Gene expression profile at single cell level of lymphocytes cells for Pan-T cell analysis

Tumour infiltrating T lymphocytes (TILs) have provided an attractive avenue for cancer treatment. Yet, the various states of TILs in tumour immune microenvironment (TIME) have not been fully characterized. Here, we present a single-cell atlas of T cells, encompassing 308,048 transcriptomes, spanning 16 different cancer types and 9 types of healthy/cancer-adjacent tissues. This allowed elucidation of 32 transcriptionally distinct T cell subsets, including cell states not reliably observable in individual datasets and those overlooked by previous studies. Our analysis revealed heterogenous subpopulations of follicular helper, regulatory, and proliferative T cells and a high degree of intertumoural heterogeneity in cell states and compositions, providing rich biological insights into the spectrum of TIL phenotypes. We identified a stress response state (TSTR) characterized by unique expression of heat-shock genes. Notably, by applying multiple spatial profiling approaches, we demonstrate TSTR cells are detectable in situ in the TIME across different cancer types, primarily in lymphocyte aggregates that are within tumor beds or surrounding tumor edges. We correlated TIL states and phenotypes with genomic, pathological, and clinical features in large cohorts of treatment-naïve patients as well as in patients treated with immune checkpoint blockade (ICB). We observed significant enrichment of intratumoural CD4/CD8 TSTR cells following ICB therapy, particularly, in non-responsive tumors, suggesting their potential role in immunotherapy resistance. Lastly, this study provides a T cell reference map via a dedicated web portal and an automatic alignment and annotation tool as a resource, supporting T cell therapy optimization and biomarker discovery. scRNA-seq profiling was performed on lungs from Gprc5a-/- mice treated with nicotine-derived nitrosamine ketone (NNK) or saline at the end of exposure (EOE, timepoint at which lungs are still clear of lesions) and at 7 months post-exposure, the time point of KM-LUAD onset in all mice (n = 4 mice per group and time point). Spatial transcriptomes of lung tissues from mice at 7 months post-exposure to NNK (n = 3) were performed to further characterize and validate our findings of the extences of KACs in mouse models. **RAW data not provided due to protection under IRB protocol**