AAV-mediated Gene Therapy Halts Retinal Degeneration in PDE6β-deficient Dogs

Abstract

We previously reported that subretinal injection of AAV2/5 RK.cpde6β allowed long-term preservation of photoreceptor function and vision in the rod-cone dysplasia type 1 (rcd1) dog, a large animal model of naturally occurring PDE6β deficiency. The present study builds on these earlier findings to provide a detailed assessment of the long-term effects of gene therapy on the spatiotemporal pattern of retinal degeneration in rcd1 dogs treated at 20 days of age. We analyzed the density distribution of the retinal layers and of particular photoreceptor cells in 3.5-year-old treated and untreated rcd1 dogs. Whereas no rods were observed outside the bleb or in untreated eyes, gene transfer halted rod degeneration in all vector-exposed regions. Moreover, while gene therapy resulted in the preservation of cones, glial cells and both the inner nuclear and ganglion cell layers, no cells remained in vector-unexposed retinas, except in the visual streak. Finally, the retinal structure of treated 3.5-year-old rcd1 dogs was identical to that of unaffected 4-month-old rcd1 dogs, indicating near complete preservation. Our findings indicate that gene therapy arrests the degenerative process even if intervention is initiated after the onset of photoreceptor degeneration, and point to significant potential of this therapeutic approach in future clinical trials.

Figures

Figure 1

Kinetics of rod function restoration and cone function preservation in treated dogs. (a,b) Rod b-wave responses (a) and cone b-wave responses (b): electroretinographic traces from a healthy control (H2) at 30 months of age and from a rcd1 dog (treated and untreated eyes) at 3, 12, and 36 months postinjection. (c,d) Averaged rod-mediated (c) and cone-mediated (d) b-wave amplitudes from healthy retinas (n = 3) (○), treated retinas ( ☐ ) (n = 3), and untreated retinas (△) (n = 3) over a 40-month period after injection. Symbols and bars represent mean ± SD. ERG, electroretinogram; mpi, month(s) postinjection.

Figure 2

Effect of gene therapy on retinal nuclear layer organization. (a) Schematic section of the entire retina. Red circles represent the different areas examined. (b) Retinal cross-sections (in peripheral or visual streak) from a healthy dog (H1, 8 months of age) and a 3.5-year-old rcd1 dog (treated and untreated eyes) indicating the extent of degeneration in the three retinal layers. All magnifications: 40× objective, zoom 1.0 (scale bars, 100 μm). GC, ganglion cells; INL, inner nuclear layer; ONL, outer nuclear layer; IS/OS, inner and outer segments.

Figure 3

Density of retinal nuclear layers in treated rcd1 dogs. Quantification of ONL nuclei (a), INL cells (b), and NeuN-positive RGCs (c) were performed in six retinal regions from 3.5-year-old rcd1 dogs (treated eyes (○) and untreated contralateral eyes (△)). Treated eyes were compared with retinas from healthy dogs (n = 2, 8 months and 8 years of age) (◯) and unaffected 4-month-old rcd1 dogs (◊). Values represent the mean ± SD. The shaded portion represents the location of the vector-exposed area. RGC, retinal ganglion cell.

Figure 4

Spatial distribution of rods and cones as observed in retinal whole mounts. Rod and cone distribution was analyzed in healthy dog (H3, 15 months of age) and 3.5-year-old-treated rcd1 dog (U10, T10). Retinal flat mounts were stained with GNAT1 and peanut agglutinin (PNA) antibody, which specifically label rods (red) and cones (green), respectively. (a–c) Rod density in healthy retina and rcd1 retinas (treated and untreated contralateral). (d,h,i) Cone density in peripheral retinas: healthy, rcd1 (treated and untreated contralateral retinas). (e,f) Cone density in the visual streak and area centralis, the site of maximum photoreceptor density, in healthy retina. (g) Preserved area in untreated rcd1 retina. (j) Cone density in the visual streak of treated rcd1 retina. (k) Cone density in treated areas.

Figure 5

Morphology and density distribution of rod and cone subtypes in retinal sections. Paraffin-embedded retina sections from healthy dogs (n = 2, 8 months and 8 years of age) and a 3.5-year-old rcd1 dogs (treated and untreated contralateral retina) were stained using rod-specific markers (GNAT1) or L/M- and S-opsin-specific antibodies. Histochemical detection of rod segments in healthy (H1, 8 months of age) and treated rcd1 (T11) retinas: peripheral retinas (a,c), and the visual streak (j,l). Absence of rods in untreated rcd1 retina (U11) (b,k). L/M- and S-opsin staining in peripheral retinas (d,f,g,i) and the visual streak (n,o,p,r) in healthy eyes and treated rcd1 eyes. No detectable L/M or S-opsin expression was detected in the peripheral retina of untreated rcd1 eyes (e,h). Immunohistochemistry detection of cone segments in the preserved area of untreated rcd1 retinas (n,q). Black arrows indicate S-opsin-positive cone outer segments. Magnifications: 40× objective, zoom 2.0 (scale bars, 50 μm). Sections were counterstained with Mayer's hemalum. Merged image shows a combination of fluorescence and Mayer's hemalum staining.

Figure 6

Density of cone subtypes in treated rcd1 retinas. L/M-cone subtypes (a) and S-cone subtypes (b) were quantified in six retinal regions from 3.5-year-old rcd1 dogs (treated (☐) and untreated contralateral (△) eyes). Treated eyes were compared with those of healthy controls (n = 2, 8 months and 8 years of age) (○). Data represent the mean ± SD. Shading indicates the location of vector-exposed areas.

Figure 7

Effect of gene therapy on inner layer and retinal glia. Image showing immunohistochemistry in the peripheral retina of healthy controls (n = 2, 8 months and 8 years of age) and rcd1 dogs (untreated and vector-exposed retinas). Bipolar cells and horizontal cells were labeled using antibodies against PKC-α and calbindin, respectively (a–f). Glial cells were stained using anti-GS and anti-GFAP antibodies (g–i). Magnifications: 40× objective, zoom 1.0, except for PKC-α staining (40× objective, zoom 2.0) (scale bars, 50 μm). GFAP, glial fibrillary acidic protein.