Abstract
The POU5F1 gene encodes one of the ‘core’ transcription factors necessary to establish and maintain pluripotency in mammals. Its function depends on its precise level of expression, so its transcription has to be tightly regulated. To date, few conserved functional elements have been identified in its 5′ regulatory region: a distal and a proximal enhancer, and a minimal promoter, epigenetic modifications of which interfere with POU5F1 expression and function in in vitro-derived cell lines. Also, its permanent inactivation in differentiated cells depends on de novo methylation of its promoter. However, little is known about the epigenetic regulation of POU5F1 expression in the embryo itself. We used the rabbit blastocyst as a model to analyze the methylation dynamics of the POU5F1 5′ upstream region, relative to its regulated expression in different compartments of the blastocyst over a 2-day period of development. We evidenced progressive methylation of the 5′ regulatory region and the first exon accompanying differentiation and the gradual repression of POU5F1. Methylation started in the early trophectoderm before complete transcriptional inactivation. Interestingly, the distal enhancer, which is known to be active in naïve pluripotent cells only, retained a very low level of methylation in primed pluripotent epiblasts and remained less methylated in differentiated compartments than the proximal enhancer. This detailed study identified CpGs with the greatest variations in methylation, as well as groups of CpGs showing a highly correlated behavior, during differentiation. Moreover, our findings evidenced few CpGs with very specific behavior during this period of development.
Introduction
The POU5F1 gene encodes one of the ‘core’ transcription factors necessary to maintain pluripotency in embryonic stem (ES) cells, alongside SOX2 and NANOG. For this reason, it has probably been one of the most widely studied transcription factors not only in pluripotent stem cells but also in embryos where its tight regulation is responsible for maintaining pluripotency in the inner cell mass (ICM) and for the first differentiation in the trophectoderm (TE) at the blastocyst stage (Niwa et al. 2005, Rizzino & Wuebben 2016). Interestingly, the level of POU5F1 expression needs to be finely tuned to maintain ES cell pluripotency because a two-fold reduction in its level of expression results in their differentiation into trophoblast cells, while a less than two-fold increase in its expression induces their differentiation into primitive endoderm and mesoderm (Niwa et al. 2000). Its transcriptional regulation has therefore been investigated in detail, which has enabled the identification of several canonical functional elements in the mouse 5′ upstream regulatory region, including a minimal promoter, a proximal enhancer first identified by (Okazawa et al. 1991) as a retinoic acid repressible enhancer, and a distal enhancer (DE) which is mostly active in the ICM cells, primordial germ cells and ES cells of the mouse blastocyst, but inactive in mouse post-implantation pluripotent epiblasts and in EpiSC, the primed pluripotent stem cells derived in vitro from epiblasts (Yeom et al. 1996, Tesar et al. 2007). The proximal enhancer contains two binding sites (PE1A and PE1B) for distinct factors present in P19 cells whatever their retinoic acid-induced state of differentiation (Minucci et al. 1996). The distal enhancer contains two binding sites for factors separated by 30 base pairs: the DE2A site first identified by Minucci et al. (1996) and the DE2B site, which binds POU5F1 and SOX2, both of which are necessary and sufficient for distal enhancer activation (Okumura-Nakanishi et al. 2005). Binding sites for other trans-activating factors have also been identified along the 5′ upstream sequence, such as a Sp1/Sp3-binding site and a hormone response element (HRE). Interspecies comparisons have demonstrated the conservation of these functional elements, most of them belonging to one of four conserved regions (CR1–4) in the 5′ upstream region of the POU5F1 gene, as evidenced by Nordhoff et al. (2001) when comparing human, mouse and bovine sequences, and confirmed by Kobolak (2009) in the dog and rabbit. Such conservation corroborates their functional role over mammalian species. How epigenetic modifications of the POU5F1 gene interact with this complex network of cis-sequences and trans-activating factors is now starting to be elucidated (for a review, see Shi & Jin 2010) mostly during studies performed in mouse stem cells, where the Pou5F1 enhancer/promoter region has been shown to be hypomethylated in ES cells but hypermethylated in trophoblast stem (TS) cells, this hypermethylation suppressing its activity (Hattori et al. 2004). Moreover, ablation of the de novo methyltransferases DNMT3A and DNMT3B results in hypomethylation of the Pou5f1 promoter and the overexpression of Pou5F1 in differentiating P19 carcinoma cells or ES cells and in the developing post-implantation mouse embryo (Li et al. 2007). Despite being considered as a master gene for early development (its lack of expression results in lethality around implantation) (Nichols et al. 1998), very few data are available concerning the precise epigenetic regulation of POU5F1 expression in the early embryo. This is mainly due to the scarcity of material available for epigenetic analyses in the preimplantation mammalian embryo. Recently, Herrmann et al. (2013) showed no differential histone post-translational modification affecting the proximal promoter of POU5F1 in the ICM and TE of bovine blastocysts, while in the post-implantation mouse embryo, (Feldman et al. 2006) evidenced the role of heterochromatinization in Pou5f1 silencing. This lack of precise data concerning the relationship between expression and epigenetic modification in early embryos is all the more surprising because POU5F1 is frequently used as a marker for normal vs altered development, and several studies have used the methylation of parts of its upstream region as a marker of ‘embryo quality’, particularly to track epigenetic alterations caused by the use of embryo biotechnologies (see for example Kawasumi et al. 2009, Al-Khtib et al. 2012, Zhao et al. 2012, Saenz-de-Juano et al. 2014). Interestingly, developmental delay in cloned mouse embryos has been shown to be associated with a high level of Pou5f1 promoter methylation due to defective epigenetic reprogramming of the somatic donor cell genome (Yamazaki et al. 2006), thus suggesting a crucial role for Pou5f1 promoter methylation during early development. In an attempt to remedy the lack of precise information concerning the epigenetic regulation of POU5F1 regulatory regions in normally developing early embryos, we decided to analyze the progressive modifications of DNA methylation that affect the POU5F1 upstream region in different compartments of the blastocyst during its development. We therefore used the rabbit embryo as a model for most mammalian blastocysts, where unlike the mouse the epiblast differentiates in a plane embryonic disk at the surface of the conceptus. The preimplantation rabbit embryo is also closer to most non-rodent mammalian embryos in terms of embryonic genome activation and related epigenetic regulations (Okamoto et al. 2011, Fischer et al. 2012). Concerning POU5F1 expression, rabbit embryo may also represent a good model for mammalian blastocysts: previous data showed a ubiquitous expression of POU5F1 protein in the early, expanded and hatched in vitro-produced rabbit blastocyst (Chen et al. 2012). Such a ubiquitous expression during blastocyst development from the early to the expanded stages was first evidenced in cattle and pigs by Kirchhof et al. (2000). Human blastocysts also contain the POU5F1 protein in both their ICM and TE (Cauffman et al. 2004). In mouse blastocysts, a ubiquitous distribution of POU5F1 protein has also been reported at the early blastocyst stage (Kirchhof et al. 2000, Szczepańska et al. 2011) but its restriction to ICM occurs at the expanded blastocyst stage, which is earlier than in non-murine species. Indeed, a more progressive restriction of POU5F1 to the ICM occurs between Day 7 and Day 9 in bovine blastocysts (Khan et al. 2012).
In this paper, we thus analyzed the POU5F1 expression and DNA methylation of four regions spanning the aforementioned functional elements and CR1-4 conserved regions in blastocysts over a 2-day period from the early expanded blastocyst stage (Day 4 blastocysts) to the Day 6 blastocyst stage, when the embryonic disk contains an epithelialized epiblast. We evidenced hypomethylation of the four regions in both the pluripotent ICM and epiblasts, together with differential DNA methylation between ICM and early TE as early as Day4, and a progressive increase in DNA methylation in differentiated embryonic layers displaying some specific features in particular regions or CpGs. In embryo analyses thus showed that DNA methylation in the 5′ regulatory region is highly dynamic and accompanies a progressive repression of POU5F1 expression.
Materials and methods
Ethical approval
All experiments were performed in accordance with the International Guiding Principles for Biomedical Research involving animals as promulgated by the Society for the Study of Reproduction and in accordance with the European Convention on Animal Experimentation. The animal experimental design was carried out under the approval of national ethic committee (APAFIS #2180-2015112615371038v2) and under the approval of the local ethic committee (Comethea n°45, registered under n° 12/107 and n°15/59).
Animals and embryo recovery
New Zealand White female rabbits (20–22 weeks old) were super-ovulated, and mating was natural. In vivo-developed embryos were collected from uteri perfused with PBS at 96, 120 and 144 h post coitum (hPC).
At 96 hpc, the mucin coat and the zona pellucida of rabbit blastocysts (96 h post coitum) were removed mechanically after brief exposure (2 min) to 5 mg/mL pronase (Sigma P-5147) at room temperature. After removing the zona pellucida, the blastocysts were cultured in TCM199 10% FBS (Sigma M4530) until they recovered a normal morphological aspect. ICMs were separated from the TE by immunosurgery: briefly, blastocysts were incubated in anti-rabbit whole goat serum (Sigma R-5131) at 37°C for 90 min washed thoroughly and then incubated (5 min) with guinea pig complement serum (Sigma S-1639). After washing in PBS serum, ICMs were mechanically dissociated from the TE by gently pipetting with a glass pipette. Samples were pooled by ten and immediately dry frozen.
