Identifying biomarkers for arsenic exposure-induced nonalcoholic fatty liver disease: An integrated proteomics, metabolomics, and machine learning approach.

Arsenic, a toxic metalloid prevalent in the environment, poses significant health risks to humans primarily via the contamination of water. Arsenic exposure has been demonstrated to cause nonalcoholic fatty liver disease (NAFLD); however, the biomarkers used for the identification of its NAFLD risk remain inadequately explored. Based on our previous finding that arsenic induces NAFLD (characterized by hepatic lipid accumulation) in rats, this study aimed to identify potential biomarkers for the risk assessment of environmentally relevant levels of arsenic-induced NAFLD using an integrated multi-omics and machine learning analysis. The results indicated that the alterations in the peroxisome proliferator-activated receptor (PPAR) signaling pathway and purine metabolism in rat liver are the primary contributors to arsenic-induced NAFLD. Notably, adenosine monophosphate (AMP) may play a key role in the interaction between these two pathways. Through a leave-one-out cross-validation, the model composed of AMP, 13-hydroxyoctadecadienoic acid (13-HODE), and hippuric acid demonstrated a robust capability to discriminate between NAFLD and controls in both rats (AUC = 0.94) and human populations (AUC = 0.86). This study provides novel insights for developing of effective metabolic biomarkers to identify NAFLD associated with environmental arsenic exposure, which could be used for early prediction and intervention.