A wave of minor de novo DNA methylation initiates in mouse 8-cell embryos and co-regulates imprinted X- chromosome inactivation

DNA methylation is extensively reprogrammed during early stage of mammalian development and is essential for normal embryogenesis. It is well established that mouse embryos acquire genome-wide DNA methylation during implantation, referred to as de novo DNA methylation, from globally hypomethylated blastocysts. However, the fact that the main de novo DNA methyltransferase 3B (DNMT3B) is initially expressed as early as the 8-cell stage during preimplantation development, contradicts the current knowledge about timing of initiation of de novo DNA methylation. Here, we reported that a previously overlooked minor wave of de novo DNA methylation initially occurs during the transition from the 8-cell to blastocyst stage, before the well-known large-scale de novo DNA methylation during implantation. Bioinformatic and functional analyses indicated that minor de novo DNA methylation preferentially occurs on the X chromosome and co-regulates imprinted X-chromosome inactivation via the interaction between DNMT3B and polycomb repressive complexes 2 core components during blastocyst formation. Furthermore, minor de novo DNA methylation also finetunes proliferation, lineage differentiation and metabolic homeostasis of preimplantation embryos, and is critical for embryonic developmental potential and pregnancy outcomes. Thus, our study updates the current knowledge of embryonic de novo DNA methylation, thereby providing a novel insight of early embryonic epigenetic reprogramming. Summary statement A minor wave of de novo DNA methylation has been initiated prior to blastocyst formation, but not during the implantation period, and co-regulates imprinted X-chromosome inactivation.


INTRODUCTION
Mammalian early embryos are co-regulated by a series of spatiotemporally restricted epigenomic events that are critical for acquiring developmental potential.In mouse, it has been widely accepted that a wave of genome-wide de novo DNA methylation occurs around the time of implantation, thus re-establishing DNA methylation patterns from a globally hypomethylated blastocysts to relatively static DNA methylation patterns in post-implantation embryos (Borgel et al., 2010;Smith et al., 2012).DNA methyltransferase 3B (DNMT3B) is the main enzyme that catalyzes embryonic de novo DNA methylation.Genetic depletion of Dnmt3b results in hypomethylation and dysregulation of pluripotency genes, gastrulation genes, germline-specific genes, etc (Auclair et al., 2014;Borgel et al., 2010), thus causing severe developmental defects and lethality after implantation (Li et al., 1992;Okano et al., 1999).However, despite its developmental importance, the exact timing of initiation of de novo DNA methylation remains largely elusive.
Although earlier results using high-throughput sequencing methods suggested that de novo DNA methylation occurs during the transition from blastocysts to early post-implantation embryos (Auclair et al., 2014;Borgel et al., 2010;Guo et al., 2014;Smith et al., 2012), our detection assays (Figure 1A,B) and other previous studies (Guenatri et al., 2013;Li et al., 2016) indicated that Dnmt3b, the main de novo DNA methyltransferase, is initially expressed as early as the 8-cell stage during preimplantation development.This fact creates a contradiction that challenges current knowledge about timing of initiation of embryonic de novo DNA methylation, and therefore raising the possibility that embryonic de novo DNA methylation may precede, but not follows, the blastocyst stage.
To test this possibility, we reanalyzed well-established genome-scale bisulfite sequencing data that profiles DNA methylation maps from the mouse preimplantation to post-implantation embryos (Smith et al., 2012), and found approximately one-third of genes gained promoter DNA methylation during the transition from the 8-cell to blastocyst stage.It means that despite global DNA demethylation throughout the preimplantation development, a small proportion of regions have initiated de novo DNA methylation prior to the blastocyst stage, which we designated as minor de novo DNA methylation.We next ask its developmental functions for embryogenesis.Of interest, minor de novo DNA methylation is more prone to occur on the X chromosome.We next confirmed that minor de novo DNA methylation plays an important role in establishing imprinted X chromosome inactivation (iXCI) via the interaction between DNMT3B and the polycomb repressive complexes 2 (PRC2) core components, which are critical for X-linked heterochromatin.Further analyses indicated that minor de novo methylation also finetunes proliferation, lineage differentiation and metabolism of preimplantation embryos, and thus affecting the developmental potential and final pregnancy outcomes.Thus, our study provides a novel understanding of de novo DNA methylation during early development, which updates the current knowledge of embryonic epigenomic hallmark events and their interactions.

