Toxicogenomics analysis following in utero exposure to di-butyl-phthalate
M.A. Ovacik1, M.G. Ierapetritou1, B. Sen2, S. Euling2, P.G. Georgopoulos3, W. Welsh3, K.W. Gaido4, I.P. Androulakis1
1Rutgers University
2USEPA National Center for Environmental Assessment
3UMDNJ - Robert Wood Johnson Medical School
4The Hamner Institute for Health Research
Toxicogenomics emerges as a powerful tool in risk assessment. As such we explore the potential of using a toxicogenomics approach to assess potential implication of exposure to Di-butyl-phthalate (DBP). The major source of exposure to DBP, a commonly used plasticizer [1], is through crop cultivation and packaging materials with minor source being air and drinking water [2]. Transcriptional studies on several phthalate ester exposure models examined gene networks and pathways involved in testicular dysgenesis; in addition to the cholesterol transport and steroidogenesis pathway and intracellular lipid and cholesterol homeostasis, insulin signaling, transcriptional regulation and oxidative stress were identified to be significantly affected [3-5]. The main goal of this study is to explore potential modes of actions, pathways, and regulatory networks using toxicogenomics data and explore whether additional pathways are affected beyond testosterone biosynthesis and insl3 signaling. Previous transcriptional studies employed the most general approach; genes’ expression are ranked based on their metric for differential expression by using one of the various statistical methods available and determine the associated enriched pathways. Recently, the concept of pathway activity [6] was introduce in order to quantify the contribution of gene expression to the overall activity of any given metabolic/signaling pathway by considering the SVD decomposition of the matrix composed of all genes in a particular pathway. We demonstrate in this presentation that this decoupling, i.e., first identify discriminating genes, then evaluate enrichment and finally evaluate pathway activity, leads to erroneous results because it fails to account for the totality of genes participating in synch in a particular pathway and focuses only of differentially expressed genes. We introduce a variation of the method which evaluates the activity of a pathway by considering the ensemble of genes in a pathway and subsequently computed p-values are assigned to each pathway by performing random permutations of artificially constructed pathways. The active pathways are subsequently ranked in terms of both activity levels and p-values. In this study, we examined DBP exposure data using two animal models. These animal studies aimed at analyzing (i) a two-state (control vs. treated) and (ii) time dose [7]. For the two-state design pregnant rats were exposed for a certain period of time and simultaneously sacrificed of gestation day 19. Testes from male fetuses were analyzed to assessed possible testicular dysgenesis. The analysis determined a number of relevant and critical active pathways that have not been previously identified, such as Valine, Leucine and Isoleucine (VLI) degradation and Glutathione metabolism which take places in cholesterol biosynthesis and xenobiotics metabolism, respectively. For the time dose exposure, animals were exposed to DBP for different time and again all simultaneously sacrificed and testes tissues from fetuses again extracted and analyzed. The experiment was designed to assess the continuum of molecular pathways that are activated over different times of exposure. We discriminate between pathways, which are active at different exposure times and showed that there is a progression from signaling pathways towards metabolic pathways suggesting that DBP is interferes with the cellular system at the transcriptional level. We further identify critical metabolites and transcription regulators that connect the active pathways into a coherent and informative network of interacting processes. In comparison with standard methods for evaluating pathway activity we demonstrate that our modification reveals ore relevant processes. We furthermore examined phylogenetic conservation of the active pathways across mouse, rat and human to assess the potential from cross-species mode of action extrapolation, as well as assessing the conservation across species of the promoter regions of highly active genes in the dominant pathways. The construction of networks of interacting pathways, their degree of conservation across species and the degree of conservation of promoter regions of critical genes within these pathways lays the foundation for the construction of an integrated picture of the putative mode of action of DBP and points towards critical cross-species extrapolation elements.