Integrative Analysis Identifies Cell-Type-Specific Genes Within Tumor Microenvironment as Prognostic Indicators in Hepatocellular Carcinoma

BackgroundHepatocellular carcinoma (HCC) is a leading cause of cancer-related mortality worldwide, but effective early detection and prognostication methods are lacking. MethodsThe Cox regression model was built to stratify the HCC patients. The single-cell RNA sequencing data analysis and gene set enrichment analysis were employed to investigate the biological function of identified markers. PLCB1 gain- or loss-of-function experiments were performed, and obtained HCC samples were analyzed using quantitative real-time PCR and immunohistochemistry assay to validate the biological function of identified markers. ResultsIn this study, we developed a model using optimized markers for HCC recurrence prediction. Specifically, we screened out 8 genes through a series of data analyses, and built a multivariable Cox model based on their expression. The risk stratifications using the Eight-Gene Cox (EGC) model were closely associated with the recurrence-free survivals (RFS) in both training and three validation cohorts. We further demonstrated that this risk stratification could serve as an independent predictor in predicting HCC recurrence, and that the EGC model could outperform other models. Moreover, we also investigated the cell-type-specific expression patterns of the eight recurrence-related genes in tumor microenvironment using single-cell RNA sequencing data, and interpreted their functional roles from correlation and gene set enrichment analyses, in vitro and in vivo experiments. Particularly, PLCB1 and SLC22A7 were predominantly expressed in malignant cells, and they were predicted to promote angiogenesis and to help maintain normal metabolism in liver, respectively. In contrast, both FASLG and IL2RB were specifically expressed in T cells, and were highly correlated with T cell marker genes, suggesting that these two genes might assist in maintaining normal function of T cell-mediated immune response in tumor tissues. ConclusionIn conclusion, the EGC model and eight identified marker genes could not only facilitate the accurate prediction of HCC recurrence, but also improve our understanding of the mechanisms behind HCC recurrence.

Automated summary

In this study, a multivariable Cox model was developed based on the expression of 8 identified genes to accurately predict HCC recurrence. The risk stratification using this model showed excellent performance in predicting HCC recurrence. Moreover, the study also investigated the cell-type-specific expression patterns of these genes in the tumor microenvironment and interpreted their functional roles.