A small proportion of promoters initiate de novo DNA methylation by the 8-cell stage
To investigate the earliest initiation of de novo DNA methylation, we first examined the expression patterns of Dnmt3s in mouse preimplantation embryos.Results from previously published transcriptome (Fan et al., 2015) and our quantitative RT-PCR (qRT-PCR) detection showed that expression of Dnmt3b, the main enzyme for embryonic de novo DNA methylation (Borgel et al., 2010;Okano et al., 1999), as well as its non-catalytic partner Dnmt3l (Gowher et al., 2005), initiated by the 8-cell stage, and were further increased upon preimplantation development (Figure 1A).By contrast, the expression of Dnmt3a, which is mainly required for gametic methylation (Kaneda et al., 2004a), was maintained at low levels concurrently (Figure 1A).The initial expression of Dnmt3b was also confirmed at the protein level.Nuclear location of DNMT3B can be clearly detected from the 8-cell stage onward (Figure 1B).Given the essential role of DNMT3B in catalyzing de novo DNA methylation, we postulated that embryonic de novo DNA methylation may initiate as early as the 8-cell stage, which challenges current knowledge that de novo DNA methylation initiates during implantation.
To test the hypothesis, we next reanalyzed a well-established genome-scale bisulfite sequencing data that profiles DNA methylation maps from the mouse preimplantation to post-implantation stage (Smith et al., 2012).We first focused our analysis on DNA methylation dynamics of promoter, the preferential target of DNMT3B (Auclair et al., 2014;Borgel et al., 2010), from the 8-cell stage onward.Despite the global loss of promoter DNA methylation (Figure 1-figure supplement 1A), more detailed characterization showed that a considerable proportion of promoters (approximately 30%) gained DNA methylation during the transition from the 8-cell to blastocyst stage, even though most promoters underwent demethylation (Figure 1C).In line with this, a small proportion of gene bodies also gained DNA methylation during the transition (Figure 1D).This observation was also supported by reanalysis results (Figure 1-figure supplement 1B) of another published bisulfite sequencing data (Smith et al., 2017).By tracking DNA methylation dynamics of two subsets of promoters that gained or lost DNA methylation from the 8-cell to blastocyst stage respectively, we found promoters that underwent demethylation were relatively hypermethylated following fertilization, and then were gradually demethylated until reaching their lowest value in ICM (Figure 1E).In contrast, promoters that gained DNA methylation by the 8-cell to blastocyst stage, were more hypomethylated following fertilization, and reached their lowest DNA methylation levels at the 8-cell stage (Figure 1F).Given the hypomethylated regions are more accessible to de novo methyltransferases (Lorincz et al., 2002), this subset of promoters may be targeted by DNMT3B at the 8-cell stage and thus initiate their de novo DNA methylation before the blastocyst stage, and then continued during implantation (Figure 1F).As exemplified by the Yy2 promoter, the results of PCR-based bisulfite sequencing further confirmed the occurrence of de novo methylation before the blastocyst stage (Figure 1G).To distinguish two waves of de novo DNA methylation before and after the blastocyst stage, we designated them as minor and major de novo DNA methylation, respectively.We noticed that, besides of many important basic processes common to two waves of de novo DNA methylation, minor de novo DNA methylation may be linked to chromosome organization (Figure 1H).Compared with conventionally accepted major de novo DNA methylation, minor de novo DNA methylation showed a smaller increase in promoter DNA methylation that tended to occur in low CpG-containing promoters (Figure 1-figure supplement 1C,D).

