elifeeLifeeLife2050-084XeLife Sciences Publications, Ltd8187510.7554/eLife.81875Research ArticleDevelopmental BiologyGenetics and GenomicsFetal growth delay caused by loss of non-canonical imprinting is resolved late in pregnancy and culminates in offspring overgrowthOberinRuby1†PetautschnigSigrid1†JarredEllen Ghttps://orcid.org/0000-0002-5394-99951QuZhipeng2TsaiTesha1‡YoungsonNeil A13PulsoniGabrielle1TruongThi T4FernandoDilini1BildsoeHeidi1BlücherRheannon O1van den BuuseMaarten5GardnerDavid Khttps://orcid.org/0000-0003-3138-82744SimsNatalie Ahttps://orcid.org/0000-0003-1421-84686AdelsonDavid Lhttps://orcid.org/0000-0003-2404-56362WesternPatrick Shttps://orcid.org/0000-0002-7587-8227patrick.western@hudson.org.au1https://ror.org/02bfwt286Centre for Reproductive Health, Hudson Institute of Medical Research and Department of Molecular and Translational Science, Monash UniversityClaytonAustraliahttps://ror.org/00892tw58Department of Molecular and Biomedical Sciences, School of Biological Sciences, University of AdelaideAdelaideAustraliahttps://ror.org/03r8z3t63School of Biomedical Sciences, University of New South WalesSydneyAustraliahttps://ror.org/01ej9dk98School of BioSciences, University of MelbourneParkvilleAustraliahttps://ror.org/01rxfrp27School of Psychology and Public Health, La Trobe UniversityMelbourneAustraliahttps://ror.org/01ej9dk98Bone Cell Biology and Disease Unit, St. Vincent’s Institute of Medical Research and Department of Medicine at St. Vincent’s Hospital, University of MelbourneFitzroyAustraliaMarstonAdèle LReviewing Editorhttps://ror.org/01nrxwf90University of EdinburghUnited KingdomMarstonAdèle LSenior Editorhttps://ror.org/01nrxwf90University of EdinburghUnited Kingdom
These authors contributed equally to this work
School of Health and Biomedical Sciences, RMIT University, Bundoora, Australia
Germline epigenetic programming, including genomic imprinting, substantially influences offspring development. Polycomb Repressive Complex 2 (PRC2) plays an important role in Histone 3 Lysine 27 trimethylation (H3K27me3)-dependent imprinting, loss of which leads to growth and developmental changes in mouse offspring. In this study, we show that offspring from mouse oocytes lacking the PRC2 protein Embryonic Ectoderm Development (EED) were initially developmentally delayed, characterised by low blastocyst cell counts and substantial growth delay in mid-gestation embryos. This initial developmental delay was resolved as offspring underwent accelerated fetal development and growth in late gestation resulting in offspring that were similar stage and weight to controls at birth. The accelerated development and growth in offspring from Eed-null oocytes was associated with remodelling of the placenta, which involved an increase in fetal and maternal tissue size, conspicuous expansion of the glycogen-enriched cell population, and delayed parturition. Despite placental remodelling and accelerated offspring fetal growth and development, placental efficiency, and fetal blood glucose levels were low, and the fetal blood metabolome was unchanged. Moreover, while expression of the H3K27me3-imprinted gene and amino acid transporter Slc38a4 was increased, fetal blood levels of individual amino acids were similar to controls, indicating that placental amino acid transport was not enhanced. Genome-wide analyses identified extensive transcriptional dysregulation and DNA methylation changes in affected placentas, including a range of imprinted and non-imprinted genes. Together, while deletion of Eed in growing oocytes resulted in fetal growth and developmental delay and placental hyperplasia, our data indicate a remarkable capacity for offspring fetal growth to be normalised despite inefficient placental function and the loss of H3K27me3-dependent genomic imprinting.
oocyteepigeneticinheritanceplacentafetal growthpolycombResearch organismMousehttp://dx.doi.org/10.13039/501100000925National Health and Medical Research CouncilGNT1144966WesternPatrick SGardnerDavid Kvan den BuuseMaartenAdelsonDavid Lhttp://dx.doi.org/10.13039/501100000925National Health and Medical Research CouncilGNT1144887WesternPatrick SGardnerDavid KAdelsonDavid Lhttp://dx.doi.org/10.13039/501100000925National Health and Medical Research CouncilGNT2021247WesternPatrick SAdelsonDavid Lhttp://dx.doi.org/10.13039/100015539Australian GovernmentResearch Training Program ScholarshipJarredEllen GOberinRubyPetautschnigSigridThe funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.Author impact statementGrowth-delayed offspring generated from eggs that lacked normal epigenetic programming exhibited a remarkable capacity to undergo late gestational fetal growth recovery, despite inefficient placental function.Introduction
Epigenetic mechanisms orchestrate tissue development, body patterning, and growth by regulating chromatin accessibility and expression of developmental genes in embryonic and extraembryonic tissues (Maccani and Marsit, 2009; Sun et al., 2021). H3K27me3 is a repressive epigenetic modification catalysed by PRC2, which is composed of the core components, EED, EZH1/2, SUZ12, and RBBP4/7, all of which are essential for histone methyltransferase activity (Pasini et al., 2004; Faust et al., 1995; O’Carroll et al., 2001; Shen et al., 2008; Glancy et al., 2021). Global deletion of any of the individual genes encoding these PRC2 subunits results in embryonic lethality in mice, while conditional deletions of Eed, Ezh2, or Suz12 result in altered patterning and organ function in a range of tissues (Pasini et al., 2004; Prokopuk et al., 2018; Boyer et al., 2006; Hemming et al., 2014; Dudakovic et al., 2015; Majewski et al., 2010; Liu et al., 2019; Sun et al., 2018; Miao et al., 2020). In humans, de novo germline mutations in EED, EZH2, or SUZ12 result in Cohen Gibson, Weaver, and Imagawa-Matsumoto syndromes, which all involve overgrowth, skeletal dysmorphologies, and cognitive abnormalities (Cohen and Gibson, 2016; Tatton-Brown et al., 2011; Cohen et al., 2016; Cooney et al., 2017; Imagawa et al., 2018; Imagawa et al., 2017; Tatton-Brown et al., 2013).
In mouse oocytes, PRC2 is required for non-canonical H3K27me3-imprinting, whereby H3K27me3 silences the maternal allele of several genes resulting in paternal allele-specific expression of the affected genes in pre-implantation embryos and extraembryonic tissues (Hanna and Kelsey, 2021; Inoue et al., 2017). Oocyte-specific deletion of Eed results in loss of H3K27me3-imprinting in preimplantation embryos and extraembryonic ectoderm causing growth restriction and male-biased lethality in mid-gestation offspring (Inoue et al., 2018). In contrast to growth restriction, another study found that postnatal offspring were overgrown following deletion of Eed in oocytes (Prokopuk et al., 2018). Loss of H3K27me3-imprinting in somatic cell nuclear transfer (SCNT) embryos results in placental hyperplasia associated with bi-allelic expression of H3K27me3 imprinted genes, including Xist, Slc38a4, a micro-RNA cluster within Sfmbt2 (C2MC), Gab1, and Smoc1 (Matoba et al., 2018; Inoue et al., 2020; Wang et al., 2020; Xie et al., 2022; Matoba et al., 2022). Moreover, deletion of Eed in oocytes caused placental hyperplasia and placental hyperplasia was partially or completely rescued by homozygous deletion of Slc38a4 or Sfmbt2 together with Eed in oocytes. In addition, homozygous maternal deletion of Xist and Eed rescued male-biased lethality and partially rescued developmental delay and survival in offspring (Matoba et al., 2022).
Here, we show that loss of EED in the oocyte results in initial fetal growth restriction followed by fetal growth recovery and perinatal offspring overgrowth. While fetuses derived from Eed-null oocytes were initially developmentally delayed and growth restricted, they underwent accelerated growth and development facilitating the birth of pups at normal weight from pregnancies of normal gestational length. Moreover, remedial fetal growth occurred in the presence of placental hyperplasia, reduced placental efficiency, low fetal blood glucose, and normal fetal blood amino acid and metabolite levels. Together, this work reveals a complicated offspring growth trajectory that involved in utero resolution of fetal growth restriction and developmental delay, followed by perinatal overgrowth, despite the loss of H3K27me3-dependent imprinting caused by Eed deletion in oocytes.
ResultsLoss of maternal EED compromised mid-gestation survival in offspring
Previous studies demonstrated that deletion of Eed in oocytes resulted in contradictory findings of early embryonic developmental delay (Inoue et al., 2018) and early postnatal offspring overgrowth (Prokopuk et al., 2018). To understand how these outcomes are realised, C57BL/6 Eedfl/fl or Eedfl/wt females were mated to males that were transgenic for Zp3Cre to generate females producing Eed wild-type (wt), Eed heterozygous (het), or Eed-homozygous null (hom) oocytes (Prokopuk et al., 2018). Mating of these females to isogenic C57BL/6 wt males allowed us to generate isogenic wild-type or heterozygous offspring as previously described (Figure 1A; Prokopuk et al., 2018). Heterozygous offspring generated from Eed heterozygous oocytes (HET-het offspring) or from Eed homozygous null oocytes (HET-hom offspring) are isogenic. However, the HET-hom offspring were generated from oocytes that completely lacked functional EED and the HET-het offspring were generated from oocytes that had one functional copy of EED and maintained essentially normal gene repression (Prokopuk et al., 2018; Jarred et al., 2022). As they are isogenic and heterozygous for Eed, comparison of HET-hom with HET-het offspring allowed the identification of differences that resulted specifically from a loss of EED in oocytes in the absence of confounding genetic differences (Figure 1A). In addition, we generated wild-type offspring from Eed-wt and Eed-het oocytes to compare WT-wt and WT-het offspring with HET-het and HET-hom offspring, providing controls for differences generated by Eed heterozygosity in offspring (Figure 1A).
(A) Model used to generate heterozygous isogenic offspring from oocytes that lacked or retained Eed. Females producing Eed wild-type (Eed-wt), Eed heterozygous (Eed-het), and Eed-homozygous (Eed-hom) oocytes were mated with wild-type males to produce offspring from oocytes with wt or het EED-dependent programming or from oocytes that lack EED-dependent programming, respectively. Eed-wt oocytes produce wild-type (WT) offspring (WT-wt control), Eed-het oocytes produce WT or HET offspring (WT-het and HET-het) from GV oocytes, and Eed-hom oocytes produce HET offspring (HET-hom). HET-het and HET-hom offspring are isogenic but were derived from oocytes that had different EED-dependent programming. (B) Cell cleavage and development times of embryos from Eed-wt or Eed-hom oocytes in ex vivo culture. Data is presented as time to reach 2-, 4-, 8-cell, morula (M), blastocyst (B), expanded blastocyst (EB), and hatched blastocyst (HB). *p<0.05, two-tailed student’s t-test, n=49 embryos from Eed-wt oocytes, 57 embryos from Eed-hom oocytes. (C) Time taken for expanded blastocysts to hatch. Data represents the time difference between hatched blastocysts and expanded blastocysts. *p<0.05, n=38 embryos from Eed-wt oocytes, 31 embryos from Eed-hom oocytes. (D) Proportion of embryos from Eed-wt or Eed-hom oocytes surviving to blastocyst. *p<0.05, Data represent the proportion of surviving embryos from each female, n=6 females. (E) Total number of cells per embryo. **p<0.005, n=17 embryos from Eed-wt oocytes, 30 embryos from Eed-hom oocytes. (F) Proportion of trophectoderm (TE) and inner cell mass (ICM) cells per embryo. Data represents the mean proportion of cells allocated to TE versus ICM for embryos from Eed-wt or Eed-hom oocytes. **p<0.05, n=17 embryos from Eed-wt oocytes, 28 embryos from Eed-hom oocytes. (G) Offspring litter size over time. ****p<0.0001. (H) Pie chart depicting the proportion of male and female HET-hom offspring at E17.5. N=32. For (B–G) a two-tailed student’s t-test was used. Error bars: mean ± SD.
(A) Proportion of embryos from Eed wild-type (Eed)-wt or Eed-homozygous (Eed-hom) oocytes surviving to expanded blastocyst (left) and hatched blastocyst (right). Data represent the proportion of surviving embryos from each female, n=6 females. (B) Representative example of a blastocyst differentially stained with propidium iodide and bisbenzimide to identify inner cell mass (ICM; purple cells) and trophectoderm (TE; pink cells). (C) Proportion of TE and ICM cells per embryo. Data represents the mean proportion of cells allocated to TE versus ICM for embryos from Eed-wt or Eed-hom oocytes. **p<0.05, n=17 embryos from Eed-wt oocytes, 28 embryos from Eed-hom oocytes. (A, C): Two-tailed student’s t-test; Error bars represent mean ± SD.
We used automated time-lapse imaging of individual embryos derived from Eed-wt and Eed-hom oocytes to track development from zygote to hatched blastocyst stages. Embryos from Eed-hom oocytes reached the two cell stage 1.08 hr earlier than embryos from Eed-wt oocytes (Figure 1B), but blastocyst expansion and hatching took 2.67 hr longer in HET-hom compared to WT-wt embryos (Figure 1C). All other developmental milestones to blastocyst stage were similar between genotypes (Figure 1B). Although, viability to blastocyst stage was lower in embryos from females producing Eed-hom compared to Eed-wt oocytes (72.17% vs 88.67%, Figure 1D), the proportion of expanded and hatched blastocyst stage embryos did not differ significantly (Figure 1—figure supplement 1A).
To determine whether the cell content of expanded blastocysts was affected, we performed cell counts in whole blastocysts at the end of the culture period using differential staining of the inner cell mass (ICM) and trophectoderm (TE) cells (Gardner and Lane, 2014; Figure 1—figure supplement 1B). This revealed that the overall cell content of embryos from Eed-hom females was lower than in Eed-wt controls (Figure 1E). The number of TE cells was not significantly altered, but the ICM contained fewer cells in HET-hom embryos compared to WT-wt controls (Figure 1F; Figure 1—figure supplement 1C), indicating that the ICM was smaller in embryos from oocytes that lacked EED compared to oocytes that maintained EED function (Figure 1F). As previous reports indicated that maternal deletion of Eed did not compromise development to the blastocyst stage, or cause apoptosis in blastocysts (Inoue et al., 2018), it is likely that lower proliferation explains the reduction we observed in blastocyst cell number, which affected the ICM rather than the TE.
To assess the effect of oocyte-specific Eed deletion on fetal survival and throughout pregnancy, we compared the number of live fetuses and litter size in offspring generated from Eed-hom, Eed-het, and Eed-wt oocytes at multiple stages during gestation and after birth. Using Zp3Cre to delete Eed in oocytes, we found no difference between genotypes in the number of implantations of live embryos in four pregnancies from Eed-null oocytes analysed at embryonic day (E)9.5 (Figure 1G), indicating that HET-hom preimplantation embryos implanted at similar rates and progressed through gastrulation. However, consistent with previous findings (Glancy et al., 2021; Hanna and Kelsey, 2021), the number of live fetuses at E12.5, E14.5, E17.5, and the number of live born pups was significantly lower for Eed-hom mothers than for Eed-het and Eed-wt control females (Figure 1G). In addition, 71.43% of live fetuses in E17.5 Eed-hom pregnancies were female, indicating male-biased lethality of HET-hom offspring (Figure 1H), consistent with previous studies (Inoue et al., 2018; Harris et al., 2019).
