Heterogeneity of CD4+ T Cells

Abstract

CD4+ T cells, which consist of T helper cells and T regulatory cells (Tregs) are indispensable for adaptive immunity. T helper cells are key positive regulators of both the humoral and cellular immune responses to infections, whereas Tregs attenuate immune responses and are important for tissue homeostasis. It has been known for over 30 years that CD4+ T cells are a heterogenous population that can be divided into several groups, which were largely defined in vitro according to their cytokine production profiles. First, we attempted to unravel the T cell heterogeneity in vivo at a single cell resolution. To this end, we performed single cell RNA-seq on colonic CD4+ T cells from healthy mice, as well as mice infected with pathogens that are known to skew the T cells towards particular cytokine responses. Upon infection, effector CD4+ T cells completely change and diversify their transcriptome. However, these transcriptional changes only partially fit with the classic T helper programs. Interestingly, Il17a-expressing cells from different conditions are transcriptionally more similar to Ifng-expressing cells from the same condition, suggesting that these are not entirely distinct cell subsets. Instead, we identified several overlapping gene modules that orchestrated a continuum of Ifng and Il17a-expressing cell states. Transcriptional profiling and ELISA of sorted populations from different “poles” of the continuum further validated these data. Moreover, we identified several new CD4+ effector T cells subpopulations, such as a population that expresses some myeloid genes. Our findings suggest that Th1/17 are not accurate classifications in vivo, but that T helper cells are a heterogenous population, which is not defined solely by their cytokine production ability. Second, we developed a set of CRISPR/Cas9 tools to study the effect of loss of function (LOF) mutations on Tregs’ heterogeneity in vivo. We employed this to validate certain genes that were found to be specifically overexpressed in tumor Tregs, which dampen immune responses to tumors. We found that LOF of Tnfrsf8, Cxcr3, and Samsn1 in Tregs led to their depletion in tumors. Thus, these genes may serve as potential targets for cancer immunotherapy.