a standard curve. The normalized vector DNA quantitation ranged from the limit of detection (0.3 copies/cell) to approximately 5 copies/cell. Vector DNA was detectable in all liver DNA samples at 5 and 11 weeks following vector administration (V4 and P3). No vector DNA was detected in multiple samples for the untreated, affected control (D3). The increased double-stranded vector DNA at 5 and 11 weeks (V4 and P3), compared with 2 weeks (P2) was consistent with increased double-stranded AAV vector DNA at later times following AAV vector administration. 27 Histological evaluation of liver sections from each dog liver showed staining of glycogen and lipids. Periodic acid-schiff (PAS) staining revealed decreased liver glyco- Figure 3 Liver G6Pase levels and glycogen content following AAV vector administration. (a) G6Pase levels; and (b) glycogen content in canine liver. The average and standard deviation are shown, and reflect the levels in Table 1 and as previously reported. 23 Three affected puppies received AAV vectors (Table 1). Affected, untreated controls (n = 3) include two affected puppies previously reported, and normal controls were previously reported. 23 (c) The box and whisker plot shows the normalized G6Pase levels in liver from three groups of dogs. Group 1 consists of affected, untreated puppy (D3); group 2 consists of affected, treated, puppies (P2, P3 and V4); and group 3 consists of unaffected, carrier puppies (carrier 1 and 2). gen 11 weeks following AAV vector administration (Figure 5a) compared with an untreated, affected control liver (Figure 5b). Lipid accumulation was particularly prominent, as expected in the affected liver in GSD Ia (Figure 5c f). 14 The affected, untreated control had the highest degree of staining for lipid (Figure 5f), whereas gene therapy with AAV vectors decreased lipid accumulation at 2, 5 and 11 weeks following vector administration (Figure 5d f). The liver for an affected dog at 11 weeks following AAV vector administration had the greatest decrease in lipid accumulation compared with the affected, untreated control (Figure 5e versus f). PAS staining revealed a similar trend of decreased glycogen in affected canines following gene therapy (not shown). Discussion We have demonstrated a potential benefit of AAV vectormediated G6Pase delivery in the canine GSD Ia model by biochemical correction of fasting glucose, cholesterol, triglycerides and lactic acid. Decreased cholesterol and triglyceride levels in blood correlated with decreased glycogen and lipid accumulation by histological analysis in liver at 11 weeks after vector administration. Affected GSD Ia canines had increased G6Pase and decreased glycogen storage following liver-targeted G6Pase delivery with AAV vectors. AAV vector DNA was present in liver

4 1018 Table 2 Gene therapy in canine GSD Ia with AAV vectors Normalized liver G6Pase for three groups of canines Clustered regression of normalized G6Pase ( mol G6P/ g protein/min) P value Groups Estimate 95%LCL 95%UCL Untreated affected canine (D3) Treated affected canines (P2, P3, V4) a Unaffected carriers (n = 2) a a The untreated, affected canine is the referent group Table 3 Normalized liver G6Pase ( mol G6P/ g protein/min) Liver No. Mean (samples analyzed) D3 (n = 5) P2 (n = 4) V4 (n = 6) P3 (n = 7) Carrier Carrier Carriers 5 and Affected + AAV (P2, P3, V4) DNA at 1 to 5 copies/cell from 6 weeks following vector administration, and G6Pase expression persisted for the duration of the experiment. Hepatic delivery of murine G6Pase with an adenoviral vector has been reported in the G6Pase-knockout mouse model for GSD Ia. 28 Increased survival of the affected mice and decreased glycogen storage in liver were demonstrated. Improvements in the persistence of expression of introduced G6Pase will likely build upon this early result, possibly through the application of AAV vectors in the murine model for GSD Ia. The effects of gene therapy with AAV vectors in the canine model are relevant to severe human GSD Ia. The initial characterization of affected puppies with GSD Ia revealed that their clinical abnormalities most closely resemble severe, neonatal onset GSD Ia in humans. 