Loss of EED in oocytes results in developmental delay followed by fetal growth recovery and postnatal overgrowth
While a previous study observed that deletion of Eed in oocytes caused embryo loss by E6.5 (Matoba et al., 2018), we observed loss of HET-hom offspring between E9.5 and E12.5, indicating that embryo survival and/or placental function was compromised. We, therefore, assessed growth and development of surviving mid-late gestation offspring. At E9.5, HET-hom offspring were small compared to WT-wt offspring (Figure 2A) and combined offspring and extraembryonic tissue weights were significantly lower than HET-het, WT-het, and WT-wt controls (Figure 2—figure supplement 1A). At E12.5 HET-hom embryos contained fewer tail somites and had delayed inter-digital tissue regression in foot plates at E14.5 compared to HET-het, WT-het, and WT-wt controls (Figure 2B, Figure 2—figure supplement 1B), demonstrating that HET-hom growth and development were retarded during early-mid gestation development.
Loss of Embryonic ectoderm development (EED)-dependent oocyte programming resulted in fetal loss and offspring developmental delay.
(A) Representative wholemount images of E9.5 embryos. Images representative of 2–3 litters/genotype. (B) Number of tail somites in E12.5 Eed WT-wt, WT-het, HET-het and HET-hom embryos. ***p<0.0005, ****p<0.0001, n=10–23/genotype. (C–F) Fetal weight at E12.5, E14.5, E17.5, and E18.5. **p<0.005, ***p<0.0005, ****p<0.0001, n=11–117/genotype. (G–H) Offspring body weights at P0 and P3 from litters of all gestational lengths. *p<0.05, **p<0.005, ****p<0.0001, n=30–50/genotype. (I) Gestational length and survival in litters from Eed-wt, Eed-het, and Eed-hom oocytes. **p<0.01, ***p<0.001. Data for each genotype include: Eed-wt 57 pups (P) from 9 litters (L) all born on E19.5; Eed-het 96 pups from 14 litters born on E19.5 and five pups from 1 litter born on E20.5; Eed-hom 43 pups from 9 litters born on E19.5, 23 pups from 7 litters born on E20.5, and 21 pups from 7 litters born on E21.5. (J) Offspring body weight for litters born on E19.5. n=19–28/genotype. (K) Offspring body weights of litters containing 5–7 pups and were born on E19.5. ns = non-significant difference, unpaired t-test, n=19–28/group. (L) Ratios of HET-hom offspring weight to WT-wt, WT-het, and HET-het control weight at E12.5, E14.5, E17.5, E18.5, and E19.5. E19.5 data included only litters born on E19.5. (B–J) Statistically analysed using one-way ANOVA plus Tukey’s multiple comparisons. Error bars: mean ± SD.
Loss of Embryonic ectoderm development (EED)-dependent oocyte programming altered fetal and neonatal development.
(A) Embryo and extraembryonic tissue weight at E9.5. ***p<0.0005, ****p<0.0001, n=17–40/genotype. (B) Representative wholemount images of E14.5 forelimbs reveal reduced inter-digital tissue regression in Eed HET-hom offspring. (C) Number of reabsorbed and/or dead embryos per litter at E17.5. ****p<0.0001, n=10–15 litters/genotype. (A, C) One-way ANOVA with Tukey’s multiple comparisons. Error bars represent mean ± SD.
As in utero developmental delay contrasted markedly with our previous observation that P2 HET-hom offspring were overgrown (Prokopuk et al., 2018) we examined the temporal trajectory of HET-hom offspring by measuring fetal weight at E12.5, E14.5, E17.5, and E18.5 and pup weight on the day of birth (P0) and at P3. HET-hom offspring were significantly lighter than HET-het, WT-het, and WT-wt controls at E12.5, E14.5, E17.5, and E18.5, but by P0 and P3 HET-hom offspring were heavier than isogenic HET-het controls (Figure 2C–H). However, by monitoring plug date and delivery day, we also observed that gestational length was extended in 60% of pregnancies from Eed-hom oocytes. Of 23 litters, 9 (40%) were born on E19.5, 7 (30%) were born on E20.5, and 7 (30%) were born on E21.5. While all pups born on E19.5 survived, 30% and 67% of offspring from pregnancies extended by one or two days respectively, were found dead. In contrast, although one litter from an Eed-het pregnancy was delivered on E20.5, 23 (9 wt and 14 het) were delivered on E19.5 (Figure 2I).
To determine if the extended gestational length of pregnancies from Eed-hom oocytes affected HET-hom offspring weight, we examined pup weight in pregnancies delivered on E19.5 only. This revealed that HET-hom offspring delivered on E19.5 were similar in weight and developmental stage to controls (Figure 2J). Moreover, comparison of data for all surviving HET-hom pups delivered on E19.5-E21.5 (Figure 2G) with those only born on E19.5 (Figure 2J) revealed that increased gestational length contributed to the overall increased pup weight of the population at P3 (Figure 2H). However, ratios of HET-hom fetal weight over WT-wt, WT-het, or HET-het fetal weight at E12.5, E14.5, E17.5, E18.5, and E19.5 (excluding litters of extended gestational length) revealed that the weight deficit in HET-hom was resolved between E14.5 and birth on E19.5 (Figure 2L). Moreover, with the exception of the modest weight deficit, E18.5 HET-hom offspring were indistinguishable from controls at the gross morphological level. Together, these data demonstrate that the gross morphological delay and the fetal weight deficit in HET-hom offspring was resolved by E19.5 and that this occurred independently of litter size. In addition, by P3 HET-hom offspring were heavier than controls. As the overgrown HET-hom offspring in the P3 weight cohort (Figure 2H) included litters born at E19.5 E20.5 and E21.5, both postnatal offspring growth and extended gestation presumably contributed to the overgrowth phenotype.
We also observed increased fetal resorptions and fetal death (Figure 2—figure supplement 1C), raising the possibility that the reduced number of fetuses/pregnancy might result in greater maternal support per fetus leading to normalisation of pup weight at E19.5 (Prokopuk et al., 2018). However, there was no difference in pup weight from Eed-hom and control litters that contained 5–7 pups and were born on E19.5 (Figure 2K), indicating that resolution of fetal growth restriction in utero was at least partly independent of litter size, a finding consistent with our earlier study (Prokopuk et al., 2018). Combined, while HET-hom offspring were initially developmentally delayed and under-weight compared to their isogenic heterozygous counterparts and wild-type controls, HET-hom offspring weight was normalised by birth, and pups were overgrown by P3 (Figure 2J and K), an outcome facilitated by fetal growth recovery and extended gestation.
Oocyte-specific loss of EED results in placental hyperplasia and reduced placental efficiency
To understand whether the normalisation of HET-hom fetal weight might be related to changes in placental development, we collected and weighed E12.5, E14.5, E17.5, and E18.5 placentas from HET-hom, HET-het, WT-het, and WT-wt offspring. At E12.5 placental weights were consistent across all genotypes, but at E14.5 HET-hom placentas were slightly heavier than HET-het controls (Figure 3A–B). By E17.5 HET-hom placentas were 58% heavier than HET-het controls (Figure 3C), and at E18.5 HET-hom placentas were 24% heavier than HET-het placentas (Figure 3D). These changes were reflected in differing HET-hom and HET-het placental growth rates over time with HET-hom placenta growth rates 0.79 and 1.81 times greater than HET-het placentas between E12.5 and E14.5 and E14.5 and E17.5, respectively (Figure 3E–F). Comparison of HET-hom placental to fetal weights revealed that placental growth preceded late-gestational fetal growth (Figure 3E–F). The HET-hom fetal growth rate was slightly lower than HET-het offspring between E12.5 and E14.5, similar between E14.5 and E18.5, but was greater between E18.5 and E19.5 so that HET-hom pups were of equivalent weight to controls at birth on E19.5 (excluding litters with extended gestation; Figures 2C–F, J–3E and G). To further understand the relationship between placental function and offspring growth, we calculated the offspring body to placental weight ratio, which is indicative of placental efficiency (Krombeen et al., 2019). By this measure, placental efficiency was reduced at E12.5, E14.5, E17.5, and E18.5 in HET-hom offspring compared to all other genotypes (Figure 3H–K).
Placental differences observed in heterozygous offspring from oocytes that lacked <italic>Embryonic ectoderm development</italic> (<italic>Eed)</italic>.
(A–D) Placental weight at E12.5, E14.5, E17.5, and E18.5, *p<0.05, **p<0.005, ****p<0.0001, n=20–110/genotype. (E) Heterozygous offspring and placenta average weight over time. ns = non-significant difference, *p<0.05, **p<0.005, ***p<0.0005, ****p<0.0001 unpaired t-test, n=14–64/genotype. (F–G) Tables depicting Eed HET-het and HET-hom offspring and placenta growth rates across three to four time points throughout gestation. (H–K) Ratio of placental weight to fetal weight at E12.5, E14.5, E17.5, and E18.5. *p<0.05, **p<0.005, ***p<0.0005, ****p<0.0001, n=11–108/genotype. (L) Placental cross-sections at E17.5 stained with periodic shiff (PAS). Scale bars = 800 μm. (A–D, H–K). One-way ANOVA with Tukey’s multiple comparisons. Error bars: mean ± SD.
Deletion of <italic>Embryonic ectoderm development</italic> (<italic>Eed)</italic> in the oocyte resulted in similar changes in male and female offspring and placental growth at E14.5 and E17.5.
(A–D) Separated female and male weight data for Eed WT-wt, WT-het, HET-het and HET-hom offspring and placenta at E14.5 and E17.5 n=12–63/genotype. (E–H) Separated female and male cross-sectional area data for placenta, junctional zone, and decidua based on assessment of hematoxylin and eosin (H&E) stained sections through the midpoints of female and male WT-wt, WT-het, HET-het, and HET-hom placenta at E14.5 and E17.5 shown in Figure 4. n=4–6/genotype. (H–J) Separated female and male cross-sectional area data and ratios for labyrinth and area occupied by CD31 positive capillaries (right images) from Figure 5c; (H) labyrinth area (I) labyrinth/placenta area ratio (J) capillary/labyrinth area ratio. (A–H) *p<0.05, **p<0.01, ***p<0.0005, ****p<0.0001, One-way ANOVA with Tukey’s multiple comparisons. Error bars represent mean ± SD.
Loss of EED in the oocyte results in altered developmental patterning of the placenta
While we observed male-biased fetal loss and that E17.5 HET-hom placentas substantially differed from controls, our data revealed that male and female fetuses and placentas were of similar weights at E14.5 and E17.5 (Figure 3—figure supplement 1A–D). To examine male and female placental development in more detail we stained sections from the midline of the placenta with periodic shiff (PAS; Figure 3L) and hematoxylin and eosin (H&E; Figure 4A–B) and used QuPath to quantify placental, junctional zone, decidua areas, and glycogen-enriched cells. To ensure comparative analyses between individuals, we examined sections from the middle of the placentas from E14.5 and E17.5 male and female WT-wt, WT-het, HET-het, and HET-hom offspring. Comparison of male and female samples revealed no sex-specific differences in placental weight or cross-sectional areas of placenta, decidua, or junctional zone, but the cross-sectional areas of placenta, junctional zone, and decidua were increased in HET-hom placentas (Figure 3—figure supplement 1C–G). Interestingly, this lack of sex-specific differences was consistent with similar fetal and placental weight between males and females (Figure 3—figure supplement 1A–D), indicating that male-biased lethality (Figure 1H) was independent of obvious sex-specific changes in the placenta or fetal growth rate.
Deletion of embryonic ectoderm development (<italic>Eed)</italic> in the oocyte altered placental morphology in offspring.
(A,B) Placental cross-sections at E17.5 stained with hematoxylin and eosin (H&E), black dotted line indicates the borders between the decidua (D), the junctional zone (JZ), and the labryrinth (L). Images are representative of four placentas from 10 biological replicates. Yellow arrows indicate expanded regions of glycogen cells in Eed HET-hom placentas. Scale bars = 500 μm. (C–E) Cross-sectional area of the placenta and percentage of placental cross-sectional area occupied by the junctional zone and decidua at E17.5. (F) Average number of glycogen cells/E17.5 placenta cross-section. (G–J) Area occupied by glycogen and non-glycogen enriched cells in the junctional zone and decidua at E17.5 For C-L: **p<0.005, ***p<0.0005, ****p<0.0001, n=10/genotype. (K) E17.5 and E18.5 fetal offspring blood glucose concentration. *p<0.05, n=5–30/genotype. All statistical analyses are one-way ANOVA plus Tukey’s multiple comparisons. Error bars: mean ± SD.
Deletion of <italic>Embryonic ectoderm development</italic> (<italic>Eed)</italic> in the oocyte caused general placental hyperplasia, with pronounced impacts on junctional zone expansion and increased glycogen cell count.
(A–E) Cross-sectional areas of placenta, junctional zone, and decidua in sections through the midpoints of Eed WT-wt, WT-het, HET-het, and HET-hom placenta at E14.5 and E17.5 (sexes combined). n=6–11/genotype. (F) Glycogen cell count in sections through the midpoints of female and male WT-wt, WT-het, HET-het and HET-hom placenta at E17.5. n=4–6/genotype. (A–F) *p<0.05, **p<0.01, ***p<0.0005, ****p<0.0001, one-way ANOVA with Tukey’s multiple comparisons. Error bars represent mean ± SD.
Given the lack of sex-specific changes in placental weight or cross-sectional areas (Figure 3—figure supplement 1), we analysed male and female samples together. H&E and PAS staining indicated that the junctional zone was expanded in HET-hom placentas, with abnormal projections of PAS-stained spongiotrophoblast cells into the labyrinth (Figures 3L and 4A). Consistent with the modest increase in weight of HET-hom compared to HET-het placentas at E14.5 (Figure 3B), there was a 19% increase in cross-sectional area of HET-hom compared to HET-het controls (Figure 4—figure supplement 1A). This difference was emphasised at E17.5, with a 56% increase in HET-hom placental cross-sectional area compared to HET-het controls (Figure 4B–C). At E14.5, junctional zone area was significantly larger in HET-hom placentas compared to the WT-wt and HET-het controls, however, there was no difference in the HET-hom decidua area compared to all other genotypes (Figure 4—figure supplement 1B–C). At E17.5, the overall HET-hom placental cross-sectional area was greater than that of HET-het and WT-wt controls (Figure 4C), including increased total area in the junctional zone and decidua (Figure 4—figure supplement 1D–E). However, as a proportion of the total placental area, the junctional zone was significantly larger, resulting in a higher junctional zone/placental ratio but a similar decidua/placental ratio in HET-hom placentas compared to other genotypes (Figure 4D–E). As H&E and PAS staining indicated that junctional zone was expanded, we used QuPath to determine glycogen cell number and areas occupied by glycogen-enriched cells and non-glycogen cells. The total glycogen cell count was increased in HET-hom placentas but was similarly higher in both sexes (Figure 4F, Figure 4—figure supplement 1F). Moreover, the area occupied by glycogen and non-glycogen cells was increased in the decidua and junctional zones (Figure 4G–J). Together, while the whole HET-hom placenta was larger than controls, the fetally-derived junctional zone was disproportionately expanded and there were significantly more PAS-stained glycogen-enriched cells in midline sections of HET-hom placentas compared to all other genotypes.
As glycogen can be converted to glucose, to determine whether the greater number of glycogen-enriched cells might impact glucose levels in HET-hom offspring, we measured blood glucose levels at E17.5 and E18.5, just prior to parturition. Fetal blood glucose level was initially relatively low and similar in all genotypes at E17.5 (Figure 4K), but increased significantly in WT-wt, HET-wt, and HET-het controls by E18.5. However, the fetal blood glucose level in HET-hom offspring did not increase between E17.5 and E18.5 and were lower than controls at E18.5 (Figure 4K). This indicated that increased circulating glucose was unlikely to explain the accelerated HET-hom fetal growth observed during late gestation.
