A tumor stemness-based prognostic model derived from single-cell and bulk transcriptomics predicts survival and therapeutic response in hepatocellular carcinoma

Abstract Hepatocellular carcinoma (HCC) is characterized by marked intratumoral heterogeneity, therapeutic resistance, and poor prognosis, in which cancer stemness plays an important role. This study aimed to develop and validate a stemness-related prognostic model and to investigate its biological and therapeutic relevance in HCC. Single-cell RNA sequencing and bulk transcriptomic data were integrated to characterize stemness-associated tumor heterogeneity. CytoTRACE analysis identified a stemness-enriched epithelial subpopulation, and a seven-gene tumor stemness-related signature (TSCMS) was subsequently constructed using LASSO-Cox regression. The model was validated across four independent external cohorts and showed favorable prognostic discrimination. High-risk tumors were characterized by activation of oncogenic and stemness-associated pathways, a higher frequency of TP53 mutations, and an immunosuppressive microenvironment. CellChat analysis further revealed extensive predicted bidirectional communication between the stemness-enriched Epi1 subpopulation and immune cells, prominently involving MIF signaling from Epi1 and VISFATIN/NAMPT signaling from immune populations. TSCMS was also associated with differential anti-PD-1 response and predicted oxaliplatin sensitivity. Further functional validation confirmed that ENO1 knockdown reduced colony formation, migration, sphere formation, stemness-marker expression, and oxaliplatin resistance, whereas ENO1 overexpression produced the opposite effects. In vivo, ENO1 depletion enhanced the antitumor efficacy of oxaliplatin in xenograft models. Furthermore, in an independent clinical cohort of 28 patients with HCC receiving combined therapy, high ENO1 expression was associated with a poorer treatment response. Collectively, TSCMS provides a stemness-related framework for prognostic stratification and therapeutic characterization in HCC, while ENO1 represents a functionally relevant component associated with tumor stemness and oxaliplatin resistance.