14,29 The original three Maltese puppies with GSD Ia died between 5 and 8 weeks of age with poor growth and lethargy, and were found to have hepatomegaly and nephromegaly. 23 Prolonged survival has been accomplished with nutritional support and gene therapy in one of the affected puppies described here. However, residual effects of GSD Ia doubtlessly contributed to the abbreviated survival encountered in these experiments. In this canine model we have replicated earlier results seen in mice following intravenous delivery of AAV vectors. Introduced protein (G6Pase in this work) was present at high levels starting at 2 weeks following vector administration. 5,7,31,32 Liver DNA analysis demonstrated 1 to 5 copies of vector DNA per cell in the liver after 5 weeks following AAV vector administration. 3,7,27,30 Vector DNA was present as head-to-tail, head-to-head, and tail-to-tail concatamers, as recently reported for AAV vectors in murine liver. 33 In this analysis the level of vector DNA varied markedly between samples from different locations in the liver. However, all liver DNA samples analyzed contained vector DNA by 5 weeks after vector administration. Further analysis of the anatomic distribution of AAV vector DNA could impact the feasibility of AAV vectors for liver-targeted gene therapy by systemic administration. The number of vector particles ( to per kg) administered was at or above the number used for liver-targeted delivery in the murine and canine hemophilia B models. 2 However, in the hemophilia B models the vector was administered via the portal vein to provide more efficient delivery to the liver. Targeted delivery of improved AAV vectors to the liver could achieve increased therapeutic effects in future trials of therapy for the canine GSD Ia model. Materials and methods Preparation of AAV vectors We have constructed AAV vectors encoding canine G6Pase, containing the CMV immediate early promoter/enhancer (AAV-CcG6PGH, not shown) or mouse albumin promoter/enhancer to drive G6Pase expression (AAV-AlbcG6PGH, Figure 1). In AAV- CcG6PGH, the CMV immediate early promoterenhancer drives canine G6Pase. In AAV-AlbcG6PGH, the mouse albumin promoter/enhancer (sequence NB in Ref. 34) drives G6Pase expression. A deletion of human growth hormone intron 4 sequence from AAV- AlbcG6PGH( ) decreased the packaging size by 541 bp compared with AAV-AlbcG6PGH (Figure 1). We and others showed that the mouse albumin promoter/enhancer can produce high level expression of introduced genes from an AAV vector in mouse liver. 7,33 The human growth hormone intron and polyadenylation sequence (pa in Figure 1) contained in these vectors further increased transgene expression in transgenic mice. 35 AAV vector stocks were prepared essentially as described. 36 Briefly, 293 cells were transfected with AAV helper plasmids containing the rep and cap genes (pmtrep and pcmv Cap), and an AAV vector plasmid. The adenovirus helper functions were provided by transduction with an (E1, E2b ) adenovirus vector encoding -galactosidase (for dogs P2 and V4) that does not replicate in 293 cells, 37 or by transfection with a plasmid containing the adenovirus E4orf6 gene (for P3). The AAV packaging plasmids contain the rep and cap genes driven by heterologous promoters, and do not typically generate detectable replication-competent AAV. 36 Cell lysate was harvested 72 h following infection and freeze-thawed four times, before precipitation with ammonium sulfate, ph 7.0, and subsequent cesium chloride gradient centrifugation and dialysis to remove the cesium chloride. 38 For purification of higher numbers of vector particles, the cell lysate from 200 plates was concentrated with a hollow fiber concentrator (CH2PR concentrator and S1Y100