Minor de novo DNA methylation preferentially occurs on the X chromosome and plays an important role in iXCI
To obtain a global view of minor de novo DNA methylation, we next detected the chromosome-wide distribution of de novo methylated promoters from the 8-cell to ICM, and found that they were globally distributed across autosomes, with the percentage ranging from 23% to 33% (Figure 2A-B).Of note, approximately half of X-linked promoters (48%) underwent minor de novo DNA methylation throughout the X chromosome (Figure 2-figure supplement 1A), much higher than those of autosomal promoters (Figure 2B).This result was similar with that of major de novo DNA methylation during implantation (Figure 2-figure supplement 1B,C).The X chromosome-biased prevalence of de novo DNA methylation is reminiscent of XCI, a female-specific epigenetic event that occurs during early development for balancing the X-linked gene dosage between males and females.In mice, early embryos undergo two waves of XCI: iXCI initiates in early preimplantation embryos and persists in extraembryonic tissues, and rXCI occurs in the embryonic cells by implantation stage (Augui et al., 2011;Chow and Heard, 2009;Lee and Bartolomei, 2013).Spatiotemporal co-occurrence of minor de novo DNA methylation and iXCI on the X chromosome, led us to ask if these two epigenetic events are functional linked during preimplantation development.
Indeed, it has been well-accepted DNA methylation is essential for maintaining X chromosome-wide transcriptional silence of rXCI during implantation (Blewitt et al., 2008;Sado et al., 2000), its role in iXCI has been documented as a long-standing open question of epigenetic programming during preimplantation development (Csankovszki et al., 2001;Kaneda et al., 2004b;Sado et al., 2000;Sado et al., 2004).Given identification of biological function of minor de novo DNA methylation requires that only minor, but not major de novo methylation, is functionally inactivated before blastocyst formation, we next transiently inhibited Dnmt3b via siRNA-mediated knockdown (KD).As expected, Dnmt3b mRNA and protein levels were significantly reduced in morulae, but not in blastocysts compared to those of the negative control (NC) group (Figure 2-figure supplement 2A-C).The transient inhibition of Dnmt3b resulted in a significant decrease in the global level of 5mC staining before blastocyst formation (Figure 2-figure supplement 2D,E).We next detected the H3K27me3, a classic marker for establishment of XCI achieving X chromosome-wide heterochromatinization of transcriptional depression (Chow and Heard, 2009;Huynh and Lee, 2005;Wutz, 2011).In line with the decreased DNA methylation levels (Figure 2-figure supplement 2D,E), we found all detected Dnmt3b-KD female blastocysts contained trophoblast cells lacking the H3K27me3 domain, with a variable frequency ranging from 10.7% to 42.4%; whereas nearly all NC female blastocysts showed a single H3K27me3 domain in each trophoblast (Figure 2C,D), indicating Dnmt3b deficiency leads to an impaired iXCI in preimplantation female embryos.Next, we further confirmed this result via Dnmt3b-knockout (KO) and chemical-induced inhibition of DNMT3B (Figure 2-figure supplement 2F,G).In addition, the involvement of minor de novo DNA methylation in iXCI establishment was further supported by the fact that many representative X-linked non-escaping genes tended to de novo methylated during the transition from the 8-cell to ICM (Figure 2-figure supplement 2H).Although previous studies have described that loss of maintenance methyltransferase Dnmt1, or de novo DNA methyltransferases Dnmt3a and Dnmt3b in the female germ line, had no effect on iXCI in the trophoblast (Kaneda et al., 2004b;Sado et al., 2000), authors also emphasized that the role of de novo methylation in iXCI, remains an open question, possibly due to heterozygous deletion of Dnmt3a/3b in the maternal germ line, but not in preimplantation embryos (Kaneda et al., 2004b).This is partially in line with another previous report showing a significantly higher expression of Pgk1, a representative X-linked non-escaping gene, in Dnmt3a/3b-deficient female embryos (Sado et al., 2004).Taken together, our data from multiple methodologies, provided direct evidence that a wave of minor de novo DNA methylation preceding the blastocyst stage, is critical for establishing iXCI.

Minor de novo DNA methylation co-regulates iXCI via the interaction between DNMT3B and PRC2 core components
Having confirmed the functional role of minor de novo methylation in iXCI, we next attempt to explore the underlying mechanism.Because iXCI is initiated by Xist RNA upregulation and coating during the 8-cell and morular stage (Penny et al., 1996), we first detected Xist mRNA expression.No significant difference in the degrees of absence of the Xist domain in each blastomere nucleus between NC and Dnmt3b-KD female morulae (Figure 3A,B), which were also confirmed by detecting Dnmt3b-KO female morulae (Figure 3-figure supplement 1).In line with this, single female embryo qRT-PCR analysis of Xist and its essential upstream activator Rnf12 during initial stage of iXCI, showed that the expression levels of Xist and Rnf12 were not affected by Dnmt3b deficiency (Figure 3C).Our results demonstrate that minor de novo DNA methylation participates in establishment, but not in initiation of iXCI.
In attempt to further understand the mechanism underlying the role of minor de novo DNA methylation in establishing iXCI, we next asked whether the Dnmt3b deficiency-induced loss of H3K27me3 domains was related to abnormal expression of histone-modifying enzymes.
We detected the expression of H3K27 histone methyltransferases or the partner (Ezh2, Suz12, Eed) and demethylase (Kdm6a, Kdm6b).The results showed Dnmt3b deficiency has no impact on the expression of these enzymes (Figure 3D).Given DNA methylation has a synergistic relationship with H3K27me3 in the process of heterochromatin formation (Hagarman et al., 2013;King et al., 2016), we reanalyzed H3K27me3 dynamics of promoters that underwent minor de novo DNA methylation during the transition from the 8-cell to ICM using a well-established H3K27me3 ChIP-seq data (Liu et al., 2016).As expected, concurrent increase in H3K27me3 enrichment can be clearly observed throughout the X chromosome and many de novo methylated X-linked promoters (Figure 3-figure supplement 2A,B).Next, we used in situ proximity ligation assays (PLA), a method that is suitable for visualizing protein interactions (Soderberg et al., 2006) to detect direct interactions between DNMT3B and PRC2 core components, i.e., EZH2, SUZ12, EED, which are essential for X-linked heterochromatin.It is clearly visible that DNMT3B-EZH2, DNMT3B-SUZ12 and DNMT3B-EED interactions can be detected in the nuclei by the blastocyst stage (Figure 3E).
In line with this, the function of de novo DNA methylation in XCI, as well as direct interaction between DNMT3B and PRC2 core components (Figure 3F,G; Figure 3-figure supplement 3A-G), were also confirmed in differentiated female ES cells.Collectively, our results support the idea that minor de novo DNA methylation plays an important role in iXCI via synergistic interaction with heterochromatic histone modifications, rather than transcriptional regulation of iXCI-regulating genes.

