Integrative transcriptional and epigenomic modeling of xenobiotic-activated gene regulatory networks

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT The environmental contaminant 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) is one of the most potent and persistent toxicants known. TCDD and other dioxin-like chemicals are primarily ingested through food, and can cause various adverse effects ranging from immune suppression to hepatotoxicity and developmental alterations. Despite gradually decreasing environmental and body burdens, dioxin exposure remains of particular concern in utero, in breastfed infants, and specific populations reliant for food on locally caught fish and wildlife. The aryl hydrocarbon receptor (AHR), a ligand-inducible transcription factor, mediates virtually all of the toxic effects of dioxins. Nearly four decades after its discovery, the AHR remains an enigmatic molecule with a variety of endogenous roles in addition to its function as an environmental sensor. Our recent analysis of the AHR signaling network in mouse liver showed that direct AHR binding to cognate sequences in promoter regions of target genes explains only about 10% of TCDD-induced gene perturbations. In addition, it is unclear why humans are much less sensitive in responses to TCDD than rodents. These gaps in our knowledge make it difficult to estimate the risks of human exposure to dioxins. Our overarching hypothesis is that tissue- and species-specific alterations in gene expression induced by AHR activation, which in turn lead to dioxin toxicity, are determined by a combination of local chromatin accessibility, AHR binding in gene regulatory regions, and AHR-mediated long-range chromatin interactions. We propose to use an innovative combination of functional genomic experiments, computational modeling, and targeted epigenome editing (CRISPRi) to develop a predictive model for AHR-mediated tissue- and species-specific gene regulation, and to reconstruct the AHR transcriptional regulatory network in human vs. mouse liver and B cells. We will draw on both AHR-specific data generated in this project, and the NIH ENCODE and Roadmap Epigenomics projects, which have collectively made available 10,000+ genomic and epigenomic data sets from more than 400 human cell lines and tissue types. In Aim 1, we will compare genome-wide chromatin accessibility of mouse and human hepatocytes and B cells in the absence and presence of TCDD. In Aim 2, a novel predictive model for genome-wide AHR binding will be developed based on ChIP-Seq, chromatin accessibility and histone modification data. In Aim 3, we will identify and predict the differential expression of AHR target genes in mouse and human from AHR binding sites in regulatory DNA and AHR-mediated long-range interactions. The overall impact of our model will be improved mechanistic understanding of tissue-and species-specific gene regulation by the AHR in unprecedented genome-wide detail. Our long-term goal is to develop a genome- based quantitative framework for human risk assessment of chemicals that dysregulate core transcriptional regulatory pathways.