Systematic evaluation of AAV vectors for liver directed gene transfer in murine models

Abstract

Vectors based on adeno-associated viruses (AAVs) are being evaluated for use in liver-directed gene therapy. Candidates have been preselected on the basis of capsid structure that plays an important role in determining performance profiles. We describe a comprehensive and statistically powered set of mouse studies designed to compare the performance of vectors based on seven novel AAV capsids. The key criteria used to select candidates for successful gene therapy are high level and stable transgene expression in the absence of toxicity. Based on these criteria, the best performing vectors, AAV8, AAVhu.37, and AAVrh.8, will be further evaluated in nonhuman primates (NHPs).

Figure 1

Phylogenetic relation of candidate AAVs to other primate AAV capsids considered for gene therapy. Dendrogram of VP1 protein sequences was derived by neighbor-joining algorithm. Candidate capsids are underlined. Scale, evolutionary distance of the number of substitutions per site. AAV, adeno-associated virus.

Figure 2

AAV-mediated transgene expression in mouse liver. Livers were harvested, sectioned, and photographed at 7, 14, 28, 90, and 120 days after intravenous (i.v.) injection of AAV.CB.EGFP. (a) EGFP expression in liver at day 90 (vector dose at 1 × 10¹¹ GC/mouse). Images representing the median expression level of the group are shown. Bar = 150 µm. (b) Time course of GFP transduction efficiency in mouse liver (vector dose at 1 × 10¹¹ GC/mouse). The intensity of the green fluorescence was quantified with ImageJ. Data are presented as median with 95% CI (confidence interval) for each group (n = 4–6). Dotted line indicates background level from a naive mouse liver. (c) GFP transduction efficiency in mouse liver injected with tenfold lower vector dose (1 × 10¹⁰ GC/mouse). AAV, adeno-associated virus; EGFP, enhanced green fluorescent protein; GC, genome copy.

Figure 3

Transient elevation of liver enzymes in mice following intravenous administration of 1 × 10¹¹ GC of AAV.CB.EGFP. (a) Time course of liver enzyme levels showing peak elevation at day 14. Data are presented as median with 95% confidence interval for each group [n = 30 (pre and day 7), 24 (day 14), 18 (days 28 and 42), and 6 (day 90)]. (b) Elevation of ALT and AST at day 14 for individual mice. AAV, adeno-associated virus; ALT, alanine aminotransferase; AST, aspartate aminotransferase; EGFP, enhanced green fluorescent protein; GC, genome copy.

Figure 4

Temporary liver inflammation at day 14 following vector administration. (a) Representative hematoxylin and eosin (H&E) staining of formalin-fixed and paraffin-embedded livers harvested at day 14 after vector injection. Bar = 100 µm. (b) Scores for portal and lobular inflammation in individual mice. Horizontal bar indicates median level of the group (n = 4–6). AAV, adeno-associated virus.

Figure 5

T-cell response to AAV capsid and transgene GFP. Splenocytes were isolated 7, 14, 28, and 90 days after intravenous administration of 1 × 10¹¹ GC of AAV.CB.EGFP. For IFN-γ ELISPOT assay, splenocytes from individual animals were stimulated in vitro with the peptides containing the dominant H-2^d-restricted CD8 T-cell epitopes of AAV capsid or GFP, respectively. Time course of the T-cell response to (a) AAV capsid and (b) GFP are shown here. Data are presented as median with 95% confidence interval for each group (n = 4–6). AAV, adeno-associated virus; EGFP, enhanced green fluorescent protein; GC, genome copy; SFC, spot forming cell.

Figure 6

Prevalence of neutralizing antibodies against different AAV types in humans evaluated by in vitro NAb assay on human sera and inhibition of in vivo gene transfer by passive transfer of pooled human IVIG. (a) Percentage of human serum samples with NAb titer >1:20. (b) Magnitude of NAb titers for each AAV type. NAb to AAV2 is included as a reference. Number of samples assayed for AAV2, 7, 8, and rh.32.33: n = 100 (data have been published previously⁶); AAV6.2: n = 50; AAVhu.37, rh.64R1, and rh.8: n = 40. (c) C57BL/6 mice received passive transfer of human IVIG at the indicated amount at 24 and 2 hours before vector administration of 1 × 10¹¹ GC of AAV.LSP.cFIX-W packaged with different AAV capsids. Canine factor IX (cFIX) expression levels in plasma of individual mice at day 28 following vector administration are shown. Horizontal bar indicates median level of the group (n = 4–5). (d) Level of inhibition shown by the ratio of median cFIX levels in mice that received IVIG to the median levels in control mice (0 mg of IVIG). Data are presented as median with 95% confidence interval for each group (n = 4–5). (e) Time course of cFIX expression levels in the plasma of control mice (0 mg of IVIG) by seven candidates. AAV, adeno-associated virus; GC, genome copy; IVIG, intravenous immunoglobulin; NAb, neutralizing antibodies.