Abstract
BACKGROUND: Metabolic dysfunction-associated steatohepatitis (MASH) remains a major clinical challenge due to the lack of effective diagnostic biomarkers and therapeutic targets. Identifying and validating key lipid metabolism-related genes may offer novel strategies for the early diagnosis and targeted treatment of MASH. METHODS: In this study, differentially expressed genes (DEGs) were identified from public databases using integrated bioinformatics approaches. Weighted gene co-expression network analysis (WGCNA) and multiple machine learning algorithms were employed to screen for hub genes closely associated with MASH. The expression levels and diagnostic potential of the candidate genes GPD1 and CEBPD were systematically evaluated through nomogram construction, immune infiltration analysis, and both in vivo and in vitro experiments. Their biological functions were further validated at the cellular level. RESULTS: The results revealed a strong association between lipid metabolism dysregulation and alterations in immune cell composition in MASH. GPD1 was significantly upregulated and CEBPD was downregulated in both the MASH animal and cell models, and both genes showed good diagnostic value. Functional experiments demonstrated that knockdown of GPD1 in HepG2 cells significantly reduced lipid accumulation, inflammatory responses, and expression of fibrosis-related markers. Similarly, overexpression of CEBPD also inhibited these pathological processes, indicating that both GPD1 and CEBPD play critical roles in MASH progression. CONCLUSION: This study highlights the importance of GPD1 and CEBPD as potential diagnostic biomarkers and therapeutic targets for MASH, providing a theoretical and experimental foundation for improving early diagnostic strategies and developing interventions targeting inflammation and lipid metabolism dysregulation in metabolic liver disease.