抄録
Systems biology provides a means to extend classical hypothesis-driven experiments from a focus on a small number of events to a more integrated understanding of the holistic response of a cell or tissue to any type of perturbation. The application of systems biology approaches requires global data sets such as those obtained from metabolomic or transcriptomic analyses. In this regard, transcriptional changes have been more widely evaluated than metabolomic profiles following toxicant exposure. However, metabolomic data ultimately reflect on the likely phenotypic consequence of altered gene expression, and represent important complementary data for determining the relevance of induction or repression of specific gene pathways. The integration of metabolomic and transcriptomic data can facilitate how we generate hypotheses, determine mechanisms of toxicity and translate the human relevance of data obtained from laboratory animals. Efforts to integrate these platforms indicate that each can inform the other, and collectively, they augment the identification of the major systemic effects of a xenobiotic or disease process. To this end, several examples will be provided of integrating data sets that lead to better definition of major biochemical events or toxic mechanisms. A very simple example is the urinary excretion of trimethylamine (TMA) and trimethylamine N-oxide (TMAO) which differs markedly between male and female mice. TMAO is a major metabolite in females but not in males, an outcome directly attributed to expression levels of flavin monooxygenases (FMO) that form this metabolite. As another example, integrating transcriptomic and metabolomic profiling from genetically-modified models is particularly helpful in establishing the phenotype of these models (usually mice). Organic anion transporting polypeptides (Oatp) are major hepatic anion uptake transporters, and Oatp1a1 null mice are viable, fertile, and appear to have minimal alterations in hepatic gene expression patterns. However, metabolomic data indicate significant perturbations in major urinary metabolites consistent with changes in resident intestinal microflora from wildtypes. Follow up studies confirmed substantive changes in constitutive abundance of Lactobacillus species in Oatp1a1 null mice, a phenotype that may contribute to alterations in bile acid homeostasis and sensitivity to toxicants in this model. Additional examples will illustrate how metabolomic data have been used to probe transcriptional changes (or vice versa) to integrate these platforms into holistic analyses of a toxic responses.