At 120 hpc, rabbit blastocysts were recovered. The embryos were then carefully hemi-sectioned using a microscalpel under a binocular microscope. The ‘embryoblast’ containing parts were separated from TE parts. Samples were pooled by ten and immediately dry frozen.
At 144 hpc, the embryos were dissected by microsurgery to obtain an epiblast, hypoblast and trophoblast. Expanded blastocysts were collected 144 h after artificial insemination and placed in FHM medium (Millipore). The zona pellucida was removed mechanically. The embryos were opened and flattened on a plastic dish to expose the embryoblast. The hypoblast was first dissociated by careful scratching with a glass needle, and the epiblast was then separated from the trophoblast with a microscalpel (Idkowiak et al. 2004). A piece of trophoblast was then dissected in the extraembryonic part of the conceptus.
In vitro-developed embryos were obtained from zygotes recovered at 19 hPC and cultured in 500 µL 199 medium plus 10% (V/V) fetal calf serum (FCS Gibco) for 53 and 77 or 79 h at 38.5°C in 5% CO2 in air to recover morula and blastocysts respectively.
DNA methylation analysis
DNA extraction and bisulfite conversion were performed using the Epitect Plus DNA Bisulfite Kit (Qiagen) following the manufacturer’s instructions. The four regions spanning the four conserved regions of the POU5F1 regulatory region defined by Kobolak (2009) were then subjected to nested PCR. The PCR products were loaded in a 1.5% agarose gel, extracted and purified using the QIAquick Gel Extraction Kit (Qiagen). Purified PCR fragments were then cloned into a PGMT-Easy vector (Invitrogen). Competent DH5α bacteria (Invitrogen) were transformed by the ligation product and grown on LB agar plates overnight. For each transformation, 10–15 clones were selected at random. Integration of the PCR fragments into plasmids was checked by PCR, and only positive clones were subjected to sequencing by Beckman Coulter Genomics. The sequences were aligned to the reference sequences and the quality control of DNA methylation data was ensured using BiQ Analyzer software (available online at http://biq-analyzer.bioinf.mpi-inf.mpg.de/). Efficiency of bisulfite conversion of unmethylated cytosine into thymine bases was about 97%. We took advantage of the sequence polymorphisms generated by the few random non-converted cytosines to select and analyze PCR-amplified fragments issued from different original alleles.
The sequences and the annealing temperature of primers specific to bisulfite-converted DNA were as follows:
Region 1–56°C
external forward: 5′-GTTTTTTTAGGGAGGGGGTAGAG-3′,
external reverse: 5′-AAAACCTTAAAAACTCAACCAAATCC-3′,
internal forward: 5′-ATGGGGTGGAAGGGATTTTAG-3′,
internal reverse: 5′-AAAATCCACCCAACCTAACTCC-3′;
Region 2–58°C
external forward: 5′-TTTAGAGGAAGAGGGAGTTGGATATTTAG-3′,
external reverse: 5′ CCCTAACTCTCCAAAAACTCCCA 3′
internal forward: 5′-TTTAGAGGAAGAGGGAGTTGGATATTTAG-3′,
internal reverse: 5′-CACTTCTACAACCCAAACCTCCA-3′;
Region 3–58°C
external forward: 5′-AGGGTTTGGGTTTTGGTTTTTTAA-3′
external reverse: 5′- CTAAATATCCAACTCCCTCTTCCTCTAAA-3′,
internal forward: 5′-TTTTAAGTTGTGGGGAGTTGTGG-3′,
internal reverse: 5′-CTCCCTCTTCCTCTAAAAAAAATCAAA-3′;
Region 4–51°C
external forward: 5′-GTTGGTTGGGTAGGAGTTTAT-3′,
external reverse: 5′-TAACCCTATCAAACTTCTAAAAAACT-3′,
internal forward: 5′-ATAAGTTAAAGAGTTTTGTTTTTGG-3′,
internal reverse: 5′-AACTTCTAAAAAACTAAATAACCTAACTCT-3′;
Allele methylation rates were analyzed using ANOVA and Fisher’s exact test. Differences were considered significant when P < 0.05.
ACP and heatmap analyses. CpG 26 and 60 were excluded from the principal component analysis (PCA) and heatmap correlation studies because their methylation status was too poorly documented. Before performing ACP on the methylation level of each CpG in each compartment at each stage, it was necessary to infer some missing data. Considering a given CpG and a given tissue, if we had at least eight methylation values available, then the proportion of methylation could be computed directly as the number of methylated alleles divided by the total number of methylation values available. Otherwise, we imputed the methylation proportion as the average proportion of methylation among all tissues/stages for this CpG so that poorly documented CpG would not bias the analyses.
Heatmap correlations were generated using the pheatmap R package (distance function: Pearson correlation coefficient/linkage method: Ward). The PCA results were produced using the FactoMineR R package (Lê et al. 2008).
LD-like plot: Methylation values at individual CpGs were binary (i.e. 0 = unmethylated, 1 = methylated). Therefore, the phi coefficient of correlation was used to assess the methylation correlation between pairs of CpGs in all samples (Warrens 2008). The phi function of the psych package in R was used to calculate the phi coefficient for each pair of CpGs.
Real-time RT-PCR
At Day 4, three biological replicates of 35 ICMs and 35 TEs were used. At Day 5, we used two biological replicates of ten embryoblasts (EBs) and ten trophoblasts (TBs) each. At Day 6, three biological replicates of ten epiblasts (EpB) 10 hypoblasts (HB) and ten TBs were used. Total RNA was extracted and DNase I treatment was performed simultaneously using PicoPur Arcturus (Excilone) and DNase I (Qiagen), according to the manufacturer’s instructions. RNA integrity was verified by a Bioanalyzer (Agilent) with a RIN >8.5 for all samples. Total RNAs were quantified by spectrophotometry based on A260 values. Reverse transcription was performed on 30 ng of total RNA using Superscript III enzyme and random hexamers (300 ng) (Life Technologies SAS), following the manufacturer’s instructions (25°C for 5 min, 50°C for 60 min, and 70°C for 15 min). Primers for POU5F1 had been published by Kobolak (2009), and the primers for the reference genes were as follows:
GAPDH F: CACGGTCAAGGCTGAGAACG (200 nM)
GAPDH R: GGATTCCACCACGTACTCGG (200 nM)
YWHAZ F: GGTCTGGCCCTTAACTTCTCTGTGTTCTA (200 nM)
YWHAZ R: GCGTGCTGTCTTTGTATGATTCTTCACTT (200 nM).
All primers were used at 200 nM final concentration. The thermal cycle profile started with a 10-min step at 95°C followed by 40 cycles consisting of a 15-s denaturation step at 95°C, and a 60-s annealing/extension step at 60°C, except for Ywhaz and Hprt1 for which the annealing temperature was 68°C. Dissociation curves were obtained after each PCR run to ensure that a single PCR product had been amplified.
The qPCR reactions were carried out using a Step One Plus machine (Applied Biosystems) in a 25 µL volume containing 12.5 µL Sybr Green PCR Master Mix (Applied Biosystems), primers and 10 µL cDNA from a 1:40 dilution of the RT product used in PCR. For each qPCR, the efficiencies were determined by serial dilutions of pre-amplified embryonic cDNA (Peynot et al. 2015) or iPS cell cDNA (Osteil et al. 2013) (kindly provided by Pierre Savatier, Inserm-SBRI, France). Each sample was quantified in duplicate. Data were analyzed using QbasePLUS software (Biogazelle, Gent, Belgium). To calculate the normalized relative quantity (NRQ), specific target and run amplification efficiencies were taken into account. Importantly, because it is difficult to find stable reference genes to compare embryonic tissues, we added several samples to each PCR run. In particular, whole Day 4, 5 and 6 embryos and mixes obtained with equal amounts of Day 6 EpB, HB and TB cDNA were included. This allowed us to define reference genes with satisfactory M and V values (GeNorm). Each NRQ value was divided by a run- and gene-specific calibration factor to determine the calibrated NRQ (CNRQ).
Statistical analysis of the RT-qPCR data was performed using non-parametric tests under R software (R Development Core Team 2010): the K-Sample-Fisher Pitman Permutation test from the ‘coin’ package and non-parametric relative contrast effect (nparcomp) from the ‘nparcomp’ package. P values of <0.05 were considered to be significant.
Immunofluorescence
Embryos were fixed in 4% paraformaldehyde (PFA, Sigma) in PBS overnight at 4°C. During immunostaining, all steps were performed at room temperature, unless otherwise specified. The embryos were first washed twice in PBS for 10 min and permeabilized with 1% Triton X-100 (Sigma) in PBS for 1 h. They were then placed in preheated citrate buffer (0.01 M citric acid; 0.01 M sodium citrate; pH 6.3) for 10 min on a heating plate at 80°C. The embryos were then rinsed twice in PBS for 5 min and incubated for 1 h in PBS containing 5% serum and then with a goat primary antibody (Santa Cruz reference sc-8628) diluted (1/50) in PBS-5% serum for 2 h. After incubation, they were washed three times in PBS for a total of 30 min and incubated for 45 min with an FITC-conjugated anti-goat antibody (Interchim Jackson ref. 305-095-003) diluted (1/200) in PBS-5% serum. After washing three times in PBS for 10 min, the embryos were counterstained for a few minutes at room temperature in 2 µg/mL DAPI (Invitrogen). The embryos were finally mounted on slides in Citifluor (AF1, BioValley) and observed using an ApoTome microscope (Zeiss).
