Metabolome analysis via an HPLC-ESI-MS-based experimental and computational pipeline for chronic nephron toxicity profiling

Untargeted metabolomics has the potential to improve the predictivity of in vitro toxicity models and therefore may aid the reduction of expensive and laborious animal models. Here we describe a chronic nephrotoxicity study conducted on a human renal proximal tubular epithelial cell line (RPTEC/TERT1) treated with low (10 µM) and high (35 µM) concentrations of chloroacetaldehyde (CAA) – a known nephrotoxic compound. The presented toxicity study followed two major strategies; the first was to establish an automated and easy to use untargeted metabolomics workflow for HPLC-MS data and second to find time- and dose dependent toxicant-induced cell responses at the metabolite level. The metabolomic changes were integrated with transcriptomics and proteomics data to obtain a more comprehensive picture of the mode of action. Automated data analysis workflows based on open-source software (OpenMS and KNIME) enable a comprehensive and reproducible analysis of the complex and voluminous metabolomics data produced by the profiling approach. For toxicity profiling of CAA, 428 and 317 metabolite features were detectable in positive and negative mode, respectively, after removal of chemical noise and unstable signals. Changes upon treatment were visually explored using principal component analysis and statistically significant differences identified using linear models (LIMMA). The analysis revealed toxic effects only for the high concentration treatment for day 3 and day 14. The most regulated metabolites were glutathione and metabolites related to the oxidative stress response of the cells. These findings are corroborated by proteomics and transcriptomics data, which show, amongst others, an activation of the Nrf2 pathway.