Machine Learning, scRNA-Seq and Biological Experiments Identify Hub Genes Responsible for Ischemia-Reperfusion Injury in Steatotic Liver Allografts.

BACKGROUND: Liver transplantation is the standard treatment for end-stage liver disease, but donor organ shortage has increased the use of marginal donor livers (eg, steatotic grafts). These livers are highly sensitive to liver ischemia/reperfusion injury (LIRI), worsening transplant outcomes, with unclear molecular mechanisms limiting targeted therapies. METHODS: We adopted an integrative bioinformatics approach combining transcriptomic analysis of public datasets (GSE151648, GSE135251, GSE15480, GSE23649 and GSE193764), machine learning (LASSO, SVM-RFE), and scRNA-Seq analysis (GSE171539, CRA004061) to identify hub genes and pathways. Key findings were validated in vitro (AML12 hepatocytes under hypoxia/reoxygenation) and in vivo (murine model of steatotic liver under I/R). RESULTS: Weighted gene co-expression network analysis (WGCNA) identified LIRI-, NAFLD-, and NASH-associated modules enriched in immune/inflammatory pathways. Machine learning pinpointed distinct hub genes for NAFLD-related LIRI (JUN, CCL2) and NASH-related LIRI (PHLDA1, PNRC1, GADD45B, NFKBIA, JUND; all AUC > 0.7). CCL2 emerged as the top predictor via SHAP analysis. scRNA-seq revealed cell-type-specific expression patterns and distinct immune infiltration signatures between NAFLD and NASH cohorts, with remodeled cell-cell communication via CCL/CXCL pathways. In vitro/in vivo models confirmed the upregulation of key genes under LIRI, and CCR2 inhibition significantly attenuated liver injury, pathological damage, and pro-inflammatory cytokine (IL-1β, TNF-α, IL-6) expression. CONCLUSION: Our study identifies distinct molecular signatures and hub genes for LIRI in NAFLD versus NASH allografts, revealing differential immune landscapes between these two conditions. The findings further support the CCL2-CCR2 axis as a therapeutically targetable pathway driving LIRI in steatotic livers and provide potential biomarkers to improve marginal liver transplantation outcomes.