To determine whether increased fetal capillary area within the labyrinth might underlie a greater potential for maternal-fetal nutrient exchange we performed IF for CD31 (PECAM-1). Sections taken from the middle of each placenta were stained and the total labyrinth area, labyrinth/placental ratio, and fetal capillary/labyrinth ratio were quantified in WT-wt, HET-het, and HET-hom placentas (Figure 5A–B). In this analysis the labyrinth was defined by the CD31-positive capillary enriched region for each genotype and the total labyrinth area was defined as capillary area plus the total space outside the capillaries. The capillary area was defined as the total CD31 positive area plus the unstained space inside the capillaries. HALO analysis revealed a reduction in overall labyrinth area in both HET-het and HET-hom placentas compared to WT-wt controls (Figure 5C), indicating that heterozygosity for Eed reduced labyrinth size in the placenta. However, when total placental size was taken into account, the labyrinth/placental ratio was decreased only in HET-hom placentas compared to WT-wt and HET-het controls (Figure 5D). Moreover, while the capillary area to labyrinth area ratio was unchanged in HET-hom placentas compared to WT-wt controls, this ratio was reduced in HET-hom compared to HET-het placentas (Figure 5E). Together, these data indicate that the area available for maternal-fetal nutrient transfer in the labyrinth of HET-hom placentas was either unaffected or decreased compared to controls.
Labyrinth area and capillary/labyrinth ratios were reduced in offspring from <italic>Embryonic ectoderm development</italic> (<italic>Eed)</italic>-null oocytes.
(A) Immunofluorescent analysis of male and female E17.5 placental sections through the middle of Eed WT-wt, HET-het and HET-hom placentas using CD31 to detect placental vasculature and DAPI to stain cell nuclei. (B) Examples of the strategy used to quantitatively analyse the total placental area and the CD31 positive staining in the labyrinth using CD31 staining (left images), the total labyrinth area, with non-capillary vasculature excluded (middle images), and the total labyrinth area and area occupied by CD31 positive capillaries (right images). (C) Labyrinth area; (D) labyrinth placenta ratio; (E) capillary/labyrinth ratio. One-way ANOVA plus Tukey’s multiple comparisons. *p<0.05, **p<0.005, ***p<0.001, n=8 WT-wt (four male and four female; nine and eight sections), eight HET-het (four male and four female; 9 and 12 sections) and 8 HET-hom (four males and four femalse; 16 and 8 sections) placentas. Error bars: mean ± SD.
Widespread gene dysregulation occurs in placentas from <italic>Eed</italic>-hom oocytes
As the developmental changes observed in the placenta indicated that transcriptional regulation may be altered in HET-hom placentas, we analysed male and female placental tissue from E17.5 HET-hom, HET-het, and WT-wt offspring using RNA-sequencing (RNA-seq). We recovered >20 million clean, mappable reads per sample and analysed 4–6 samples for all genotypes and sexes except male HET-hom, for which only three samples were available. Principal component analysis (PCA) revealed that female HET-hom placental gene expression differed from HET-het and WT-wt placental transcriptomes (Figure 6A). Supporting this, differential gene expression analysis using an FDR <0.05 (false discovery rate) revealed 2083 differentially expressed genes (DEGs; together referred to as Eed placental DEGs) between HET-hom vs HET-het placentas (Figure 6B, Supplementary file 1A) and only four DEGs (Eed, Fibin, Hapln4, & Dtx1) between HET-het and WT-wt placentas. Together, this indicated that loss of Eed in oocytes rather than heterozygosity of Eed in offspring caused the vast majority of transcriptional dysregulation in HET-hom placentas. Gene ontology analysis indicated that the Eed DEGs identified between female HET-hom and HET-het placentas regulate morphogenesis, system development, and multicellular organism development (Figure 6C), suggesting that the gene dysregulation observed influences a diverse range of systems and processes. Surprisingly, PCA revealed that male HET-hom placental transcriptomes were similar to HET-het and WT-wt controls (Figure 6—figure supplement 1A), with only one sequence (CAAA01077340.1) differentially expressed between male HET-hom and HET-het placentas (FDR <0.05). However, comparison of the male and female RNA-seq data without FDR correction identified 478 male DEGs and 3964 female DEGs between HET-het and HET-hom placentas. Of these 227 were common to male and female placentas indicating that they were commonly dysregulated in male and female placentas but the impact in male placentas was potentially less pronounced (Figure 6—figure supplement 1B; Supplementary file 1B-D).
Loss of Embryonic ectoderm developmen (EED)-dependent oocyte programming altered placental transcription.
(A) Principal Component Analysis (PCA) of E17.5 female Eed HET-hom, HET-het, and WT-wt placenta bulk RNA-seq data. n=4–5/genotype. (B) Differential gene expression analysis of E17.5 HET-het vs HET-hom placenta represented by a volcano plot showing logFC against statistical significance. Genes with false discovery rate (FDR)-adjusted FDR <0.05 are coloured in red. Deletion of Eed in the oocyte resulted in 2083 differentially expressed genes (DEGs) in HET-hom placenta (Eed placental DEGs). (C) GO enrichment analysis of HET-het vs HET-hom DEGs representing the top 10 significantly different biological processes impacted. (D) Venn analysis of DEGs identified between E17.5 female HET-het and HET-hom placentas showing non-imprinted, classically imprinted, and H3K27me3 imprinted genes. (E–G) Relative transcription levels of female E17.5 HET-hom placental DEGs including (E) the top 25 non-imprinted genes that were higher and lower in HET-hom vs HET-het samples, (F) H3K27me3 imprinted genes identified (G) Classically imprinted genes identified. Only DEGs with significant expression differences and an FDR <0.05 are shown. In (F) and (G) the typical maternal (red) /paternal (blue)/bi-allelic (gray) expression pattern is indicated below each graph. The dotted black line shows the level for a twofold increase in expression. (H) Representative immunohistochemical analysis of PLAC1 (encoded by the X-linked gene Plac1) in mid-cross sections of E17.5 HET-het and HET-hom placentas. The black dotted lines show boundaries between the decidua (D), the junctional zone (JZ), and the labryrinth (L). n=4/genotype. Scale bars = 500 um (top), 100 um (bottom). (I) Mean methylation difference vs significance for Differentially methylated CpG-rich regions (DMRs) in E17.5 female HET-het vs HET-hom placentas with DMRs within 1 kb of classically imprinted (blue), histone 3 Lysine 27 trimethylation (H3K27me3)-imprinted (red), and differentially expressed non-imprinted genes (green) shown.
Transcriptional analysis using a confidence level of p<0.05 revealed commonly altered genes between E17.5 female vs male <italic>Eed</italic> HET-hom placentas.
(A) Principal Component Analysis (PCA) of E17.5 male Eed HET-hom, HET-het and WT-wt placenta RNA-seq data. n=3–4/genotype. (B) Venn diagram showing overlap between E17.5 female and male placental differentially expressed genes (DEGs) using p<0.05 confidence level, without False Detection Rate applied.
Comparison of female placenta differentially expressed genes (DEGs) revealed differential transcriptional impacts in the junctional zone, labyrinth, and decidua in <italic>Eed</italic> HET-hom placentas.
(A–C) Overlap of placental DEGs with genes preferentially expressed in the junctional zone (Eed junctional zone DEGs), labyrinth (Eed labyrinth DEGs) or decidua (Eed decidua DEGs) in E14.5 wild-type placentas. E14.5 tissue-specific data from Han et al., 2018, Cell (Ref 36). (D–F) Number of Eed junctional zone DEGs, Eed labyrinth DEGs, and Eed decidua DEGs with increased or decreased expression in HET-hom placentas.
Loss of Embryonic ectoderm development (EED) in the oocyte increased PLAC1 expression in the developing placenta, but did not result in overrepresentation for increased expression of X-linked genes.
Representative images showing immunohistochemical analysis of PLAC1 in mid-cross sections of female Eed WT-wt, HET-het and HET-hom placentas. The black dotted lines show boundaries between the decidua (D), the junctional zone (JZ), and the labyrinth (L). n=4/genotype. Scale bars = 500 um (top), 100 um (bottom).
By E14.5 the majority of mature placental cell types and their key cell specific transcriptional programs are established (Woods et al., 2018). As single-cell RNA-seq data from E14.5 C57BL/6 placentas was publicly available (Han et al., 2018), we used these data to gain an indication of the placental cell types in which gene dysregulation occurred in female HET-hom placentas. Of the 2083 Eed placental DEGs identified in female HET-hom placentas, 543 genes were found in the E14.5 mouse placental data set (Han et al., 2018; Supplementary file 1E). Cell-specific comparisons revealed that of these 543 genes, 166 Eed female placental DEGs included genes preferentially expressed in the junctional zone, 57 Eed female placental DEGs were preferentially transcribed in the labyrinth and 104 Eed female placental DEGs were preferentially transcribed in the maternally derived decidua. For ease of reference, these were respectively defined as ‘Eed junctional zone DEGs,’ ‘Eed labyrinth DEGs,’ and ‘Eed decidua DEGs’ (Figure 6—figure supplement 2A–C, Supplementary file 1E). Together these data revealed that gene dysregulation in HET-hom placentas was not isolated to a single region, and that both maternal and fetal-derived cells were altered. Of the Eed junctional zone DEGs 86% were upregulated, whereas 86% and 97% of the Eed labyrinth DEGs and Eed decidua DEGs were downregulated, respectively (Figure 6—figure supplement 2D–F, Supplementary file 1E). Together, these data demonstrate a strong bias for up-regulation of genes expressed in the junctional zone layer of the Eed placenta, which likely reflects the significant expansion observed in the junctional zone as a proportion of the total placental size.
<italic>Eed</italic> placental DEGs included non-imprinted genes, H3K27me3 and classically imprinted genes, and a small number of X-linked genes
The majority (2029/2083; 97.4%) of placental DEGs were bi-allelically expressed autosomal genes that are associated with a range of developmental processes (Figure 6C–E). In addition, of 76 candidate non-canonical H3K27me3 imprinted genes identified by Inoue et al., 2017, 16 were dysregulated in HET-hom vs HET-het placentas, though eight of these genes were also listed as classically imprinted genes in the list we examined (Figure 6C and F). Notably, Slc38a4, Slc38a1, Sfmbt2, Gab1, and Smoc1 are normally paternally expressed (maternally imprinted/silenced), but were upregulated in female HET-hom placentas consistent with de-repression of the maternal allele (Figure 6F, Supplementary file 1F). Loss of H3K27me3-dependent imprinting at these genes has been functionally linked to placental hyperplasia, including in offspring derived from SCNT (Inoue et al., 2020; Wang et al., 2020; Xie et al., 2022; Matoba et al., 2022). Given that both Eed HET-hom offspring and SCNT offspring lack H3K27me3 imprinting, we compared E14.5 and E19.5 fetal and placental weight data collected in this study with SCNT offspring data extracted from Matoba et al., 2018. This revealed very similar fetal and placental growth trajectories in the Eed HET-hom and SCNT offspring models (Figure 7).
<italic>Eed</italic> HET-hom offspring demonstrate similar growth profiles to somatic cell nuclear transfer (SCNT) mice.
Comparison of fetal weights of SCNT mice from Xie et al., (Xie et al., 2022) and offspring from the Eed-ZP3-Cre mouse model at (A) E14.5 and (B) E19.5. N values - E14.5: IVF 9; SCNT 12; Eed WT-wt 64; WT-het 26; HET-het 45; HET-hom 43. E19.5 IVF 17; SCNT 18; Eed WT-wt 21; WT-het 20; HET-het 19; HET-hom 28. **p<0.005, ****p<0.0001. One-way ANOVA with Tukey’s multiple comparisons. Error bars: mean ± SD.
Of the remaining DEGs listed as H3K27me3 imprinted genes (Figure 6F), only Slc38a1 is usually paternally expressed. The remainder are either maternally or bi-allelically expressed indicating that they are unlikely to be regulated by H3K27me3 imprinting in oocytes. In addition to H3K27me3 imprinted genes, classically imprinted genes were also dysregulated in HET-hom placentas. Of 325 candidate or confirmed classically imprinted genes we examined (Andergassen et al., 2017; Wanigasuriya et al., 2020), 22 were increased in expression including 15 paternally expressed (maternally imprinted/silenced) and seven maternally expressed (paternally imprinted/silenced) genes (Figure 6G, Supplementary file 1F). Three (Gab1, Sfmbt2, and Smoc1) of the upregulated paternally expressed genes were also identified in our analysis of H3K27me3 imprinted genes. Similarly, of 24 genes that were decreased in expression, 17 are normally maternally expressed and seven are paternally expressed genes (Figure 6G, Supplementary file 1F).
Maternal EED is required for regulating X-inactivation in pre-implantation embryos and loss of EED in oocytes causes male-biased HET-hom fetal lethality (Figure 1H; Inoue et al., 2018; Harris et al., 2019). Moreover, loss of H3K27me3 imprinting increases maternal Xist expression and deletion of maternal Xist partially rescued growth delay caused by loss of EED in oocytes (Matoba et al., 2022). Xist was not significantly increased in HET-hom placentas in this study. However, the X-linked genes Plac1 and Ldoc1 were increased in female HET-hom compared to HET-het placentas but were not differentially expressed in male placentas raising the possibility that these genes escape X-inactivation in Het-hom placentas (Supplementary file 1B-D). In line with its increased transcription, PLAC1 protein was also increased in cells within the junctional zone (Figure 6H, Figure 6—figure supplement 3). These data are of interest as together with the imprinted genes Ascl2 and Peg3, Plac1, and Ldoc1 are known to regulate placental glycogen stores (Tunster et al., 2020) and have been associated with junctional zone development (Tunster et al., 2016; Tunster et al., 2018; Lee et al., 2015; Allas et al., 2019; Thiaville et al., 2013; Naruse et al., 2014). Despite the increased expression of this subset of X-linked genes, placental DEGs occurred at a similar rate on the X-chromosome (p=0.26626) and autosomes (Supplementary file 1G). Moreover, while there were 81 X-linked DEGs, 32 had increased and 49 had reduced expression (Supplementary file 1A) indicating that there was no bias for increased expression of X-linked genes in HET-hom placentas. However, there was a significant overrepresentation of genes located on chromosomes 6 (p=0.000375) and 15 (p=0.03088) genes encoded by the mitochondrial genome (p=0.00202) in the DEG list (Supplementary file 1G). Included in the mitochondrial sequences were mt-Co1, mt-Co2, and mt-Co3 (mitochondrial cytochrome oxidase C I-III), mt-Nd3, mt-Nd4, and mt-Nd4l (mitochondrial NADH dehydrogenase 3 and 4), and mt-Atp6 (mitochondrial ATP synthase 6). These are of interest given that mitochondrial function is dysregulated in placentas with compromised function (Colson et al., 2021).
DNA methylation changes in E17.5 placentas are found at a subset of DEGs
While loss of EED in oocytes disrupts H3K27me3 imprints (Hanna and Kelsey, 2021; Inoue et al., 2017), our data suggest that other genes may also be affected, perhaps through altered DNA methylation (Jarred et al., 2022). We, therefore, analysed DNA methylation in placentas of E17.5 female WT-wt, HET-het, and HET-hom offspring using RRBS. This identified 81 DMRs between HET-het and HET-hom placentas, 87 DMRs between WT-wt and HET-hom placentas, and 27 DMRs between WT-wt and HET-het placentas (FDR <0.05; Figure 6I; Supplementary file 1H-J). Of the 81 HET-het vs HET-hom DMRs, two were located within or nearby the H3K27me3 imprinted genes Sfmbt2 and Mbnl2. The DMR associated with Sfmbt2 was 25% methylated in the HET-het placentas but 10% methylated in the HET-hom group (Supplementary file 1H), which may explain its increased transcript levels in HET-hom placentas. In addition, of the 81 HET-het vs HET-hom placental DMRs, seven were within 1 kb of the orthologous human region that had H3K27me3 peaks in human oocytes (Xia et al., 2019), including one 201 bp upstream of the SFMBT2 transcriptional start site.
