Extended Data Fig. 7: PMN-MDSC impair PS^high CD8⁺ T cells through CD300ld.

a: Flow cytometric quantification of PS^high CD8⁺ T cells in MC38 tumors from WT or KO mice at day 18 post tumor inoculation (a, n = 5 mice per group). b: Percentage of activated or exhausted CD8⁺ T cells in PS^high and PS^low CD8⁺ T cell populations in MC38 tumors from WT or KO mice at day 18 post tumor inoculation, as determined by flow cytometry (b, Gzmb⁺ n = 8 mice, IFN-γ⁺ n = 8 mice, PD-1⁺ Tim3⁺ n = 5 mice). c: Gene expression profiles of lineage-defining markers from single-cell RNA-seq, used to annotate major cell clusters. d: Merged UMAP embedding and density plots showing the distribution of PS^high and PS^low CD8⁺ T cells from WT and KO tumors. e: Pathway enrichment analysis of effector CD8⁺ T cells from WT and KO tumors. f: T cells are colored by pseudotime, with color intensity increasing from naive (dark) to terminally differentiated (bright) cells. g: UMAP plots showing the clustering of myeloid-lymphocyte doublets detected in the scRNA-seq dataset. h: Dot plot showing the expression of myeloid- or lymphocyte-specific genes across the indicated clusters. i: Box plots displaying the number of genes expressed per cell in doublet populations compared to their singlet counterparts (i, lymphocyte n = 27071, myeloid-lymphocyte n = 763, myeloid n = 2343). j: Frequency of myeloid-lymphocyte doublets among total CD8⁺ cell plus myeloid-lymphocytes in each experimental group. k: Flow cytometry dot plots (left) and quantification (right) showing PS expression levels in CD8⁺ T cells after 48-h co-culture with WT or KO PMN-MDSCs (k, n = 3 independent samples per group). l: Flow cytometric analysis (left) and quantification (right) of CD300ld expression levels on PMN-MDSCs immobilized with Fc or CD300ld-ECD protein (l, n = 3 independent samples per group). m: Flow cytometric analysis (left) and quantification (right) of SA⁺ labeling of Fc or CD300ld-ECD-immobilizatd PMN-MDSCs (SA⁺) following co-incubation with SrtA⁺ CD8⁺ T cell (m, n = 3 independent samples per group). The scRNA-seq dataset was generated by pooling cells from five mice per group into one sequencing sample per group (c, d, e, f, g, h, i, j). Data are presented as mean ± s.e.m (k, l, m). For i, the central line represents the mean and the upper and lower edges of the box correspond to the first and third quartiles, respectively; the whiskers extend to the minimum and maximum values. Statistical analysis was performed using ordinary one-way ANOVA (b, i, k, m), and Student’s two-sided unpaired t-test (a, l). For e, the GO enrichment is reported as GeneRatio with BH-adjusted P value, based on one-sided hypergeometric over-representation analysis Data are representative of three (a, b, k, l, m) independent experiments.

Source data