Gene therapy using a novel G6PC-S298C variant enhances the long-term efficacy for treating glycogen storage disease type Ia

Abstract

The current phase I/II clinical trial for human glycogen storage disease type-Ia (GSD-Ia) (NCT03517085) uses a recombinant adeno-associated virus (rAAV) vector expressing a codon-optimized human glucose-6-phosphatase-α (G6Pase-α or G6PC). DNA sequence changes introduced by codon-optimization can negatively impact gene expression. We therefore generated a novel variant in which a single amino acid change, S298C, is introduced into the native human G6PC sequence. Short term gene transfer study in G6pc-/- mice showed that the rAAV-G6PC-S298C vector is 3-fold more efficacious than the native rAAV-G6PC vector. We have shown previously that restoring 3% of normal hepatic G6Pase-α activity in G6pc-/- mice prevents hepatocellular adenoma/carcinoma (HCA/HCC) development and that mice harboring <3% of normal hepatic G6Pase-α activity are at risk of tumor development. We have also shown that G6Pase-α deficiency leads to hepatic autophagy impairment that can contribute to hepatocarcinogenesis. We now undertake a long-term (66-week) preclinical characterization of the rAAV-G6PC-S298C vector in GSD-Ia gene therapy. We show that the increased efficacy of rAAV-G6PC-S298C has enabled the G6pc-/- mice treated with a lower dose of this vector to survive long-term. We further show that mice expressing ≥3% of normal hepatic G6Pase-α activity do not develop hepatic tumors or autophagy impairment but mice expressing <3% of normal hepatic G6Pase-α activity display impaired hepatic autophagy with one developing HCA/HCC nodules. Our study shows that the rAAV-G6PC-S298C vector provides equal or greater efficacy to the codon optimization approach, offering a valuable alternative vector for clinical translation in human GSD-Ia.

Keywords: Autophagy impairment; Clinical translation; Glucose-6-phosphatase-α variant; Recombinant adeno-associated virus vector.

Conflict of interest statement

Declaration of competing interest None declared.

Copyright © 2020 Elsevier Inc. All rights reserved.

Figures

Fig. 1.

Biochemical analysis of 62–66-week-old rAAV-G6PC-S298C-treated G6pc−/− mice. (A) Hepatic microsomal G6Pase-α activity in rAAV-G6PC-S298C-treated G6pc−/− mice is shown at the indicated ages in weeks (W). The mice were grouped based on viral dosages: 1013 vp/kg (n = 2), 3 × 1012 vp/kg (n = 6), and 1012 vp/kg (n = 9). Two major subgroups emerge for mice restoring 3.3–35% (S298C/≥3% mice, n = 9) and 0.3–2.9% (S298C/<3% mice, n = 8), respectively of normal hepatic G6Pase-α activity. Hepatic microsomal G6Pase-α activity in 62–66-week-old wild-type mice (n = 15) averaged 240.2 ± 15.8 units, representing 100% normal hepatic G6Pase-α activity. The grey area denotes 3% of normal hepatic G6Pase-α activity. (B) Fasting blood glucose tolerance profiles. (C) Histochemical analysis of hepatic G6Pase-α activity. Each image represents an individual mouse. (+/+), wild-type, (−/−), untreated G6pc−/−, and (S298C), rAAVG6PC-S298C-treated mice. Scale bar = 200μm. The numbers in percentage represent hepatic G6Pase-α activity restored in the rAAV-G6PC-S298C-treated mice.

Fig. 2.

Phenotypic analysis of 62–66-week-old rAAV-G6PC-G6PC-S298C-treated G6pc−/− mice. (A) H&E stained liver sections and hepatic glycogen contents. Each plate represents an individual mouse. Numbers in parentheses represent % of hepatic G6Pase-α activity restored in the mice. The representative hematoxylin and eosin (H&E) stained non-tumor and tumor lesions in the tumor-bearing S298C/<3% mouse are shown. The arrow denotes HCA/HCC. Scale bar = 100μm. (B) Body weight (BW) and body fat values. Both values were similar between S298C/≥3% and S298C/<3% mice and were grouped as the S298C (n = 17) mice. (C) Liver weight (LW) and LW/BW values. Data represent the mean ± SEM. *p< 0.05, **p< 0.005.

Fig. 3.

Analysis of 62–66-week-old rAAV-G6PC-S298C-treated G6pc−/− mice. The data were analyzed from wild-type (+/+, n = 15), S298C/≥3% (n = 9) and S298C/<3% (n = 8) mice. When values between S298C/≥3% and S298C/<3% mice were similar, they were grouped together as the S298C (n = 17) mice. (A) Blood insulin levels. (B) Insulin tolerance profiles. (C) Hepatic metabolites values. (D) Hepatic levels of G6pt transcript. (E) Western-blot and densitometry analyses of hepatic GCK and β-actin. (F) The serum antibodies against human G6Pase-α. Lanes 1: anti-human G6Pase-α antiserum; lanes 2, 4, 6, 8, 10, 12: serum samples (1: 50 dilution) from wild-type mice, or lanes 3, 5, 7, 9, 11, 13: serum samples (1: 50 dilution) from rAAV-G6PCS298C-treated G6pc−/− mice. Numbers in parenthesis represent % of normal hepatic G6Pase-α activity restored in the mice. Data represent the mean ± SEM. *p< 0.05, **p< 0.005.

Fig. 4.

Impaired hepatic autophagy in the S298C/P < 0.05, **P < 0.005.