Identification of mitochondria-related biomarkers in liver fibrosis via interpretable machine learning and WGCNA: transcriptomic analysis and In Vivo validation.

BACKGROUND

Hepatic fibrosis is a key pathological stage in the progression of many chronic liver diseases; timely intervention is critical to preventing cirrhosis and hepatocellular carcinoma. Mitochondria regulate energy metabolism, lipid homeostasis, and redox balance, and their dysfunction is increasingly recognized as a driver of fibrogenesis.

OBJECTIVE

To identify key mitochondria-related genes associated with liver fibrosis and explore their mechanistic roles and therapeutic potential using a multi-omics mining strategy.

METHODS

Fibrosis-related bulk RNA-seq datasets (GSE152329, GSE167216, GSE119953, GSE254610) and scRNA-seq datasets (GSE145086, GSE233084) were retrieved from GEO. WGCNA-derived modules were intersected with DEGs and a mitochondrial gene set to obtain candidate genes. GO and KEGG enrichment analyses were performed with clusterProfiler, and TF activity was inferred with decoupleR. An XGBoost algorithm was utilized to prioritize critical mitochondrial targets. Cell-cell communication was analyzed using CellChat. A CCl4-induced C57BL/6 mouse model was established for HE/Masson staining, Western blotting, and TSA-IF. Crucially, functional validation was performed in human LX-2 hepatic stellate cells via ACOT9 knockdown to assess its regulatory role in fibrogenesis.

RESULTS

Bulk RNA-seq and WGCNA identified 38 mitochondria-related DEGs in CCl4-induced fibrosis. Machine learning prioritization highlighted Acot9, Aldh1b1, and Pck2 as key targets. scRNA-seq revealed specific expression patterns (predominantly Aldh1b1 in hepatocytes; Pck2 in cholangiocytes/HSCs; Acot9 in endothelial subsets), and CellChat analysis demonstrated remodeling of TGF-β and COLLAGEN signaling networks. In vivo, ACOT9, ALDH1B1, and PCK2 were upregulated in fibrotic liver tissue, consistent with transcriptomic changes. In the human cirrhosis dataset (GSE254610), ACOT9 was also significantly upregulated, confirming its relevance to advanced human disease. In vitro, silencing ACOT9 in human LX-2 cells significantly downregulated α-SMA, COL1A1, and TGF-β, indicating that ACOT9 functions as an upstream regulator of classical fibrotic markers.

CONCLUSIONS

Acot9 was identified as a key mitochondrial target associated with liver fibrosis. Its consistent upregulation in fibrotic liver tissue and, notably, the ACOT9-dependent modulation of fibrosis markers in hepatic stellate cells highlight its mechanistic relevance and potential as a therapeutic target for further study.