An evolutionarily conserved DNA-encoded logic shapes CpG island formation

DNA methylation is an epigenetic modification of the vertebrate genome that contributes to transcriptional repression, imprinting, and X-chromosome inactivation. While the majority of the genome is blanketed in DNA methylation, regions known as CpG islands (CGIs) remain remarkably refractory to this modification. CpG islands are associated with roughly two thirds of gene promoters, are evolutionarily conserved, and play central roles in gene regulation, yet how they are protected from DNA methylation remains enigmatic. Based on the conserved nature of CpG islands, we have exploited genomic approaches and a transchromosomic model system to ask if DNA sequence is sufficient to specify the hypomethylated state at CpG islands when a human chromosome is newly introduced into mouse. Interestingly, this approach revealed that promoter-associated CGIs remain immensely refractory to DNA methylation regardless of the host species, in fitting with their conservation across vertebrate species and revealing that DNA sequence is a central driver in this outcome. In contrast, the methylation state of distal elements is highly variable between species and is host nucleus dependent. These alterations in methylation state at distal elements are defined by DNA nucleotide frequency and occupancy of DNA binding transcription factors, uncovering a widespread role for these features in defining the how this aspect of the epigenome forms away from gene promoters. These central principles are further supported by transplantation of mouse DNA sequences into the evolutionarily distant zebrafish genome, revealing the existence of a highly conserved and DNA encoded logic that shapes the vertebrate epigenome. Overall design: Bio-CAP was used to identify hypomethylated regions of DNA in the transchromosomic mouse model Tc1, which contains an almost entire copy of human chromosome 21, and wildtype Tc0 mice in liver and testis tissues, and in zebrafish embryos in which mouse BAC DNA has been introduced into the genome.

DNA甲基化（DNA methylation）是脊椎动物基因组的一类表观遗传修饰，参与转录抑制、基因组印记与X染色体失活。尽管基因组的绝大多数区域都覆盖有DNA甲基化，但被称为CpG岛（CpG islands, CGIs）的区域却显著抵抗该修饰。CpG岛与约三分之二的基因启动子相关，在进化上高度保守，并在基因调控中发挥核心作用，但其如何免受DNA甲基化修饰的机制仍未明晰。
基于CpG岛的进化保守特性，我们借助基因组学研究手段与跨染色体模型系统，探究当人类染色体被全新引入小鼠体内后，DNA序列是否足以决定CpG岛的低甲基化状态。有趣的是，该研究发现，无论宿主物种为何，与启动子相关的CGI仍极度抵抗DNA甲基化，这与其在脊椎动物物种间的保守性相符，同时证实DNA序列是该现象的核心驱动因素。
与之相反，远端调控元件的甲基化状态在物种间差异极大，且依赖于宿主细胞核环境。远端元件的甲基化状态改变由DNA核苷酸频率与DNA结合转录因子的占据情况所决定，这揭示了这些特征在定义基因启动子以外的表观基因组形成模式中具有广泛作用。
上述核心原则还通过将小鼠DNA序列移植到进化距离较远的斑马鱼基因组中得到进一步验证，表明存在高度保守且由DNA编码的逻辑机制，用以塑造脊椎动物的表观基因组。
实验设计概述：本研究采用Bio-CAP技术，在携带几乎完整人类21号染色体的跨染色体小鼠模型Tc1、野生型Tc0小鼠的肝脏与睾丸组织，以及向基因组中导入了小鼠细菌人工染色体（Bacterial Artificial Chromosome, BAC）DNA的斑马鱼胚胎中，鉴定DNA的低甲基化区域。