Whole-Genome ERalpha Binding in Mouse Uterine Tissue.

Abstract

The ovariectomized mouse uterus exhibits rapid biochemical and biological responses to estrogens that have been extensively studied and characterized. Using this model we previously evaluated transcript responses by microarray and have identified several thousand transcripts that are increased or decreased following estrogen treatment. To extend these findings we have mapped ERalpha binding sites in chromatin prepared from uterine tissue one hour after injection of saline or estradiol (E2) using chromatin immunoprecipitation (ChIP) with an anti ERalpha antibody followed by sequencing of enriched chromatin fragments using an Illumina system (ChIP-seq). More than 5000 ERalpha binding sites were apparent in the saline treated uterine chromatin, while more than 17000 were seen after the 1 hour E2 treatment. The ERalpha binding sites in E2 treated samples were mapped to almost 10000 Entrez genes while the ERalpha binding sites in the saline sample could be mapped to more than 5000 genes. The average number of sites per gene was greater in the E2 treated chromatin than in the saline treated chromatin, and almost all of the ERalpha bound genes in the saline sample overlap with those in the E2 samples, indicating that some sites are occupied by unliganded ERalpha, and that ERalpha binding is increased by E2. Other studies in MCF7 breast cancer cells have indicated that most ERalpha binding sites are not proximal to promoter regions but are primarily found in distal (>100 kb) or intronic regions. Similarly, in our analysis, fewer than 10% of the uterine ERalpha binding sites in the E2 chromatin were within 500 bp of transcription start sites and more that 80% of sites are within genes or more that 100 kb distal from mapped genes. A de novo motif analysis of sequences in the ERalpha bound chromatin confirmed that EREs were significantly enriched; in addition other transcription factor motifs were enriched as well. Interestingly, in areas of binding without predicted ERE motifs, homeodomain transcription factor binding sites were significantly enriched. To assess the relationship between ERalpha binding and E2 responsive gene regulation, the uterine ChIP-seq data was compared with data from microarrays of uterine RNA after treatment with saline or with E2 for 2 or 24 hours. About 20% of the genes with ERalpha binding sites in the ChIP-seq analysis also had E2 dependent changes in transcription after 2h, and after 24h more than 70% of genes with ERalpha binding encoded regulated transcripts. E2 regulation of uterine transcripts by tethered DNA binding-independent interaction of ERalpha was previously demonstrated using the KIKO mouse model, which has a DNA binding deficient ERalpha. Candidate ERalpha binding sites for several such transcripts were validated by ChIP-PCR of WT and KIKO chromatin. Overall, these findings have the potential to greatly enhance our understanding of mechanisms governing estrogen response in uterine tissue. (platform)