ABSTRACT: The healthy eye limits destructive inflammation through the phenomenon known as immune privilege, but despite that, it is subject to autoimmune uveitis driven by retina-specific effector T cells. Using an in vivo immune privilege model based on introducing allotypically marked, TCR-transgenic naive retina-specific T cells into eyes of healthy mice, we found that after one week all the injected cells became primed within the eye. Among them, about 30% had converted to functional Foxp3+ regulatory T cells (Tregs), whereas the remaining ~70% did not convert to Tregs, but nevertheless failed to elicit pathology. It has long been known that ocular fluids in vitro, as well as the ocular environment in vivo, support conversion of T cells to Tregs. However, in-depth investigation of the cell fate of the non-Foxp3-converted T cells, and why they fail to elicit pathology, was hindered by the limited number of cells retrievable from mouse eyes. Therefore, to gain insights into this, we tracked changes in gene expression of retina-specific T cells responding to their cognate antigen in the living eye, at the single-cell level. Our data revealed unique transcriptional signatures not only in the Foxp3+ Tregs, but also in the non-Foxp3-converted T cell populations. Of particular interest, most non-Foxp3-converted T cell populations appeared phenotypically anergic with high levels of inhibitory regulators (Ctla4, Lag3, Pdcd1, Cblb, and Dgkz), and expressed higher levels of some immunosuppressive genes, e.g. Tgfb1, than the converted Foxp3+ Tregs. Moreover, trajectory analyses suggested a branched pattern showing that the Tregs and anergic T cells differentiate simultaneously, rather than in tandem, from a common proliferative precursor. Our findings offer novel insights into ocular immune privilege and its complex underlying network of inhibitory mechanisms and could suggest therapeutic approaches to ocular inflammatory diseases.