In situ hybridization
Embryos were fixed in 4% PFA overnight at 4°C, dehydrated and stored in 100% methanol at −20°C until further processing. The samples were then rehydrated and processed as described by Püschel and Jouneau (2015). Briefly, hybridization was performed with Dig-labeled riboprobes. After incubation with anti-Dig antibody, the embryos were stained in BM-Purple (Roche) until the color developed. All processed samples were photographed under a stereomicroscope using a digital camera (Olympus). The antisense POU5F1 probe was a generous gift from B. Püschel. When necessary whole-mount samples were then included in Technovit 7100 resin (Heraeus) and 6 µm depth sections were obtained on superfrost glass blades. Sections were counter-colored by nuclear red, then rinsed and mounted in Mowiol before observation.
Results
POU5F1 expression in rabbit blastocyst
We followed the expression and methylation of POU5F1 in rabbit blastocysts over the period from Day 4 to Day 6. At Day 4, the early expanded blastocysts contained an early ICM and a single-layer differentiated TE, whereas at Day 6, the blastocysts contained a plane embryonic disk and a bilayered trophoblast. After the disappearance of Rauber’s membrane (monolayered TE lining the outer surface of the ICM) between Days 4 and 5, the Day 6 embryonic disk comprised an epithelialized epiblast in direct contact with the embryo microenvironment, covered on its inner surface with a monolayered hypoblast. At the same stage, the trophoblast was composed of TE and an extraembryonic mural hypoblast that migrated starting from the inner surface of the embryonic disk, throughout the inner surface of the TE from Day 5 onward. Importantly, for the molecular analyses performed during this study, the Day 4 ICM and TE were isolated using a moderate immunosurgery protocol. This was not possible at Day 5 due to the loss of Rauber’s membrane and direct exposure of the ICM cells to the external environment. Day 5 blastocysts were therefore hemi-sectioned so that their TE samples would be pure, while the ‘embryoblast’ samples were enriched in embryonic disk cells, very probably slightly contaminated by a few trophectodermal cells. Day 6 blastocysts were micro-dissected in such a way that both epiblast and hypoblast samples were pure, and indeed the trophoblast sample contained both the extraembryonic hypoblast and the TE layers.
Immunolocalisation of the POU5FI protein
The POU5F1 protein was localized by immunofluorescence in all the nuclei (both inner and outer cells) of rabbit morulae developed in vitro or in vivo (Fig. 1A and B). At Day 4 (96–98 h post coitum) in early blastocysts, the POU5F1 protein was localized in the nuclei of both the TE and ICM cells (Fig. 1C and D). In later blastocysts, 112 h post coitum, the POU5F1 protein appeared to be restricted to the embryonic disk without any labeling in TE cells (Fig. 1E and F).
Immunolocalization of the POU5F1 protein in rabbit embryos. The embryos were either developed in vitro (B, C and D) from the one-cell stage onwards or developed in vivo (A, E and F). They were fixed at 67 (A), 72 (B), 96 (C), 98 (D) and 112 (E and F) hours post coitum. F is an enlargement of E focused on the embryonic disk.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Immunolocalization of the POU5F1 protein in rabbit embryos. The embryos were either developed in vitro (B, C and D) from the one-cell stage onwards or developed in vivo (A, E and F). They were fixed at 67 (A), 72 (B), 96 (C), 98 (D) and 112 (E and F) hours post coitum. F is an enlargement of E focused on the embryonic disk.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Immunolocalization of the POU5F1 protein in rabbit embryos. The embryos were either developed in vitro (B, C and D) from the one-cell stage onwards or developed in vivo (A, E and F). They were fixed at 67 (A), 72 (B), 96 (C), 98 (D) and 112 (E and F) hours post coitum. F is an enlargement of E focused on the embryonic disk.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
POU5F1 transcript detection
In situ hybridization experiments were performed to investigate whether this limitation of the POU5F1 protein was due to a restriction in the POU5F1 transcript (Fig. 2). POU5F1 transcripts were clearly restricted to the embryonic disk from Day 5 onward (120 and 144 h post coitum), and more precisely to epiblasts at 144 h post coitum (Fig. 2B, C and D). However, in earlier blastocysts (Day 4, 96 h post coitum), the POU5F1 transcripts mostly appeared to be expressed in ICM cells, but their presence in the TE could not be totally excluded (Fig. 2A). We therefore decided to quantify POU5F1 transcripts by qRT-PCR in the different compartments of rabbit blastocysts from Day 4 (96 h post coitum) to Day 6 (144 h post coitum). The RT-qPCR results (Fig. 3 and Table 1) showed that POU5F1 transcripts were indeed markedly present in the TE at Day 4 but significantly enriched in the ICM (about 1.6-fold higher). At Day 5, the relative expression of POU5F1 decreased throughout the embryo, but the difference between the embryonic and TE compartments increased (3.3-fold higher expression in the embryoblast than in the TE). At Day 6, POU5F1 transcripts were restricted to the epiblast, with clearly reduced expression in the hypoblast (30-fold lower than in the epiblast) and hugely reduced expression in the trophoblast (145-fold lower than in the epiblast). It was also noticeable that POU5F1 expression gradually but significantly decreased in the TE/trophoblast compartment over the Day 4–Day 6 period (7.5-fold decrease between Day 4 and Day 5 and a further 10-fold decrease between Day 5 and Day 6).
POU5F1 transcript localization analyzed using in situ hybridization in Day 4 (A) and Day 5 (B) whole blastocysts and in dissected embryonic disks with pieces of their surrounding trophoblast at Day 6 (C) and on a transversal slide of the Day 6 embryonic disk (D).
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
POU5F1 transcript localization analyzed using in situ hybridization in Day 4 (A) and Day 5 (B) whole blastocysts and in dissected embryonic disks with pieces of their surrounding trophoblast at Day 6 (C) and on a transversal slide of the Day 6 embryonic disk (D).
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
POU5F1 transcript localization analyzed using in situ hybridization in Day 4 (A) and Day 5 (B) whole blastocysts and in dissected embryonic disks with pieces of their surrounding trophoblast at Day 6 (C) and on a transversal slide of the Day 6 embryonic disk (D).
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
POU5F1 transcript quantification by RT-qPCR in embryonic and extraembryonic tissues at Day 4 and Day 5. The results are presented as mean ± s.e.m. Statistical differences between compartments with the same embryonic age are indicated on the figure (***P < 0.001), while statistical differences between samples of different ages are reported in Table 1.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
POU5F1 transcript quantification by RT-qPCR in embryonic and extraembryonic tissues at Day 4 and Day 5. The results are presented as mean ± s.e.m. Statistical differences between compartments with the same embryonic age are indicated on the figure (***P < 0.001), while statistical differences between samples of different ages are reported in Table 1.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
POU5F1 transcript quantification by RT-qPCR in embryonic and extraembryonic tissues at Day 4 and Day 5. The results are presented as mean ± s.e.m. Statistical differences between compartments with the same embryonic age are indicated on the figure (***P < 0.001), while statistical differences between samples of different ages are reported in Table 1.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Pou5f1 transcripts quantification by RT-qPCR: statistical analyses of the comparisons between two embryonic compartments of different stages.
| Embryoblast D5 | Epi D6 | Hypo D6 | Troph D5 | Troph D6 | |
|---|---|---|---|---|---|
| ICM D4 | *** | – | *** ‡ |
*** ‡ |
*** ‡ |
| Troph D4 | *** ‡ |
*** ‡ |
*** ‡ |
*** | *** |
| Embryoblast D5 | ‡ | *** | *** | ‡ | *** ‡ |
| Troph D5 | ‡ | ***‡ | – ‡ |
‡ | *** |
–, no statistical difference; ***, statistically different with P < 0.001; ‡, corresponds to comparisons without biological relevance and cells without this symbol correspond to compartments of the same stage: see results of the comparison in Fig. 3.
Methylation of the POU5F1 upstream region
To analyze DNA methylation of the POU5F1 gene’s upstream region, we used bisulfite conversion followed by the sequencing of PCR products from DNA extracted from the different compartments of Day 4, Day 5 and Day 6 rabbit blastocysts. We therefore used four series of primers (for both primary and nested PCRs) enabling the analysis of four regions spanning the four interspecies conserved domains (CR1–4) and the most important functional domains for POU5F1 regulation by trans-acting factors. Four different regions were thus analyzed, containing respectively: the minimal promoter, partially included in CR1 and containing an HRE site and an Sp1/Sp3-binding site, and the first 75 bases of exon 1 for the R1 region, the proximal enhancer1B (PE1B) sequence included in CR2 together with 120 bases in 5′ and 46 bases in 3′ for the R2 region, the proximal enhancer 1A (PE1A) sequence and the CR3 plus 23 bases in 3′ for the R3 region and the distal enhancer region including the DE2A sequence and DE2B (or POU5F1/SOX binding site), both located in CR4 plus 180 bases in 3′ for the R4 region. These four regions contained respectively 25, 8, 10 and 17 CpGs (Fig. 4).
