It has long been known that methylation of genomic DNA correlates with gene expression. However, the structural mechanisms that underlie these observations remain obscure. In this project, we will pursue several innovative strategies for studying how methylation affects transcription factor (TF) binding. First, we will use Methyl-SELEX-seq ? a novel experimental method developed in the previous cycle of this grant that uses barcoded mixtures of methylated and unmethylated DNA ligands ? to create detailed maps of the effect of methylation on binding affinity for a broad panel of human transcription factors from various structural families. Second, we will perform detailed computational analyses and follow-up experiments to test the hypothesis that methylation causes local changes in DNA shape, which in turn modify TF binding affinity. We have shown that adding a methyl group in the major groove changes the geometry of the minor groove and enhances the electrostatic interaction between negative charges in the DNA minor groove and positively charged amino- acids in the TF. We will extend these analyses to other DNA modifications, as well as a wider range of DNA shape parameters and associated flexibility parameters. By building interpretable TF-DNA recognition models that integrate base, shape, and flexibility features using a powerful new machine learning framework developed in the previous funding cycle, we will make specific predictions regarding sequence and methylation readout mechanisms, and validate these using SELEX experiments with mutated TFs. To assess to what extent our quantitative models for binding to naked DNA built from SELEX data are predictive of binding to genomic DNA in the context of the living cell, we will perform detailed parallel analyses of SELEX and ChIP- seq data for Hox proteins and other TFs. Finally, to study the relationship between DNA binding and gene expression control in human cell lines, we will exploit Survey of Regulatory Elements (SuRE-seq), a novel massively parallel reporter assay that provides unique information about the autonomous transcriptional activity for each of >108 overlapping genomic fragments.
Epigenetic modifications of the human genome are known to play a key role in development and disease, but the underlying molecular mechanisms are largely unknown. The main goal of this proposal is to develop and apply new experimental and computational tools for analyzing how the binding of regulatory proteins to DNA is influenced by cytosine methylation. This will help us better understand its impact on gene expression and disease.
Showing the most recent 10 out of 55 publications
