Long-term evolutionary DNA methylation dynamic of protein-coding genes and its underlying mechanism

DNA methylation is an important type of epigenetic modifications for the maintenance of genome functionality and stability. Although there are many studies on DNA methylation patterns, mechanisms, and functions, no study has focused on the evolutionary dynamic of DNA methylation. Here, we present the first genome-wide pattern of evolutionary DNA methylation dynamic in protein-coding genes, by grouping the Arabidopsis thaliana protein-coding genes into several conservation levels representing different evolutionary ages, and by investigating their DNA methylation features for three methylation contexts in both genic and flanking regions. The main results include: in a long-term evolutionary period, (1) genic CHG and CHH methylation levels tend to be decreased over time, which is mainly due to the reductions in the number of siRNA target sites in genes; (2) genic CG methylation levels are firstly reduced and then increased on average over evolutionary time, which is the interactional result of increased proportion and decreased CG methylation level of CG methylated genes; and (3) increased gene length and the stochastic methylation mechanism in CG context may further account for genic CG methylation trend in evolution. The diverse DNA methylation mechanisms in different contexts, together with altered gene length in evolution, could interpret the methylation dynamic of protein-coding genes over evolutionary time. This evolutionary perspective provides a dynamic understanding of the intrinsic relationship between DNA methylation and its functional and evolutionary effects on the genomes.