Persistent transcriptional responses show the involvement of feed-forward control in a repeated dose toxicity study

抄録

One of the major challenges in developing in vitro cell models for predicting repeated dose toxicity in humans as an alternative to the 28 (90) days rodent models for assessing chronic toxicity relates to demonstrating that read outs from such in vitro models which are in general characterized by relatively short toxicant treatment protocols, are actually capable of expressing chronic adverse effects. Within this context, chemical-induced molecular changes which appear persistent after terminating toxicant treatment, may generate the desired biomarkers for predicting chronic toxicity. In particular, it has been suggested that toxicant-induced epigenomic alterations may cause such long-lasting perturbations of cell function. To test this hypothesis, we created a multi-donor primary human hepatocyte model which we challenged repeatedly across several days with prototypical agents known to induce several phenotypes of liver injury, followed by a 3 days wash out period. Epigenomic responses were monitored during and after toxicant treatment, by applying microarray or next generation sequencing methods. Downstream effects were analyzed by using multiple 'omics platforms comprising transcriptome, microRNA and proteome analyses. Results disclosed a multitude of novel gene-gene interactions caused by toxicant treatment, and also demonstrated to some degree, depending on liver toxicity class, that cross-omics changes actually persist upon terminating the toxic challenge, thus opening a novel venue for exploring biomarkers for predicting repeated dose toxicity in vitro.