Abstract
BACKGROUND: Aberrant DNA methylation plays a pivotal role in cancer progression by enhancing oncogene activation or silencing tumor suppressor genes, contributing to malignant phenotypes. Methylation driver genes (MDGs) are characterized by an inverse correlation between DNA methylation levels and mRNA expression, making them critical targets for cancer research. METHODS: We analyzed the liver hepatocellular carcinoma (LIHC) dataset from The Cancer Genome Atlas (TCGA) using the R package MethylMix to identify MDGs. Prognostic models were developed through univariate Cox regression, least absolute shrinkage and selection operator (LASSO) regression, and multivariate Cox regression to identify core genes. We further evaluated the associations of these genes with the tumor immune microenvironment, immune checkpoint inhibitors (ICIs), and chemotherapeutic sensitivity. Finally, liver cancer tissue organoid culture experiments combined with DNA methylation sequencing were conducted to validate predictions of drug sensitivity. RESULTS: A total of 21 MDGs were identified, among which GNA14, glutaminase (GLS), and GNG4 were selected to construct a prognostic risk score model. The model demonstrated robust predictive performance, with Receiver Operating Characteristic (ROC) values of 0.723, 0.764, and 0.716 for 1-, 3-, and 5-year survival, respectively. Among these, GLS emerged as a key gene, showing low methylation levels and high mRNA expression, which were associated with poor prognosis, significant alterations in the tumor immune microenvironment, and differential sensitivity to ICIs and chemotherapeutic agents. CONCLUSION: The three-gene MDG-based prognostic model effectively predicts survival outcomes in LIHC patients. Moreover, the methylation status of GLS serves as a biomarker for assessing immune microenvironment characteristics, responsiveness to immunotherapy, and chemotherapy sensitivity, highlighting its potential as a therapeutic target in liver cancer.