Minor de novo DNA methylation participates in embryonic proliferation, differentiation and affects developmental potential
It has been known that de novo DNA methylation during implantation stage plays an essential role in transcriptional regulation of developmental genes that are critical for embryonic differentiation and survival (Auclair et al., 2014;Borgel et al., 2010), thus we next attempted to test the additional short-and long-term developmental consequences of minor de novo DNA methylation.Thus, we performed an integrated analysis using the published bisulfite sequencing data (Smith et al., 2012) and transcriptome data (Wu et al., 2016)  Given identification of long-term developmental consequence of minor de novo DNA methylation requires that only minor, but not major de novo methylation, is functionally inactivated before blastocyst formation, we next used transient Dnmt3b-KD (Figure 2-figure supplement 2A-E) to examine whether minor de novo methylation is functionally associated with these basic biological processes before blastocyst formation and in turn affects developmental potential and pregnancy outcomes.Dnmt3b deficiency not only inhibited proliferation (Figure 4A), but also changed lineage differentiation of blastocysts (Figure 4B).
Next, we examined the energy metabolism of Dnmt3b-KD embryos.Despite the unchanged total ATP production, Dnmt3b deficiency led to a significantly disruption in oxidative phosphorylation (OXPHOS)-dependent-ATP production (Figure 4C).These observations were further confirmed via chemical-induced inactivation of de novo DNA methylation (Figure 4-figure supplement 2A-C).Finally, we transferred Dnmt3b-KD embryos into pseudopregnant recipient females.Compared with that in NC group, Dnmt3b-KD embryos resulted in a significantly reduced implantation rate and live birth rate (Figure 4D-E), suggesting the important role of minor de novo DNA methylation in influencing embryonic developmental potential.Since knockdown-mediated Dnmt3b deficiency has been restored to normal levels at the blastocyst stage, the possibility can be excluded that the reduced implantation rate and live birth rate may be due to defective major de novo DNA methylation.
Similarly, previous studies using Tet1-deficient or Tet1/3-deficient embryos, indicated the critical role of TET-mediated DNA demethylation in these processes, as well as embryonic and postnatal development (Dawlaty et al., 2011;Ito et al., 2010;Kang et al., 2015).Thus, it can be presumed that minor de novo DNA methylation, and TET-mediated demethylation, may synergistically co-regulate DNA methylation homeostasis before blastocyst formation, thus fine-tuning embryonic developmental potential and long-term outcomes.
It is also worth mentioning that minor de novo DNA methylation is definitely distinct from previously reported DNA re-methylation in monkey and human preimplantation embryos, which occurs during the 2-to 8-cell and 4-to 8-cell transition respectively (Gao et al., 2017;Guo et al., 2014;Zhu et al., 2018).The so-called DNA re-methylation does not occur in naturally conceived mouse embryos.However, an aberrant wave of DNA re-methylation has been identified in cloned mouse preimplantation embryos as an epigenetic barrier that impairs full-term developmental potential (Gao et al., 2018).
Collectively, our study presents an update on the current knowledge of embryonic de novo DNA methylation.We identified that a wave of minor de novo DNA methylation has initially occurred during the transition from the mouse 8-cell to blastocyst stage, but not during the implantation stage.Of interest, we provide the first evidence that DNA methylation is important for establishing iXCI in preimplantation female embryos.Our data, together with previous results, also support a model where minor de novo DNA methylation and demethylation co-orchestrate DNA methylation homeostasis before blastocyst formation, and serve as epigenetic mechanisms affecting embryonic developmental potential by fine-tuning the processes of lineage commitment, proliferation and metabolic homeostasis (Figure 4F).Thus, our work provides a novel insight for understanding epigenetic programming and reprogramming during early embryonic development.

