Research on the Cholesterol Metabolism-Related Genes in the Progression of Metabolic Fatty Liver Disease

Abstract Objective Cholesterol metabolic disturbance accompanies metabolic fatty liver disease (MAFLD) progression. This study aimed to identify cholesterol metabolism-related core genes and explore their roles in MAFLD. Methods Four liver microarray datasets were analyzed, with GSE135251 as the training set and GSE48452, GSE89632, and GSE126848 as validation sets. ssGSEA quantified cholesterol metabolic activity. Differential expression and WGCNA identified candidate genes, and LASSO, SVM-RFE, and XGBoost screened core biomarkers. Candidate drugs were predicted via DSigDB and validated by molecular docking. A high-fat diet (HFD) mouse model was used for in vivo verification. Results MAFLD samples showed elevated cholesterol metabolic activity (P < 0.001). WGCNA and differential analysis yielded 274 candidate genes enriched in lipid/steroid metabolism, inflammation, and detoxification. From 93 PPI hub genes, the three algorithms identified INHBE, LGALS4, and PPP1R3C, with AUCs of 0.940, 0.921, and 0.939 in the training set and good diagnostic performance in validation sets. Co-Dergocrine Mesilate, Etoposide, and Trichostatin A stably bound these targets. In HFD mice, hepatic expression of all three genes was upregulated, with elevated cholesterol, triglycerides, and marked lipid accumulation. Conclusion INHBE, LGALS4, and PPP1R3C are potential diagnostic biomarkers and therapeutic targets for MAFLD, whose progression is driven by cholesterol disturbance-mediated lipid deposition and inflammation.