Deconstructing T cell transcriptional heterogeneity and clonal dynamics in response to immune checkpoint blockade

T cells can fight cancer, but an immunosuppressive tumor microenvironment (TME) disallows them from carrying out their function over time. Upregulation of inhibitory checkpoint molecules such as programmed cell death protein 1 (PD1) and cytotoxic T-lymphocyte-associated protein 4 (CTLA4) can lead to such an immunosuppressive TME. Despite their widespread use, immune checkpoint blockade (ICB) antibodies targeting checkpoint molecules remain ineffective in most cancer patients. We do not understand why some patients respond to ICB better than others. To understand the heterogeneity of ICB response, we must understand the heterogeneity of the T cell subsets acted upon by such therapies. Here, we ask how T cell subsets change in the presence and absence of ICB. We track T cell clones through their T cell receptor sequences and link phenotypes with T cell receptor specificities. Through multiplexed single cell TCR sequencing, single cell RNA sequencing, and the use of cell- surface CITE-seq antibodies, coupled with surgical biopsy, we longitudinally tracked the fate of individual T cell clones within tumors at baseline and in response to ICB in an immunogenic mouse tumor model. Furthermore, computational clustering of T cells solely based on their gene expression profiles may ignore upstream regulatory mechanisms that control T cell gene expression. Hence, we employed Virtual Inference of Protein-activity by Enriched Regulon (VIPER) analysis to cluster CD8+ and CD4+ T cell phenotypes. VIPER leverages inference of gene regulatory networks to allow full quantitative characterization of protein activity for transcription factors, co-factors, and signaling molecules by assessing the enrichment of their transcriptional targets cell-by-cell among expressed genes. This gave us a window into the transcriptional states and their inferred protein activity. We next developed a computational analysis toolkit to study TCR clonality incorporating sub-sampling of TCR clonotypes, forward and back tracing of shared clones between ...