Extended Data Fig. 3: Boosting BMP while blocking TGFβ1 recovers effector function in dysfunctional CD8+ T cells.
(a) IGV Snapshots showing: DNA methylation levels (top panel) at individual CpG sites within mouse naïve (black), antigen-specific exhausted WT (red) or Dnmt3a-deficient (green) CD8+ T cells on day 35 post-chronic LCMV infection. Blue-to-red ratio indicates % of unmethylated versus methylated reads (GSE99450); open chromatin peaks (middle panel) detected in naïve (black), progenitor (PD-1+ Tim-3-; green), or terminally exhausted (PD-1+ Tim-3 + ; red) CD8+ T cells from chronically infected mice (PRJNA546023); and levels of H3K27 acetylation marks (bottom panel) in the progenitor (green), cytolytic (blue; Cx3cr1+PD-1+), and terminally exhausted (Cx3cr1- PD-1+ Tim-3+) CD8 T cells during chronic LCMV infection (GSE149810) at the Smad1, Smad5, and Tgfbr3 loci. (b) Heatmap of RNA levels in exhausted T cell subsets during chronic LCMV infection (GSE84105). (c) tSNE plot visualization of human CD8+ T cells for the described conditions in Fig. 4a after PMA/ionomycin stimulation on day 28, and representative plots showing protein expression levels of IFNG, TNFα, CD107a, T-bet, and CD103 in PMA/Ionomycin-stimulated CD8+ T cells on day 28 (n = 4 wells per condition). (d) Bar graphs showing expression levels of TOX in adult blood-derived CD8+ T cells on day 21 of chronic ‘Strong’ or (e) ‘Weak’ TCR stimulation. (f) Bar graphs showing MFI of IL-7R and (g) CD103 in adult CD8+ T cells on day 21 of ‘Weak’ TCR stimulation. For panels d-g, n = 3–4 biological replicates per group per experiment, data pooled from 2 independent experiments. Adjusted P-value ** p < 0.01, *** p < 0.001, **** p < 0.0001 comparisons were made using one-way ANOVA with Tukey’s multiple comparisons as indicated (panels d,f) or one-way ANOVA with Bonferroni’s multiple comparisons relative to Dysf. group (panels e,g). Error bars = mean ± SEM.