Also included in the 81 HET-het vs HET-hom placental DMRs were four within or nearby the classically imprinted genes Igf2r, Airn, Peg10, Sgce, Jade1, and 2410003L11Rik, and 74 DMRs at non-imprinted loci. The latter included five DMRs within or nearby eight non-imprinted genes (Ntn1, Unc45b, Fam83h, Tiam1, Pcdhga1, Pcdhga5, Pcdhgb5, and Pcdhgb7) that were differentially expressed between HET-het and HET-hom placentas (Figure 6I; Supplementary file 1G). Of these, NTN1 (Netrin1) is of particular interest as it is reduced in the placentas of women with fetal growth restriction and potentially influences placental size by increasing the viability of placental microvascular endothelial cells (Qian-hua et al., 2011; Wang et al., 2011). While Ntn1 is not imprinted, it had reduced DNA methylation, consistent with its increased transcription in HET-hom compared to HET-het placentas (Supplementary file 1H).
Loss of EED in the oocyte did not alter late gestational fetal blood metabolite levels
A previous study of SCNT-derived mice indicated that loss of non-canonical imprinting increased Slc38a4 expression, resulting in placental hypertrophy and increased amino acid transport (Xie et al., 2022). To determine whether altered placental function changed fetal blood amino acid levels and/or fetal or maternal metabolomic state, we collected blood samples from male and female E17.5 WT-wt, HET-het, and HET-hom fetal offspring and their mothers and assessed 356 metabolites using mass spectrometry. One hundred and sixty metabolites were detected across the sample sets. Principal component analysis revealed three separate clusters corresponding to maternal and fetal blood samples and 19 pooled biological quality control samples (PBQCs: included to ensure data reproducibility), demonstrating sensitivity to detect metabolic differences (Figure 8A). The maternal samples clustered separately from the fetal samples, revealing clear metabolic differences between the fetal and maternal bloods. However, HET-hom and WT-wt or HET-het of both sexes were inseparable using PCA (Figure 8B) and there were no significant differences that were consistent between HET-hom and WT-wt or HET-het fetal blood samples of either sex (FDR <0.05; Figure 8C). For example, while 14 metabolites (10 with a mean difference >0.5) were statistically significant between female HET-hom and HET-het fetal blood samples (Figure 8C, Supplementary file 1K), none of the same metabolites differed between female HET-hom and WT-wt samples (Figure 8C, Supplementary file 1L-M) and there were no differences between male HET-hom and HET-het samples (FDR <0.05; Figure 8C). Given we expect similar changes between HET-het and WT-wt controls compared to HET-hom samples, we concluded that the differences detected were unlikely to be due to altered placental function in HET-hom offspring. Moreover, although differences in levels of several amino acids have been reported in fetal blood samples of SCNT-derived animals (Xie et al., 2022), we detected no consistent differences in amino acids in HET-hom fetal blood samples and their counterpart HET-het and WT-wt controls in either sex (Figure 8D). Phenylalanine was modestly increased in female and male HET-hom vs WT-wt samples, but was unchanged between HET-hom and HET-het samples of either sex (Figure 8D).
Loss of Embryonic ectoderm development (EED) in the oocyte did not overtly affect the fetal blood metabolomic state in late gestation offspring.
(A) Principal component analysis (PCA) of male and female E17.5 fetal serum samples, matched maternal serum samples, and pooled serum quality controls (B) PCA plots for metabolites in male and female fetal serum samples from Eed WT-wt, HET-het and HET-hom offspring. (C) Volcano plots showing significant differences and difference in mean levels for metabolites and relative amino acid levels in male and female HET-hom vs HET-het fetal serum samples (D) Relative serum levels for all amino acids in female and male WT-wt, HET-het, and HET-hom offspring. Students t-test. *p<0.05; N=8–12. Error bars: mean ± SD. (A–D) Total sample set: 85 samples: n=8–12 for fetal serum samples (Female: 12 WT-wt; 10 HET-het; 12 HET-hom. Male: 12 WT-wt; 10 HET-het; 8 HET-hom). n=5–9 for maternal serum samples (5 wt; 9 het; 7 hom). Of 356 metabolites tested, 160 were reliably detected in the fetal serum samples.
Finally, serum metabolites of mothers carrying WT-wt, WT-het/HET-het, and HET-hom pregnancies clustered together. Moreover, no significant differences in metabolites were detected between the genotypes indicating that the blood metabolomes of these females were similar late in pregnancy, despite the substantial differences in placental and fetal growth profiles of the HET-hom offspring (Figure 8A). Together, these data indicate that an increased supply of amino acids or other metabolites across the placenta is unlikely to account for the accelerated growth we observed in late gestational HET-hom offspring.
Discussion
Fetal growth restriction is commonly caused by placental insufficiency investigated using surgical or nutritional interventions in animal models (Palliser et al., 2010; Louey et al., 2000; Swanson and David, 2015; Vuguin, 2007). In this study similar outcomes were caused by deleting Eed in the oocyte, resulting in the production of isogenic HET-hom offspring that are epigenetically different to HET-het controls. Initially, offspring were characterised by embryonic and fetal developmental delay, but this was resolved late in gestation. Despite lower placental efficiency indicated by fetal/placental weight ratio, low fetal glucose levels, and unaffected amino acid and metabolomic profiles in late gestation fetal and maternal serum, HET-hom offspring underwent a period of late fetal growth recovery and early postnatal overgrowth. Although placental transcription and DNA methylation was altered at important H3K27me3-imprinted genes and some classically imprinted genes, transcription and DNA methylation were also altered at many non-imprinted sites. In addition, several X-linked genes that have been associated with placental development and function were transcriptionally increased, but we found no evidence for widespread dysregulation of X-inactivation in late-stage placentas. While the underlying cause remains unclear, our data indicate that fetal growth is normalised in HET-hom offspring in the absence of any obvious increase in nutritional support from the placenta and that loss of EED in oocytes causes programming effects on the fetus that are unlikely to be explained by loss of only H3K37me3-dependent imprinting or increased maternal Xist.
A previous study reported that blastocysts from Eed-null oocytes formed at normal rates and were not affected by increased cell death, indicating normal pre-implantation development (Inoue et al., 2018). However, we found that while preimplantation development progressed at similar rates in embryos from Eed wild-type and Eed-null oocytes, blastocysts derived from Eed-null oocytes contained low cell numbers, primarily due to fewer inner cell mass cells. This deficit may be due to low cell proliferation during pre-implantation development caused by loss of EED-dependent oocyte programming or to loss of maternal EED supplied in the mature oocyte (Inoue et al., 2017; Inoue et al., 2018; Jarred et al., 2022; Harris et al., 2019; Erhardt et al., 2003). While the latter is consistent with the established role of PRC2 in driving cell division in stem cells and other cell types (Bracken et al., 2003), other EED-dependent effects in the oocyte may also contribute.
Consistent with previous studies (Matoba et al., 2018; Xie et al., 2022; Matoba et al., 2022), later development of offspring generated from oocytes lacking EED was characterised by developmental delay of the fetus and placental hyperplasia. Placental hyperplasia results from loss of H3K27me3 imprinting in SCNT and from maternal deletion of Eed (Matoba et al., 2018; Xie et al., 2022; Matoba et al., 2022), demonstrating a common impact on placental development in both models. Moreover, comparison of our data with SCNT offspring (Matoba et al., 2018) indicated that fetal growth restriction and subsequent fetal growth recovery occur in Eed-null oocyte and SCNT-derived offspring, and that this phenotype was resolved during pregnancy. While both models involve loss of H3K27me3-dependent imprinting (Inoue et al., 2018; Wang et al., 2020; Xie et al., 2022), the HET-hom offspring in this study were derived from oocytes that lacked maternal Eed/EED (RNA or protein) and SCNT embryos were derived from wild-type enucleated oocytes that presumably contained maternal PRC2. Therefore, while the lack of maternal EED could explain the lower number of cells observed in Eed HET-hom preimplantation offspring, this is less likely in SCNT embryos. However, early development of SCNT embryos is restricted by the low efficiency of oocyte-driven reprogramming (Wang et al., 2020; Matoba et al., 2014; Liu et al., 2016; Cao et al., 2013), and comparison of the earliest stages of development between these models is challenging. Despite this, growth and developmental delay appear to be resolved late in gestation in both models without correction of H3K27me3-imprinting, indicating a level of plasticity that permits delivery of fully grown pups. Even though prolonged gestation decreased pup survival in the Eed model, an interesting conjecture may be that an unknown sensing mechanism in these growth-restricted pregnancies supports adaptation(s) that lead to late fetal growth and extended gestation, favouring the delivery of viable pups. While speculative, such a mechanism could involve epigenetic adaptation during early or late development and/or altered epigenetic control of imprinted or non-imprinted genes in the placenta and/or fetus.
Placental hyperplasia in HET-hom offspring was characterised by an increased number of glycogen-enriched cells and increased junctional zone and decidua area occupied by glycogen-enriched cells and non-glycogen cells. While increased size was observed particularly in the fetally-derived junctional zone, the maternal side of the placenta also became larger, perhaps facilitating overall tissue balance in the placenta in response to the hyperplastic junctional zone. The number of glycogen-enriched trophoblasts usually peaks at E16.5 and declines by approximately 60% by E18.5 (Tunster et al., 2020), potentially releasing glycogen stores to the mother and/or fetus. Consistent with this, fetal blood glucose levels increased between E17.5 and E18.5 in WT-wt, WT-het, and HET-het control offspring. However, despite the increased numbers of glycogen-enriched cells in HET-hom placentas, fetal blood glucose levels did not significantly increase in HET-hom offspring. Moreover, the decreased fetal/placental weight ratio observed in HET-hom offspring suggested decreased placental efficiency and we did not observe any increase in labyrinth area, or the fetal labyrinth/placenta or capillary/labyrinth ratios, suggesting that the area devoted to maternal-fetal nutrient exchange was not substantially affected. Consistent with this, we found no change in levels of fetal blood amino acids or other metabolites, also suggesting that placental function was not obviously enhanced. Together, increased maternal-fetal placental exchange, glucose, or amino acid supply seem unlikely to explain the ability of substantially growth-restricted HET-hom offspring to attain normal weight by birth, indicating that unknown mechanism(s) contribute.
In addition to facilitating increased glucose release and fetal growth, glycogen-producing trophoblasts are considered to inhibit the release of placental factors such as oxytocin and prolactins, which prepare the pregnancy for parturition (Harris et al., 2019; Lee et al., 2015). Therefore, one possibility may be that the increased glycogen cell number in late-gestation HET-hom placentas inhibits the release of oxytocin, prolactins, or other hormones, causing the delay in parturition observed in 60% of the pregnancies from Eed-hom oocytes. Moreover, while extended gestation may also allow additional time for growth delay to resolve, HET-hom pup mortality increased in pregnancies extended by one or two days, indicating that there is a substantial cost for lengthening pregnancy in this model.
As well as morphological alterations observed in HET-hom placentas, we found extensive transcriptional dysregulation with 2083 DEGs identified between HET-hom and HET-het placentas. Consistent with the increased size of fetal and maternal tissues in HET-hom placentas, comparison of these DEGs with single-cell data from E14.5 normal placentas (Han et al., 2018) indicated that altered gene expression was not restricted to genes normally expressed in a single region of the HET-hom placenta, but affected fetal and maternal layers. While one might expect oocyte-specific loss of EED to affect only fetally-derived placental tissue, this ignores the possibility that increased size of the maternal tissue may occur in response to hyperplasia in the fetally-derived layers. Indeed, it seems plausible that the placenta may not properly support late fetal growth without remodelling maternally-derived placental tissue and inducing associated transcriptional change.
Another study demonstrated that loss of H3K27me3 imprinting of Xist contributes to the fetal developmental delay and male-biased fetal death of offspring derived from Eed-null oocytes (Matoba et al., 2022). Oocyte-specific deletion of both Xist and Eed increased survival from approximately 3–6 pups/litter, but a significant deficit in litter size remained compared to the average of ~8.5 pups in wild-type control litters. Moreover, while Xist/Eed maternal double deletion improved fetal body weight, it remained marginally, but significantly lower than wild-type control. Therefore, while deletion of maternal Xist rescued male-biased fetal death observed as a result of Eed deletion in oocytes, it did not completely restore offspring growth or survival, demonstrating that other mechanisms contribute (Matoba et al., 2022). Moreover, the fetal growth recovery of HET-hom offspring described in our study cannot be explained by loss of Xist, indicating that the mechanism involved in the late gestational resolution of HET-hom fetal growth and development remains unresolved.
The DEGs identified in HET-hom hyperplastic placentas included increased expression of H3K27me3 imprinted genes Slc38a4, Sfmbt2, Gab1, and Smoc1, nine classically imprinted genes, X-linked genes including Plac1, Wdr1, and Ldoc1 and many bi-allelically expressed autosomal genes. Consistent with increased Plac1 transcription, immunohistochemistry confirmed that PLAC1 expression was increased in junctional zone cells in HET-hom placentas. However, rather than increased expression, deletion of Plac1 has been associated with male-biased lethality and a similar placental phenotype to that observed in HET-hom offspring (Jackman et al., 2012). Notwithstanding this difference, it seems likely that placental hyperplasia in HET-hom offspring could involve Plac1, Wdr1, Ldoc1, and/or autosomal genes, perhaps in conjunction with Slc38a4 or Sfmbt2/C2MC, which clearly contribute to placental hyperplasia caused by loss of EED in the oocyte (Matoba et al., 2022).
Comparison of HET-hom with HET-het placentas using RRBS identified 81 DMRs, including H3K27me3 imprinted loci Sfmbt2/C2MC and Mbnl2, five classically imprinted loci and 74 non-imprinted loci, indicating that loss of EED in oocytes affects epigenetic status at both imprinted and non-imprinted genes. However, while HET-hom and HET-het offspring are isogenic and these data demonstrate that DNA methylation changes occur in the placenta as a result of deletion of Eed in oocytes, it is unclear that this is a direct effect. Moreover, as we observed 27 DMRs between HET-het and WT-wt placentas heterozygosity for Eed may also contribute, although only four differentially expressed genes, including Eed, were identified in the same comparison. This supports the idea that perturbation of H3K27me3 in the oocyte leads to altered DNA methylation and transcription of imprinted and non-imprinted genes that are influential in placental function and embryo growth. However, further work is required to determine the extent to which epigenetic changes in the oocyte directly affect embryo and placental development in offspring.
In summary, we demonstrate that loss of PRC2 function in the oocyte results initially in embryonic developmental delay and fetal growth restriction, followed by placental hyperplasia and late fetal growth that results in normalisation of offspring weight at birth and perinatal overgrowth (Figure 9). Moreover, it appears that offspring derived from either Eed-null oocytes or by SCNT have an innate ability to correct fetal growth restriction during the late stages of pregnancy despite maternally inherited impacts mediated by Xist or H3K27me3-dependent imprinting. In offspring from Eed-null oocytes normalisation of fetal growth occurs despite low placental efficiency and in the absence of enhanced glucose, amino acid, and metabolite levels, indicating that unknown mechanisms contribute. Together, this work reveals that altered PRC2-dependent programming in the oocyte elicits a complex intrauterine response that supports compensatory fetal growth despite apparently negative impacts on placental development and function. As fetal growth restriction and fetal catch-up growth have been linked with negative health outcomes later in life, including metabolic conditions (Singhal, 2017; Singhal and Lucas, 2004), this model may provide opportunities for understanding the physiological basis of such outcomes. Moreover, given that mutations in EED, EZH2, and SUZ12 have all been associated with overgrowth and a range of co-morbidities in Cohen-Gibson, Weaver, and Imagawa-Matsumoto syndrome patients (Cohen and Gibson, 2016; Tatton-Brown et al., 2011; Cooney et al., 2017; Imagawa et al., 2017; Tatton-Brown et al., 2013), and similar outcomes have been observed in mouse offspring lacking EED in oocytes (Prokopuk et al., 2018), further work may provide insights into these rare human conditions.
