Mechanisms of epiallelic variation in mouse

Abstract

Transposable elements (TEs) dominate the eukaryotic genome landscape, accounting for half of the mammalian genome. Over evolutionary time, TEs are co-opted by the host and contribute to genome regulatory networks and chromatin architecture. TEs are also the target of elaborate epigenetic regulatory mechanisms by the host, with DNA methylation and repressive histone marks usually coinciding with TEs. By studying TEs, we gain new insights into mechanisms of epigenetic regulation and the evolution of genome function. While TEs generally exist in a repressive chromatin state, a subset of elements exhibits interindividual DNA methylation variation which impacts neighbouring gene expression. The best known example is the Agouti viable yellow mutant mouse, where variable DNA methylation at a TE insertion leads to coat colour differences among genetically identical mice. Such loci are termed “epialleles”. In mice, the intracisternal A-particle (IAP) subclass of TEs account for most variably methylated TEs. Over 50 constitutively variably methylated IAPs (VM-IAPs) have been identified in the C57BL/6J mouse strain. This thesis investigates the mechanisms driving interindividual methylation variation at TEs in the C57BL/6J mouse strain. First, I deconvolute a genetic model of abnormal folate metabolism. Initial results suggest that methylation is lost at VM-IAPs in mice with altered folate metabolism. However, further assessment of the model reveals that the mutant allele also introduces a cluster of 129P2/OlaHsd mouse strain-specific modifiers which may alter methylation at VM-IAPs. I perform a series of hybrid crosses and determine that while VM- TEs behave differently in inter-strain contexts, genetic effects alone do not explain the change in methylation observed in the abnormal folate metabolism model. Next, I assess the role of the multifunctional DNA binding protein CTCF at VM-IAPs. CTCF is enriched at VM-IAPs but not at IAPs in general. I postulate that CTCF antagonises the DNA methylation machinery during early development, thus driving variable methylation. I use CRISPR/Cas9 mutagenesis in mouse embryonic stem cells (mESCs) to discretely disrupt the CTCF binding site within a subset of VM-IAPs. These small mutations successfully lead to the loss of CTCF at the targeted VM-IAPs. Strikingly, in the absence of CTCF, the IAPs become hypermethylated. This suggests that molecular antagonism is a necessary force in the establishment and/or maintenance of variably methylated epialleles. The mutant mESCs are also used to generate transgenic mouse lines to study the in vivo effects of CTCF displacement. My work reinforces the role of genetic background as a key driver of epigenetic variability and introduces CTCF as a putative player in the establishment of variably DNA methylated epialleles. Altogether, I demonstrate how TEs can be used to shape our understanding of early epigenetic programming.