|Developmental Epigenetics Unlocked - “A unique regulatory phase of DNA methylation in the early mammalian embryo”|
DNA methylation during mouse development is known to be a dynamic process; fertilization involves the removal of paternally contributed DNA, and this loss of DNA methylation continues until the blastocyst stage, when the inner cell mass (ICM) forms (Reik et al). After this stage, embryo implantation occurs alongside global re-methylation of the genome that is believed to contribute to lineage restriction and the loss of cellular potency (Kafri et al and Borgel et al). However, while this model is informative, no base-resolution maps covering stages of mouse development have been constructed to allow for its precise analysis. Now in a report published in Nature from the laboratory of Alexander Meissner at the Broad Institute of MIT and Harvard, USA, researchers have generated such maps, providing provide a genome-scale, base-resolution timeline of DNA methylation where they demonstrate the dynamic nature of this epigenetic mark, before it returns to an expected somatic pattern (Smith and Chan et al).
Early DNA methylation dynamics were analysed through creating global, high resolution maps of mouse cells at important developmental stages; in oocytes, sperm, zygotes, 2-, 4- and 8-cell cleavage stage embryos, the early ICM and embryonic day (E) 6.5/7.5 post-implantation embryos, using reduced representation bisulphite sequencing (RRBS)(Meissner et al). This technique was used due to its suitability for low cell numbers, and therefore low quantities of genomic DNA (0.5-10 ng), whilst giving good genomic coverage and high reproducibility. One caveat with this type of analysis is the indistinguishability of the cytosine methylation (5mC) and cytosine hydroxymethylation (5hmC) modification, which has been suggested to be an intermediate step in the active demethylation of DNA (Ficz et al, Wu et al, and Xu et al).
Examination of 100-base-pair tiles covering the genome found that oocytes were hypo-methylated relative to sperm, and that oocyte DNA methylation more closely resembled early embryonic time points than the methylation levels in sperm, post-implantation embryos or adult tissues. A further reduction in DNA methylation was also observed from oocytes to early ICM. Sperm and post-implantation embryo’s shared a characteristic of somatic cells; DNA methylation tended to occur in regions of low CpG density, a correlation which was weaker between oocyte and pre-implantation embryos. The two most dramatic observed changes in DNA methylation levels occurred at two key developmental time points: the sperm to zygote (reduction in methylation) and the early ICM to post-implantation embryo (re-methylation) transitions. Notably, zygotes displayed a decrease in paternal methylation in contrast to maternally contributed CpGs, which remained unmethylated, overall suggesting that the oocyte largely reflects the zygotic/pre-implantation methylome and prescribes its architecture. The authors note that the reduction in methylation levels at this stage could be due to several mechanisms; passive, replication-based methylation removal, active methylation removal or methylation removal through the coupling of base-excision repair and DNA replication (Wossidlo et al).
One of the major roles of DNA methylation is in the silencing of endogenous retrotransposon activity (Jaenisch & Bird). Regions which had the largest methylation changes between sperm and zygote were found to be enriched for long interspersed elements (LINEs), which showed a bimodal reduction, with 18% of LINEs reducing their methylation by over 45%, long terminal repeat (LTR) retroelements and short interspersed elements (SINEs), neither of which demonstrated a clear bimodality. LINEs with the largest decrease in DNA methylation consisted of two closely related families (L1Md_T and L1Md_Gf) and several LTR families showed considerable loss of methylation within the zygote, whereas the class II intracisternal A-particles (IAPs) did not. Interestingly all retrotransposons resolved identically, reaching minimal DNA methylation values at the ICM stage before increasing to more somatic levels by E6.5/7.5.
However, it is well understood that there are some regions, such as imprinted regions, which must maintain specific methylation patterns (Edwards and Ferguson-Smith) and the authors identified 376 oocyte-contributed differentially methylated regions (DMRs), mainly residing in CpG island-containing promoters, and 4,894 sperm-contributed DMRs, which were predominantly intergenic. Oocyte DMRs were associated with genes that are not expressed during later stages of oogenesis, such as Dnmt3b and the somatic isoform of Dnmt1 which suggests a possible regulatory function for some of these DMRs. In general, oocyte-contributed promoter DMRs retained intermediate methylation values from the zygote through to the ICM, followed by resolution to hypo-methylation in the specified embryo, while sperm-contributed DMRs also retain intermediate methylation values through to the ICM, before being hyper-methylated post-implantation to typical somatic levels.
Finally, non-CpG methylation was also studied and it was demonstrated that oocytes had the highest level of CpA (cytosine followed by adenine, rather than cytosine followed by a guanine as for CpG) methylation across pre-implantation development, and that this level decreased by approximately 50% in the zygote stage, indicating that non-CpG methylation is inherited as part of oocyte-contributed methylated alleles but is likely lost more rapidly.
Overall, the data provides a clear picture; pre-implantation development represents a unique developmental period where DNA methylation is dynamic before being restored and maintained in a somatic fashion. The oocyte, which is the major influence on DNA methylation in the early embryo, is globally hypo-methylated, particularly at specific families of LINEs and LTRs and contributes a unique set of differentially methylated regions, which, importantly, may mediate gene expression changes at this stage, whereafter the DMRs are lost. Lastly, this study provides a useful resource for those studying early development, in that it confirms previous data, has uncovered new insights and will be important for future studies in this area.
Borgel, J. et al.
Edwards, C. A. & Ferguson-Smith, A.
Ficz, G. et al.
Jaenisch, R. & Bird, A.
Kafri, T. et al.
Meissner, A. et al.
Reik, W., Dean, W. & Walter, J.
Smith, ZD. And Chan, MM. et al.
Wossidlo, M. et al.
Wu, H. et al.
Xu, Y. et al.
STEM CELLS correspondent Stuart P. Atkinson reports on those studies appearing in current journals that are destined to make an impact on stem cell research and clinical studies.
Original article from Nature.