Localization of regions 1–4 and conserved regions (CR) 1–4 on the rabbit POU5F1 sequence. Forward and reverse primers used for PCR amplification after bisulfite treatment and thus delimitation of the four analyzed regions are indicated by colored arrows. Conserved regions are underlined. Sites for trans-factor binding are highlighted or framed. All CpGs in the rabbit sequence are indicated in red, while those analyzed in the paper are numbered from 1 to 60.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Localization of regions 1–4 and conserved regions (CR) 1–4 on the rabbit POU5F1 sequence. Forward and reverse primers used for PCR amplification after bisulfite treatment and thus delimitation of the four analyzed regions are indicated by colored arrows. Conserved regions are underlined. Sites for trans-factor binding are highlighted or framed. All CpGs in the rabbit sequence are indicated in red, while those analyzed in the paper are numbered from 1 to 60.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Localization of regions 1–4 and conserved regions (CR) 1–4 on the rabbit POU5F1 sequence. Forward and reverse primers used for PCR amplification after bisulfite treatment and thus delimitation of the four analyzed regions are indicated by colored arrows. Conserved regions are underlined. Sites for trans-factor binding are highlighted or framed. All CpGs in the rabbit sequence are indicated in red, while those analyzed in the paper are numbered from 1 to 60.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
The proportions of methylated, unmethylated or undetermined (so-called missing) CpGs in each category of samples are illustrated in Supplementary Fig. 1 (see section on supplementary data given at the end of this article). To obtain further detail, among the documented CpGs, we analyzed the percentage of CpG methylation of the alleles in the different embryonic compartments at the three stages, focusing on each region and each conserved region (Fig. 5 and Table 2). Generally speaking, the conserved regions behaved the same way as the region they belong to. However, at Day 4, CR2, CR3 and CR4 were not statistically differentially methylated in ICM and TE while their neighboring regions displayed a higher methylation level in TE than in ICM (P < 0.01). It was clear that all four regions appeared to be under-methylated in both Day 4 ICM and Day 6 epiblast compared to all the other samples. In the TE/trophoblast samples, methylation significantly (P < 0.05) increased in the four regions as well as in CR1, CR2 and CR3 between Day 4 and Day 5 (Table 2). It then remained stable between Day 5 and Day 6 in regions R1, R3 and R4 and CR3. Interestingly, in these compartments, the methylation level of CR4 remained unchanged over the period, and its increase in R2, CR1 and CR2 between Day 4 and Day 5 was followed by a decrease between Day 5 and Day 6. Comparing Day 5 embryoblasts with Day 6 hypoblasts also revealed a major increase in R3, CR3 and R4 methylation with differentiation. In the pluripotent compartment, methylation levels increased between Day 4 ICM and Day 5 embryoblast in regions R1, R2 and R4 and in CR4, but then decreased between Day 5 embryoblast and Day 6 epiblast so that between Day 4 ICM and Day 6 epiblasts – the two pluripotent compartments without any putative contamination by few trophectodermal cells – no statistical difference appeared in R4, CR4 and a slight decrease in methylation was observed in R2, CR2 and CR3. In Day 4 early blastocysts, methylation levels were higher in the TE than in the ICM in all regions but R1 (Fig. 5). This tendency toward increased methylation in trophoblasts compared to the ICM or embryoblasts persisted at Day 5 especially for CR1, R3 and CR3 but not for R2, CR2 and CR4. It finally appeared significant for all regions and conserved regions at Day 6 (Fig. 5). At that stage, only epiblasts retained a very low level of methylation.
Mean percentages of methylation in alleles of the four regions and conserved regions (bars indicate s.e.). Stars indicate statistical differences between compartments of the same stage: **P < 0.01, ***P < 0.001. (D4 ICM = Day 4 Inner cell Mass, D4 Troph = Day 4 trophectoderm, D5 Embryoblast = Day 5 embryoblast, D5 Troph = Day 5 Trophoblast, D6 Epi = Day 6 Epiblast, D6 Troph = Day 6 Trophoblast, D6 Hypo = Day 6 hypoblast). Statistical differences between stages are reported in Table 2.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Mean percentages of methylation in alleles of the four regions and conserved regions (bars indicate s.e.). Stars indicate statistical differences between compartments of the same stage: **P < 0.01, ***P < 0.001. (D4 ICM = Day 4 Inner cell Mass, D4 Troph = Day 4 trophectoderm, D5 Embryoblast = Day 5 embryoblast, D5 Troph = Day 5 Trophoblast, D6 Epi = Day 6 Epiblast, D6 Troph = Day 6 Trophoblast, D6 Hypo = Day 6 hypoblast). Statistical differences between stages are reported in Table 2.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Mean percentages of methylation in alleles of the four regions and conserved regions (bars indicate s.e.). Stars indicate statistical differences between compartments of the same stage: **P < 0.01, ***P < 0.001. (D4 ICM = Day 4 Inner cell Mass, D4 Troph = Day 4 trophectoderm, D5 Embryoblast = Day 5 embryoblast, D5 Troph = Day 5 Trophoblast, D6 Epi = Day 6 Epiblast, D6 Troph = Day 6 Trophoblast, D6 Hypo = Day 6 hypoblast). Statistical differences between stages are reported in Table 2.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
POU5F1 upstream regions methylation: statistical analyses of the percentage in allele methylation between two embryonic compartments of different stages.
| Embryoblast D5 | Epi D6 | Hypo D6 | Troph D5 | Troph D6 | |
|---|---|---|---|---|---|
| ICM D4 | ↗ R1** ↗ R2*** ↗ R4*** ↗ CR4** |
↘ R2* ↘ CR2* ↘ CR3* |
‡ | ‡ | ‡ |
| Troph D4 | ‡ | ‡ | ‡ | ↗ R1** ↗ R2* ↗ R3*** ↗ R4** ↗ CR1** ↗ CR2*** ↗ CR3*** |
↗ R1* ↗ R3*** ↗ R4*** ↗ CR3*** |
| Embryoblast D5 | ‡ | ↘ R1** ↘ R2*** ↘ R4*** ↘ CR2** ↘ CR4*** |
↗ R3*** ↗ R4** ↗ CR3*** |
‡ | ‡ |
| Troph D5 | ‡ | ‡ | ‡ | ‡ | ↘ R2* ↘ CR1* ↘ CR2* |
For one given comparison regions (R) and conserved regions (CR) with statistical difference (P < 0.05*, P < 0.01**, P < 0.001***) are indicated in the corresponding case. ↗ (and ↘) indicate an increase (or decrease) in methylation between the embryonic sample referenced as the head of the line and the one as the head of the column.
‡Corresponds to comparisons without biological relevance.
The rabbit DE identified by Kobolak (2009) during a functional test covers the whole sequence of R4 region as defined in this study, plus about 40 bases in 5′, which indeed are devoid of any CpG. We therefore considered that our R4 region corresponded to the functionally tested rabbit DE that contains 17 CpGs (Fig. 5). The proximal enhancer has not been functionally characterized in the rabbit but contains both PE1A and PE1B and covers 243 bases in the mouse (Okazawa et al. 1991), so we considered that it extended from PE1A to PE1B and contained eight CpGs (CpGs 24 to 31). The rabbit minimal promoter was identified by sequence alignment with mouse, human, bovine and dog sequences by Kobolak (2009) and contains ten CpGs (CpGs 41–50). Statistical analysis of minimal promoter and exon 1 (Fig. 6 and Table 3) showed that both of these sequences are poorly methylated in Day 4 ICM and Day 6 epiblast, but the difference between pluripotent and differentiated compartments of the blastocyst became statistically significant only at Day 5 for minimal promoter and six for exon 1 (Fig. 6). In pluripotent compartment, both sequences became significantly more methylated between Day 4 and Day 5, then methylation decreased, although in a significant way only for minimal promoter. As a result, at Day 6, methylation of both sequences in epiblast is similar to the ICM at Day 4. In TE, both minimal promoter and exon 1 methylation increased but with different patterns: methylation of the minimal promoter appeared transient: it significantly increased between Day 4 and Day 5 and then decreased between Day 5 and Day 6. In the same time, methylation of exon 1 progressively increased so that it became significantly different only between Day 4 and Day 6 (Table 3).
Mean percentages of methylation in alleles of minimal promoter (MP) and exon 1 (bars indicate standard errors). Stars indicate statistical differences between compartments of the same stage: **P < 0.01, ***P < 0.001. (D4 ICM = Day 4 Inner cell Mass, D4 Troph = Day 4 trophectoderm, D5 Embryoblast = Day 5 embryoblast, D5 Troph = Day 5 Trophoblast, D6 Epi = Day 6 Epiblast, D6 Troph = Day 6 Trophoblast, D6 Hypo = Day 6 hypoblast). Statistical differences between stages are reported in Table 3.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Mean percentages of methylation in alleles of minimal promoter (MP) and exon 1 (bars indicate standard errors). Stars indicate statistical differences between compartments of the same stage: **P < 0.01, ***P < 0.001. (D4 ICM = Day 4 Inner cell Mass, D4 Troph = Day 4 trophectoderm, D5 Embryoblast = Day 5 embryoblast, D5 Troph = Day 5 Trophoblast, D6 Epi = Day 6 Epiblast, D6 Troph = Day 6 Trophoblast, D6 Hypo = Day 6 hypoblast). Statistical differences between stages are reported in Table 3.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Mean percentages of methylation in alleles of minimal promoter (MP) and exon 1 (bars indicate standard errors). Stars indicate statistical differences between compartments of the same stage: **P < 0.01, ***P < 0.001. (D4 ICM = Day 4 Inner cell Mass, D4 Troph = Day 4 trophectoderm, D5 Embryoblast = Day 5 embryoblast, D5 Troph = Day 5 Trophoblast, D6 Epi = Day 6 Epiblast, D6 Troph = Day 6 Trophoblast, D6 Hypo = Day 6 hypoblast). Statistical differences between stages are reported in Table 3.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
POU5F1 minimal promoter (MP) and 5′ end of exon 1 methylation: statistical analyses of the percentage in allele methylation between two embryonic compartments of different stages.
| Embryoblast D5 | Epi D6 | Hypo D6 | Troph D5 | Troph D6 | |
|---|---|---|---|---|---|
| ICM D4 | ↗ Exon 1* ↗ MP* |
‡ | ‡ | ‡ | |
| Troph D4 | ‡ | ‡ | ‡ | ↗ MP** | ↗ Exon 1** |
| Embryoblast D5 | ‡ | ↘ MP** | ‡ | ‡ | |
| Troph D5 | ‡ | ‡ | ‡ | ‡ | ↘ MP* |
For one given comparison MP and exon 1 with statistical difference (P < 0.05*, P < 0.01**) are indicated in the corresponding case. ↗ (and ↘) indicates an increase (or decrease) in methylation between the embryonic sample referenced as the head of the line and the one as the head of the column.
