Bioinformatics-Based Construction of Immune-Related microRNA and mRNA Prognostic Models for Hepatocellular Carcinoma.

Introduction: The development and progression of Hepatocellular Carcinoma (HCC) is more relevant to immune regulation. Therefore, there is an urgent need to find immune-related molecular markers that can predict the prognosis and immune status of HCC.

Methods: RNA-seq and clinical HCC data from the Cancer Genome Atlas (TCGA) were analyzed for differential expression of microRNA (miRNAs), mRNAs, and lncRNAs. MiRNAs associated with immune scores were identified by Spearman analysis and Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment. MiRNAs and mRNAs were screened for prognosticity using COX regression. Kaplan-Meier survival analysis, risk scores, and correlation with clinical features were performed. Immune infiltration, Tumor Immune Dysfunction and Exclusion (TIDE), and chemotherapy prediction analyses were performed for high and low risk groups. Finally, prognostic mRNA expression was validated in cell lines.

Results: Five prognostic miRNAs (hsa-miR-145-3p, hsa-miR-150-3p, hsa-miR-153-3p, hsa-miR-223-3p, hsa-miR-424-3p) were identified in the study. A risk score model based on these prognostic miRNAs accurately predicted overall survival and was validated in GSE31384. Six mRNAs (KCTD17, MAFG, RAB10, SFPQ, TRMT6, UBE2D2) were further identified as prognostic. A risk model including these mRNAs also accurately predicted overall survival, and higher risk scores were associated with lower survival. Univariate and multivariate Cox regression analyses confirmed that both miRNA and mRNA risk scores were independent prognostic factors. The TIDE results showed lower TIDE scores and T-cell exclusion scores in the low risk score group. Chemotherapeutic drug sensitivity analysis revealed that the high-risk group was more sensitive to multiple chemotherapeutic agents. In addition, real-time quantitative PCR (RT-qPCR) results of the cell lines supported the results of the public database analysis.

Conclusion: This study validated immune-related prognostic miRNAs and mRNAs and identified risk signatures for HCC, potentially advancing HCC prognosis and treatment.