Regulation of T cell responses during chronic neuroinflammation: from regulatory T cells to ICOS

Control of chronic central nervous system (CNS) infection with the parasite Toxoplasma gondii requires ongoing inflammatory T cell responses in the brain. With this long-lived inflammatory response required for control of the parasite, understanding how this response occurs without causing debilitating immunopathology, particularly in a largely non-regenerative tissue, is of tremendous interest. Similar to other infection models, regulatory or immunosuppressive mechanisms are in play during chronic T. gondii infection that limit immune-mediated pathology. To explore some of these regulatory mechanisms, we began by focusing on the canonical immunosuppressive cell type, regulatory T cells (Tregs). Tregs have been shown to play an integral role in balancing inflammatory immune responses both in CNS autoimmune disorders as well as CNS infections. The precise mechanism by which the Treg population in the CNS carries out this regulation in vivo, however, remains unclear. We began by generally characterizing the phenotype of Tregs recruited to the CNS during chronic infection. We found that the Tregs recruited to the inflamed brain were largely Th1-polarized, expressing Tbet, CXCR3, and IFNγ, as well as IL-10. We also interrogated the TCR clonality of the effector CD4+ T cells and Tregs in the CNS during chronic infection. Interestingly, while we found CD4+ effector T cells with TCR specificity for a T. gondii-specific MHCII tetramer, we found no Tregs in the CNS specific for this same reagent. Upon further characterization of the TCR clonality of these populations using TCR sequencing, we observed minimal overlap between the TCR sequences of effector CD4+ T cells and Tregs in the inflamed CNS, suggesting that these two populations have distinct lineages and may be recognizing largely separate antigen pools during chronic CNS infection.
We also observed differential localization of the effector CD4+ T cells and Tregs within the CNS. While CD4+ effector T cells were abundantly found in the brain parenchyma, Tregs were mainly relegated to the meninges and perivascular spaces. The meninges and perivascular spaces during chronic infection were also enriched for MHCIIhiCD11c+ APCs. We hypothesized that Treg interaction with these APCs was important for their local regulation of infiltrating T cell responses, perhaps serving as a “gatekeeper” to the CNS. To begin to understand what is involved in the interaction between these two cell types that might support Treg suppression in the inflamed CNS, we blocked either the adhesion molecule LFA-1 or MHCII during chronic infection. Both blockade of LFA-1 and MHCII led to increased Treg velocity and less extensive contact time with APCs in the CNS, suggesting that both are important for maintenance of Treg:APC contact and, perhaps, continued Treg suppression. Interestingly, we found that LFA-1 blockade also rapidly led to a significant depletion of APCs from the CNS, suggesting that LFA-1:ICAM interactions play a role in recruitment or maintenance of APC populations in the inflamed CNS in addition to maintaining Treg:APC contact. MHCII blockade on the other hand, did not lead to loss of APCs from the CNS, but only affected Treg:APC contact. These results suggest that, upon recruitment to the inflamed CNS, Tregs are anatomically restricted, and their interaction with APCs, through both TCR and adhesion molecule interactions, regulates their local behavior, and perhaps their suppressive capacity, during chronic CNS infection.
In addition to Treg-mediated suppression of the ongoing immune response to T. gondii in the CNS, we also sought to understand other signals involved in regulation of chronic T cell responses in the inflamed brain. To explore the loss of suppressive cytokine exclusively during the chronic phase of infection, we blocked IL-10 receptor (IL-10R) in chronically infected mice. Consistent with previous reports, IL-10R blockade led to severe, fatal pathology associated with widespread changes in the inflammatory response, including increased antigen presenting cell (APC) activation, expansion of CD4+ T cells, and neutrophil recruitment to the brain. We then sought to identify regulatory mechanisms contributing to IL-10 production, focusing on ICOS (inducible T cell costimulator), a molecule implicated in IL-10 production in other models. Unexpectedly, ICOS-ligand (ICOSL) blockade during chronic infection led to a local expansion of effector T cells in the brain without affecting IL-10 production or APC activation. Instead, we found that ICOSL blockade led to changes in T cells associated with their proliferation and survival. In particular, we observed increased expression of IL-2 associated signaling molecules CD25, phosphorylated STAT5, Ki67, and Bcl-2 in effector T cells in the brain, along with decreased apoptosis. Interestingly, increases in CD25 and Bcl-2 were not observed in effector T cell populations following IL-10R blockade. Also unlike IL-10R blockade, ICOSL blockade led to an expansion of both CD8+ and CD4+ effector T cells in the brain, with no expansion of peripheral T cells or neutrophil recruitment to the CNS. Overall, these data suggest that IL-10 and ICOS differentially regulate T cell responses in the brain during chronic T. gondii infection. Taken together, the above results suggest multiple levels of regulation are necessary to maintain a balanced immune response during chronic infection, and each of these regulatory signals must be constantly integrated to limit immune-mediated pathology.