‡Corresponds to comparisons without biological relevance.
The methylation levels of all regions and conserved regions in each embryonic compartment over the period are recapitulated in Supplementary Fig. 2. Furthermore, the proportions of methylated and unmethylated CpGs among those determined in each functional region (DE, proximal enhancer, minimal promoter and first exon), are also compared to their neighboring CpGs. Of note, proximal enhancer spans over two different regions (R2 and R3) so that its methylation cannot be statistically analyzed on a per allele basis. Moreover, all these regions and functional elements have different CpG numbers and densities, and functional meaning of different methylation levels in different regions remains largely unknown. Therefore, we did not compare their methylation levels on a statistical point of view but only described major features (Supplementary Fig. 2). Among the most differentiated samples (Day 5 and Day 6 trophoblast and Day 6 hypoblast), we observed a lower level of methylation in the R1 region and a higher level in the R2 and R3 regions. The lower methylation in R1 was even more pronounced in CR1 at Day 6. Interestingly, PE methylation remained very low in Day 6 epiblasts, while it progressively increased in differentiated compartments but to a lesser extent than its neighboring CpGs from the R2 and R3 regions. The lowest methylation level of the DE (R4 in Supplementary Fig. 2) was found in Day 4 ICM and Day 6 epiblasts. Its methylation also increased gradually with differentiation.
We then focused on methylation along single alleles and analyzed the bisulfite sequencing results obtained for different alleles from different embryos (at least two embryos per stage and compartment) at each stage and for each compartment (Fig. 7). A total absence of methylation on CpGs located in CR1 around the HRE and Sp1/Sp3-binding site and in the first exon appeared to be a feature of Day 4 ICM. Although the CR1 region was very poorly documented in our results relative to Day 6 epiblasts, thus preventing any conclusion, a slight methylation appeared in exon 1 at that stage. Also, the R4 region appeared to be partitioned into two sub-regions: CR4 (CpGs1 to 7) which contained the DE2A and DE2B sites remained poorly methylated in Day 5 trophoblasts and Day 6 trophoblasts and hypoblasts, and other CpGs in the R4 region, which became more methylated in these differentiated samples.
DNA methylation profile of individual CpG sites in the four regions of the POU5F1 upstream region in the different embryonic compartments at Days 4, 5 and 6. The status, either methylated (closed circle) or unmethylated (open circle), is indicated at each CpG site. Conserved regions, functional elements and binding sites are surrounded in color.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
DNA methylation profile of individual CpG sites in the four regions of the POU5F1 upstream region in the different embryonic compartments at Days 4, 5 and 6. The status, either methylated (closed circle) or unmethylated (open circle), is indicated at each CpG site. Conserved regions, functional elements and binding sites are surrounded in color.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
DNA methylation profile of individual CpG sites in the four regions of the POU5F1 upstream region in the different embryonic compartments at Days 4, 5 and 6. The status, either methylated (closed circle) or unmethylated (open circle), is indicated at each CpG site. Conserved regions, functional elements and binding sites are surrounded in color.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
To analyze how far the methylation of one specific CpG is correlated to methylation of its neighboring CpGs, we built the LD-like plot of correlations for methylation status between each pair of CpG within each region (Fig. 8). This analysis revealed very few negative correlations. In the R4 region, CpGs 8–17 behaved in a more highly correlated way than CpGs 1–7 (belonging to CR4). Also, in the R3 region, CpG 18–24, together with CpGs 26 and 27, appeared to be highly correlated, while CpG 25 located in PE1A behaved independently of the other CpGs in the region. This was probably mainly due to its particular methylation observed in Day 5 ICM, but not solely (Fig. 7). The LD-like plot (Fig. 8) also revealed a relatively low correlation between the CpGs of the R2 region; in particular, CpG 33 located in CR2 was very poorly correlated with CpGs 34 and 35 which were outside CR2, and still poorly correlated with CpG 32 which belonged to CR2. Indeed, it was most frequently unmethylated despite the methylation of its neighboring CpGs in Day 4 and Day 6 trophoblasts, as well as in Day 6 hypoblasts (Fig. 7). In R1 region, CpGs 36 to 47 had the most correlated behavior, whereas CpGs 48–59 (60 was too poorly documented in our data to be taken into account here) appeared to poorly correlate to other CpGs in the same region.
LD-like plot. Methylation correlation status between pairs of CpGs in each region for all samples.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
LD-like plot. Methylation correlation status between pairs of CpGs in each region for all samples.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
LD-like plot. Methylation correlation status between pairs of CpGs in each region for all samples.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Because we were mostly focusing on variations in methylation between compartments and stages, we also analyzed the data using PCA (Fig. 9). PCA was thus performed on the percentage of methylated alleles for each CpG in each compartment at each developmental stage. However, because of the scarcity of the biological material, some CpGs were poorly documented despite several attempts to sequence more alleles. Methylation percentages were only calculated for CpGs with at least eight documented alleles. For poorly documented CpGs, the values were inferred so that they would not interfere with the analysis of methylation variations between compartments and stages (see ‘Materials and methods’ section). PCA performed on the whole set of data showed that the first axis displayed 64.6% of the variability and separated Day 4 ICM and Day 6 epiblasts on the one hand from Day 5 and Day 6 trophoblasts and Day 6 hypoblasts on the other. Such overriding first axes were also found when performing the PCA on each region separately. They reflected 57.4, 56.6, 84.5 and 77.3% of variability in R1, R2, R3 and R4, respectively. Variations in methylation between pluripotent ICM and epiblasts and differentiated trophoblasts and hypoblasts thus spanned the four analyzed regions, and this was even more pronounced for the R3 and R4 regions encompassing respectively the proximal enhancer 1A and the DE. Interestingly, in the R3 region, Day 5 embryoblast segregated with Day 4 ICM and Day 6 epiblasts along the first axis and therefore contributed to the opposition between pluripotent and differentiated compartments; this was not the case in R1, R2 and R4 where the D5 embryoblast sample was middle-of-the-road regarding this opposition. It should also be noted that the TE at Day 4 did not segregate along these first axes in regions 1 to 4, displaying intermediate methylation (particularly in the R3 and R4 regions) between pluripotent and differentiated lineages. Since proximal enhancer activity has been shown to be recapitulated by a 243bp fragment extending from PE1A to PE1B in the mouse, we performed the same type of PCA analysis on the subgroup of CpGs 24–31 in the rabbit (Fig. 9). The first axis (55% variability) was found to segregate Day 4 ICM, Day 4 TE and Day 6 epiblasts from Day 5 samples, Day 6 differentiated samples being intermediate. Proximal enhancer region (including PE1A and PE1B) thus behaved quite differently from region 3 that contains PE1A and CR3 and whose all CpGs except CpG25, highly contributed to the separation between differentiated and pluripotent compartments. Gathering the three samples with the highest expression of POU5F1 (Day 4 TE, Day 4 ICM and Day 6 epiblasts) in a single group along the first PCA axis evidenced a specific feature of the proximal enhancer region.
Principal component analyses of the percentages of methylation of each CpG in each sample. Graphical representation of the first two axes of the PCA performed on all CpGs, on CpGs in each analyzed region, and on CpG in the proximal enhancer region. CpG numbers are colored according to the region to which the corresponding CpG belongs.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Principal component analyses of the percentages of methylation of each CpG in each sample. Graphical representation of the first two axes of the PCA performed on all CpGs, on CpGs in each analyzed region, and on CpG in the proximal enhancer region. CpG numbers are colored according to the region to which the corresponding CpG belongs.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Principal component analyses of the percentages of methylation of each CpG in each sample. Graphical representation of the first two axes of the PCA performed on all CpGs, on CpGs in each analyzed region, and on CpG in the proximal enhancer region. CpG numbers are colored according to the region to which the corresponding CpG belongs.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
At a more detailed level, and in an attempt to determine which CpG mainly contributed to the difference between the pluripotent and differentiated compartments, we identified CpGs 8–10, 12–13 and 15–17 for R4, CpGs 18–24 for R3, CpGs 30–32 and 34–35 for R2 and CpGs 36, 39, 44, 54 and 56 for R1 (Fig. 9).
