Fig. 7

Single-cell RNA-Seq from matched PBMC, NT, and T sites of NSCLC patients identifies different PD1/CD28-associated clusters. A Flowchart of the overall study design. Lymphocytes from two NSCLC patients were purified from three matched districts using FACS-sorting and employed for single-cell RNA-Seq, performed with the BD Rhapsody platform. B UMAP projection showing the main immune subsets within total cells (n = 10,944 cells, right panel) and in the three different districts (middle panels), color-coded according to the cell type. Right, barplot showing the distribution of immune cell types within the different districts in the two patients. C Gating strategy to select PD1⁺CD28⁻ and PD1⁺CD28⁺ within a CD8⁺CD4⁻CD19⁻ T-cell population on merged patients. D-E UMAP projection of concatenated (upper left panel) or differently distributed cells in the three districts (bottom panels), gated on PD1⁺CD28⁻ (D) (n = 558) or PD1⁺CD28⁺ (E) (n = 541) cells, showing the identification of five color-coded clusters for each PD1/CD28 subset. Upper right, Frequency of each cluster across the three districts. The two patients have been grouped together. F Heatmap from single-cell analysis showing gene expression levels in cells distributed into clusters (indicated on the top by the same color as in D-E). The top ten marker genes, differentially expressed for each cluster (listed by fold change), are shown on the y-axis. The complete list of DEG is reported in Table S4 and in Table S5. Panel A was created with Biorender