Mouse embryo preparation
All experimental procedures were approved by and performed in accordance with the guidelines of the Institutional Animal Care and Use Committee of China Agricultural University.Mice were fed ad libitum and housed under controlled lighting (12 light:12 dark) and specific pathogen-free conditions.ICR female mice were superovulated with 5 IU pregnant mare serum gonadotrophin (PMSG; Ningbo Hormone Product Co., Ltd, Ningbo, China) and a further 5 IU human chorionic gonadotropin (hCG; Ningbo Hormone Product CO., Ltd) 46-48 h later.Then, superovulated ICR females were cogaged individually with ICR males.In vivo zygote, 2-cell embryos, 4-cell embryos, 8-cell embryos, morulae, or blastocysts were flushed from oviducts and uterus at 16 h, 42 h, 54 h, 66 h, 78 h, or 93 h post-hCG, respectively.

Quantitative real-time PCR analysis
Total RNA was extracted from embryos using TRIzol TM (Invitrogen, Carlsbad, CA, USA).
RNA was treated with DNase, and reverse-transcribed into cDNA using the HiScript II Q Select RT Kit (Vazyme, Nanjing, China).qRT-PCR was performed using SsoFast Eva Green Supermix (Bio-Rad, Hercules, CA, USA).The relative expression data were calculated using the 2 -△ Ct method.Relative expression level was normalized to the housekeeping gene H2afz, and each experiment was repeated at least three times, as previously reported.Primers used are listed in Table S1.
For qRT-PCR on single female or male embryo, a blastomere was isolated and treated with 2 μL Cells-to-cDNA™ II Cell Lysis Buffer (Invitrogen) at 75°C for 15 min.Next, the lysate was digested with proteinase K (Merck, Kenilworth, NJ, USA) at 55°C for 2 h, and used for sex determination according to the manufacturer's instruction of Ex Taq Hot Start Version (Takara Bio, Shiga, Japan).The amplification was conducted as follows: 95 °C for 3 min; 40 cycles of 95 °C for 10 s, 52 °C for 15 s, 72 °C for 15 s; and extension at 72 °C for 5 min.The remaining part of embryo was treated with 5 μL Cells-to-cDNA™ II Cell Lysis Buffer (Invitrogen), and then reverse-transcribed into cDNA with the HiScript II Q Select RT Kit (Vazyme).
For 5-methylcytosine (5-mC) staining, permeabilized embryos were treated with 4 M HCl for 10 min and 100 mM Tris-HCl for 10 min at room temperature, and then blocked overnight at 4 °C.Next, embryos were incubated with primary anti-5mC antibody (1:200; Active Motif, Carlsbad, CA, USA) at room temperature for 2 h, followed by secondary antibody for 1 h.Furthermore, embryos for sex determination were removed from the slides and digested with Cells-to-cDNA™ II Cell Lysis Buffer (Invitrogen) and proteinase K (Merck) as described above, followed by sex determination and genotyping using PCR.The primer sequences are listed in Table S1.

Bisulfite sequencing
The DNA methylation profile was analyzed by PCR-based bisulfite sequencing.40 IVO blastocysts or 80 8-cell embryos were digested and bisulfite converted according to the manufacturer's instruction of EZ DNA Methylation-Direct Kit (Zymo, Irvine, CA, USA).

Microinjection of siRNA
siRNA oligo sequences were synthesized by GenePharma (Shanghai, China).The sequences of Dnmt3b siRNA used in the present study were as follows: sense 5'-CCUCAAGACAAAUAGCUAUTT-3', antisense 5'-AUAGCUAUUUGUCUUGAGGTT-3'.siRNA was diluted to 100 μM and stored at -80 °C.For siRNA microinjection, 5 pL siRNA was microinjected into the cytoplasm of zygotes.Then embryos were cultured in KSOM at 37 °C under 5% CO 2 and collected at the 8-cell, morula or blastocyst stage for further analyses.

