Engagement of ICOS in tissues promotes establishment of CD8+ tissue-resident memory T cells

Recirculating and tissue-resident memory CD8+ T cells provide distinct modes of immune protection, yet the signals that dictate differentiation of these populations are ill-defined. In particular, the interactions within tissues that promote generation of resident memory T cells (TRM) are unclear. Here, we show that the inducible costimulatory molecule ICOS, well known to regulate differentiation of CD4+ T cell populations, is required for CD8+ TRM but not recirculating memory subsets. Furthermore, ICOS engagement during CD8+ T cell recruitment to non-lymphoid tissues is critical for efficient TRM establishment: ICOS/ICOS-L interactions are dispensable throughout CD8+ T cell priming and for TRM maintenance, while ICOS-L expression by radioresistant cells is key for optimal TRM generation. This role for ICOS depends on its ability to signal through PI3K. Together, our data illustrate that specific local costimulatory cues promote production of tissue-resident populations, with potential implication for therapeutic manipulation. CD8+ P14 T cells from WT or Icos knockout (KO) mice were mixed, co-adoptively transferred into recipient mice, and the mice were infected with LCMV the following day. CD8+ T cells were isolated from recipient mouse spleen, small intestine intraepithelial lymphocytes (SI-IEL), and kidney in two different cohorts (i.e. day 5 or day 32 post infection). Cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq) single cell RNA sequencing (scRNA-seq) analysis was completed. Cells were stained with antibody-oligo derived tags (ADTs) against CD45.2 and ICOS to identify cells derived from WT and Icos KO mice, and three hashtag oligos (HTOs) to identify cells derived from different tissues. For each timepoint, cells from different tissues were mixed at equal proportion and captured in the same library using the 10x Genomics Chromium Single Cell 3' (ver. 3.1) reagent kits, generating three independent gene expression (GEX) and feature barcode (ADT and HTO) sequencing libraries. The metadata.txt (representing 14 samples) contains each sample information and additional details.