The second axes of PCA analyses revealed different patterns of sample segregation depending on the region analyzed. The analysis of R1 showed that Day 5 trophoblasts globally differed from Day 6 trophoblasts (23.41% variability), and this was mainly due to methylation of CpGs 54, 56 and 58, all located in exon 1 (Fig. 9). The third axis of PCA performed on all pooled regions evidenced the high methylation of CpG 58, which is specific to Day 6 trophoblasts (Supplementary Fig. 2). In R3, CpG 25 located in proximal enhancer 1A displayed a very particular methylation in Day 5 embryoblast, shown on axis 2 (9.04% variability).
Lastly, heatmaps of correlation regarding the level of methylation for each CpG over the four regions (Fig. 10) once again emphasized the highly correlated methylation behavior of specific groups of CpGs. Five of these groups gathered CpGs from the same region. This is the case for CpGs 54, 56, 58, 59 in exon 1, as well as for CpGs 37, 38 and 40 in R1 region, these two groups differing in their relative methylation in Day 5 and Day 6 differentiated compartments. Also CpGs 34 and 35 in R2 region and CpGs 1, 12 and 17 in R4 region formed highly correlated groups. However, the most obvious group gathered CpGs 18–23 from R3 region, which did cluster very closely and were demethylated in pluripotent samples, intermediate in Day 4 TE and highly methylated in Day 5 and Day 6 differentiated samples. CpG 25 from PE1A in the same region displayed a very different behavior, mainly due to its high level of methylation in Day 5 embryoblasts and its lower methylation in Day 5 and Day 6 differentiated compartments, while CpG24 differed by a slightly lower methylation in Day 5 and Day 6 trophoblasts. Interestingly, this heatmap also evidenced groups of CpGs from different regions sharing similar behavior: especially CpGs 47, 51, 52 and 53 from CR1 and exon 1 and CpGs 11, 13, 15 and 16 from R4 region (DE). The same was observed for CpGs 6, 7 and 14 from R4 region and CpGs 42 and 45 from R1 region and minimal promoter.
Heatmap correlation based on the Pearson correlation coefficients of the methylation of individual CpGs in the different samples.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Heatmap correlation based on the Pearson correlation coefficients of the methylation of individual CpGs in the different samples.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Heatmap correlation based on the Pearson correlation coefficients of the methylation of individual CpGs in the different samples.
Citation: Reproduction 156, 2; 10.1530/REP-17-0689
Discussion
The rabbit blastocyst: a model for the regulation of POU5F1 in mammalian blastocysts
Our results revealed ubiquitous expression of the POU5F1 protein in early (Day 4) in vivo-developed rabbit blastocysts and its restriction to the embryonic disk by Day 6. Such spatiotemporal pattern of expression thus appears to be a feature of all mammalian blastocysts. POU5F1 rabbit transcripts were also progressively restricted to the pluripotent cells of the blastocyst with a 1.6-fold enrichment between pluripotent and TE compartment at Day 4 (confirming the 1.3-fold enrichment reported by Kobolak 2009) and 145-fold enrichment at Day 6. Interestingly, this appeared also to represent the situation in cattle, where 1.7-fold, 2-fold, 7-fold and 190-fold levels of enrichment were reported in the ICM compared to Day 7 trophoblasts (Berg et al. 2011), Day 8 expanded trophoblasts (Herrmann et al. 2013) and Day 9 and Day 11 blastocysts (Berg et al. 2011) respectively. It was also similar to findings in the pig, where POU5F1 transcripts have been reported to be 5000-fold enriched in Day 10 epiblasts compared to the trophoblast (Gao et al. 2011). We therefore considered that the rabbit blastocyst represents the development characteristic of most mammals in terms of POU5F1 expression.
Direct analysis of the entire Pou5f1 regulatory region in the developing embryo
To our knowledge, our study offers the first extensive (spanning four conserved and functional regions) and dynamic (three different stages) analysis of methylation of the POU5F1 regulatory region during mammalian blastocyst development. Indeed, most previous studies used ES cells and compared them with TS cells (Hattori et al. 2004) or analyzed them during in vitro-induced differentiation into trophoblasts (Carey et al. 2014) or embryoid bodies (Sato et al. 2006, Athanasiadou et al. 2010). Whatever the case, these models may not entirely reflect the ‘in embryo’ situation since both the epigenome and transcriptome of such in vitro-cultured cells may differ from the embryo compartments they are supposed to mimic (Schmaltz-Panneau et al. 2014). Moreover, Pou5F1 silencing in different tissues may be associated with different methylation patterns (Sato et al. 2006). Another series of studies focused on a single region, and particularly the promoter and first exon in the embryo (Yamazaki et al. 2006, Nakanishi et al. 2012, Zhao et al. 2012, Saenz-de-Juano et al. 2014). Further, Gao et al. (2011) focused on a single stage of the porcine conceptus, in order to analyze the POU5F1 regulatory region in Day 10 epiblasts, hypoblasts and TE. However, their analysis focused on CpG islands that did not cover the four conserved regions containing functional elements related to POU5F1 expression that we documented, so that their results showing a low level of methylation in differentiated lineages cannot be compared directly with our results. Interestingly, an analysis of the 400 bp region upstream of the ATG codon, and of the 5′ end of exon 1, produced the conclusion that TE differentiation in the mouse blastocyst precedes differential methylation (Nakanishi et al. 2012). By contrast, our analysis of the four regions in the rabbit TE compared to ICM showed an increase in methylation as early as Day 4. Although differences in blastocyst development between the mouse and rabbit cannot be excluded to explain these contradictory results, it is worth emphasizing that the R1 region (mainly analyzed in the mouse study and containing CR1, minimal promoter and first exon) was the only region without differential methylation between ICM and TE at Day 4 in our study. Differences in conclusions are thus most probably due to the restriction of the mouse analysis to this specific region. These examples reinforce the interest of our approach, which analyzed the most functional regions in the upstream regulatory region of POU5F1 – starting from the embryo itself – in a dynamic and developmental manner, even though the scarcity of biological material rendered this more difficult and made it necessary to deal with some missing data.
Methylation occurs in the four regions during differentiation, but with differences
Whatever the approach adopted to analyze progressive modifications to the methylation of POU5F1 regulatory regions during differentiation in the developing blastocyst, our results clearly showed that differentiation was accompanied by an increase in DNA methylation over the four analyzed regions, but to different extents and in different ways. According to PCA analyses R3 and R4 regions appeared to be those with the greatest increase in methylation related to differentiation. This relied mainly on CpGs 18-23 in R3 region and CpGs 8–17 in the 3′ end of R4 region. The R1 and R4 regions showed a similar distribution of the biological samples along the first axis of PCA, but variations in R1 region methylation were less extensively related to differentiation and R1 had the lowest methylation level whatever the stage and compartment. However, R1 and R4 regions have few CpGs with correlated behaviors (CpG 6, 7, 14 with 42, 45 and CpGs 11, 13, 15, 16 with 47, 52, 53) during development, which may reveal functional links between these two regions. CpG methylation in the R2 region also increased in line with differentiation, about to the same extent as R1 region. It however presented some specific features. In particular, R2 and CR2 methylation decreased between Day 5 and Day 6 in the trophoblasts. Despite CpGs 34 and 35 displayed a much correlated behavior, the correlation in methylation behavior between pairs of CpG in this region was lower than that in other regions. This may have been linked to a few CpGs with a highly specific behavior; in particular, CpG 33 in CR2 remained unmethylated in the most differentiated samples (Day 6 trophoblasts and hypoblasts).
CpG methylation and lineage differentiation
Focusing on the TE/trophoblast compartment, we evidenced a very progressive methylation of CpGs in the four regions. Methylation of CpGs 8, 10, 12 and 17 contributed markedly to the specific features of early Day 4 TE, together with CpGs 18-23, CpGs 34 and 35 (among the most highly methylated in this sample), and CpGs 39 and 44 from R3, R2 and R1 regions respectively. Later on, methylation still increased in trophoblast between Day 4 and Day 5 for all these CpGs except for CpGs 8, 10, 12 and 17 suggesting different roles in TE differentiation. Early methylation of these few CpGs located in R4 region could be involved in the very beginning of POU5F1 repression in this compartment. This early step was followed by methylation of few CpGs especially CpGs 33 and 46 at Day 5. Interestingly, CpG 46 is located in the Sp1/Sp3 binding site, so that its methylation could interfere with the basal transcription machinery function and participate in the progressive reduction of POU5F1 expression (Tenayuca et al. 2017). Between Day 5 and Day 6, methylation abruptly increased on CpGs 54, 56, 58 all located in exon 1, suggesting that methylation of exon 1 was coming to an end and might participate to the final repression of POU5F1 in this compartment. Methylation appears to be very similar in both Day 6 trophoblast and hypoblast, suggesting a non-lineage-specific pattern so that difference in methylation cannot be responsible for the higher residual expression of POU5F1 in hypoblast at that stage.
