| Description |
Gene regulatory enhancers play key roles in development and disease. Enhancers vastly outnumber protein coding genes in the human genome, indicating that most genes are likely regulated by multiple enhancers. How multiple enhancers combine to produce a transcriptional response remains poorly understood, largely due to technical limitations. To determine the functional roles of enhancers individually and in combinations, they must be removed from the endogenous genomic locus, a low-throughput task. These genetic manipulations at a handful of loci have revealed that enhancers can collaborate to regulate gene expression. How exactly these collaborations are carried out in the context of the genome is unclear. Estrogen signaling provides an ideal model system to untangle relationships between enhancers and identify the molecular underpinnings of enhancer relationships. Estrogen receptor a (ER) is a ligand-activated transcription factor that binds to thousands of primarily distal enhancer regions and up-regulates hundreds of genes, suggesting that many genes are likely regulated by multiple enhancers. To identify functional relationships between enhancers, we first developed a CRISPR/Cas9-based tool enabling higher-throughput manipulation of enhancers, such that they could easily be dissected individually and in combination. We used this tool to dissect enhancer relationships at 3 genes up-regulated by estrogen in endometrial cancer cells and found that pairs of neighboring enhancers (within 5 kb of one another) tend to collaborate to produce the estrogen response of their respective targets. We identified two iv modes of enhancer collaboration: 1) synergistic, where two enhancers are equally necessary for an estradiol response, and 2) hierarchical, wherein one enhancer contributes the majority of the estradiol response and its neighbor can contribute only when the predominant site is active. We next investigated the molecular underpinnings of these relationships and found that ER-bound enhancers can affect their neighbor's ability to bind ER and maintain active chromatin. Further genome engineering revealed that highaffinity ER binding sites (ERBSs) are required for a transcriptional response to estrogen, but the genomic location of high-affinity ERBSs could also influence its contribution to the estrogen response. Our data suggest a model in which a pair of neighboring ERBSs must have a high-affinity ERBS and the presence of pre-existing histone acetylation in order for estrogenic gene regulation to occur. |
