A metabolomic and multivariate statistical process to assess the effects of genotoxins in Saccharomycescerevisiae

Abstract

There is an increased need to develop robust cellular model systems which could replace or reduce the need for animals in toxicological testing. Current in vitro strategies for genotoxicity testing suffer from a high irrelevant positive rate, requiring the need for the development of new in vitro tools. Saccharomyces cerevisiae is used widely to study DNA damage and repair, and a high-throughput green fluorescent protein based assay has been developed to detect genotoxic-induced DNA damage. In this study a combined high resolution ¹H NMR spectroscopy and gas chromatography mass spectrometry based metabolomic approach has been used to monitor and distinguish different genotoxic compounds from other types of toxic lesion using the multivariate classification tool partial least squares-discriminate analysis (PLS-DA). The metabolic profiles of extracts of yeast (W303α strain) readily distinguished the individual toxins from control cells across 22 different treatments. In addition, these metabolic profiles also demonstrated dose and time responses for selected compounds (methyl methane sulfonate and nocodazole). Finally, predictive models were built for distinguishing the genotoxic carcinogens from the control group according to the metabolic profile of the cell culture media.