RNA-FISH
RNA-FISH was performed using ViewRNA™ ISH Cell Assay Kit (Invitrogen) according to the manufacturer's instructions.Fixed embryos were treated with Detergent Solution QC at room temperature for 5 min, and digested with protease QS at room temperature for 10 min.
Next, the embryos were hybridized with probes against Xist (Invitrogen) at 40 °C for 3 h.
Then the embryos were hybridized with PreAmplifier and Amplifier Mix at 40 °C for 30 min, respectively.Label Probe Mix was used to produce a signal.The embryos were counterstained with DAPI, and imaged using an attached digital microscope camera (DP72; Olympus).

Proximity ligation assays
The in situ proximity ligation assays (PLA) was performed using the Duolink® In Situ Red Starter Kit according to the manufacturer's instruction (MilliporeSigma).Embryos were fixed and permeabilized as described for immunofluorescence.After blocking, embryos were incubated with primary antibody.Embryos were then incubated with PLA probes, followed by ligation and amplification reaction.Finally, embryos were counterstained with Duolink® In Situ Mounting Medium with DAPI (MilliporeSigma).Furthermore, DNMT3B, EZH2, SUZ12, or EED single primary antibody, and secondary antibodies alone were used as technical negative controls.
Knockout of Dnmt3b or Dnmt3a for ES cells was performed using CRISPR/Cas9n.sgRNA was designed using Benchling ( https://www.benchling.com/ ).sgRNAs were constructed to pSpCas9n(BB)-2A-Puro vector (PX462, Addgene plasmid no.62987).4 μg of the plasmid was transfected using Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol.2 μg/ml puromycin (Solarbio, Beijing, China) was used to selected the positive colonies.After selection, colonies were picked out and expanded.Cells were verified by genotyping and Sanger sequencing.The sgRNA sequences and knockdown cell detection primers are listed in Table S1.

Co-immunoprecipitation
Co-immunoprecipitation (Co-IP) was performed with Dynabeads™ Protein G Immunoprecipitation Kit (Invitrogen) according to the manufacturer's protocol.Briefly, 3 μg DNMT3B antibody (GeneTex) or mouse IgG (MilliporeSigma) was incubated to Protein G beads with rotation for 10 minutes at room temperature.The beads were washed with Washing Buffer.Then cell lysates were then incubated with the beads-antibody complex with rotation for 4h at 4 °C.Then the beads were washed 3 times with Washing Buffer, and were eluted with loading buffer by heating at 98 °C for 10min.Co-IP products were analyzed by western blot using the indicated primary antibodies and secondary antibodies.And protein bands were detected with Tanon 5200 detection systems (Tanon, Shanghai, China).

Bioinformatics analysis
Gene ontology (GO) annotation analysis was performed using Database for Annotation, Visualization and Integrated Discovery (DAVID 6.8: https://david.ncifcrf.gov/).Heatmaps were plotted using the R software "pheatmap" package (version 1.0.12).The integrative genomics viewer (IGV) was used to visualize the ChIP-seq results.The phenotype of genes was analyzed on the Mouse Genome Informatics (MGI; http://www.informatics.jax.org/) database.

EdU assay
The proliferation of embryos was detected with a BeyoClick™ EdU-594 Cell Proliferation Kit (Beyotime, Shanghai, China) according to the manufacturer's instruction.Embryos were incubated in EmbryoMax® KSOM Mouse Embryo Media (MilliporeSigma) containing 10 μM EdU at 37 °C for 2 h, and then fixed and permeabilized.Afterwards, embryos were incubated with the Click Reaction mixture for 2 min at room temperature before stained with DAPI.Images were acquired under an upright microscope (BX51; Olympus) using an attached digital microscope camera (DP72; Olympus).

ATP measurement
ATP level was measured using an ATP determination kit according to the manufacturer's instruction (Beyotime).Briefly, blastocysts were incubated with M2 medium, and 1 μM oligomycin (Abmore, Houston, USA) or 1 mM sodium oxamate (MilliporeSigma) in M2 medium at 37 °C for 15 min, respectively.Next, embryos were digested with 20 μL ATP lysis buffer on ice, and the lysate was added to 96-well plates containing 100 μL ATP detection working dilution.The luminescence was detected by a multifunctional microplate reader (Infinite F200; TECAN, Mannedorf, Switzerland).

Embryo transfer
Pseudo-pregnant ICR females were mated with ICR males 3.5 d before embryo transfer.
Blastocysts derives from NC or Dnmt3b-KD group were transferred to the uterine horn of pseudo-pregnant female mice.

