Identification of regulatory cell death-related genes during MASH progression using bioinformatics analysis and machine learning strategies.

Background: Metabolic dysfunction-associated steatohepatitis (MASH) is becoming increasingly prevalent. Regulated cell death (RCD) has emerged as a significant disease phenotype and may act as a marker for liver fibrosis. The present study aimed to investigate the regulation of RCD-related genes in MASH to elucidate the role of RCD in the progression of MASH. Methods: The gene expression profiles from the GSE130970 and GSE49541 datasets were retrieved from the Gene Expression Omnibus (GEO) database for analysis. A total of 101 combinations of 10 machine learning algorithms were employed to screen for characteristic RCD-related differentially expressed genes (DEGs) that reflect the progression of MASH. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were conducted to explore the enrichment pathways and functions of the feature genes. we performed cell classification analysis to investigate immune cell infiltration. Consensus cluster analysis was performed to identify MASH subtypes associated with RCD. The GSE89632 dataset was utilized to analyze the correlation of characteristic genes with clinical features of MASH. The DGIdb database was employed to screen for potential therapeutic drugs and compounds targeting the feature genes. In addition, we established mouse liver fibrosis models induced by methionine-choline-deficient (MCD) diet or CCl4 treatment, and further validated the expression of characteristic genes through quantitative real-time PCR (q-PCR). Lastly, we knocked down EPHA3 in LX2 cells to explore its effect on TGFb-induced activation of LX2 cells. Results: This study discovered a total of 11 RCD-associated DEGs, which predicted the progression of MASH. Advanced MASH has higher levels of immune cell infiltration and is significantly correlated with the RCD-related DEGs expression. MASH can be classified into two subtypes, cluster 1 and cluster 2, based on these feature genes. Compared with cluster 1, cluster 2 has highly expressed RCD-related DEGs, shows an increase in the degree of fibrosis. Furthermore, We discovered that the expression levels of feature genes were positively correlated with AST and ALT levels. Subsequently, We also evaluated the expression of these 11 feature genes in the liver tissues of mice with fibrosis induced by MCD or CCl4, and the results suggested that these genes may be involved in the development of fibrosis. WB results showed that the protein level of EPHA3 significantly increased in both mouse models of liver fibrosis. Conclusion: Our study sheds light on the fact that RCD contribute to the progression of MASH, high lighting potential therapeutic targets for treating this disease.