Summary of the placental and offspring growth response due to Embryonic ectoderm development (EED) loss in the oocyte.
During oocyte growth, all three subunits of polycomb repressive complex 2 (PRC2) are present at the primary to secondary stages, which is important for the silencing of developmental genes and Histone 3 Lysine 27 trimethylation (H3K27me3)-dependent imprinting (Inoue et al., 2017; Inoue et al., 2018; Matoba et al., 2022; Jarred et al., 2022). Deletion of Eed in oocyte results in loss of maternal H3K27me3 and PRC2, early growth restriction, followed by placental hyperplasia and late gestation fetal catch-up growth, outcomes consistent with loss of H3K27me3-dependent imprinting observed in SCNT offspring (Matoba et al., 2018; Inoue et al., 2020; Wang et al., 2020). Placentas generated from oocytes lacking EED have expanded glycogen-enriched cells in the junctional zone and significant gene dysregulation in the placenta. Despite placental hyperplasia and reduced placental efficiency late in gestation, offspring catch-up growth observed in this model may be explained by loss of imprinting for Slc38a4, increased placental amino acid transport, and extended gestational length, explaining why these offspring are overgrown immediately after birth (Prokopuk et al., 2018).
MethodsMouse strains, animal care, and ethics
Mice were housed using a 12 hr light-dark cycle at Monash Medical Centre Animal Facility, as previously reported (Prokopuk et al., 2018). Room temperature was maintained at 21–23°C with controlled humidity, and food and water were provided ad libitum. All animal work was undertaken in accordance with Monash University Animal Ethics Committee (AEC) approvals. Mice were obtained from the following sources: Zp3Cre mice C57BL/6-Tg 93knw/J; Jackson Labs line 003651, constructed and shared by Professor Barbara Knowles (de Vries et al., 2000), Eed floxed mice (Eedfl/fl) B6; 129S1-Eedtm1Sho/J; Jackson Labs line 0022727; constructed and shared by Professor Stuart Orkin (Yu et al., 2009). The Eed line was backcrossed to a pure C57BL6/J and shared with us by Associate Professor Rhys Allen and Professor Marnie Blewitt, Walter and Eliza Hall Institute for Medical Research, Melbourne.
Genotyping
Genotyping was performed by Transnetyx (Cordova, TN) using real-time PCR assays (details available upon request) designed for each gene as described previously (Prokopuk et al., 2018).
Collection and culture of pre-implantation embryos
Eight to twelve-week-old female mice were superovulated and mated to C57BL/6 males for one night. Zygotes were collected in handling media (G-MOPS PLUS, Vitrolife) at 37 °C (Gardner and Lane, 2014; Gardner and Truong, 2019) denuded of cumulus cells with G-MOPS PLUS containing hyaluronidase. All embryos were washed in G-MOPS PLUS and embryo development kinetics was assessed using the EmbryoScope (Vitrolife) time-lapse imaging system. Embryos were cultured individually in 25 μl of medium, with time-lapse images generated at 15 minute (min) intervals throughout the culture period.
Cell allocation in blastocysts
Following EmbryoScope culture, differential staining was performed (Hardy et al., 1989) in hatched blastocysts using propidium iodide to label TE nuclei, while leaving the ICM unlabelled. After fixation, embryos were treated with bisbenzimide to stain ICM and TE, whole-mounted in glycerol, and imaged using an inverted fluorescence microscope (Nikon Eclipse TS100). Nuclei were counted using ImageJ.
Collection of post-implantation embryos, placenta, and postnatal offspring
Mice were time mated for two-four nights, with females plug checked daily for copulation plugs. Positive plugs were noted as day E0.5 and all females for which a plug was discovered were immediately separated from the male. Gestational length was measured in days post copulation by recording the morning of a copulation plug was detected as E0.5 and visually monitoring females twice daily (morning and afternoon) for births from late gestation (E18.5) until pups were delivered (E19.5-E21.5, depending on oocyte genotype). Pregnant females were euthanised and embryos were collected at E9.5, E12.5 E14.5, E17.5, and E18.5. E9.5 whole sacs were weighed, fixed in 4% PFA for 72 hr at 4 °C, processed, paraffin-embedded, and sectioned at 5 µm. Whole-mount images of E9.5 embryos were taken using a LEICA M80 light microscope with LEICA MC170 HD camera attachment. Embryos and placentas were collected from the same offspring at E12.5, E14.5, E17.5, and E18.5 and weighed separately. Placentas were then bisected, and half was fixed in 4% PFA for 72 hr at 4 °C, processed, and paraffin-embedded with the cut side of the placenta facing the front of the block. The other half of the placenta was rinsed in PBS, snap-frozen on dry ice, and stored at –80 °C for RNA analysis and RRBS. P0 (day of birth) or P3 pups were weighed, euthanised by decapitation, and samples collected as required.
Placental histology
Each block was trimmed to the minimum extent possible to allow collection of full placental sections. Blocks were sectioned in compound series at 5 µm using a Leica microtome and sections transferred to Superfrost plus slides (Thermo Fisher Scientific). With as much accuracy as possible, every section was collected, starting from the midline of each placenta. Two slides at the start of the series were used for H&E and PAS staining and slides from the remaining set were used for PLAC1 immunohistochemistry and CD31 immunofluorescence. Periodic antigen-shiff (PAS) and hematoxylin and eosin (H&E) staining were performed by the Monash Histology Platform (MHTP node) and the H&E and PAS-stained slides and were scanned using an Aperio slide scanner. Quantitative histological analysis was conducted on one complete, intact section located as close to the midline of each placenta as possible. Decidua and junctional zones were histologically identified and the area of each calculated in sections of E14.5 and E17.5 H&E-stained placentas using QuPath v0.2.3 (Bankhead et al., 2017). QuPath was also trained to identify glycogen-enriched and non-enriched cells using a small subsection WT-wt, HET-het, and HET-hom placental sections. Once a robust protocol had been established, machine-learning assisted image analysis provided by QuPath v0.2.3 (Bankhead et al., 2017) was used to quantify Glycogen cell and non-glycogen enriched cells in the junctional zone and decidua in fully intact midline sections of 10 (five male and five female) E14.5 and E17.5 placentas from WT-wt, WT-het, HET-het, and HET-hom mice. Investigators were blinded for sample genotypes throughout quantitative scoring of placental samples analysed.
Immunohistochemistry and image analysis
Slides with 5 µm thick placental sections (described in placental histology) were baked at 60 °C for 20 min. Tissue sections were dewaxed in three changes of xylene and rehydrated in three changes of ethanol then rinsed in distilled water. Antigen retrieval was performed in DAKO PT Link in a DAKO Target Retrieval (Low pH) Solution (DAKO, Cat# S1699) at 98 °C for 30 min. Slides were then washed in DAKO EnVision Flex Wash Buffer (Cat# K8000) for 5 min. IHC was then performed on a DAKO Autostainer Plus in the following steps. Sections were washed once in EnVision Flex Wash Buffer following each subsequent step. Peroxidase Blocking Solution (DAKO, Cat# S2023) was applied for 10 min and non-specific binding was prevented with AffiniPure Fab Fragment Goat Anti-Mouse IgG for 1 hr. Mouse anti-PLAC1 (G-1) (Santa Cruz, Cat# sc-365919) primary antibody was diluted 1/100 in PBS containing 0.1% Triton X-100 (PBST) and 1% Bovine Serum Albumin (BSA, Merck) and applied to the sections for 1 hr. EnVision System-HRP Labelled Polymer Anti-Mouse (DAKO, Cat# K4001) was applied for 1 hr. Immunostaining was visualised using DAKO Liquid DAB + Substrate Chromogen System (Cat# K3468). A counterstain DAKO Automation Haematoxylin Staining Reagent was then applied for 10 min. Slides were removed from the Autostainer, transferred to a slide staining rack, and rinsed in distilled water. In a fume-hood, slides were then washed in Scott’s Tap water and distilled water. Finally, slides were dehydrated in three changes of 100% Ethanol, cleared in three changes of Xylene, and mounted in DPX. Slides were scanned using a VS120 slide scanner (Olympus).
Immunofluorescence and image analysis
Placental sections were de-waxed and processed for antigen retrieval as described above. Non-specific binding was blocked using 5% Bovine Serum Albumin (BSA, Merck) containing 10% Donkey Serum (Merck) for 1 hr at room temperature (RT). Block solution was replaced with PBST containing 1% BSA and goat anti-CD31 antibody (1/100, R&D Systems) and incubated overnight at 4 °C. Samples were washed three times in PBS before incubation with PBST containing 1% BSA and Donkey anti-Goat IgG (H+L) Cross-Adsorbed secondary antibody (Alexa Fluor 555, Thermo Fisher Scientific) for 1 hr at RT. Samples were washed three times in PBS, followed by incubation with TrueView Autofluorescence Quenching Kit (Vector Laboratories) for 5 min at RT according to the manufacturer’s instructions. Following three washes in PBS, samples were incubated in dH2O containing 5 µg/ml DAPI (4',6-Diamidino-2-Phenylindole, Dilactate, Thermo Fisher Scientific) for 15 min at RT. Samples were washed a final three times in PBS, rinsed in dH2O, and mounted in Vectashield Vibrance Antifade Mounting Medium (Vector Laboratories). Slides were scanned using the VS120 Slide scanner (Olympus). Quantification of blood vessels, labyrinth area, and capillary area was achieved using machine learning-assisted image analysis using HALO (Indica Labs). Initially, HALO was trained to recognise the labyrinth area based on CD31 positive capillary staining and to exclude larger blood vessels and other tissue (e.g. the chorionic plate), which were easily distinguishable (Figure 5B). Then HALO was trained to calculate the area of the labyrinth occupied by CD31-positive blood vessels. Once training was complete, HALO was used to analyse all sections across 24 slides (described in placental histology) from male and female placentas of WT-wt, HET-het, and HET-hom placentas. Data was annotated in excel and statistically analysed using One-way ANOVA plus Tukey’s multiple comparisons in GraphPad Prism.
Placental RNA-sequencing and data analyses
Placental isolation is described above. RNA was extracted from 3 to 5 E17.5 placentas of both sexes for each genotype using NucleoSpin RNA Plus columns. RNA quality was assessed on an Agilent Bioanalyser and samples with RIN >7.5 used for library preparation and sequencing on the BGI Genomics platform (BGI Genomics, Hong Kong). Adaptor and low-quality sequences in raw sequencing reads were trimmed using Trimmomatic (Bolger et al., 2014) (v0.39) with the following parameters: LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:20. Clean reads were mapped to the mouse reference genome (GRCm38) using STAR (v2.7.5c) with the following settings: outFilterMismatchNoverLmax 0.03 --alignIntronMax 10000. Raw counts for mouse reference genes (ensembl-release-101) were calculated using STAR (v2.7.5c) with parameter ‘--quantMode GeneCounts’ simultaneously when doing the genome mapping. Differential gene expression analysis was carried out using the R package ‘limma’ (Ritchie et al., 2015) with ‘treat’ function and parameter ‘lfc = log(1.1).’ Statistically significantly differentially expressed genes were identified using ‘FDR <0.05.’ Gene Ontology (GO) enrichment analysis for significantly differentially expressed genes was carried out using The Database for Annotation, Visualisation, and Integrated Discovery (DAVID) with the following settings: GO term level 3, minimum gene count 5, and FDR <0.05 (Dennis et al., 2003).
5 WT-wt, 5 HET-het, and 6 HET-hom placentas from female E17.5 mice were chosen for RRBS. DNA was extracted from a quarter of each placenta with a Qiagen DNeasy Blood & Tissue Kit and eluted in AE buffer (10 mM Tris-HCl, 0.5 mM EDTA). 4–13 μg of DNA was sent to CD Genomics (NY, USA) and underwent sodium bisulphite conversion, library preparation, and Illumina PE150 sequencing at 10 Gb raw data per sample. Sequence analyses were performed on the Galaxy Australia Bioinformatics Platform (https://usegalaxy.org.au/) (Afgan et al., 2022). Sequence files were aligned to the GRC39 (mm39) mouse genome with BWA-meth. Per-base CpG methylation metrics were extracted with MethylDackel with the output limited to CpG sites with a minimum of five reads coverage with output presented as CpG methylation fractions. Differentially methylated CpG-rich regions (DMRs) between the groups were identified with Metilene with settings of a minimum of 10 CpGs per DMR and a minimum of 10% methylation difference between groups. Statistically significant DMRs which were below the threshold of Bonferroni adjusted p-value of q<0.05 were examined further. DMR regions were visualised and nearby genes and orthologous genome regions identified with the UCSC Genome Browser (https://genome.ucsc.edu/ with Table Browser and LiftOver tools) or Ensemble Biomart (https://asia.ensembl.org/info/data/biomart/index.html).
Metabolomics
E17.5 fetuses were isolated from pregnant females and decapitated on ice. ~50 ul of blood was collected from each fetus using a pipette and transferred into tubes containing 2 ul 0.5 M EDTA. collected by centrifugation and frozen at –80 C before metabolic analysis. Cardiac blood and serum samples were also collected from the mothers of each litter using the same approach. In addition, a spot of blood collected in the same way from each fetus and mother was directly applied to an Accu-Check Blood Glucose Test Strips and glucose concentration quantified using an Accu-Check Monitor. Mass spectrometry was performed on 10 ul of serum by Metabolomics Australia and included targeted profiling of 356 metabolites using the Shimadzu GCMS 8050 system. 1 ul of each sample was injected and analysed using internal standards 13C5, 15N1 Valine, and 13C6 Sorbitol. 85 samples and 19 pooled blood quality controls (PBQCs) were analysed, with PBQCs used to ensure technical consistency of the instrument and sample quantifications. Samples were analysed in randomised order, with a PBQC run after every five samples. 160 metabolites were detected with 85% resulting in a coefficient of variation of <30%. Data was analysed using Metaboanalyst 5.0 to assess data quality and multivariate analyses were used to generate principal component analyses, heatmap, and volcano plots. Statistical differences between samples were assessed using an FDR <0.05 to ensure differences detected with 95% confidence after multiple comparisons correction.
Additional informationCompeting interests
No competing interests declared
Author contributions
Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing – original draft, Writing – review and editing
Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing – original draft, Writing – review and editing
Conceptualization, Data curation, Formal analysis, Validation, Investigation, Methodology, Writing – original draft, Project administration, Writing – review and editing
Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review and editing
Conceptualization, Data curation, Formal analysis, Investigation, Methodology
Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review and editing
Data curation, Formal analysis, Investigation, Methodology
Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing – original draft, Writing – review and editing
Data curation, Formal analysis
Formal analysis, Investigation
Data curation, Investigation
Resources, Supervision, Funding acquisition, Writing – review and editing
Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Investigation, Methodology, Writing – review and editing
Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Methodology, Writing – original draft, Project administration, Writing – review and editing
Ethics
All animal work was undertaken in accordance with Monash University Animal Ethics Committee (AEC) approvals issued by Monash University and Hudson Institute Animal Ethics Committees (AEC), approval numbers MMCB/2018/16 and MMCB/2020/37.
Additional files
Summary of genomic and metabolomic data collected from offspring generated from Eed wild-type (<italic>Eed</italic>-wt), Eed heterozygous (<italic>Eed</italic>-het), and Eed homozygous (<italic>Eed</italic>-hom) oocytes.
