Comparative whole genome DNA methylation profiling across cattle tissues reveals global and tissue-specific methylation patterns
Yang Zhou,Shuli Liu,Yan Hu,Lingzhao Fang,Yahui Gao,Han Xia,Steven G. Schroeder,Benjamin D. Rosen,Erin E. Connor,Cong-jun Li,Ransom L. Baldwin,John B. Cole,Curtis P. Van Tassell,Liguo Yang,Li Ma,George E. Liu
DOI: https://doi.org/10.1186/s12915-020-00793-5
IF: 7.364
2020-07-06
BMC Biology
Abstract:Abstract Background Efforts to improve animal health, and understand genetic bases for production, may benefit from a comprehensive analysis of animal genomes and epigenomes. Although DNA methylation has been well studied in humans and other model species, its distribution patterns and regulatory impacts in cattle are still largely unknown. Here, we present the largest collection of cattle DNA methylation epigenomic data to date. Results Using Holstein cattle, we generated 29 whole genome bisulfite sequencing (WGBS) datasets for 16 tissues, 47 corresponding RNA-seq datasets, and 2 whole genome sequencing datasets. We did read mapping and DNA methylation calling based on two different cattle assemblies, demonstrating the high quality of the long-read-based assembly markedly improved DNA methylation results. We observed large differences across cattle tissues in the methylation patterns of global CpG sites, partially methylated domains (PMDs), hypomethylated regions (HMRs), CG islands (CGIs), and common repeats. We detected that each tissue had a distinct set of PMDs, which showed tissue-specific patterns. Similar to human PMD, cattle PMDs were often linked to a general decrease of gene expression and a decrease in active histone marks and related to long-range chromatin organizations, like topologically associated domains (TADs). We tested a classification of the HMRs based on their distributions relative to transcription start sites (TSSs) and detected tissue-specific TSS-HMRs and genes that showed strong tissue effects. When performing cross-species comparisons of paired genes (two opposite strand genes with their TSS located in the same HMR), we found out they were more consistently co-expressed among human, mouse, sheep, goat, yak, pig, and chicken, but showed lower consistent ratios in more divergent species. We further used these WGBS data to detect 50,023 experimentally supported CGIs across bovine tissues and found that they might function as a guard against C-to-T mutations for TSS-HMRs. Although common repeats were often heavily methylated, some young Bov-A2 repeats were hypomethylated in sperm and could affect the promoter structures by exposing potential transcription factor binding sites. Conclusions This study provides a comprehensive resource for bovine epigenomic research and enables new discoveries about DNA methylation and its role in complex traits.
biology