Extended Data Fig. 10: Autoimmune NOD pLN IGRP-specific CD8 T cells in T1D are distinct from CD8 T cells generated during acute and chronic infections.
From: An autoimmune stem-like CD8 T cell population drives type 1 diabetes

scRNA-seq datasets from pLN NRP-V7+ CD8 T cells (Fig. 4) were compared to scRNA-seq datasets of memory precursor effector cells (MPEC) and central memory CD8 T cells (TCM) generated during acute LCMV infection (day 7 and day 129 respectively), as well as to progenitor exhausted CD8 T cells (TPEX) from day 7 during chronic (clone 13) LCMV infection. a, Principal component analysis (PCA) of pseudo-bulk RNA-seq samples representing TCF1hiCD62Lhi, TCF1hiCD62Llo (intermediate), and TCF1loCD62Llo NOD pLN NRP-V7+ CD8 T cells from 3 technical replicates from Gearty et al (Fig. 4; T1D), TCM cells from Schauder et al (PNAS 2021; GSE130130; acute LCMV, day 129) and MPEC as well as TPEX as defined by Yao et al (Nat. Immunology, 2019; GSE119943; acute and chronic LCMV, day 7). The bar charts to the left and on the bottom of the PCA plot represent the genes that are most strongly (anti)correlated with a given principal component (PC). b, Dendrogram representing the results of hierarchical clustering based on the pseudo-bulk RNA-seq samples. c, Heatmap displaying row-normalized expression values of individual cells for genes identified as marker genes when comparing the single cells of the different populations (TCF1hiCD62Lhi, intermediate, and TCF1lo, MPEC, TPEX, TCM) to each other. d, Top, GO terms (biological processes) that are enriched based on genes that are specifically overexpressed in the different populations compared to all others. Bottom, the network-like representations display the genes (small circles) that belong to the respective GO terms (large circles), highlighting genes shared among individual GO terms and upregulated in indicated cell populations.