The role of non-coding regulatory elements and PRC2 proteins in disease and development

Abstract

Nucleosomes are the basic repeating units of chromatin. Not only do they provide a way to package the DNA into a tight nucleus, but they also contribute to the regulation of our genome. Composed of histone proteins, nucleosomes can be post-translationally modified by the addition of chemical groups on the histone tails. Specific histone modifications are enriched at different genomic regions, for example at gene promoters, intergenic chromatin, gene bodies, enhancers. The addition of specific chemical groups to histone tails ensures the correct transcriptional activation and prevents the aberrant expression of genes. In this thesis, I investigate mutations that occur at non-coding genomic sites marked by histone modifications and chromatin-binding proteins. In the context of breast cancer, I find a subset of non-coding regulatory elements that are significantly enriched in mutations. In particular, active promoters, active enhancers and insulators are the most frequently mutated regulatory elements of this cancer type. I also investigate coding mutations in the EZH2 protein that catalyses the deposition of H3K27 methylation, a histone modification associated with repression. In the context of Weaver syndrome, heterozygous mutations in the Ezh2 gene cause a global reduction in the H3K27me3 modification, in particular at intergenic chromatin. This research finds the first line of evidence that these mutations could dominant-negatively impair the wild-type EZH2 protein, representing a step towards understanding a mechanism that could underlie the aetiology of this disease. Lastly, I investigate the AEBP2 protein, which, similarly to EZH2, is involved in the deposition of H3K27 methylation within the Polycomb repressive complex 2. I provide evidence that the different isoforms of AEBP2 have distinct chromatin binding abilities and that the short isoform is important for the PRC2 subcomplex classification and early differentiation. In summary, I believe that these results advance our understanding of non-coding mutations in breast cancer, provide molecular and mechanistic insights for Weaver syndrome, and underline the importance of investigating alternative protein isoforms in development.