Illuminating chromatin organization at nuclear landmarks using chemical genomics

Abstract

The three-dimensional conformation of chromatin is key to regulating gene expression and maintaining cell states. Protein cofactors are essential for the formation and maintenance of this DNA structure, but the ways in which nuclear proteins and DNA folding are interconnected are incompletely understood. Furthermore, the effects of epigenomic modifications, which regulate protein binding, on nuclear structure remain unclear. In this thesis, I describe the use of acute chemical and genetic perturbations coupled with new technologies to provide a deeper understanding of chromatin structure as it relates to the key architectural proteins Lamin B1 and CTCF.

In Chapter 1 I briefly review chromatin architecture, beginning with a basic overview of hierarchical chromatin structure and current models of how the genome folds. I then focus on three key nuclear landmarks and protein cofactors that are central to this thesis: the nuclear lamina, nuclear speckles, and CTCF. I end with an overview of how chemical biology tools have been essential for uncovering the causes and consequences of genome folding.

In Chapter 2, I describe the development of Lamina-Inducible Methylation and Hi-C (LIMe- Hi-C) to simultaneously map DNA’s conformation, methylation status, and localization relative to the nuclear lamina in a single workflow. Applying LIMe-Hi-C to study changes in DNA structure upon inhibition of DNMT1 or EZH2 reveals distinct changes in genome organization. While DNMT1 inhibition leads to largescale compartment mixing, EZH2 inhibition causes the movement of previously H3K27me3-marked B-compartment genomic regions to the nuclear periphery. This demonstrates a previously unknown antagonism between H3K27me3 and the nuclear lamina with implications for gene repression. I conclude this chapter with a brief vignette about the preliminary development of locus-specific LIMe-Hi-C variant workflows.

In Chapter 3, I describe the use of a selective DNMT1 inhibitor to study the looping patterns and functions of methylation-sensitive CTCF peaks. Treating cells with a DNMT1 inhibitor caused the enrichment of new CTCF peaks and loops on gene bodies, leading us to hypothesize that one function of gene body methylation is to protect against aberrant CTCF occupancy. We connect CTCF reactivation to gene expression by annotating a cluster of genes enriched for CTCF reoccupancy that turn on upon DNA demethylation in a CTCF-dependent manner. Moreover, we demonstrate that reactivated CTCF peaks often connect genes to stripe anchors in active regulatory regions, suggesting one way in which methylation-sensitive CTCF peaks could contribute to transcription. Finally, we observe that methylation-sensitive CTCF peaks and their highly-looping partners are located close to nuclear speckles, bodies of protein and RNA implicated in splicing and transcription. Using a first-in-class system to acutely ablate speckles, we note an immediate transcriptional response without any concomitant changes in CTCF binding or looping, thus contributing to the growing model that CTCF is robust to transcriptional perturbations.

In Chapter 4, I conclude with perspectives and future directions revalent to each project. Collectively, these studies illustrate how selective perturbations can help uncover new features that regulate genome folding. By integrating chemical genomics with methods to detect chromatin structure, we dissected the roles of Polycomb, DNA methylation, and nuclear speckles in DNA folding at multiple length scales and revised existing models of genome structure-function relationships.