Statistical analysis
All data are presented as mean ± SD and analyzed with the student's t test or correct χ 2 procedure by using SPSS v.25.0 (IBM, Armonk, NY, USA).Values of P < 0.05 were considered statistically significant.(F) A model illustrating a wave of minor de novo DNA methylation that initiates during the transition from the mouse 8-cell to blastocyst stage.Minor de novo DNA methylation co-regulates iXCI via the interaction between DNMT3B and PRC2 core components, and fine-tunes the processes of lineage commitment, proliferation and metabolic homeostasis before blastocyst formation.Supplemental Tables (separate files) Supplemental Table S1.Primers sequences used in this study.
that profiles DNA methylation and gene expression maps from the 8-cell to ICM, respectively.Putative genes transcriptionally regulated by minor de novo DNA methylation (Figure 4-figure supplement 1A) were functionally related to many basic processes that were critical for normal embryonic survival, development and postnatal growth (Figure 4-figure supplement 1B,C), while genes subject to DNA demethylation also participate in similar processes (Figure 4-figure supplement 1D), suggesting the developmental significance of DNA methylation homeostasis.

Figure 1 .
Figure 1.A small proportion of promoters initiate de novo DNA methylation by 8-cell stage.(A) The mRNA expression levels of Dnmt3s from the 2-cell to blastocyst stage.The left ordinate represents FPKM values of Dnmt3s from published transcriptome data (Fan et al., 2015), and the right ordinate represents relative expression levels detected by qRT-PCR.(B) Immunofluorescent staining of DNMT3B in mouse preimplantation embryos.Scale bars: 50 µm.(C-D) DNA methylation changes of promoters (C) and gene bodies (D) during the transition from the 8-cell to blastocyst stage.Left panel: heat map of DNA methylation levels at promoters or gene bodies.Right panel: general trends of promoters or gene bodies undergoing de novo methylation or demethylation, and their relative contribution to differentially methylated regions.(E-F) DNA methylation dynamics of promoters subject to demethylation (E) or de novo methylation (F) during the transition from the 8-cell to blastocyst stage (red and green lines, average methylation levels; colored space, 10th/ 90th percentile).(G) Bisulfite sequencing analysis of Yy2 in the 8-cell embryos and blastocysts.Black and white circles represent methylated cytosines and unmethylated cytosines respectively.(H) A model illustrating two waves of de novo DNA methylation before and after the blastocyst stage, and their potential functions.

Figure 2 .
Figure 2. Minor de novo methylation preferentially occurs on the X chromosome and plays an important role in iXCI.(A) The chromosome-wide distribution of de novo methylated promoters with different fold changes (FC) during the transition from the 8-cell to blastocyst stage.(B) Percentage of de novo methylated promotors on each chromosome during the transition from the 8-cell to blastocyst stage.(C) Representative immunostaining for H3K27me3 (red) in the nuclei (DAPI) of NC and Dnmt3b-KD female blastocysts colabeled with CDX2 (green)-positive trophoblast cells.Scale bars: 50 µm.(D) The ratio of trophoblast cells classified by the number of H3K27me3 domains in each NC and Dnmt3b-KD female blastocysts.Each bar represents one female embryo.All data are presented as the mean ± SD of at least three independent experiments.**P < 0.01.ns, not significant.

Figure 3 .
Figure 3. Minor de novo DNA methylation co-regulates iXCI via the interaction between DNMT3B and PRC2 core components.(A) Representative localization of Xist expression

Figure 4 .
Figure 4. Minor de novo DNA methylation participates in embryonic proliferation, differentiation and affects developmental potential.(A) EdU staining of NC and Dnmt3b-KD blastocysts.Right panel: the percentage of EdU-positive cells.(B) The total cell number (left panel) and the ratio of ICM/TE (right panel) in NC and Dnmt3b-KD blastocysts.(C) The total ATP production (left panel) and OXPHOS-dependent ATP production (right panel) in NC and Dnmt3b-KD blastocysts.(D-E) Implantation rate (D) and live birth rate (E) in NC and Dnmt3b-KD groups.All data are presented as the mean ± SD of at least three independent experiments.*P < 0.05, **P < 0.01.ns, not significant.Scale bars: 50 µm (A, B).

Figure 1 -
Figure 1-figure supplement 1. (A) Dynamics of global methylation during preimplantation embryos (blue line, average methylation levels; colored space, 10th/ 90th percentile).(B) General trends of CpG islands undergoing de novo methylation or demethylation, and their relative contribution to differentially methylated regions.(C) Number of de novo methylated promoters with different fold changes (FC) at each consecutive transition.(D) DNA methylation dynamics of promoters with different CpG-densities during consecutive transitions.