Methylation of conserved regions and functional elements during blastocyst development
Interspecies conserved regions are supposed to have functional roles in the regulation of gene expression. We wondered whether the four interspecies conserved regions previously identified had a particular behavior regarding methylation during differentiation. Our analyses revealed two major features for these conserved regions: first, the CpGs mostly involved in the differentiation-linked methylation in each region are localized out of the conserved sequences with CR3 being the only exception. Noticeably, CR3 is also the only conserved region devoid of any identified functional element. Second, differential methylation between pluripotent and differentiated compartments occurs later in the conserved regions than in their surrounding regions. Going into more detail, the functionally tested rabbit DE (which corresponds to the R4 region) became methylated in differentiated cells as soon as the early Day 4 TE. But its 5′ end (CR4 region including the DE2A and DE2B sites reported in the mouse as being necessary and sufficient to activate the distal enhancer) (Okumura-Nakanishi et al. 2005) remained clearly less methylated than its 3′ end during differentiation. Further, the region surrounding PE1B (in CR2) did not appear to be highly methylated in the most differentiated compartments of Day 6 blastocysts. The methylation of CpGs 24 and 25 included in PE1A obviously increased in line with differentiation, but to a lesser extent than the surrounding CpGs of R3 region. Another very interesting point was the low methylation of the minimal promoter region during differentiation but the considerable increase in exon 1 methylation between Day 5 and Day 6 in the trophoblast compartment. Thus, methylation of functional elements is not a priority during differentiation of the blastocyst.
Low methylation in pluripotent compartments
R2, R3 and R4 methylation in pluripotent compartments remains lower than in differentiated ones whatever the stage, and this was also true for R1 region but only from Day 5 onward. Indeed Day 5 embryoblast appeared transiently methylated. This could be due either to the heterogeneity of embryoblast containing precursors of both epiblast and hypoblast at that stage, or to few trophectodermal cells contaminating the hemi-sectioned sample at that stage. Alleles from contaminating trophoblast cells would thus behave as those analyzed in the Day 5 trophoblast. This however can be formally excluded for the R3 region, whose alleles show exclusive methylation of CpG 25 only in embryoblast sample. Methylation of this CpG increased hugely in Day 5 embryoblasts, and then decreased abruptly in Day 6 epiblasts and to a lesser extent in Day 6 hypoblasts. Transient methylation of specific CpGs (also observed for CpG 33 in Day 5 trophoblast) thus occurs during development, the function of which remains to be elucidated.
Considering only the Day 4 and Day 6 pluripotent samples, our data reveal new information about DE function. This enhancer has been reported to be active in mouse ICM, primordial germ cells and naïve ES cells, but not in mouse primed epiblasts and EpiSC cells (Yeom et al. 1996, Tesar et al. 2007). Our data showed that there was no specific methylation of this enhancer sequence in epiblasts, so that its loss of activity in primed pluripotent cells of the blastocyst, if conserved in the rabbit, was not reliant on its DNA methylation.
Blastocyst vs stem cell differentiation and the methylation of POU5F1 regulatory regions
Our results revealed a gradual but moderate increase in methylation of the DE region during differentiation in the blastocyst, reduced methylation of the proximal enhancer (from PE1A to PE1B) in compartments with a high level of POU5F1 expression, and differential methylation of the proximal enhancers 1A and 1B, with a gradual increase in methylation in PE1A sequence during differentiation until the Day 6 stage, while methylation in the PE1B region (CR2) initially increased and then decreased between Day 5 and Day 6. There was also a relatively low methylation level in the minimal promoter but a huge increase in exon 1 methylation at Day 6. To some extent, these findings agreed with the higher resistance to methylation of the DE and proximal promoter described by Athanasiadou et al. (2010) in a retinoic acid-induced ES cell differentiation model. Also, Sato et al. (2006) evidenced a lower methylation of the DE compared to the proximal enhancer during ES cell differentiation. By contrast, our results differed from the high methylation of the DE (especially its 5′ end) and promoter observed by Hattori et al. (2004) in TS cells. Although interspecies differences cannot be formally excluded, they are not very probable since our results were in agreement with those found by Athanasiadou. This difference was more likely due to either an effect of the in vitro culture of TS cells on their DNA methylation or to the fact that TS cells are not representative of the early trophoblast differentiation observed during the Day 4–Day 6 period of rabbit blastocyst development. In accordance with this assumption, Senner et al. (2012) evidenced in the mouse a generally higher methylation level in TS cells compared to embryonic trophoblasts, and they found it even for the Pou5f1 gene. While the proximal enhancer has been shown to be gradually methylated during the differentiation of ES cells (Athanasiadou et al. 2010, Park et al. 2013), the differential behavior of the PE1A and PE1B surrounding region has not been analyzed in these models. It therefore appears that the behavior of PE1A in blastocysts is close to the proximal enhancer methylation reported in ES analyses, whereas the specific behavior of the PE1B surrounding region has not been evidenced in previous ES studies.
Finally, our study showed that methylation of the DE region did not occur during the transition from naïve to primed pluripotency in the rabbit, which appeared as a model embryo for most mammalian blastocysts. For the first time, it evidenced an overall progressive methylation of the POU5F1 regulatory region and the 5′ part of exon 1 which started early in the Day 4 TE and accompanied the gradual transcriptional repression of POU5F1 during differentiation. This detailed and exhaustive study also identified individual CpGs with the greatest increases in methylation, as well as groups of CpGs showing correlated methylation behavior, during differentiation. Our findings thus highlighted the specific behavior of certain sub-regions and even single CpGs, whose involvement in gene silencing could now be further investigated by means of genome/epigenome editing.
Supplementary data
This is linked to the online version of the paper at https://doi.org/10.1530/REP-17-0689.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This work was supported by an ‘AMP Diagnostic Prénatal et Diagnostic Génétique’ 2012 grant from the French Agence de la Biomédecine and the ANR LABEX ‘REVIVE’ (ANR-10-LABX-73). Authors are members of RGB-Net (TD 1101) and Epiconcept (FA 1201) COST actions.
Acknowledgements
The authors would like to thank members of Unité Commune d’Expérimentation Animale (UCEA) responsible for our rabbit facility. They are grateful to Hélène Jammes for training E Canon to bisulfite treatment protocols, to Christian Poirier for his help in preparation of the figures. They also thank Pierre Adenot and the MIMA2 facility for access to the ApoTome microscopy and the Région Ile-de-France for funding this system.
References
Al-Khtib M, Blachère T, Guérin JF & Lefèvre A 2012 Methylation profile of the promoters of Nanog and Oct4 in ICSI human embryos. Human Reproduction 27 2948–2954. (https://doi.org/10.1093/humrep/des284)
Athanasiadou R, de Sousa D, Myant K, Merusi C, Stancheva I & Bird A 2010 Targeting of de novo DNA methylation throughout the Oct-4 gene regulatory region in differentiating embryonic stem cells. PLoS ONE 5 e9937. (https://doi.org/10.1371/journal.pone.0009937)
Berg DK, Smith CS, Pearton DJ, Wells DN, Broadhurst R, Donnison M & Pfeffer PL 2011 Trophectoderm lineage determination in cattle. Developmental Cell 20 244–255. (https://doi.org/10.1016/j.devcel.2011.01.003)
Carey TS, Choi I, Wilson CA, Floer M & Knott JG 2014 Transcriptional reprogramming and chromatin remodeling accompanies Oct4 and Nanog silencing in mouse trophoblast lineage. Stem Cells and Development 23 219–229. (https://doi.org/10.1089/scd.2013.0328)
Cauffman G, Van de Velde H, Liebaers I & Van Steirteghem A 2004 Oct-4 mRNA and protein expression during human preimplantation development. MHR: Basic Science of Reproductive Medicine 11 173–181.
Chen C-H, Xu J, Chang W-F, Liu C-C, Su H-Y, Chen YE, Du F & Sung L-Y 2012 Dynamic profiles of Oct-4, Cdx-2 and acetylated H4K5 in in-vivo-derived rabbit embryos. Reproductive Biomedicine Online 25 358–370. (https://doi.org/10.1016/j.rbmo.2012.07.001)
Feldman N, Gerson A, Fang J, Li E, Zhang Y, Shinkai Y, Cedar H & Bergman Y 2006 G9a-mediated irreversible epigenetic inactivation of Oct-3/4 during early embryogenesis. Nature Cell Biology 8 188–194. (https://doi.org/10.1038/ncb1353)
Fischer B, Chavatte-Palmer P, Viebahn C, Navarrete Santos A & Duranthon V 2012 Rabbit as a reproductive model for human health. Reproduction 144 1–10. (https://doi.org/10.1530/REP-12-0091)
Gao Y, Jammes H, Rasmussen MA, Oestrup O, Beaujean N, Hall V & Hyttel P 2011 Epigenetic regulation of gene expression in porcine epiblast, hypoblast, trophectoderm and epiblast-derived neural progenitor cells. Epigenetics 6 1149–1161. (https://doi.org/10.4161/epi.6.9.16954)
Hattori N, Nishino K, Ko Y, Hattori N, Ohgane J, Tanaka S & Shiota K 2004 Epigenetic control of mouse Oct-4 gene expression in embryonic stem cells and trophoblast stem cells. Journal of Biological Chemistry 279 17063–17069. (https://doi.org/10.1074/jbc.M309002200)
Herrmann D, Dahl JA, Lucas-Hahn A, Collas P & Niemann H 2013 Histone modifications and mRNA expression in the inner cell mass and trophectoderm of bovine blastocysts. Epigenetics 8 281–289. (https://doi.org/10.4161/epi.23899)
Idkowiak J, Weisheit G, Plitzner J & Viebahn C 2004 Hypoblast controls mesoderm generation and axial patterning in the gastrulating rabbit embryo. Development Genes and Evolution 214 591–605.
