Introduction Hepatocellular carcinoma (HCC) remains a leading cause of cancer‐related mortality worldwide, with cellular senescence playing a context‐dependent role in tumor progression and the immunosuppressive microenvironment. This study is aimed at identifying senescence‐related gene signatures through integrated single‐cell and transcriptomic analyses to construct a robust prognostic model for predicting survival and immunotherapy response in HCC patients. Methods We obtained single‐cell RNA sequencing (scRNA‐seq) data from the Gene Expression Omnibus (GEO) database and transcriptomic data from The Cancer Genome Atlas (TCGA). The scRNA‐seq data were processed using the Seurat and Harmony packages for cell clustering and batch correction. Senescence scores were calculated via the AUCell package, and differentially expressed genes were identified using the limma package. Prognostic genes were selected through univariate and LASSO Cox regression (glmnet package) to construct a risk model, which was validated in multiple independent cohorts. Immune infiltration was assessed with single‐sample gene set enrichment analysis (ssGSEA), TIMER, and MCPCounter algorithms, and response to immune checkpoint blockade was predicted using the tumor immune dysfunction and exclusion (TIDE) platform. Experimental validation included qRT‐PCR, Cell Counting Kit‐8 (CCK‐8), wound healing, and Transwell assays in HCC cell lines. Results A total of 80,997 identified cells were allocated to eight clusters, with an evidently higher percentage of natural killer (NK) cells in HCC samples. A higher senescence score was also seen in HCC samples, and poor prognosis was noticed in the patients of high senescence score group. Further, the DEGs were intersected with the genes highly expressed in Population 4 of NK cells to reveal their enrichment in cell cycle and cell division. Further, eight genes ( TMEM106C , BSG , COPE , CDCA8 , KPNA2 , LIG1 , UQCRH , and CCT5 ) with differential expression in HCC were applied to construct the risk model, which could stratify HCC patients into different risks and predict the prognosis. Besides, the high immune infiltration and expression levels of immune checkpoint–relevant genes yet poor immunotherapy response were noticed in HCC patients of high risk. Further validation tests have suggested that the knockdown of CDCA8 repressed the malignant phenotypes of HCC cells. Discussion This integrated analysis establishes a senescence‐related gene signature as a robust tool for prognostic stratification and immunotherapy response prediction in HCC. The model highlights the complex interplay between cellular senescence and the immunosuppressive tumor microenvironment, offering insights for personalized treatment strategies. Furthermore, the identified biomarker CDCA8 represents a promising therapeutic target warranting further investigation. Conclusion These discoveries provide novel evidence on senescence in HCC, which may tailor the pharmacological interventions to improve the clinical management.
Yu et al. (Thu,) studied this question.