Figure 2 -
Figure 2-figure supplement 1. (A) The distribution of de novo methylated and demethylated promoters cross the X chromosome during the transition from the 8-cell to blastocyst stage.(B) The chromosome-wide distribution of de novo methylated promoters with different fold changes (FC) during the transition from the blastocyst to post-implantation stage.(C) Percentage of de novo methylated promoters on each chromosome during the transition from the blastocyst to post-implantation stage.

Figure 2 -
Figure 2-figure supplement 2. (A) Relative expression levels of Dnmt3b in NC and Dnmt3b-KD morulae and blastocysts.(B-C) Immunofluorescent staining (C) and quantification (D) of DNMT3B in NC and Dnmt3b-KD morulae and blastocysts.(D-E) Immunofluorescent staining (D) and quantification (E) of 5-mC in NC and Dnmt3b-KD morulae and blastocysts.(F) The ratio of trophoblast cells classified by the number of H3K27me3 domains in each wild-type and Dnmt3b-KO female blastocysts.Each bar represents one female embryo.(G) The ratio of trophoblast cells classified by the number of H3K27me3 domains in female blastocysts exposed or not exposed to 5-aza-2'-deoxycytidine (5-aza-dC) from the 8-cell to blastocyst stage.Each bar represents one female embryo.(H) The fold changes of promoter methylation levels of X-linked non-escaping genes from the 8-cell to blastocyst stage.All data are presented as the mean ± SD of at least three independent experiments.*P < 0.05, **P < 0.01.ns, not significant.Scale bars: 50 µm (B, D).

Figure 3 -
Figure 3-figure supplement 1.The ratio of blastomere with Xist signal to the total number of blastomeres in each wild-type and Dnmt3b-KO female morulae.Each bar represents one female embryo.

Figure 3 -
Figure 3-figure supplement 2. (A) H3K27me3 enrichment cross the X chromosome from

Figure 3 -
Figure 3-figure supplement 3. (A) Schematic illustration of generation of Dnmt3b-KO female ES cells via the CRISPR/Cas9n system.(B) Detection of deletion introduced by sgRNA-Cas9n targeting Dnmt3b via PCR with genomic DNA from wild-type and Dnmt3b-KO female ES cells.(C) Relative expression levels of Dnmt3b in wild-type and Dnmt3b-KO female ES cells.(D) Schematic illustration of generation of DKO female ES cells via the CRISPR/Cas9n system.(E) Detection of deletion introduced by sgRNA-Cas9n targeting Dnmt3a/3b via PCR with genomic DNA from wild-type and DKO female ES cells.(F) Relative expression levels of Dnmt3a and Dnmt3b in wild-type and DKO female ES cells.(G) Representative immunostaining for H3K27me3 (red) in the nuclei (DAPI) of wild-type, Dnmt3b-KO, and DKO female ES cells differentiated for 5 or 9 days, as well as the ratio of ES cells classified by the number of H3K27me3 domains.All data are presented as the mean ± SD of at least three independent experiments.**P < 0.01.Scale bars: 20 µm.

Figure 4 -
Figure 4-figure supplement 1. (A) A scatter plots of changes in promoter DNA methylation and gene expression from the 8-cell embryos to ICM.Red plots indicate putative genes that undergo minor de novo DNA methylation (FC > 2) and are downregulated (FC < 0.5).(B-C) Gene ontology analysis (B) and phenotype annotation (C) of genes transcriptionally regulated by minor de novo DNA methylation.Blue squares indicate genes that are related to developmental or lethal phenotypes.(D) Gene ontology analysis of genes transcriptionally regulated by DNA demethylation from the 8-cell embryos to ICM.

Figure 4 -
Figure 4-figure supplement 2. (A) EdU staining of blastocysts exposed to 5-aza-dC from the 8-cell to blastocyst stage.Right panel: the percentage of EdU-positive cells.(B) The total cell number (left panel) and the ratio of ICM/TE (right panel) of blastocysts treated with or without 5-aza-dC.(C) The total ATP production (left panel) and OXPHOS-dependent ATP production (right panel) of blastocysts treated with or without 5-aza-dC.All data are presented as the mean ± SD of at least three independent experiments.*P < 0.05, **P < 0.01.ns, not significant.Scale bars: 50 µm (A, B).