(A) Eed Placenta differentially expressed genes (DEGs) - List of significant Eed female Placenta DEGs (HET-hom vs HET-het; false discovery rate, FDR <0.05; 2083 DEGs) (B) Differential Gene Expression Analysis Male HET-hom Placenta vs HET-het Placenta (No FDR; p<0.05) (C) Differential Gene Expression Analysis Female HET-hom vs HET-het Placenta (No FDR; p<0.05) (D) Genes Commonly Differentially Expressed in Male and Female HET-hom vs HET-het Placenta (No FDR; P<0.05) (E) Eed Placenta DEGs vs E14.5 single-cell RNA-seq: Comparison of Eed Placenta DEGs list with published single-cell RNA-seq data from E14.5 C57BL/6 placentas (F) Eed Placenta DEGs vs Imprinted genes: Comparison of Eed Placenta DEGs vs Mouse Imprinted genes (G) Eed Placenta DEG relative enrichment per chromosome (H) Eed Placental DMRs HET-het vs HET-hom: List of significant Eed female placenta differentially methylated regions (FDR/q<0.05). (I) Eed Placental DMRs WT-wt vs HET-hom: List of significant Eed female placenta differentially methylated regions (FDR/q<0.05). (J) Eed Placental DMRs WT-wt vs HET-het: List of significant Eed female placenta differentially methylated regions (FDR/q<0.05). (K) Metabolites detected at significantly different levels between E17.5 female HET-hom and HET-het fetal blood samples (FDR <0.05) (L) Metabolites detected at significantly different levels between E17.5 female HET-hom and WT-wt fetal blood samples (FDR <0.05) (M) Metabolites detected at significantly different levels between E17.5 female HET-het and WT-wt fetal blood samples (FDR <0.05).
Data availability
All RNA sequencing and RRBS data have been deposited to the Gene Expression Omnibus (GEO) and are publicly available with accession number GSE210398. The metabolomics data are available at the NIH Common Fund’s National Metabolomics Data Repository (NMDR) website, the Metabolomics Workbench, https://www.metabolomicsworkbench.org where it has been assigned Study ID ST003211. The data can be accessed directly via its Project DOI: http://doi.org/10.21228/M8TR5T. Data for Figures 6 and 8 are included in Supplementary file 1.
The following datasets were generated:
OberinRPetautschnigSJarredEGQuZTsaiTYoungsonNAPulsoniGTruongTTFernandoDBildsoeHBlucherROvan den BuuseMGardnerDKSimsNAAdelsonDLWesternPS2024Fetal growth delay caused by loss of non-canonical imprinting is resolved late in pregnancy and culminates in offspring overgrowthNCBI Gene Expression OmnibusGSE210398
OberinRPetautschnigSJarredEJQuZTsaiTYoungsonNAPulsoniGTruongTTFernandoDBildsoeHBlucherROvan den BuuseMGardnerDKSimsNAAdelsonDLWesternPS2024Fetal growth delay caused by loss of non-canonical imprinting is resolved late in pregnancy and culminates in offspring overgrowthMetabolomics Workbench10.21228/M8TR5T
Acknowledgements
We thank A/Prof John McBain for generously supporting this work. We also thank Prof. Marnie Blewitt for critical comments on the manuscript, the Hudson Institute Animal Research Platform staff for assistance with mouse care, the Monash Histology Platform for assistance with sample preparation and slide scanning, and the Monash Micro Imaging Facility and Hudson Genomics Facility for assistance and technical advice. This project used NCRIS-enabled Metabolomics Australia infrastructure at the University of Melbourne and funded through BioPlatforms Australia. We would like to thank Dr. David De Souza and Dr. Nadeem Elahee Doomun for their advice and for running the metabolomics analyses in this study. This work was supported by grants and research funds from: National Health and Medical Research Project and Ideas Grants GNT1144966 (PSW, DKG, MvdB, DLA), GNT1144887 (PSW, DKG, DLA), and GNT2021247 (PSW, DLA), Hudson Institute of Medical Research, Victorian Government’s Operational Infrastructure Support Program, Australian Government Research Training Program Scholarship support to EGJ, RO, and SP, and a philanthropic donation from Associate Professor John McBain. Metabolomics Workbench is supported by NIH U2C-DK119886 and OT2-OD030544.
ReferencesAfganENekrutenkoAGrüningBABlankenbergDGoecksJSchatzMCOstrovskyAEMahmoudALonieAJSymeAFouillouxABretaudeauANekrutenkoAKumarAEschenlauerACDeSantoADGuerlerASerrano-SolanoBBatutBGrüningBALanghorstBWCarrBRaubenoltBAHydeCJBromheadCJBarnettCBRoyauxCGallardoCBlankenbergDFornikaDJBakerDBouvierDClementsDde Lima MoraisDATaberneroDLLariviereDNasrEAfganEZambelliFHeylFPsomopoulosFCoppensFPriceGRCuccuruGCorguilléGLVon KusterGAkbulutGGRascheHHotzHREguinoaIMakuninIRanawakaIJTaylorJPJoshiJHillman-JacksonJGoecksJChiltonJMKamaliKSudermanKPoterlowiczKYvanLBLopez-DelisleLSargentLBassettiMETangaroMAvan den BeekMČechMBerntMFahrnerMTekmanMFöllMCSchatzMCCrusoeMRRoncoroniMKucherNCoraorNStolerNRhodesNSoranzoNPinterNGoonasekeraNAMorenoPAVidemPMelaniePMandreoliPJagtapPDGuQWeberRJMLazarusRVordermanRHPHiltemannSGolitsynskiySGargSBraySAGladmanSLLeoSMehtaSPGriffinTJJaliliVYvesVWenVNagampalliVKBaconWAde KoningWMaierWBriggsPJThe Galaxy Community2022The galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2022 updateNucleic Acids Research50W345W35110.1093/nar/gkac24735446428AllasLBoumédieneKBaugéC2019Epigenetic dynamic during endochondral ossification and articular cartilage developmentBone12052353210.1016/j.bone.2018.10.00430296494AndergassenDDotterCPWenzelDSiglVBammerPCMuckenhuberMMayerDKulinskiTMTheusslHCPenningerJMBockCBarlowDPPaulerFMHudsonQJ2017Mapping the mouse allelome reveals tissue-specific regulation of allelic expressioneLife6e2512510.7554/eLife.2512528806168BankheadPLoughreyMBFernándezJADombrowskiYMcArtDGDunnePDMcQuaidSGrayRTMurrayLJColemanHGJamesJASalto-TellezMHamiltonPW2017QuPath: open source software for digital pathology image analysisScientific Reports71687810.1038/s41598-017-17204-529203879BolgerAMLohseMUsadelB2014Trimmomatic: a flexible trimmer for Illumina sequence dataBioinformatics302114212010.1093/bioinformatics/btu17024695404BoyerLAPlathKZeitlingerJBrambrinkTMedeirosLALeeTILevineSSWernigMTajonarARayMKBellGWOtteAPVidalMGiffordDKYoungRAJaenischR2006Polycomb complexes repress developmental regulators in murine embryonic stem cellsNature44134935310.1038/nature0473316625203BrackenAPPasiniDCapraMProsperiniEColliEHelinK2003EZH2 is downstream of the pRB-E2F pathway, essential for proliferation and amplified in cancerThe EMBO Journal225323533510.1093/emboj/cdg54214532106CaoFFukudaAWatanabeHKonoT2013The transcriptomic architecture of mouse Sertoli cell clone embryos reveals temporal–spatial-specific reprogrammingReproduction14527728810.1530/rep-12-043523580949CohenASGibsonWT2016EED-associated overgrowth in a second male patientJournal of Human Genetics6183183410.1038/jhg.2016.5127193220CohenASAYapDBLewisMESChijiwaCRamos-ArroyoMATkachenkoNMilanoVFradinMMcKinnonMLTownsendKNXuJVan AllenMIRossCJDDobynsWBWeaverDDGibsonWT2016Weaver syndrome-associated ezh2 protein variants show impaired histone methyltransferase function in vitroHuman Mutation3730130710.1002/humu.2294626694085ColsonASonveauxPDebièveFSferruzzi-PerriAN2021Adaptations of the human placenta to hypoxia: opportunities for interventions in fetal growth restrictionHuman Reproduction Update2753156910.1093/humupd/dmaa05333377492CooneyEBiWSchlesingerAEVinsonSPotockiL2017Novel EED mutation in patient with weaver syndromeAmerican Journal of Medical Genetics. Part A17354154510.1002/ajmg.a.3805527868325DennisGShermanBTHosackDAYangJGaoWLaneHCLempickiRA2003DAVID: database for annotation, visualization, and integrated discoveryGenome Biology4P312734009de VriesWNBinnsLTFancherKSDeanJMooreRKemlerRKnowlesBB2000Expression of cre recombinase in mouse oocytes: a means to study maternal effect genesGenesis2611011210686600DudakovicACamilleriETXuFRiesterSMMcGee-LawrenceMEBradleyEWParadiseCRLewallenEAThalerRDeyleDRLarsonANLewallenDGDietzABSteinGSMontecinoMAWestendorfJJvan WijnenAJ2015Epigenetic control of skeletal development by the histone methyltransferase ezh2The Journal of Biological Chemistry290276042761710.1074/jbc.M115.67234526424790ErhardtSSuIHSchneiderRBartonSBannisterAJPerez-BurgosLJenuweinTKouzaridesTTarakhovskyASuraniMA2003Consequences of the depletion of zygotic and embryonic enhancer of zeste 2 during preimplantation mouse developmentDevelopment1304235424810.1242/dev.0062512900441FaustCSchumacherAHoldenerBMagnusonT1995The eed mutation disrupts anterior mesoderm production in miceDevelopment12127328510.1242/dev.121.2.2737768172GardnerDKLaneM2014Mammalian Preimplantation embryo cultureLewandoskiMMouse Molecular Embryology: Methods and ProtocolsBoston, MASpringer US16718210.1007/978-1-60327-292-6_11GardnerDKTruongTT2019Culture of the mouse preimplantation embryoMethods in Molecular Biology2006133210.1007/978-1-4939-9566-0_231230270GlancyECiferriCBrackenAP2021Structural basis for PRC2 engagement with chromatinCurrent Opinion in Structural Biology6713514410.1016/j.sbi.2020.10.01733232890HanXWangRZhouYFeiLSunHLaiSSaadatpourAZhouZChenHYeFHuangDXuYHuangWJiangMJiangXMaoJChenYLuCXieJFangQWangYYueRLiTHuangHOrkinSHYuanGCChenMGuoG2018Mapping the mouse cell atlas by microwell-seqCell1721091110710.1016/j.cell.2018.02.00129474909HannaCWKelseyG2021Features and mechanisms of canonical and noncanonical genomic imprintingGenes & Development3582183410.1101/gad.348422.12134074696HardyKHandysideAHWinstonRM1989The human blastocyst: cell number, death and allocation during late preimplantation development in vitroDevelopment10759760410.1242/dev.107.3.5972612378HarrisCCloutierMTrotterMHintenMGayenSDuZXieWKalantryS2019Conversion of random X-inactivation to imprinted X-inactivation by maternal PRC2eLife8e4425810.7554/eLife.4425830938678HemmingSCakourosDIsenmannSCooperLMenicaninDZannettinoAGronthosS2014EZH2 and KDM6A act as an epigenetic switch to regulate mesenchymal stem cell lineage specificationStem Cells3280281510.1002/stem.157324123378ImagawaEHigashimotoKSakaiYNumakuraCOkamotoNMatsunagaSRyoASatoYSanefujiMIharaKTakadaYNishimuraGSaitsuHMizuguchiTMiyatakeSNakashimaMMiyakeNSoejimaHMatsumotoN2017Mutations in genes encoding polycomb repressive complex 2 subunits cause weaver syndromeHuman Mutation3863764810.1002/humu.2320028229514ImagawaEAlbuquerqueEVAIsidorBMitsuhashiSMizuguchiTMiyatakeSTakataAMiyakeNBoguszewskiMCSBoguszewskiCLLerarioAMFunariMAJorgeAALMatsumotoN2018Novel SUZ12 mutations in weaver-like syndromeClinical Genetics9446146610.1111/cge.1341530019515InoueAJiangLLuFSuzukiTZhangY2017Maternal H3K27me3 controls DNA methylation-independent imprintingNature54741942410.1038/nature2326228723896InoueAChenZYinQZhangY2018Maternal Eed knockout causes loss of H3K27me3 imprinting and random X inactivation in the extraembryonic cellsGenes & Development321525153610.1101/gad.318675.11830463900InoueKOgonukiNKamimuraSInoueHMatobaSHiroseMHondaAMiuraKHadaMHasegawaAWatanabeNDodoYMochidaKOguraA2020Loss of H3K27me3 imprinting in the Sfmbt2 miRNA cluster causes enlargement of cloned mouse placentasNature Communications11215010.1038/s41467-020-16044-832358519JackmanSMKongXFantME2012Plac1 (placenta-specific 1) is essential for normal placental and embryonic developmentMolecular Reproduction and Development7956457210.1002/mrd.2206222729990JarredEGQuZTsaiTOberinRPetautschnigSBildsoeHPedersonSZhangQHStringerJMCarrollJGardnerDKVan den BuuseMSimsNAGibsonWTAdelsonDLWesternPS2022Transient polycomb activity represses developmental genes in growing oocytesClinical Epigenetics1418310.1186/s13148-022-01400-w36544159KrombeenSKShankarVNooraiRESaskiCASharpJLWilsonMEWilmothTA2019The identification of differentially expressed genes between extremes of placental efficiency in maternal line gilts on day 95 of gestationBMC Genomics2025410.1186/s12864-019-5626-030925895LeeSYeAKimJ2015DNA-binding motif of the imprinted transcription factor PEG3PLOS ONE10e014553110.1371/journal.pone.014553126692216LiuWLiuXWangCGaoYGaoRKouXZhaoYLiJWuYXiuWWangSYinJLiuWCaiTWangHZhangYGaoS2016Identification of key factors conquering developmental arrest of somatic cell cloned embryos by combining embryo biopsy and single-cell sequencingCell Discovery21601010.1038/celldisc.2016.1027462457LiuPPXuYJDaiSKDuHZWangYYLiXGTengZQLiuCM2019Polycomb protein eed regulates neuronal differentiation through targeting sox11 in hippocampal dentate gyrusStem Cell Reports1311513110.1016/j.stemcr.2019.05.01031204298LoueySCockMLStevensonKMHardingR2000Placental insufficiency and fetal growth restriction lead to postnatal hypotension and altered postnatal growth in sheepPediatric Research4880881410.1203/00006450-200012000-0001811102551MaccaniMAMarsitCJ2009Epigenetics in the placentaAmerican Journal of Reproductive Immunology62788910.1111/j.1600-0897.2009.00716.x19614624MajewskiIJRitchieMEPhipsonBCorbinJPakuschMEbertABusslingerMKosekiHHuYSmythGKAlexanderWSHiltonDJBlewittME2010Opposing roles of polycomb repressive complexes in hematopoietic stem and progenitor cellsBlood11673173910.1182/blood-2009-12-26076020445021MatobaSLiuYLuFIwabuchiKAShenLInoueAZhangY2014Embryonic development following somatic cell nuclear transfer impeded by persisting histone methylationCell15988489510.1016/j.cell.2014.09.05525417163MatobaSWangHJiangLLuFIwabuchiKAWuXInoueKYangLPressWLeeJTOguraAShenLZhangY2018Loss of h3k27me3 imprinting in somatic cell nuclear transfer embryos disrupts post-implantation developmentCell Stem Cell2334335410.1016/j.stem.2018.06.00830033120MatobaSKozukaCMiuraKInoueKKumonMHayashiROhhataTOguraAInoueA2022Noncanonical imprinting sustains embryonic development and restrains placental overgrowthGenes & Development3648349410.1101/gad.349390.12235483741MiaoXSunTBarlettaHMagerJCuiW2020Loss of RBBP4 results in defective inner cell mass, severe apoptosis, hyperacetylated histones and preimplantation lethality in mice†Biology of Reproduction103132310.1093/biolre/ioaa04632285100NaruseMOnoRIrieMNakamuraKFuruseTHinoTOdaKKashimuraMYamadaIWakanaSYokoyamaMIshinoFKaneko-IshinoT2014Sirh7/Ldoc1 knockout mice exhibit placental P4 overproduction and delayed parturitionDevelopment1414763477110.1242/dev.11452025468940O’CarrollDErhardtSPaganiMBartonSCSuraniMAJenuweinT2001The polycomb-group gene Ezh2 is required for early mouse developmentMolecular and Cellular Biology214330433610.1128/MCB.21.13.4330-4336.200111390661PalliserHKZakarTSymondsIMHirstJJ2010Progesterone receptor isoform expression in the guinea pig myometrium from normal and growth restricted pregnanciesReproductive Sciences1777678210.1177/193371911037151720595710PasiniDBrackenAPJensenMRLazzerini DenchiEHelinK2004Suz12 is essential for mouse development and for EZH2 histone methyltransferase activityThe EMBO Journal234061407110.1038/sj.emboj.760040215385962ProkopukLStringerJMWhiteCRVossenRWhiteSJCohenASAGibsonWTWesternPS2018Loss of maternal EED results in postnatal overgrowthClinical Epigenetics109510.1186/s13148-018-0526-830005706Qian-huaWShao-pingZJian-wenZYunYLiZ2011Reduced expression of netrin-1 is associated with fetal growth restrictionMolecular and Cellular Biochemistry350818710.1007/s11010-010-0684-221193949RitchieMEPhipsonBWuDHuYLawCWShiWSmythGK2015limma powers differential expression analyses for RNA-sequencing and microarray studiesNucleic Acids Research43e4710.1093/nar/gkv00725605792ShenXLiuYHsuYJFujiwaraYKimJMaoXYuanGCOrkinSH2008EZH1 mediates methylation on histone H3 lysine 27 and complements EZH2 in maintaining stem cell identity and executing pluripotencyMolecular Cell3249150210.1016/j.molcel.2008.10.01619026780SinghalALucasA2004Early