Canonical and non-canonical DNA methylome dynamics during vertebrate development

Abstract

Upon fertilisation of vertebrate embryos, the epigenomes of the responsible gametes need to be reconfigured into a state that is compatible with totipotency and zygotic transcriptional programs. Furthermore, the epigenomes of differentiating cells then need to be remodelled again in order to form the complex structures of the body, such as the vastly intricate nervous system. This includes, but is not limited to, the remodelling of DNA methylation, the most abundant DNA modification in vertebrates with critical roles in embryogenesis and neurodevelopment. In mammals, methylation of cytosines in cytosine-guanine dinucleotides (mCG) is almost completely erased after fertilization before it is re-established during gastrulation. Similarly, methylation of cytosines outside the CG context (mCH; H = A,T,C) is diluted in the early mammalian embryo before it is re-established mainly in the nervous system. However, in non-mammalian vertebrates, it appears that no global erasure of mCG takes place, raising questions about their propensity for transgenerational epigenetic inheritance. Additionally, the conservation of mCH in non-mammalian vertebrates is largely unexplored. In this thesis, I look to expand our knowledge on the developmental dynamics, evolutionary conservation and the molecular components of DNA methylome remodelling in vertebrates by studying methylome dynamics in two distantly related teleost species (ray-finned, protruding jawed fish). I functionally explore how DNA methylation is regulated during the development of zebrafish (Danio rerio), medaka (Oryzias latipes), and zebrafish-medaka hybrids, in both the CG and CH context. I employ CRISPR/cas9 technology, whole-genome bisulfite sequencing (WGBS), reduced representation bisulfite sequencing (RRBS), and RNA sequencing (RNA-seq), to interrogate a wide range of developmental time points and adult tissues. Overall, I have: i) developed a system to functionally test for regulators of developmental DNA methylation; ii) revealed a novel form of developmentally remodelled mCH in zebrafish and medaka which is deposited by the teleost specific DNMT3BA enzyme, iii) demonstrated evolutionary conservation of mammalian-like mCH features in the developing zebrafish nervous system, and iv) shown that DNA methylome dynamics in medaka and zebrafish embryos are highly comparable and compatible during the first 24 hours of zebrafish-medaka hybrid development. Altogether, this work greatly expands our understanding of the form and function of a critical DNA modification during development.