2,4-Dinitrotoluene (2,4DNT), which has been found to contaminate both soil, surface water and groundwater, is used in production of polyurethane foam, propellants and as a plasticizer in explosives . 2,4DNT has been found to be toxic to reproductive organs in rats  and cause genetic toxicity in munitions facility workers and copper miners using explosives [3, 4]. DNTs including 2,4DNT are listed as a priority pollutant by the U.S. Environmental Protection Agency . It is therefore important to develop methods to biomonitor people and animals exposed to nitrotoluenes to prevent such potential harmful effects.
Toxicogenomics using microarray technology has recently been applied to 2,4DNT and other military compounds to understand its molecular mechanisms[5–10]. While little information exists on the molecular pathways affected by 2,4DNT in mammals, mechanisms of 2,4DNT toxicity have been explored in the ecotoxicological fish model Pimephales promulas. Expression analysis of the effects of 2,4DNT on P. promulas indentified molecular pathways involving oxidation of hemoglobin and alteration of lipid metabolism via peroxisome proliferative activator receptor alpha signaling as being involved in 2,4DNT induced toxicity [8, 11].
In toxicogenomics, molecular expression pattern changes of cells occur as a result of exposure to toxicants and give insight into how toxicants act and cause disease. Nonetheless, many of these patterns are complex, interconnected, and reflect a dynamic process evolving from exposure to disease. Therefore, it is necessary to develop a systems approach to studying toxicogenomics by integrating ‘omics' measurements with public domain knowledge and interpreting them with a computational approach that encodes such a comprehensive biological context . Examples of such knowledge integration applied to systems biology include gene ontology analysis , protein-protein interaction network analysis , biological pathway analysis , and visualization of complex biological networks . However, few successful examples have been demonstrated in toxicogenomics.
In this work, we aim to study systems-level functional toxicogenomic changes induced by 2,4DNT exposure, using rat liver as an experimental model. We employ a new comprehensive pathway database resource, the Human Pathway Database (HPD) (Chowbina, Wu et al. 2009), which contains a comprehensive collection of human annotated and predicted pathway data. We mapped significantly-changed expression patterns of rat genes to human genes, and used HPD pathways to help us interpret complex interconnected gene expression patterns. We compared different dose effects of 2,4-DNT and the broad pathway-level changes the compound induced. To connect changes brought into different pathways, we further constructed a pathway- connecting-network (PCN) specific to differentially expressed genes induced by 2,4-DNT. This systems-level toxicogenomics study enables us to identify four major pathways implicated with 2,4-DNT induced cellular toxicity including long-term depression, breast cancer regulation by stathmin1, retinol metabolism, and WNT signaling. Genes significantly involved in these processes include PPP2R2B, PLCB1, CACNA1D, PTPRD, PTPRG, and RDH16.