Preliminary study to classify mechanisms of mitochondrial toxicity by in vitro metabolomics and bioinformatics.

AIM:Mitochondrial toxicity is one of the causes for drug-induced liver injury, and the classification of phenotypes or mitochondrial toxicity are highly required though there are no molecular-profiling approaches for classifying mitochondrial toxicity. Therefore, the aim of this study was to classify the mechanisms of mitochondrial toxicity by metabolic profiling in vitro and bioinformatics. MAIN METHODS:We applied an established gas chromatography tandem mass spectrometry-based metabolomics to human hepatoma grade 2 (HepG2) cells that were exposed to mitochondrial toxicants, whose mechanisms are different, such as rotenone (0.1 μM), carbonyl cyanide-3-chlorophenylhydrazone (CCCP, 0.5 μM), nefazodone (20 μM), perhexiline (6.25 μM), or digitonin (positive cytotoxic substance, 4 μM). These concentrations were determined by the Mitochondrial ToxGlo Assay. Galactose medium was used for suppressing the Warburg effect in HepG2 cells, and the metabolome analysis successfully identified 125 metabolites in HepG2 cells. Multivariate, metabolic pathway and network analyses were performed by the R software. KEY FINDINGS:Metabolic profiling enabled the classifying the mitochondrial toxicity mechanisms of RCC inhibition and uncoupling. The metabolic profiles of respiratory chain complex (RCC) inhibitors (rotenone and nefazodone) and an uncoupler (CCCP) were fully differentiated from those of other compounds. The metabolic pathway analysis revealed that the RCC inhibitors and the uncoupler mainly disrupted TCA-cycle and related metabolic pathways. In addition, the correlation-based network analysis revealed that succinic acid, β-alanine, and glutamic acid were potential metabolic indicators for RCC inhibition and uncoupling. SIGNIFICANCE:Our results provided new insights into classifying mechanisms of mitochondrial toxicity by in vitro metabolomics.