5 Gene therapy in canine GSD Ia with AAV vectors 1019 Figure 4 Southern blot analysis of vector DNA. Each lane represents 10 g liver DNA, and ethidium staining showed approximately equal loading of each sample. The time following vector administration for each sample in weeks (wk) is indicated. The DNA contained high levels of triglycerides as expected for GSD Ia liver, which necessitated an additional phenol/chloroform extraction and possibly contributed to uneven electrophoretic mobility between samples. (a) Liver DNA from each puppy, either digested with BamHI or uncut, including J5 that received vector particles of AAV- CG6PGH. (b) Fragment size predicted for the vector AAV-AlbG6PGH as detected by Southern blotting for the BamHI digest of concatamerized vector DNA in different configurations, including head-to-tail, head-to-head, or tail-to-tail. Fragment size for AAV-AlbG6PGH( ) are smaller and are shown in parentheses. (c) Multiple liver DNA samples for each puppy, each extracted from a different 2 g liver sample, digested with EcoRI. EcoRI digested vector plasmid DNA was loaded in indicated amounts for a calibration curve. Signals for vector DNA were quantitated by phosphorimaging (Molecular Dynamics, Sunnyvale, CA, USA), and normalized to the genomic DNA signal in each lane. cartridge; Amicon, Beverly, MA, USA). The number of vector DNA containing-particles was determined by DNase I digestion, DNA extraction, and Southern blot analysis as described. 33 Contaminating wt AAV particles were undetectable in recombinant AAV vector preparations by Southern blot analysis of extracted vector DNA ( 1% of AAV particles). All viral vector stocks were handled according to Biohazard Safety Level 2 guidelines published by the NIH. Polymerase chain reaction (PCR) analysis of mutation status The point mutation at position 450 in the canine glucose- 6-phosphatase cdna eliminates an NcoI recognition site. 24 PCR amplification with primers flanking the mutation generated an 82 bp DNA fragment. Following NcoI digestion PCR fragments were electrophoresed on a 2.0% agarose gel. Digestion with NcoI generated 63 bp and 19 bp fragments in the normal dog, and did not cut

6 Gene therapy in canine GSD Ia with AAV vectors 1020 Figure 5 Histologic analysis. PAS staining of formalin-fixed liver: (a) 11 weeks following AAV vector administration (P3 in Table 1), and (b) an untreated, affected control (D3 in Table 1). Oil red-o staining of frozen liver: (c) Two weeks (P2 in Table 1), (d) 5 weeks (V4 in Table 1), and (e) 11 weeks (P3 in Table 1) following AAV vector administration, and for (f) an untreated, affected control (D3 in Table 1). Magnification 100. the 82 bp fragment representing the mutant allele in affected puppies. Similar analysis for heterozygous dogs resulted in detection of both the 82 bp fragment representing the mutant allele, and 63 bp and 19 bp fragments representing the normal allele. Enzyme activity and glycogen analysis Enzyme analysis was performed as described. 23 Briefly, tissues were flash-frozen and stored at 70 C. Specific G6Pase activity was measured by using glucose-6-phosphate as substrate after subtraction of nonspecific phosphatase activity as estimated by -glycerophosphate. For normalized G6Pase activity, the mean G6Pase level for each sample was normalized to protein content. G6Pase activity was assayed three times independently for each sample. Each assay represented mg liver from individual, approximately 2 g canine liver samples. Phosphorylase was measured by the determination of inorganic phosphate released from 0.1 M glucose-1-phosphate in the presence of 1% glycogen and 2 mm adenosine monophosphate. Glycogen content was assayed by complete digestion of polysaccharide using amyloglucosidase (Sigma Chemical, St Louis, MO, USA). The structure of the polysaccharide was inferred by using phosphorylase free of the debranching enzyme to measure the yield of glucose-1-phosphate. Statistical analysis We performed statistical analyses using Stata (version 7.0, College Station, TX, USA). The regression analysis presented uses clustered regression technique. This technique accounts for the fact that the observations are not entirely independent of one another (multiple samples were taken from the same subject liver). Because there were a substantial percentage of zero values in the data, we evaluated the data using censored normal regression. The results were not substantially different from clustered regression (data not shown). We present a box and whisker plot (Figure 3c). The boxes for the plot show the median (middle line of the box), 25th% and 75th% (the interquartile range or the bottom and top of the box). The whiskers represent 