Cell-type specific gene signatures reveal novel immune checkpoints and prognostic markers in lung cancer

The tumor microenvironment (TME) is a complex interplay of immune, stromal, and malignant cells whose interactions shape cancer progression and therapeutic responses. In this study, we performed an integrative single-cell transcriptomic analysis to define cell-type-specific gene signatures with emphasis on immune-tumor communication, exhaustion states, and their prognostic implications. We derived 30-gene signatures for B cells, CD8⁺ T cells, fibroblasts, macrophages, NK cells, T cells, Tregs, tumor cells, and unclassified clusters. Ligand-receptor mapping revealed widespread communication, including macrophage-fibroblast and Treg-tumor axes. Pseudotime analysis further showed immune exhaustion as a dynamic process enriched with checkpoint genes such as PDCD1, CTLA4, LAG3, and TIGIT, particularly within Tregs and exhausted CD8⁺ T cells. Survival analysis of representative genes revealed contrasting effects of immune and stromal activity: MS4A1 (B-cell signature) and TPPP3 (tumor signature) correlated with improved prognosis, whereas fibroblast-specific COL1A1 predicted poor outcomes. Incorporation of less-characterized genes highlighted novel prognostic signals. Correlation networks among signatures underscored the functional interdependence of immune and stromal compartments. Together, this study provides a systems-level framework linking transcriptional with survival outcomes. By combining established immune checkpoints with novel candidates, our findings expand the biomarker landscape for prognostic stratification and therapeutic targeting in lung cancer.