Deciphering the transcriptomic landscape of tumor-infiltrating CD8 lymphocytes in B16 melanoma tumors with single-cell RNA-Seq

Abstract

Recent studies have proposed that tumor-specific tumor-infiltrating CD8⁺ T lymphocytes (CD8 TIL) can be classified into two main groups: "exhausted" TILs, characterized by high expression of the inhibitory receptors PD-1 and TIM-3 and lack of transcription factor 1 (Tcf1); and "memory-like" TILs, with self-renewal capacity and co-expressing Tcf1 and PD-1. However, a comprehensive definition of the heterogeneity existing within CD8 TILs has yet to be clearly established. To investigate this heterogeneity at the transcriptomic level, we performed paired single-cell RNA and TCR sequencing of CD8 T cells infiltrating B16 murine melanoma tumors, including cells of known tumor specificity. Unsupervised clustering and gene-signature analysis revealed four distinct CD8 TIL states - exhausted, memory-like, naïve and effector memory-like (EM-like) - and predicted novel markers, including Ly6C for the EM-like cells, that were validated by flow cytometry. Tumor-specific PMEL T cells were predominantly found within the exhausted and memory-like states but also within the EM-like state. Further, T cell receptor sequencing revealed a large clonal expansion of exhausted, memory-like and EM-like cells with partial clonal relatedness between them. Finally, meta-analyses of public bulk and single-cell RNA-seq data suggested that anti-PD-1 treatment induces the expansion of EM-like cells. Our reference map of the transcriptomic landscape of murine CD8 TILs will help interpreting future bulk and single-cell transcriptomic studies and may guide the analysis of CD8IL subpopulations in response to therapeutic interventions.

Keywords: B16-F10 melanoma; CD8 TIL classifier; Single-cell RNA-Seq; effector memory T cells; tumor-infiltrating CD8 T cells.

Figure 1.

Defining CD8 TIL states. (a) Experimental design. (b) tSNE plots indicating global transcriptomic similarities of CD8 TILs, and colored by unsupervised clustering (upper-left panel, clusters C1 to C4) or by expression of specific marker genes (other panels). Cycling cells are marked in magenta. (c) Dotplot indicating average expression of a panel of marker genes (x-axis, associated with naïve/memory, exhaustion and effector T cell phenotypes) for the four T cell clusters (y-axis). Color scale indicates scaled and centered log-normalized UMI counts. (d) Projection of T cell states onto three phenotypic score axes (stemness, inhibition/exhaustion and cytotoxicity, see Methods). Phenotypic scores are relative to each other, varying from 0% (inner circle) for the lowest score to 100% (outermost circle) for the highest score. (e) Gene-signature enrichment analysis against reference CD8 T cell subtypes signatures observed in cancer and chronic infection. UP and DOWN refer to the sets of up- or down-regulated genes associated with each comparison (e.g. Memory-like vs exhausted UP: genes up-regulated in memory-like and down-regulated in exhausted; Memory-like vs exhausted DOWN: genes down-regulated in memory-like and up-regulated in exhausted). Color scale indicates statistical significance of signature overlap (FDR corrected p-values, Fisher’s exact test). Details about the reference signatures in Methods. (f) Volcano plot showing top differentially expressed genes between EM-like vs exhausted and memory-like states. (g) Relative T cell cluster composition for each mouse. (h) Percentage of cycling cells in each cluster, as defined by high expression levels of cell-cycle genes in all TILs (left) or PMEL-specific TILs (expressing PMEL TCR, right). See methods for details.

Figure 2.

Flow cytometry validation of CD8 TIL populations. (a) Violin plots showing the decreasing and increasing expression levels (log-transformed normalized UMI counts +1) for Tcf7 and Pdcd1 for naïve, EM-like, memory-like and exhausted states. (b) Flow cytometry analysis of endogenous CD8 TILs from one representative tumor. Histograms show cell counts normalized by mode for naïve (CD44 low) T cells (red), Tcf1^high PD-1^low (light violet), Tcf1^high PD-1^intermediate (violet), Tcf1^high PD-1^high (blue) and Tcf1^low PD-1^high (green). (c) geometric Mean Fluorescence Intensity (MFI) for Ly6c and Cxcr3 and percentage of ITGB7^high cells for three tumors. Representative of two independent experiments. * denote statistically significant differences (Dunnett’s multiple comparisons test p-value <0.05).

Figure 3.,

Clonal relatedness of CD8 TIL states. (a) Percentage of cells with productive TCR paired alpha/beta chains obtained in each of the seven mice. (b) Percentage of T cells with non-unique (clonally expanded) TCR clonotype in each CD8 TIL cluster in endogenous responses in wild-type mice. (c) tSNE map distribution of (non-PMEL) clones with a clonal size of at least 10 cells per clonotype. (d) TCR repertoire overlap (Morisita-Horn index) between TIL states in wild-type mice. Ninety-five percent confidence intervals are shown in brackets. E Examples of expanded clones in wild-type mice. In each sub-panel, T cells sharing alpha and beta TCR sequences detected in individual mice are shown. Cell colors represent corresponding CD8 TIL states.

Figure 4.

CD8 TIL landscape modulation upon anti-PD-1 therapy. (a) Gene signatures of EM-like, memory-like and exhausted derived from our scRNA-seq analysis. For visualization, 100 random cells were sampled from each TIL state, and top differentially expressed genes are displayed. Horizontal lines separate groups of genes with similar expression patterns. (b) TIL state gene-signature enrichment analysis (GSEA) for the transcriptomic response of bulk endogenous CD8 TILs to PD-1 blockade (raw from Gubin et al.²⁷). NES = Normalized Enrichment Score. (c) TILPRED classification of total (left) and PD-1^low (right) CD8 TIL from MC38 colon adenocarcinoma (datasets from Xiong et al.³⁷ and Kurtulus et al.³⁸ expression of markers in Supplemental Figure 8). (d) States composition (TILPRED classification) of CD8 TILs upon ICB in MC38 (anti-PD-1 +anti-TIM3 vs isotype control) and sarcoma (anti-PD-1 vs isotype control)