A prognostic gene signature and subtype-specific drug sensitivity in TNBC revealed by single-cell and bulk RNA sequencing: Insights into stemness and tumor heterogeneity

Abstract

Triple-negative breast cancer (TNBC) remains one of the most aggressive Triple-negative breast cancer (TNBC) remains one of the most aggressive and therapeutically challenging breast cancer subtypes, largely due to its lack of targetable receptors and its intrinsic chemoresistance. In this study, we applied an integrative multi-omics approach − combining single-cell RNA sequencing (scRNA-seq) with bulk transcriptomic, epigenomic, and mutational analyses, to investigate the cellular heterogeneity and underlying mechanisms of drug resistance in TNBC. Analysis of the scRNA-seq dataset (GSE176078) revealed a complex tumor microenvironment with a highly plastic cancer epithelial subpopulation (Cluster C4) exhibiting elevated multipotency and distinct intercellular communication patterns. Concurrently, unsupervised clustering of TCGA-BRCA data delineated three molecular subtypes (CS1, CS2, and CS3) with unique biological and metabolic profiles, where CS3 notably exhibited unique molecular features associated with chromatin remodeling and high proliferative activity, suggesting distinct therapeutic vulnerabilities. An overlap analysis between scRNA-seq and bulk RNA-seq data identified 220 common differentially expressed genes (DEGs), from which a four-gene prognostic signature (CTSF, GBP1, BCL2A1, and EMP1) was derived. This signature robustly stratified patients by overall survival across both internal and external cohorts. Overall, our findings provide critical insights into the molecular drivers of chemoresistance in TNBC and offer a foundation for personalized therapeutic strategies.