1.5 times the interquartile range. Nutritional and gene therapy administration to affected dogs At birth the puppies were weighed, photographed and sketched for initial identification. Jugular venipuncture was performed within the first h for determination of blood glucose concentration and PCR analysis of mutation status. Physical examination was performed and all puppies were monitored every 2 h for nursing activity and neurologic reflexes. Body weights were obtained twice daily for the first 3 weeks of life. Puppies which were weak, hypoglycemic or which had obvious hepatomegaly were given supplemental formula at 1 ml per 100 g every 2 h (Esbilac, 60 ml (PetAg, Hampshire, IL, USA) with Polycose, 1.8 g (Ross, Columbus, OH, USA)). Once identified by genetic screening, affected puppies were allowed to continue to nurse until they showed overt clinical signs of hypoglycemia. At this time they received supplemental formula as described above. Affected puppies required hand rearing intermittently in a neonatal incubator, beginning as early as the first day of life. The incubator was equipped with temperature and

7 humidity control, oxygen and anesthetic gas supply, and nebulizing chambers. Puppies with profound or refractory clinical signs or hypoglycemia, or intercurrent illness that precluded oral feeding were given additonal glucose in the form of 12.5% dextrose by subcutaneous, intravenous or intraosseous routes as required. The goal of nutritional intervention was to provide 12 to 14 mg/kg/min glucose in the form of glucose or cornstarch (the latter after 2 months of age). The composition of the diet was 65% carbohydrates, 15% protein and 20% fat, as recommended for human GSD Ia. 14 Postmortem evaluation was performed immediately after death, and selected tissues were fixed in 10% buffered neutral formalin or frozen at 80 C. Frozen sections of unfixed and fixed liver were stained with oil-red-o for fat and PAS for glycogen. The cause of death was related to complications of GSD Ia in each case, precipitated by a mild infection or other stressful event. Each puppy was hypoglycemic at the time of death, which occurred at the following ages: J5, 61 days; V4, 39 days; P2, 20 days; P3, 86 days; and D3, 27 days. Acknowledgements We thank Ibrahim Bori, Eric Faulkner, Kathy Frid, Kwang Ok Shin and Joe Zeidner for excellent technical support, and Andrea Amalfitano and Dieksha Bali for insightful comments. We are very grateful to Dr Steve Van Camp for his contributions to the establishment of the canine GSD Ia colony. This work was supported a Howard Hughes Young Investigator Award (DDK), and by the Association for Glycogen Storage Disease. References 1 Koeberl DD et al. Persistent expression of human clotting factor IX form mouse liver after intravenous injection of adeno-associated virus vectors. Proc Natl Acad Sci USA 1997; 94: Snyder RO et al. Correction of hemophilia B in canine and murine models using recombinant adeno-associated viral vectors. Nat Med 1999; 5: Snyder RO et al. Persistent and therapeutic concentrations of human factor IX in mice after hepatic gene transfer of recombinant AAV vectors. Nat Genet 1997; 16: Nakai H et al. Adeno-associated viral vector-mediated gene transfer of human blood coagulation factor IX into mouse liver. Blood 1998; 91: Wang L et al. Sustained correction of bleeding disorder in hemophilia B mice by gene therapy. Proc Natl Acad Sci USA 1999; 96: Kessler PD et al. Gene delivery to skeletal muscle results in sustained expression and systemic delivery of a therapeutic protein. Proc Natl Acad Sci USA 1996; 93: Xiao W et al. Adeno-associated virus as a vector for liverdirected gene therapy. J Virol 1998; 72: Fisher KJ et al. Recombinant adeno-associated virus for muscle directed gene therapy. Nat Med 1997; 3: Herzog RW et al. Stable gene transfer and expression of human blood coagulation factor IX after intramuscular injection of recombinant adeno-associated virus. Proc Natl Acad Sci USA 1997; 94: Herzog RW et al. Long-term correction of canine hemophilia B by gene transfer of blood coagulation factor IX mediated by adeno-associated viral vector. Nat Med 1999; 5: Kay MA et al. Evidence for gene transfer and expression of factor