PD-1 function on T cells in spontaneous and anti-PD-1-mediated autoimmunity

Abstract

Immune-related adverse events (irAEs) are a potentially life-threatening complication of PD-1 cancer immunotherapy in patients. A deeper understanding of irAEs is essential to inform strategies to treat and monitor irAEs. This knowledge may also uncover information about the pathogenesis of spontaneously arising autoimmune diseases. In this thesis, I investigated how anti-PD1 affects autoimmunity by undertaking two independent studies. First, I studied how the cellular states of diabetogenic T cells in anti-PD-1-induced T1D differ from those in spontaneous T1D in non-obese diabetic (NOD) mice. Using a combination of scRNA-seq, TCR-seq, and flow cytometry studies, I identified significant differences in the phenotype of diabetogenic T cells mediating anti-PD-1-induced T1D compared to spontaneous T1D. Anti-PD-1-induced T1D showed a significant increase in proliferative and terminally exhausted/effector-like CD8+ T cells, while spontaneous T1D is largely dominated by memory T cell subsets. Moreover, I determined that anti-PD-1 increases TCR sharing between the pancreas and periphery, which suggests that analysis of blood T cells may be an effective strategy to monitor irAEs rather than relying on the autoimmune target organ. Secondly, I examined the effect of interactions of PD-1 and TIGIT on T cells in autoimmunity. Given the critical roles of conventional CD4+ T cells (Tcon) and Treg in promoting and suppressing EAE, respectively, I analyzed the transcriptional profiles of CD4+ T cells in the spleen and CNS of mice with EAE treated with anti-PD-1 at peak of disease and used the TCR as a molecular barcode to identify individual clonotypes. I found that anti-PD-1 exacerbated disease by promoting clonotype expansion of Th1 in the central nervous system (CNS) and reduced the proportion of Treg in the CNS that have shared TCRs in the spleen. These findings suggest that anti-PD-1 reduces Treg in the CNS which allows for clonotype expansion of encephalitogenic Tcon. Thus, using a combination of in vivo mouse models, transcriptomic, and phenotypic analyses, I have revealed the blood as a potential site for monitoring irAEs, identified differences in diabetogenic T cells in anti-PD-1-induced and spontaneous T1D; and determined differences in Treg and Tcon clonotypes in EAE in mice treated with anti-PD-1 versus controls.