A prognostic model based on DNA methylation-related gene expression for predicting overall survival in hepatocellular carcinoma

Hepatocellular carcinoma (HCC) continues to increase in morbidity and mortality among all types of cancer. DNA methylation, an important epigenetic modification, is associated with cancer occurrence and progression. The

Through bioinformatics analysis and experimental validation, we obtained three DNA methylation marker: GLS, MEX3B, and GNA14. This helps to predict the prognosis and may be a potential therapeutic target for HCC patients.

Transcriptomic, clinical, and DNA methylation data on 374 tumor tissues and 50 adjacent normal tissues were downloaded from The Cancer Genome Atlas-Liver Hepatocellular Carcinoma database. The gene expression profiles of the GSE54236 liver cancer dataset, which contains data on 161 liver tissue samples, were obtained from the Gene Expression Omnibus database. We analyzed the relationship between DNA methylation and gene expression levels after identifying the differentially methylated and expressed genes. Then, we developed and validated a risk score model based on the DNA methylation-driven genes. A tissue array consisting of 30 human hepatocellular carcinoma samples and adjacent normal tissues was used to assess the protein and mRNA expression levels of the marker genes by immunohistochemistry and qRT-PCR, respectively.

Three methylation-related differential genes were identified in our study: GLS, MEX3B, and GNA14. The results revealed that their DNA methylation levels were negatively correlated with local gene expression regulation. The gene methylation levels correlated strongly with the prognosis of patients with liver cancer. This was confirmed by qRT-PCR and immunohistochemical verification of the expression of these genes or proteins in tumors and adjacent tissues. These results revealed the relationship between the level of relevant gene methylation and the prognosis of patients with liver cancer as well as the underlying cellular and biological mechanisms. This allows our gene signature to provide more accurate and appropriate predictions for clinical applications.