Kawasumi M, Unno Y, Matsuoka T, Nishiwaki M, Anzai M, Amano T, Mitani T, Kato H, Saeki K & Hosoi Y et al. 2009 Abnormal DNA methylation of the Oct-4 enhancer region in cloned mouse embryos. Molecular Reproduction and Development 76 342–350. (https://doi.org/10.1002/mrd.20966)
Khan DR, Dubé D, Gall L, Peynot N, Ruffini S, Laffont L, Le Bourhis D, Degrelle S, Jouneau A & Duranthon V 2012 Expression of pluripotency master regulators during two key developmental transitions: EGA and early lineage specification in the bovine embryo. PLoS ONE 7 e34110. (https://doi.org/10.1371/journal.pone.0034110)
Kirchhof N, Carnwath JW, Lemme E, Anastassiadis K, Scholer H & Niemann H 2000 Expression pattern of Oct-4 in preimplantation embryos of different species. Biology of Reproduction 63 1698–1705. (https://doi.org/10.1095/biolreprod63.6.1698)
Kobolak J, Kiss K, Polgar Z, Mamo S, Rogel-Gaillard C, Tancos Z, Bock I, Baji AG, Tar K & Pirity MK et al. 2009 Promoter analysis of the rabbit POU5F1 gene and its expression in preimplantation stage embryos. BMC Molecular Biology 10 88. (https://doi.org/10.1186/1471-2199-10-88)
Lê S, Josse J & Husson F 2008 FactoMineR: an R package for multivariate analysis. Journal of Statistical Software 25 1–18.
Li J-Y, Pu M-T, Hirasawa R, Li B-Z, Huang Y-N, Zeng R, Jing N-H, Chen T, Li E & Sasaki H et al. 2007 Synergistic function of DNA methyltransferases Dnmt3a and Dnmt3b in the methylation of Oct4 and Nanog. Molecular and Cellular Biology 27 8748–8759. (https://doi.org/10.1128/MCB.01380-07)
Minucci S, Botquin V, Yeom YI, Dey A, Sylvester I, Zand DJ, Ohbo K, Ozato K & Scholer HR 1996 Retinoic acid-mediated down-regulation of Oct3/4 coincides with the loss of promoter occupancy in vivo. EMBO Journal 15 888.
Nakanishi MO, Hayakawa K, Nakabayashi K, Hata K, Shiota K & Tanaka S 2012 Trophoblast-specific DNA methylation occurs after the segregation of the trophectoderm and inner cell mass in the mouse periimplantation embryo. Epigenetics 7 173–182. (https://doi.org/10.4161/epi.7.2.18962)
Nichols J, Zevnik B, Anastassiadis K, Niwa H, Klewe-Nebenius D, Chambers I, Schöler H & Smith A 1998 Formation of pluripotent stem cells in the mammalian embryo depends on the POU transcription factor Oct4. Cell 95 379–391. (https://doi.org/10.1016/S0092-8674(00)81769-9)
Niwa H, Miyazaki J & Smith AG 2000 Quantitative expression of Oct-3/4 defines differentiation, dedifferentiation or self-renewal of ES cells. Nature Genetics 24 372–376. (https://doi.org/10.1038/74199)
Niwa H, Toyooka Y, Shimosato D, Strumpf D, Takahashi K, Yagi R & Rossant J 2005 Interaction between Oct3/4 and Cdx2 determines trophectoderm differentiation. Cell 123 917–929. (https://doi.org/10.1016/j.cell.2005.08.040)
Nordhoff V, Hübner K, Bauer A, Orlova I, Malapetsa A & Schöler HR 2001 Comparative analysis of human, bovine, and murine Oct-4 upstream promoter sequences. Mammalian Genome 12 309–317. (https://doi.org/10.1007/s003350010279)
Okamoto I, Patrat C, Thépot D, Peynot N, Fauque P, Daniel N, Diabangouaya P, Wolf J-P, Renard J-P & Duranthon V et al. 2011 Eutherian mammals use diverse strategies to initiate X-chromosome inactivation during development. Nature 472 370–374. (https://doi.org/10.1038/nature09872)
Okazawa H, Okamoto K, Ishino F, Ishino-Kaneko T, Takeda S, Toyoda Y, Muramatsu M & Hamada H 1991 The oct3 gene, a gene for an embryonic transcription factor, is controlled by a retinoic acid repressible enhancer. EMBO Journal 10 2997–3005.
Okumura-Nakanishi S, Saito M, Niwa H & Ishikawa F 2005 Oct-3/4 and Sox2 regulate Oct-3/4 gene in embryonic stem cells. Journal of Biological Chemistry 280 5307–5317. (https://doi.org/10.1074/jbc.M410015200)
Osteil P, Tapponnier Y, Markossian S, Godet M, Schmaltz-Panneau B, Jouneau L, Cabau C, Joly T, Blachère T & Gócza E et al. 2013 Induced pluripotent stem cells derived from rabbits exhibit some characteristics of naïve pluripotency. Biology Open 2 613–628. (https://doi.org/10.1242/bio.20134242)
Park Y, Lee J-M, Hwang M-Y, Son G & Geum D 2013 NonO binds to the CpG island of oct4 promoter and functions as a transcriptional activator of oct4 gene expression. Molecules and Cells 35 61–69. (https://doi.org/10.1007/s10059-013-2273-1)
Peynot N, Duranthon V & Khan DR 2015 Gene expression analysis in early embryos through reverse transcription quantitative PCR (RT-qPCR). Methods in Molecular Biology 1222 181–196.
Püschel B & Jouneau A 2015 Whole-mount in situ hybridization to assess advancement of development and embryo morphology. Methods in Molecular Biology 1222 255–265.
R Development Core Team 2010 R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing.
Rizzino A & Wuebben EL 2016 Sox2/Oct4: a delicately balanced partnership in pluripotent stem cells and embryogenesis. Biochimica et Biophysica Acta 1859 780-791. (https://doi.org/10.1016/j.bbagrm.2016.03.006)
Saenz-de-Juano M, Marco-Jimenez F, Viudes-de-Castro M, Lavara R & Vicente J 2014 Direct comparison of the effects of slow freezing and vitrification on late blastocyst gene expression, development, implantation and offspring of rabbit morulae. Reproduction in Domestic Animals 49 505–511. (https://doi.org/10.1111/rda.12320)
Sato N, Kondo M & Arai K 2006 The orphan nuclear receptor GCNF recruits DNA methyltransferase for Oct-3/4 silencing. Biochemical and Biophysical Research Communications 344 845–851. (https://doi.org/10.1016/j.bbrc.2006.04.007)
Schmaltz-Panneau B, Jouneau L, Osteil P, Tapponnier Y, Afanassieff M, Moroldo M, Jouneau A, Daniel N, Archilla C & Savatier P et al. 2014 Contrasting transcriptome landscapes of rabbit pluripotent stem cells in vitro and in vivo. Animal Reproduction Science 149 67–79. (https://doi.org/10.1016/j.anireprosci.2014.05.014)
Senner CE, Krueger F, Oxley D, Andrews S & Hemberger M 2012 DNA methylation profiles define stem cell identity and reveal a tight embryonic-extraembryonic lineage boundary. Stem Cells 30 2732–2745. (https://doi.org/10.1002/stem.1249)
Shi G & Jin Y 2010 Role of Oct4 in maintaining and regaining stem cell pluripotency. Stem Cell Research and Therapy 1 39. (https://doi.org/10.1186/scrt39)
Szczepańska K, Stańczuk L & Maleszewski M 2011 Oct4 protein remains in trophectoderm until late stages of mouse blastocyst development. Reproductive Biology 11 145–156.
Tenayuca J, Cousins K, Yang S & Zhang L 2017 Computational modeling approach in probing the effects of cytosine methylation on the transcription factor binding to DNA. Current Topics in Medicinal Chemistry 17 1778–1787. (https://doi.org/10.2174/1568026617666161116142031)
Tesar PJ, Chenoweth JG, Brook FA, Davies TJ, Evans EP, Mack DL, Gardner RL & McKay RDG 2007 New cell lines from mouse epiblast share defining features with human embryonic stem cells. Nature 448 196–199. (https://doi.org/10.1038/nature05972)
Warrens MJ 2008 On association coefficients for 2 × 2 tables and properties that do not depend on the marginal distributions. Psychometrika 73 777–789. (https://doi.org/10.1007/s11336-008-9070-3)
Yamazaki Y, Fujita TC, Low EW, Alarcón VB, Yanagimachi R & Marikawa Y 2006 Gradual DNA demethylation of theOct4 promoter in cloned mouse embryos. Molecular Reproduction and Development 73 180–188. (https://doi.org/10.1002/mrd.20411)
Yeom YI, Fuhrmann G, Ovitt CE, Brehm A, Ohbo K, Gross M, Hubner K & Scholer HR 1996 Germline regulatory element of Oct-4 specific for the totipotent cycle of embryonal cells. Development 122 881–894.
Zhao X-M, Du W-H, Hao H-S, Wang D, Qin T, Liu Y & Zhu H-B 2012 Effect of vitrification on promoter methylation and the expression of pluripotency and differentiation genes in mouse blastocysts. Molecular Reproduction and Development 79 445–450. (https://doi.org/10.1002/mrd.22052)
