Variation and Plasticity of DNA Methylation in Genome Regulation

DNA methylation is an important and well-studied epigenetic mark. It plays a pivotal role in imprinting, X-chromosome inactivation and genome stability and regulation in many normal development processes. DNA methylation, along with histone modifications, orchestrates the cell type- and developmental stage-specific chromatin landscapes and influences gene expression in vertebrates. In this thesis, I utilized the latest next-generation sequencing based methods to map DNA methylation levels at a genome-wide scale to investigate two areas of interest: the variation of DNA methylation across a large number of sample types and its functional potentials; the plasticity of DNA methylation in response to environmental stimuli including both the commensal gut microbiota and pathogenic Helicobactor pylori. In the first section of the thesis, I analyzed the largest collection of complete human DNA methylomes at single CpG resolution with 54 normal human samples representing 21 cell types to better understand the variation of DNA methylation at a genome-wide scale. I uncovered the general pattern of DNA methylation for normal somatic cells with a near-constant ratio of methylated and unmethylated CpGs, but the specific CpGs that are methylated or unmethylated can be dynamic and cell type-specific. I segmented the genome into regions with distinct DNA methylation signatures and focused the analysis on 22.6% of autosomal CpGs that can be variably methylated across cell types. These variably methylated regions (VMRs) are associated with enhancer chromatin states, and some have been validated as enhancers. They are also associated with transcription factor binding sites and GWAS variants enrichment sites. The evidence suggests a regulatory role of variable DNA methylation in modulating cell type specificity. In the second section of the thesis, I examined the plasticity of DNA methylation in response to environmental stimuli in two projects where I utilized H. pylori infection of a gastric cell line and the gnotobiotic mouse as two model systems. In both projects, commensal gut microbiota and pathogenic H, pylori were considered as microbial environmental factors, and I conducted experiments to collect samples with and without the influence of the environmental factor. I then applied two sequencing based assays, MeDIP-seq and MRE-seq, to profile the genome-wide DNA methylation and compared their patterns to identify genomic regions that show significant differences in DNA methylation. The overall DNA methylation patterns remain similar in both cases upon the impact of microbes. Focusing on local DNA methylation, for the mouse project, I identified hundreds of differentially methylated regions in the tissues examined and the colon is the site with the biggest difference. Some of these regions are associated with enhancer histone modification signatures, and genes near these regions are enriched for functions relevant for the tissue type. In the H. pylori project, I did not observe dramatic differences in DNA methylation between untreated and treated gastric cells, which might be due to insufficient infection time and conditions. In summary, I applied the latest high-throughput sequencing technologies for DNA methylation profiling to the study of the variation and plasticity of DNA methylation in cell type specificity and in response to environmental stimuli. These studies demonstrated the power of high-throughput epigenomic data integration in uncovering novel insights into the role of DNA methylation at unprecedented scales, and provided a foundation for additional studies on the role of DNA methylation in multicellular organism development and in genome-environment interaction