Study and treatment of intraocular inflammation by anti-inflammatory gene transfer to the retina

We herein propose to study different therapeutical paradigms for intraocular inflammation using anti-inflammatory gene transfer to the retina.

Transfer of immuno-modulatory genes in RPE cells prior to their transplantation into the subretinal space could be useful to reduce immune rejection. We thus compared in vitro adeno-associated viral (AAV) gene transfer to a human immortalised RPE cell-line (ARPE-19) and primary cells (hRPE), to modify their genetic properties. We investigated 3 different serotypes and promoters in vitro, before evaluating a SOCS1 gene transfer to decrease immunogenicity of ARPE-19 cells in a xenograft rat model. We showed that AAV2 efficiently transduced at least 60% of ARPE-19 and hRPE cells, by comparison with the AAV1 and 5. In dividing ARPE-19 cells, mean-fluorescent intensity of CMV-driven gene expression was higher as compared to chicken beta-actin (CAG) and tetracycline inducible (TetON) promoters, but quickly decreased with time whereas CAG was more stable. AAV2-CAG-SOCS1 infection of ARPE-19 cells significantly decreased IFNγ-induced MHC II expression. In a last experiment, we infected in vitro ARPE-19 cells, using AAV2-CAG-SOCS1, prior to their delivery into the SRS of Lewis rats, and compared it with AAV2-CAG-eGFP-infected cells or non-infected cells. Since our preliminary results were not conclusive due to technical limitations, more extended investigations are necessary.

In another part, we developed a clinical grading system (CGS) to efficiently score EAU development in mice fundus. Particularly, we introduced the concept of active and inactive inflammation. However, some differences between CGS and histological (HGS) grading systems were pointed out to better characterise weaknesses of each method. We thus enhanced our CGS to reduce discrepancies with HGS but will need further investigations to obtain comparable grading systems.

Finally, we examined in vivo effects of a SOCS1 overexpression on EAU development, following AAV2-CAG-SOCS1 intravitreal (IVit) delivery in right eyes. We first tried two different intraocular routes of injections in this inflammatory model and showed IVit delivery to be the less traumatic. Due to important animal variabilities in EAU, SOCS1 overexpression did not lead to a significant reduction of inflammation when compared to GFP as a whole. However, our design study, allowing to compare injected versus non injected eyes, furthermore revealed IVit injection side effects with pro-inflammatory reaction due to the injection of AAV2-CAG-eGFP itself. In order to reduce the impact of inter-animal variability, we standardized the data by comparing the mean of ratios of injected over non-injected eyes (I/NI) for each animal rather than absolute values. We showed a significant reduction of the clinical and histological scores of the SOCS1 group as compared to the GFP group that was even stronger in the AAV2-targeted parts of the eyes. However, we missed a saline control to corroborate using our GFP group as a control and will need to introduce in a close future some bilateral injections to validate the use of the mean of grading ratios of I/NI in our experiments. Particularly, we showed a different pattern of MHC II positive invading cells in the ciliary body between SOCS1 treated and non-treated eyes. Further investigations are necessary to confirm and characterise SOCS1 protective mechanism in EAU.