Mechanism-informed identification of causal enhancer variants - PROJECT SUMMARY Enhancers are the genomic elements that encode the instructions for the timing, location and levels of gene expression. The majority of variants leading to disease are thought to reside in enhancers. However, we do not understand which changes in enhancer sequence are inert variants and which impact gene regulation and cellular integrity. Identifying the causal variants underlying these changes is critical for understanding the mechanisms driving the disease, development of novel treatments, better diagnosis and stratification of patients for different treatment strategies. It is therefore critical that we identify the causal variants underlying enhanceropathies. However, identifying which variants within an enhancer contribute to disease is a huge challenge. It is not experimentally or financially feasible to test all likely causal variants alone or in combinations in every disease and tissue. Therefore, we urgently need fundamental and generalizable principles to predict likely causal variants. Here we integrate our mechanistic understanding of enhancers into current approaches to predict causal variants in eQTL, GWAS, germline and somatic mutation datasets. We will then rigorously test our predictions to determine if this approach would enable systematic identification of causal enhancer variants. Enhancers control gene expression by binding transcription factors to specific sequences. My lab has previously shown that low affinity or suboptimal affinity binding sites are a prevalent feature of enhancers and that such sites are often overlooked. We have recently shown that the prevalent use of low affinity sites within enhancers creates a vulnerability in our genomes, whereby single nucleotide changes can increase binding affinity leading to aberrant gene expression and changes in molecular and organismal level phenotypes including extra digits and a second beating heart. We now wish to use our approaches to pinpoint causal variants that contribute to phentoypes within the context of disease and response to treatment. We will predict causal enhancer variants in two diverse systems: cardiac conduction traits and within melanoma metastasis and drug resistance. We will test our predictions using highly parallel functional reporter assays in relevant cell types along with genotype to phenotype studies. We will apply our approach to eQTL,GWAS, germline and somatic variants. In addition to prioritize causal varaints, and testing our predictions, we will provide our computational tools for the community to predict and prioritize causal enhancer variants within their diseases or tissues of interest. The successful completion of this project will move us closer to understanding the types of changes within enhancers that contribute to disease and treatment outcomes and could provide a generalizable systematic approach to identify causal variants that underly enhanceropathies.
