Gene therapy following subretinal AAV5 vector delivery is not affected by a previous intravitreal AAV5 vector administration in the partner eye

PURPOSE In an earlier study we found normal adeno-associated viral vector type 2 (AAV2)-mediated GFP expression after intravitreal injection to one eye of normal C57BL/6J mice. However, GFP expression was very poor in the partner eye of the same mouse if this eye received an intravitreal injection of the same vector one month after the initial intravitreal injection. We also found both injections worked well if they were subretinal. In this study, we tested whether the efficiency of subretinal AAV vector transduction is altered by a previous intravitreal injection in the partner eye and more importantly whether therapeutic efficiency is altered in the rd12 mouse (with a recessive RPE65 mutation) after the same injection series. METHODS One microl of scAAV5-smCBA-GFP (1 x 10(13) genome containing viral particles per ml) was intravitreally injected into the right eyes of four-week-old C57BL/6J mice and 1 microl of scAAV5-smCBA-hRPE65 (1 x 10(13) genome containing viral particles per ml) was intravitreally injected into the right eyes of four-week-old rd12 mice Four weeks later, the same vectors were subretinally injected into the left eyes of the same C57BL/6J and rd12 mice. Left eyes of another cohort of eight-week-old rd12 mice received a single subretinal injection of the same scAAV5-smCBA-hRPE65 vector as the positive control. Dark-adapted electroretinograms (ERGs) were recorded five months after the subretinal injections. AAV-mediated GFP expression in C57BL/6J mice and RPE65 expression and ERG restoration in rd12 mice were evaluated five months after the second subretinal injection. Frozen section analysis was performed for GFP fluorescence in C57BL/6J mice and immunostaining for RPE65 in rd12 eyes. RESULTS In rd12 mice, dark-adapted ERGs were minimal following the first intravitreal injection of scAAV5-smCBA-RPE65. Following subsequent subretinal injection in the partner eye, dramatic ERG restoration was recorded in that eye. In fact, ERG b-wave amplitudes were statistically similar to those from the eyes that received the initial subretinal injection at a similar age. In C57BL/6J mice, GFP positive cells were detected in eyes following the first intravitreal injection around the injection site. Strong GFP expression in both the retinal pigment epithelium (RPE) and photoreceptor (PR) cells was detected in the partner eyes following the subsequent subretinal injection. Immunostaining of retinal sections with anti-RPE65 antibody showed strong RPE65 expression mainly in the RPE cells of subretinally injected eyes but not in the intravitreally injected eyes except minimally around the injection site. CONCLUSIONS These results show that an initial intravitreal injection of AAV vectors to one eye of a mouse does not influence AAV-mediated gene expression or related therapeutic effects in the other eye when vectors are administered to the subretinal space. This suggests that the subretinal space possesses a unique immune privilege relative to the vitreous cavity.