IX in haemophilia B patients treated with an AAV vector. Nat Genet 2000; 24: von Gierke E. Hepato-nephro-megalia glycogenica Gene therapy in canine GSD Ia with AAV vectors (Glykogenspeicher-krankheit der Leber und Nieren). Beitr Pathol Anat 1929; 82: Cori GT, Cori CF. Glucose-6-phosphatase of the liver in glycogen stoarage disease. J Biol Chem 1952; 199: Chen YT. Glycogen storage diseases. In: Scriver CR, Beaudet AL, Sly WS, Valle D (eds). The Metabolic and Molecular Basis of Inherited Disease. McGraw-Hill: New York, 2001, pp Kishnani PS, Bengur AR, Chen YT. Pulmonary hypertension in glycogen storage disease type I. J Inher Metab Dis 1996; 19: Ohura T et al. Progressive pulmonary hypertension: a fatal complication of type I glycogen storage disease. J Inher Metab Dis 1995; 18: Talente GM et al. Glycogen storage disease in adults. Ann Int Med 1994; 120: Chen YT et al. Type I glycogen storage disease: nine years of management with cornstarch. Eur J Pediatr 1993; 152: S56 S Wolfsdorf JI. Crigler Jr. JJ. Effect of continuous glucose therapy begin in infancy on the long-term clinical course of patients with type I glycogen storage disease. J Ped Gastrol Nutr 1999; 29: Wolfsdorf JI. Crigler Jr. JJ. Biochemical evidence for the requirement of continuous glucose therapy in young adults with type I glycogen storage disease. J Inher Metab Dis 1994; 17: Lee PJ et al. Glomerular and tubular function in glycogen storage disease. Ped Nephrol 1995; 9: Wolfsdorf JI. Crigler Jr. JJ. Metabolic control and renal dysfunction in type I glycogen storage disease. J Inher Metab Dis 1997; 20: Brix AE et al. Glycogen storage disease type Ia in two littermate Maltese puppies. Vet Pathol 1995; 32: Kishnani P et al. Isolation and nucleotide sequence of canine glucose-6-phospatase mrna: identification of mutation in puppies with glycogen storage disease type Ia. Biochem Mol Med 1997; 61: Kishnani P et al. Canine model and genomic structure organization of glycogen storage disease type Ia (GSD Ia). Vet Pathol 2001; 38: Matern D et al. Liver transplantation for glycogen storage disease types I, III and IV. Eur J Ped 1997; 158: (Suppl. 2) S43 S Miao C et al. The kinetic of raav integration in the liver. Nat Genet 1998; 19: Zingone A et al. Correction of glycogen storage disease type 1a in a mouse model by gene therapy. J Biol Chem 2000; 275: Perlman M et al. Successful treatment of severe type I glycogen storage disease with neonatal presentation by nocturnal intragastric feeding. J Ped 1979; 94: Miao CH et al. Nonrandom transduction of recombinant adenoassociated virus vectors in mouse hepatocytes in vivo: cell cycling does not influence hepatocyte transduction. J Virol 2000; 74: Jung S-C et al. Adeno-associated viral vector-mediated gene transfer results in long-term enzymatic and functional correction in multiple organs of Fabry mice. Proc Natl Acad Sci USA 2001; 98: Koeberl DD et al. Persistent, therapeutically relevant levels of human granulocyte colony-stimulating factor in mice after systemic delivery of adeno-associated virus vectors. Hum Gene Ther 1999; 10: Nakai H et al. Identification of intermediates for recombinant adeno-associated virus vector genome concatemerization and integration in hepatocytes in vivo: evidence for double-stranded linear monomer genomes not circles as reactive intermediates. Mol Ther 2001; 3: S Pinkert CA et al. An albumin enhancer located 10 kb upstream functions along with its promoter to direct efficient, liver-specific expression in transgenic mice. Genes Dev 1987; 1: Brinster RL et al. Introns increase transcriptional efficiency in transgenic mice. Proc Natl Acad Sci USA 1988; 85: Allen JM, Halbert CL, Miller DM. Improved adeno-associated 1021

8 1022 Gene therapy in canine GSD Ia with AAV vectors virus vector production with transfection of a single helper adenovirus gene, E4orf6. Mol Ther 2000; 1: Amalfitano A et al. Production and characterization of improved adenovirus vectors with the E1, E2b and E3 genes deleted. J Virol 1998; 72: Snyder RO et al. Vectors for gene therapy, unit 12.1: production of recombinant adeno-associated viral vectors. In: Dracopoli NC et al (eds). Current Protocols in Human Genetics. John Wiley: Chichester, 1996, pp