Dynamics of DNA methylation reprogramming at the single-cell level in early human embryos†

With boundaries in epigenetic mapping methods and technologies being constantly pushed further, major advances in our understanding of these regulatory mechanisms are being made. The investigation of DNA methylation has particularly benefited from these innovations, rendering genome-wide distribution of DNA methylation marks at a single-base resolution easily obtainable. DNA methylation plays a key role in cellular differentiation and development, and is mainly recognized for its involvement in processes such as transcriptional repression, genomic imprinting, and X-inactivation [1, 2]. DNA methylation marks are mediated by the combined action of mainly three DNA methyltransferases (DNMTs), the de novo enzymes DNMT3A and DNMT3B as well as the maintenance enzyme DNMT1. The absence of DNMT1 or DNMT3A/B during embryo development is lethal, as methylation profiles are not established or properly maintained. In somatic cells, tissue-specific DNA methylation patterns are relatively stable and inherited through cell divisions; however, during germ cell and early embryo development, these patterns are extensively dynamic. Efforts have been particularly devoted to defining how DNA methylation profiles are reprogrammed in mammalian cleavage stage embryos, as it is critical for normal and future development [3–6]. These studies have identified key features regarding the erasure process (i.e., demethylation) occurring after fertilization and the reestablishment of marks (i.e., remethylation) during peri-implantation. However, because of technical limitations, the precise dynamics and kinetics of DNA methylation profiles during this reprogramming wave remained elusive on a genome-wide scale of single blastomeres especially in human embryos. The emergence of single-cell approaches developed for whole-genome bisulfite sequencing has overcome the restrictions associated with the low cell number of early embryos [7, 8]. Using a post-bisulfite adaptor tagging strategy, a collaborative group led by Fuchou Tang, Jie Qiao, and Liying Yan recently established for the first time genome-wide and single-base resolution DNA methylomes of individual cells across the entire first week of human pregnancy [9]. They mapped DNA methylation profiles for 480 individual human preimplantation cells (i.e., paired male and female pronuclei from zygotes, 2-, 4- and 8-cell embryos, morula, blastocyst trophoblast (TE), and inner cell mass (ICM)), oocytes (i.e., germinal vesicle (GV), MII), and sperm cells. Their approach was innovative in many ways. First, compared to previous studies using pooled oocytes or embryos, the sequencing data obtained from a single oocyte or blastomere allowed them to detect aneuploid cells. Since they found that aneuploid cells contained abnormal levels of DNA methylation compared to euploid cells, they were excluded from future analyses to minimize disparity. Secondly, by sequencing the genome of each sperm donor they were able to determine the parental origin of tens of thousands of heterozygous CpG sites in the DNA methylome datasets for individual embryos and blastomeres produced by ICSI (intracytoplasmic sperm injection). The established DNA methylation roadmap revealed a much more complex reprogramming wave than the formerly presumed unidirectional demethylation process. In this landmark paper, Zhu et al. uncovered that the epigenetic reprogramming actually consists of three waves of demethylation interspersed between two remethylation waves with allele-specific variation in DNA methylation levels (Figure 1). The first wave of demethylation occurs independently of DNA replication within the first ∼12 h of fertilization and is especially directed towards CpGs located in enhancers and gene bodies. During this period, the oocyte-derived and sperm-derived pronucleus, respectively, lose ∼4% (54.5%–50.7%) and ∼30% (82.0%–52.9%) of their genomic methylation. A second wave of global demethylation is observed between the late zygote and the 4-cell stage embryo, followed by a third demethylation wave between the 8-cell embryo and ICM/TE of the blastocyst, settling genomic methylation levels to ∼24%. Here, CpGs located in introns and short interspersed nuclear elements were specially targeted. Interestingly, they found two notable and transitory waves of de novo methylation. The first one being a small but significant increase occurring in the male pronucleus of the zygote between the early to mid-pronuclei stages. The second one was a major surge ensuing from the 4-cell to 8-cell progression, similar to the one recently observed during the 2-cell to 8-cell transition in monkeys [5]. Although a very large amount of de novo methylated fragments (n = 73 000 of ≥300 bp in length) that were strongly enriched for major families of transposable elements were affected, most of these regions were rapidly demethylated in the following developmental stages. Subsequently, the authors established that a DNA methylation asymmetry in profiles exists between parental alleles following the first embryonic division until the blastocyst stage. The paternal derived genome is consistently and considerably less methylated than the maternal genome. Surprisingly, single-cell RNA sequencing identified very few genes with allele-specific expression even though asymmetry was clearly present in the ICM and TE cells of the blastocyst. Remarkably, this preferential hypermethylation of the maternal genome was further observed in both the embryonic and extra-embryonic derived lineages. Amongst other interesting findings, they found that single-cell DNA methylation patterns from single blastomeres of a 4-cell embryo could be linked with each parental cell from 2-cell embryos. Dynamics of DNA methylation reprogramming in the maternal and paternal genomes throughout early human embryogenesis. Adapted from Zhu et al. [9]. In this study, Zhu et al. provided clear evidence that the early human embryo reprogramming wave is a tightly regulated process that is even more dynamic than once believed. Curiously, this deep fundamental genome-wide demethylation phase is interrupted by short yet focused remethylation periods. However, it continues to be ambiguous how predetermine regions of the genome are targeted for demethylation or de novo methylation at specific embryonic stages and in an allele-specific manner. If de novo transposable elements targeted methylation potentially explains the maintenance of genome stability during the reprogramming process, the reasoning behind the methylation asymmetry between the maternal and paternal allele genomes and their long-term consequences on gene expression are still unclear. Future work derived from this epigenetic roadmap could determine using DNA methylation editing tools [10] if interfering with the kinetics or parental asymmetry of DNA methylation will lead to abnormal development. It will also be noteworthy to see how key differences between mammals in this epigenetic reprogramming wave contribute to species-specific particularities in early cell fate determination and development. We thank Maude Hoffmann-Belisle for figure design and Elizabeth Maurice-Elder for editing.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.