origins of cardiovascular disease: is there a unifying hypothesis?Lancet3631642164510.1016/S0140-6736(04)16210-715145640SinghalA2017Long-term adverse effects of early growth acceleration or catch-up growthAnnals of Nutrition & Metabolism7023624010.1159/00046430228301849SunBChangEGerhartlASzeleFG2018Polycomb protein eed is required for neurogenesis and cortical injury activation in the subventricular zoneCerebral Cortex281369138210.1093/cercor/bhx28929415247SunLFuXMaGHutchinsAP2021Chromatin and epigenetic rearrangements in embryonic stem cell fate transitionsFrontiers in Cell and Developmental Biology963730910.3389/fcell.2021.63730933681220SwansonAMDavidAL2015Animal models of fetal growth restriction: considerations for translational medicinePlacenta3662363010.1016/j.placenta.2015.03.00325819810Tatton-BrownKHanksSRuarkEZachariouADuarteSDVRamsayESnapeKMurrayAPerdeauxERSealSLovedayCBankaSClericuzioCFlinterFMageeAMcConnellVPattonMRaithWRankinJSplittMStrengerVTaylorCWheelerPTempleKIColeTDouglasJRahmanN2011Germline mutations in the oncogene EZH2 cause weaver syndrome and increased human heightOncotarget21127113310.18632/oncotarget.38522190405Tatton-BrownKMurrayAHanksSDouglasJArmstrongRBankaSBirdLMClericuzioCLCormier-DaireVCushingTFlinterFJacquemontMLJossSKinningELynchSAMageeAMcConnellVMedeiraAOzonoKPattonMRankinJShearsDSimonMSplittMStrengerVStuurmanKTaylorCTitheradgeHVan MaldergemLTempleIKColeTSealSRahmanN2013Weaver syndrome and EZH2 mutations: clarifying the clinical phenotypeAmerican Journal of Medical Genetics. Part A161A2972298010.1002/ajmg.a.3622924214728ThiavilleMMHuangJMKimHEkramMBRohTYKimJ2013DNA-binding motif and target genes of the imprinted transcription factor PEG3Gene51231432010.1016/j.gene.2012.10.00523078764TunsterSJMcNamaraGICreethHDJJohnRM2016Increased dosage of the imprinted Ascl2 gene restrains two key endocrine lineages of the mouse PlacentaDevelopmental Biology418556510.1016/j.ydbio.2016.08.01427542691TunsterSJBoqué-SastreRMcNamaraGIHunterSMCreethHDJJohnRM2018Peg3 deficiency results in sexually dimorphic losses and gains in the normal repertoire of placental hormonesFrontiers in Cell and Developmental Biology612310.3389/fcell.2018.0012330320110TunsterSJWatsonEDFowdenALBurtonGJ2020Placental glycogen stores and fetal growth: insights from genetic mouse modelsReproduction159R213R23510.1530/REP-20-000732191912VuguinPM2007Animal models for small for gestational age and fetal programming of adult diseaseHormone Research6811312310.1159/00010054517351325WangQZhuJZouLYangY2011Role of axonal guidance factor netrin-1 in human placental vascular growthJournal of Huazhong University of Science and Technology. Medical Sciences3124625010.1007/s11596-011-0261-221505994WangLYLiZKWangLBLiuCSunXHFengGHWangJQLiYFQiaoLYNieHJiangLYSunHXieYLMaSNWanHFLuFLLiWZhouQ2020Overcoming intrinsic h3k27me3 imprinting barriers improves post-implantation development after somatic cell nuclear transferCell Stem Cell2731532510.1016/j.stem.2020.05.01432559418WanigasuriyaIGouilQKinkelSATapia Del FierroABeckTRoperEABreslinKStringerJHuttKLeeHJKeniryARitchieMEBlewittME2020Smchd1 is a maternal effect gene required for genomic imprintingeLife9e5552910.7554/eLife.5552933186096WoodsLPerez-GarciaVHembergerM2018Regulation of placental development and its impact on fetal growth-new insights from mouse modelsFrontiers in Endocrinology957010.3389/fendo.2018.0057030319550XiaWXuJYuGYaoGXuKMaXZhangNLiuBLiTLinZChenXLiLWangQShiDShiSZhangYSongWJinHHuLBuZWangYNaJXieWSunYP2019Resetting histone modifications during human parental-to-zygotic transitionScience36535336010.1126/science.aaw511831273069XieZZhangWZhangY2022Loss of Slc38a4 imprinting is a major cause of mouse placenta hyperplasia in somatic cell nuclear transferred embryos at late gestationCell Reports3811040710.1016/j.celrep.2022.11040735196486YuMRivaLXieHSchindlerYMoranTBChengYYuDHardisonRWeissMJOrkinSHBernsteinBEFraenkelECantorAB2009Insights into GATA-1-mediated gene activation versus repression via genome-wide chromatin occupancy analysisMolecular Cell3668269510.1016/j.molcel.2009.11.0021994182710.7554/eLife.81875.sa0Editor's evaluationMarstonAdèle LReviewing Editorhttps://ror.org/01nrxwf90University of EdinburghUnited Kingdom
This important study shows that a lack of Polycomb-dependent epigenetic programming in the oocyte and early embryo influences the developmental trajectory through gestation in the mouse. The authors provide convincing evidence for a two-phase outcome of early growth restriction followed by enhancement, addressing previous inconsistencies in the field. The work establishes a link between the function of a protein in oocytes and programming of fetal growth and placental function in late gestation, though the underlying functional relationshsip between increased fetal growth and placental function will require further investigation. This manuscript will interest scientists within the fields of developmental biology and epigenetics.
10.7554/eLife.81875.sa1Decision letterMarstonAdèle LReviewing Editorhttps://ror.org/01nrxwf90University of EdinburghUnited KingdomCharalambousMarikaReviewerhttps://ror.org/0220mzb33King's College LondonUnited Kingdom
Our editorial process produces two outputs: i) public reviews designed to be posted alongside the preprint for the benefit of readers; ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.
Decision letter after peer review:
Thank you for submitting your article "Loss of EED in the oocyte causes initial fetal growth restriction followed by placental hyperplasia and offspring overgrowth" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Carlos Isales as the Senior Editor. The following individual involved in the review of your submission has agreed to reveal their identity: Marika Charalambous (Reviewer #3).
The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.
Essential revisions (for the authors):
The three reviewers recognized the merit of the work and its potential for reconciling previously conflicting results about the role of maternally inherited H3K27me3 on a growth trajectory. However, they also shared strong concerns about the interpretation of the data and in particular, the existence of a placenta-dependent catch-up phase. Additionally, it was not formally demonstrated that H3K27me3 patterns are altered in the placenta of embryos derived from EED-null oocytes, making it difficult to understand direct and indirect gene expression changes.
Please find below four points that require necessary revision. In addition, answer the individual comments of the reviewers in a detailed rebuttal letter and tone down or revise conclusions in your text accordingly.
1. Provide a more in-depth characterization of the placenta phenotype. Please first revise placental efficiency measurements as stated by Reviewer #3. Then, are there some sex-specific differences in placental alterations (considering that the placental transcriptome is differentially affected in males and females from EedKO oocytes)? Is the effect limited to the junctional zone only, and in particular to glycogen cells? The increase in the endocrine compartment may have been emphasized by the chosen methodology. Improved calculation of the different layers, within the same gestational age (E17.5) is required.
2. Whether the catch-up phase may be due to placental compensation prior to birth requires better support, especially in face of the increased gestational length and different post-coitum days of birth between mutants and WT. So far, enhanced growth is reported to happen between E17.5 and P3. The catch-up phase may equally happen prior to birth, in link with a potential placental role, or after birth, in relation to increased feeding, for example. It is really necessary to document when the catch-up phase occurs relative to birth. A measure at E19.5 was done but is only presented in Sup material: there, there is no difference between WT and mutants, which would suggest that extended gestation itself may lead to overgrowth. Please be clearer and report data in post-coitum days rather than postnatal days for WT and mutants, and add additional time points if necessary. Also, there seems to be no former evidence that placentomegaly can enhance fetal growth, and there could equally be an intrinsic fetal contribution to the phenotype.
3. Provide more clarity in the methods section, as descriptions are currently too limited to allow a proper understanding of the work carried out.
4. To claim potential epigenetic programming of placental phenotype and gene expression via Polycomb-dependent mechanisms, H3K27me3 mapping (ChIP-seq or CUT&RUN) would be required. This should be performed on placental samples at the stage when RNA-seq was done. Correlation analyses between gene expression and H3K27me3 changes (if any) are needed. It would be also interesting to compare with H3K27me3 patterns present in the oocyte (for which ChIP-seq data are publicly available).
Reviewer #1 (Recommendations for the authors):
In this current version, the weaknesses of the paper far outweigh the strengths. I believe further investigations are needed to improve the manuscript, in particular by validating several of the claims and conclusions. Accordingly, the authors are encouraged to consider the following (please see the justification for most of these under 'specific points'):
a) Provide a more in-depth characterization of the placenta hyperplasia phenotype to answer unequivocally the following questions:
- Is the surface area for nutrient transport increased in hom mutants vs het mutants? Suggested type of experiments – stereological measurements of the labyrinthine layer looking at fetal capillaries, maternal blood spaces, trophoblast area, and interhemal membrane thickness (a thinner membrane might indicate increased diffusion of nutrients). (Note: The need for these experiments depends on the outcome of other preliminary measurements such as volume calculations as suggested in specific points)
- Is transplacental flux increased in the hyperplastic placentas? This can be done by placental transfer assays with radiolabelled isotopes (glucose; amino acids – e.g. MeAIB; markers of diffusion such as mannitol injected to the maternal tail vein). Measuring amino-acid concentrations in fetal plasma and fetal/maternal glucose gradients would also help to provide evidence that the placenta expansion is indeed responsible for fetal catch-up growth.
- Is the junctional zone hyperplasia restricted to glycogen cells? Measuring the area occupied by spongiotrophoblasts would answer this question and add value to the paper.
- Is there evidence for sexual dimorphism related to the placenta hyperplasia phenotype? The authors did a transcriptomic analysis on male and female placentas and found very few changes in male placentas but widespread alterations in female placentas. It will be important to show that male and female placentas are hyperplastic to the same degree, so providing weights and crude histology for male vs female placentas is suggested.
- Do hyperplastic placentas show dysregulated epigenetic marking? Given that the gist of the paper is to show oocyte epigenetic programming impacting placental development, it would be important to show if there is an association with altered epigenetic marking, H3K27me3/DNA methylation (e.g. global levels, or on target genes thought to cause hyperplasia)
b) Provide alternative explanations and increase critical appraisal of data
- Why is the impact of oocyte epigenetic programming on fetal organs dismissed? It is possible that fetal organs become hyperplasic after the placenta at E17.5 and that epigenetic de-regulation also affects fetal organs? In that regard, is there evidence for disproportionate organ growth at P0 and P3?
- Could fetal and postnatal overgrowth be a secondary outcome of the increased gestational length, and therefore unrelated to placental hyperplasia and hypothetical increase in nutrient supply? It would help to show the fetal and placental weights, as well as P0 body weights, of the litters with normal gestational length vs +1 day, +2 days. If the effect on fetal catch-up growth is driven by placental hyperplasia it might be expected that offspring with normal gestational length are indistinguishable in weight from those with increased gestation length. A stratification of the weight data by gestational length and sex would clarify some of these issues.
Specific points
Data presented in Figure 4B-E shows that the junctional zone (Jz) is disproportionately bigger at E17.5 in Homozygous mutants in relation to the labyrinth (Lz) and maternal decidua, compared to control genotypes. Figure Supp Figure 3G suggests that the labyrinth may also show evidence of hyperplasia when comparing het vs hom as Lz raw area is increased but this is not the case when compared to wt (wt-het) controls. Therefore, the question of whether the phenotype is restricted to the Jz, and the glycogen cells, in particular, remains unresolved and requires more in-depth investigations by calculating volumes of the different layers, adding Jz to Lz ratios as graphs, and depending on the results then proceed with further stereological measurements within the Lz to confirm hyperplasia of this layer.
Please explain why only two groups were compared for the data that generated Figure 1B-F, why 3 groups were used in Figure 1G, and then for most of the remaining figures comparisons with 4 groups were made.
Figure 1A (bottom panel F1) – where is the evidence that WT-wt and WT-het are epigenetically different? Moreover, this study does not directly show that HET-het and HET-hom are epigenetically different, and as such a reference to those studies should be provided next to 'epigenetically different' text in panels.
Figure 1C- The variability of the data is quite considerable (4- 5 fold) but yet the error bars are rather small for those to represent SD. Could a mistake have been made and these represent SEM instead? If yes, please check carefully all other figures and figure legends. In general, it is good practice that the primary data for each figure panel, when appropriate, is provided as supplementary data.
Figure 3 (g-i) – the graph shows placental to fetal weight ratios, which does not match the text on the Y axis (i.e. body weight/placental weight). Most importantly, the graph should show fetal to placenta (F/P) ratios (and not P/F as shown)
Figure 4G – It is important to establish the extent to which the defect is limited to glycogen cells, i.e. are the number or area occupied by spongiotrophoblasts also affected (this can be done with staining for Tpbpa)? A new graph with this information should be added to this figure.
Figure 5 – I find these calculations (D-I) potentially misleading as these only contain a subset of all the potential E17.5 DEGs. I am of the opinion that this data does not merit being in the main figure and should be interpreted with caution. Instead in the main figure, it is more important to present the top DEG genes and validations by QPCR in independent samples of those and the key genes that are thought to contribute to hyperplasia. Epigenetic analysis for these selected genes, could/should be added to improve the paper in line with providing further novelty.
