Combined rod and cone transduction by adeno-associated virus 2/8

Abstract

Gene transfer to both cone and rod photoreceptors (PRs) is essential for gene therapy of inherited retinal degenerations that are caused by mutations in genes expressed in both PR types. Vectors based on the adeno-associated virus (AAV) efficiently transduce PRs of different species. However, these are predominantly rods and little is known about the ability of the AAV to transduce cones in combination with rods. Here we show that AAV2/8 transduces pig cones to levels that are similar to AAV2/9, and the outer nuclear layer (mainly rods) to levels that are on average higher, although not statistically significant, than both AAV2/5 and AAV2/9. We additionally found that the ubiquitous cytomegalovirus (CMV), but not the PR-specific GRK1 promoter, transduced pig cones efficiently, presumably because GRK1 is not expressed in pig cones as observed in mice and humans. Indeed, the GRK1 and CMV promoters transduce a similar percentage of murine cones with the CMV reaching the highest expression levels. Consistent with this, the AAV2/8 vectors with either the CMV or the GRK1 promoter restore cone function in a mouse model of Leber congenital amaurosis type 1 (LCA1), supporting the use of AAV2/8 for gene therapy of LCA1 as well as of other retinal diseases requiring gene transfer to both PR types.

FIG. 1.

Porcine retinal transduction after AAV subretinal delivery. LW pigs were subretinally injected with 1×10¹⁰ GC/eye of AAV2/5 (a), 2/8 (b), or 2/9-CMV-EGFP (c). Retinal cryosections were obtained 4 weeks after injection, and EGFP was analyzed using fluorescence microscopy. (d) Fluorescence intensity in the PR layer was quantified for each group of animals; p=0.3, one-way ANOVA. Scale bar=25 μm. AAV, adeno-associated virus; ANOVA, analysis of variance; AU, arbitrary units provided by the LAS AF lite software (see Materials and Methods); CMV, cytomegalovirus; EGFP, enhanced green fluorescent protein; LW, Large White; n, number of eyes; ONL, outer nuclear layer; PR, photoreceptor; RPE, retinal pigment epithelium. Color images available online at www.liebertpub.com/hum

FIG. 2.

AAV-mediated cone transduction in pig retina. LW pigs were subretinally injected with AAV2/5, 2/8, and 2/9-CMV-EGFP (1×10¹⁰ GC/eye of each vector). (a) Retinal cryosections were obtained 4 weeks after injection, and anti-LUMIf-hCAR immunostaining (red) was performed to label cones; EGFP+ cones are indicated by white arrows. (b) Higher magnification of the transduced cones. The confocal microscopy images are representative of single plans of Z-stack taken at 63×magnification. (c) Quantification of cone transduction efficiency of AAV2/5, 2/8, and 2/9-CMV-EGFP. The histograms represent the percentage of EGFP+/CAR+ cells counted on a total of 150–200 CAR+ cells. p=0.63, one-way ANOVA. CAR, cone arrestin; DAPI, 4′,6′-diamidino-2-phenylindole; INL, inner nuclear layer.

FIG. 3.

AAV2/8-mediated retinal transduction in pig retina. LW pigs were subretinally injected with AAV2/8-CMV-EGFP (a), AAV2/8-CMV-EGFP-miR204T (b), and AAV2/8-GRK1-EGFP (c, 1×10¹⁰ GC/eye of each vector). Retinal cryosections were obtained 4 weeks after injection, and EGFP was analyzed using fluorescence microscopy. The exposure time is the same for the eyes injected with AAV2/8-CMV-EGFP and AAV2/8-CMV-EGFP-miR204T, while for AAV2/8-GRK1-EGFP the representative image was taken at an exposure time that was double the other vectors. (d) Fluorescence intensity in the PR layer was quantified for each group of animals at the same exposure time; p=0.13, one-way ANOVA. Scale bar=25 μm. Color images available online at www.liebertpub.com/hum

FIG. 4.

Effect of different regulatory elements on AAV2/8-mediated cone transduction in pig retina. LW pigs were subretinally injected with AAV2/8-CMV-EGFP, AAV2/8-CMV-EGFP-miR204T, and AAV2/8-GRK1-EGFP (1×10¹⁰ GC/eye of each vector). Retinal cryosections were obtained 4 weeks after injection, and anti-LUMIf-hCAR immunostaining was performed to label cones (see Fig. 2a and b). Quantification of cone-transduction efficiency is shown. The histograms represent the percentage of EGFP+/CAR+ cells counted on a total of 150–200 CAR+ cells. For the eyes injected with AAV2/8-GRK1-EGFP, the quantification of cone transduction was done at both low and high exposure times. *p≤0.05; **p≤0.01, one-way ANOVA.

FIG. 5.

AAV2/8-mediated retinal transduction in mouse retina. Four-week-old C57BL/6 mice were subretinally injected with 1.7×10⁹GC/eye each of either AAV2/8-CMV-EGFP (a) or AAV2/8-GRK1-EGFP (b and c, at low and high exposure times, respectively). Retinal cryosections were obtained 4 weeks after injection, and EGFP was analyzed using fluorescence microscopy. (d) Fluorescence intensity in the PR layer was quantified for each group of animals; scale bar=40 μm. **p≤0.01; Student's t-test. (e) Retinal cryosections were immunostained using an anti-LUMIj-mCAR antibody (red) to label cones. Representative confocal microscopy image from a single plan (magnification 63×) of a retina injected with AAV2/8-GRK1-EGFP: EGFP+ cones are indicated by white arrows. (f) Higher magnification of the transduced cones. (g) Quantification of cone transduction efficiency of AAV2/8-CMV-EGFP and AAV2/8-GRK1-EGFP in mouse retina. The histograms represent the percentage of EGFP+/CAR+ cells counted on a total of 150–200 CAR+ cells. For the eyes injected with AAV2/8-GRK1-EGFP, the quantification of cone transduction was done at both low and high exposure times. **p≤0.01, one-way ANOVA.

FIG. 6.

Retinal function assessment in Gucy2e^−/− mice after AAV2/8-mediated gene transfer. Gucy2e^−/− mice were treated at P10 with 1.5×10⁹ GC/eye of either AAV2/8-GUCY2D in one eye or AAV2/8-EGFP (or untreated) in the contralateral eye. Six-hertz flicker results of Gucy2e^−/− animals treated with either AAV2/8-CMV-GUCY2D (a) or AAV2/8-GRK1-GUCY2D (b) were analyzed 1 month postinjection. (c) Retinal function was also assessed by paired flash ERG. The histograms represent the average amplitudes of the responses to either the first (black bar) or the second (white bar) flash. n is the ratio between the number of represented eyes and the total number of injected eyes. The numbers of represented eyes in the CMV and GRK1 groups are those in which improvement of cone activity was evident, which is additionally indicated as % of total in parentheses. *p≤0.05; **p≤0.01; ***p≤0.001, one-way ANOVA.