Deciphering early molecular responses to aristolactam I associated with hepatocellular carcinoma: Computational prediction of a core gene signature and identification of transcription-translation uncoupling under acute aristolactam I exposure.

Abstract

Aristolochic acid I (AAI) is a potent hepatocarcinogen mainly activated to aristolactam I (ALI). The molecular mechanisms driving ALI‑associated hepatocellular carcinoma (HCC), particularly post‑transcriptional events linking acute liver injury to malignant transformation, remain poorly defined. Here, we integrated pharmacological ALI-target prediction, HCC transcriptomic datasets, weighted gene co-expression network analysis (WGCNA), and SHAP‑based interpretable machine learning to establish a ten-gene signature for ALI-related HCC. Signature dysregulation was validated in a chronic AAI‑carbon tetrachloride (CCL₄) pre-neoplastic mouse model. In vitro phenotypic and molecular responses were investigated in ALI-exposed HepG2 and Huh7 cells, supplemented by molecular docking and 100 ns molecular dynamics simulations. Functional enrichment indicated that core signature genes participate in cell-cycle modulation, metabolic reprogramming, and epigenetic regulation. Consistent upregulation SAE1/AURKA and repressed MAT1A were observed in human HCC and murine pre-neoplastic liver tissues. Acute ALI exposure induced widespread transcription‑translation uncoupling with cell-line-specific patterns. In HepG2, cell-cycle genes were transcriptionally activated without protein elevation, while MAT1A protein increased despite stable mRNA levels. In Huh7, suppressed SAE1/AURKA transcription did not alter protein abundance, and MAT1A protein was reduced with unchanged mRNA expression. Simulations predicted stable ALI binding to SAE1/AURKA but a weak transient the ALI‑MAT1A interaction, explaining the divergent post-transcriptional responses. We propose a biphasic model of ALI hepatotoxicity. Acute ALI induces early post-transcriptional disturbances, whereas chronic AAI injury causes stable signature dysregulation during HCC progression. This signature provides candidate prognostic biomarkers and supports the safety evaluation of aristolochic acid‑containing herbal medicines.