Early diagnosis of liver injuries caused by drugs or occupational exposures is necessary to enable effective treatments and prevent liver failure. Whereas histopathology remains the gold standard for assessing hepatotoxicity in animals, plasma aminotransferase levels are the primary measures for monitoring liver dysfunction in humans. In this study, using Sprague Dawley rats, we investigated whether integrated analyses of transcriptomic and metabolomic data with genome-scale metabolic models (GSMs) could identify early indicators of injury and provide new insights into the mechanisms of hepatotoxicity. We obtained concurrent measurements of gene-expression changes in the liver and kidneys, and expression changes along with metabolic profiles in the plasma and urine, from rats 5 or 10 h after exposing them to one of two classical hepatotoxicants, acetaminophen (2 g/kg) or bromobenzene (0.4 g/kg). Global multivariate analyses revealed that gene-expression changes in the liver and metabolic profiles in the plasma and urine of toxicant-treated animals differed from those of controls, even at time points much earlier than changes detected by conventional markers of liver injury. Furthermore, clustering analysis revealed that both the gene-expression changes in the liver and the metabolic profiles in the plasma induced by the two hepatotoxicants were highly correlated, indicating commonalities in the liver toxicity response. Systematic GSM-based analyses yielded metabolites associated with the mechanisms of toxicity and identified several lipid and amino acid metabolism pathways that were activated by both toxicants and those uniquely activated by each. Our findings suggest that several metabolite alterations, which are strongly associated with the mechanisms of toxicity and occur within injury-specific pathways (e.g., of bile acid and fatty acid metabolism), could be targeted and clinically assessed for their potential as early indicators of liver damage.
Keywords: Acetaminophen; Biomarker; Bromobenzene; Genome-scale models; Hepatotoxicants; Pathways.
Copyright © 2020 Elsevier B.V. All rights reserved.