Line 116 – specify the blastocyst stage: all hatched blastocysts?
Line 125 – Measuring cell proliferation would confirm the findings and add value to the paper,
Line 131 – how does Figure 1D relate to Figure 1G? i.e. this is confusing – in vitro data suggests loss of blastocyst numbers but there is no difference in the number of implantation sites in vivo. please clarify in the main text.
Line 148 – please rephrase, it gives the impression that extra-embryonic tissues are lighter, but these were not measured in isolation.
Line 166 (and throughout the text) – placental efficiency is not enhanced but instead decreased. The definition of placental efficiency is 'gram of fetus produced by gram of placenta' – if Hom placentas are overgrown from E14.5 onwards and fetuses growth restricted at least until E17.5, how can this be evidence of increased placenta efficiency if fewer grams of fetus are being produced per gram of placenta?
The authors would also require a later time point between E17.5 and E19.5 if they want to validate any transient rise in placental 'efficiency' – that data is missing in the current version. In any case, it seems clear to me that placentae up to and including E17.5 are not more efficient.
Lines 189-190 – How prevalent are the spongiotrophoblast abnormal 'projections' compared to controls? Also, Figure 3J seems redundant with Figure 4A. What is the added value of 3J?
Line 198 – specify if these are midline sections (this could be in Materials and methods).
Line 207 – Where is the evidence that the overall cross-sectional area of the Labyrinth in hom-placentas is higher compared to other genotypes? At E14.5 that is not the case and at E17.5 it does not seem to be different from WTs (wt and het; as shown in Supp Figure 3G). Further analyses are thus required. This is an important point because if the Lz layer turns out not to be increased over wild-type there is no structural basis to substantiate the claim that the hyperplastic placentas are primary determinants of the catch-up growth (and also having in mind that there is no established role for glycogen stores in the placenta in determining fetal growth).
Line 208 – Should read Supp Figure 3 (E-H)?
Lines 217-220 – this is pure speculation, better suited for discussion and limitations of the study. There is no attempt in this paper to link the increased glycogen storage with glucose handling and altered metabolic profile.
Line 227 – please provide an estimate of 'total' neonatal lethality (i.e. including gestation lengths of +1 day, + 2 days, + normal length).
Lines 240-141 – I can't follow the rationale for this. Are male placentas hyperplastic to the same extent as females? One possibility to explain the lack of transcriptomic changes is that they are not? Why not perform QPCR in a number of male placentas for those genes thought to contribute to the hyperplasic phenotype (e.g. Slc38a4, Peg 3, Sfmbt2, etc) to help rule out the issues that were raised?
Lines 271-273 – as stated before, this comparison is an indirect one and only holds strictly true if comparing layers within the same gestational age at E17.5 (not against E14.5).
Lines 281-282 and Supp table 2 – please also provide the number of 'robust' imprinting, i.e. those genes that have been unequivocally shown to be imprinted, and highlight those in Supp table 2. References should also be provided in Supp Table 2. so that the reader can check for the imprinting status of the genes. For example, Slc38a1 is annotated as an imprinted gene – where is the evidence for this and is it strong enough to be credible?
Line 850 (RE: legend to Figure 1e) – please specify the stage of the embryos.
Line 859 (RE: legend to Figure 2a)- how many embryos were analysed?
Figure 6B, C and Supp Figure 5 – please add statistical indicators of variability to graphs
Supp Figure 2D – can't find an entry for this figure in the main text?
Materials and methods
The description of the methods is not detailed enough, e.g.:
Cell allocation experiments – how was the counting/i.e.segmentation of the cells done?; please provide further details on the Propidium Iodide procedure including how the staining leaves the ICM unlabelled – is this related to the timing of staining?
Placental histology and area/proportion measurements of layers – how many cross-sections were analysed per placenta? Were these midline? Did they represent different locations within the placenta?
What is meant by raw area? Would the total area be a better description? How was the 'raw' area calculated, for example in the cases where there are significant spongiotrophoblast projections into the labyrinth?
Weight measurements – please specify that these correspond to wet weights in material and methods; Use more specific terminology throughout the manuscript – consider replacing 'offspring weight' or 'body weight' with 'fetal weight' where appropriate.
Typos – should read labyrinth instead of Labryinth (Figure 4D); epigenetically instead of epigentically (Figure 1A).
Reviewer #2 (Recommendations for the authors):
Overall this is a thorough assessment of the fetal growth and placental characteristics of this highly novel model in which the role of Eed in the oocyte is experimentally distinguished. The experimental plan is highly innovative and clever. Considerable new information is presented. My main concern is the interpretation of the data presented. The authors strongly conclude that this is a model of late fetal growth catch-up driven by placental hyperplasia. While this may be the case, for the reasons given, there are other interpretations. For example, a different interpretation is that the observed placental changes prolong gestation allowing for longer development in utero.
Line 69 "several genes". Could the authors name the relevant genes?
Line 71 "We proposed that loss of EED in the oocyte sets up a developmental trajectory that involves initial fetal growth restriction that is resolved by placentally driven catch-up growth", "proposed" suggests that the authors have previously made this hypothesis but there is no reference. More importantly, fetal growth is intrinsically determined. This means that the fetus has the genetic potential to reach a certain size. Fetuses that do not reach their genetic potential are considered extrinsically growth restricted. This may be due to placental insufficiency or reduced maternal nutrient availability ie catch-up growth cannot be "placentally-driven". Rather, the placenta is able to support the catch-up growth of the fetus.
Line 162 "they underwent rapid catch-up growth between E17.5 and birth and were overgrown by P3" P0 is the day of birth – were the pups fed or unfed when weighed? If possible, it might be worth separating fed and unfed to ask whether there is a difference in their relative weights ie is catch-up postnatal due to feeding or prenatal due to increased placental capacity late in gestation? The numbers may not be sufficient or fed status may not be available but it is important to make this distinction given the authors' conclusion that catch-up is all about the placenta. As mentioned later in the Results section, gestation is longer which could also account for the "catch-up".
Line 165 "and enhanced placental efficiency".
Without knowing precisely when catch-up occurred and without any functional data, it is not possible to conclude that there is "enhanced placental efficiency". Please remove or rephrase.
Line 183 "To further investigate the relationship between placental function".
The authors are not experimentally determining function. Please rephrase.
Line 184 "placental weight ratio, which is indicative of placental efficiency (30)".
Although the term "placental efficiency" is used in the literature, it is not really meaningful. There are plenty of examples where the placenta is overgrown but fetal growth is still restricted – for example when the placenta is overgrowth but disorganised or where placental overgrowth is sequestering nutrients. The authors should refer only to the F:P ratios in their Results section and then discuss the possibility of increased placental efficiency underlying their observed catch-up growth. As previously stated, without weight data from either very late in gestation or from unfed pups, it is not possible to conclude the catch-up is due to the placental changes.
Line 199 "19% increase in HET-hom placental area'.
Be clear here what is meant by "placental area". Midline section of the whole placenta?
Line 217 "As placental glycogen is converted to glucose to support fetal growth, the increased glycogen cell number in HET-hom placentas is consistent with a greater glucose requirement in the late gestational fetal catch-up growth observed in the HET-hom offspring."
Although it seems intuitive that placental glycogen is used to support fetal growth, this has not been demonstrated experimentally in any study. It is possible these stores are for the dam in support of parturition or possibly to support placental hormone production. This sentence should be rephrased. In addition, the Results section should report results rather than inferring what the results might mean.
Line 223 "Consistent with this and the greater numbers of glycogen enriched cells in HET-hom placentas, gestational length was extended by 1 day".
Why "consistent with"? ie why would "supporting support the production of hormones that promote parturition" result in a longer gestation? It has been shown that an expanded junctional zone is associated with a longer gestation but not that the presumed increase in placental hormones prolongs gestation.
Line 223 "gestational length was extended by 1 day".
Could the weight gain reported at P0 simply reflect the fact that gestation is longer in the HET-hom model?
Line 229 and elsewhere "fetal catch-up growth".
The authors need to be cautious in concluding that there is fetal catch-up growth as this has not been demonstrated.
Line 235 "we analysed male and female placental tissue from HET-hom, HET-het and WT-wt offspring" and "Surprisingly".
The authors should be commended for recognising sexual dimorphism in the placenta. Were any differences in placental regions or fetal/placental weights detected between males and females at any stage? The absence of substantial differences in gene expression would only be surprising if the authors show a similar placental HET-hom phenotype for males and females
Line 271 "Of the Eed junctional zone DEGs the majority (86.1%) were upregulated, whereas a large proportion of Eed labyrinth DEGs and Eed decidua DEGs were downregulated"
As the placental phenotype involves an expansion of the junctional zone, it is to be expected that Jz-expressed genes will be expressed at overall higher levels and Lz at lower levels simply as a consequence of the relative changes in the number of cells expressing these genes ie there is no "upregulation" or "down-regulation". Please rephrase. The authors do acknowledge that the gene changes are consistent with the relative changes in the proportion of these regions in their last sentence but it is still important to avoid active terms such as upregulation/downregulation.
Line 292 "nine were commonly dysregulated.
Do the authors mean that the direction of aberrant expression was the same? i.e. expressed at higher levels in both scenarios?
Line 300 "dysregulation detected in Eed HET-hom placenta.
Rephrase as "some of the gene expression differences detected in female Eed HET-hom placenta.
Line 308 "similar placental phenotypes".
Loss of expression of Plac1 results in Jz overgrowth. Both overexpression and loss of expression of Ascl2 result in a decreased Jz. The term "similar phenotypes" should be removed here and instead simply state that a number of imprinted genes have been genetically demonstrated to regulate Jz development.
Line 322 "and inter-related fetal and placental growth trajectories
Inter-related – what is implied by this term?
Line 332 "indicating a role for the placenta in responding to fetal growth restriction to drive offspring catch-up growth prior to birth".
This cannot be concluded from the data presented. The authors have not shown that the placenta is responding. Critically, the authors have not definitively shown catch-up prior to birth.
Line 333 "placental efficiency increased late in gestation.
F:P ratio is not a true indicator of placental efficiency. If the authors were to show fetal growth catch-up in utero, the change in F;P ratio would be "consistent with" increased placental efficiency.
Line 338 "growth restriction can be corrected by placental expansion at later stages".
It may be true that fetal growth restriction as a consequence of placental insufficiency can be corrected by placental expansion but there are plenty of examples where the opposite is true ie overgrowth of the placenta alongside fetal growth restriction.
Line 373 "this was resolved by birth".
Again, not quite demonstrated by the data presented particularly given the increased length of gestation.
Line 394 "the maternal side of the placenta also expands to accommodate greater placental function".
Again, not experimentally demonstrated. It may be the case that changes in the trophoblast lineages/placental hormone production drive increased decidualisation – not quite the same thing.
Line 400 "sustained high levels of glycogen stores in the E17.5 placentas of HET-hom offspring increase glycogen release to the embryo and drive rapid growth catch-up growth."
Firstly, the authors have not measured placental glycogen stores biochemically. More importantly, the term "drive" is over-interpreting the data presented.
Line 413 "It may be that the increased number of glycogen cells remaining in the late-gestation HET-hom placenta delays late gestational hormonal release from the placenta and prolongs pregnancy. Consistent with this, extended gestational length occurred in about two-thirds of the pregnancies from Eed-hom oocytes."
Here the authors are retrofitting their data. It may be the case that prolonged gestation might be explained by delays in hormone release but how the increased number of glycogen cells would lead to such a result is not clear at all.
Line 416 "with no apparent birthing difficulties for the mothers".
Did the authors observe or video the births? If not, they cannot conclude there were no birth difficulties. This would be possible if, for example, all the pups were observed to be viable immediately after the birth and subsequently died. Is there any data on whether the pups that died were fed or unfed?
Line 471 "fetal catch-up growth".
Again, not proven.
Discussion generally
The reduced number of males is of interest. Nothing in the discussion?
Line 534 "Junctional zone, labyrinth and decidual area".
Area or mid-line area? Please clarify in Materials and methods.
Line 516 Were males removed during the day – specify in Materials and methods.
Reviewer #3 (Recommendations for the authors):
Methods – more clarity is requested in this section, descriptions are often very brief.
1) Placental CSA and cell proportions – how were the sections chosen or was the placental half exhaustively sectioned and then cell counts performed? The proportions of cells in the lateral part of the placenta are different from those at the midline – authors should describe their method in more detail including the location of sections chosen. How was GlyT cell number estimated and what are the units?
2) Presentation of the placental proportion data – there is clearly an expansion of all of the placental zones, but a proportionate increase in the JZ. In terms of the interpretation of the data, this is important. Because the LZ is increased in size then there is an increased exchange volume between mother and offspring. Can the JZ expansion be accounted for by the increase in GlyT cells alone?
3) How are cells in placental zones deduced from sc-RNAseq data – this section needs more detail.
4) RNAseq – how was the list of imprinted genes derived (there seems like a lot)? Is there functional enrichment for imprinted genes in their DEG list?
5) How was cell viability performed in the live tracking experiment (Figure 1D)?
[Editors' note: further revisions were suggested prior to acceptance, as described below.]
Thank you for resubmitting your work entitled "Fetal growth delay caused by loss of non-canonical imprinting is resolved late in pregnancy and culminates in offspring overgrowth" for further consideration by eLife. Your revised article has been evaluated by Adèle Marston (Senior Editor) and a Reviewing Editor.
The manuscript has been improved but there are some remaining issues raised by reviewer 2 that need to be addressed, as outlined below:
In particular, the point about normalization of fetal growth in mutants being a consequence of longer gestation period or the proposed intra-uterine catch-up growth has not been adequately addressed. Evidence should be provided that timing of birth is identical between genotypes to make this claim. Without this, conclusions about fetal catch-up should be toned down, removed from abstract and the two hypotheses (gestational length/catch-up) would need to considered alongside in results & discussion, highlighting that future studies are needed to distinguish between the two.
Please also re-write the Results section for clarity, following the suggestions of Reviewer 3 below, and move the interpretations of results to the Discussion section.
Reviewer #3 (Recommendations for the authors):
Main points about the claims of the paper:
1) Offspring of mouse oocytes lacking EED are initially developmentally delayed and growth restricted.
Well founded and supported by the data, though this was previously established.
2) Placentas from EED-deleted oocytes are dysmorphic with increased junctional zone and glycogen cell volume.
Histologically well demonstrated, but the molecular evaluation is incomplete and muddled. There are considerable differences in gene expression between EED-mutant and control placentas. This would be expected since the cell proportions are very different. The transcriptional data does not provide much additional explanatory power, at least as currently presented.
3) Offspring undergo accelerated development and growth in late gestation.
Evidence for this claim is incomplete. I remain unconvinced that the normalisation in size between Het-hom and control offspring on the day of birth does not reflect a difference in the length of gestation. All embryo weights up to e18.5 show growth restriction in the mutants (Figure 2B-F). It is only in post-natal animals that normalisation is observed.
As far as I can tell from the methods, the estimation of gestation length is based on the observation of a copulation plug on the morning following a mating. It is not clear if animals were separated in between mating days and what time they were set up. In addition, the time of birth was not recorded. Therefore, gestational length could reflect an interval of up to 24 hours. This is unlikely since mice usually mate at the beginning of the dark cycle and give birth at the end of it. However, without monitoring the time of birth there could be a significant delay in parturition in the mutant pregnancies that could still account for the small amount of 'growth recovery' observed in this study. Without a more thorough study of birth timing, I don't feel that the authors can hang the major conclusions of their paper on the single finding that the P0 weights are similar between genotypes.
10.7554/eLife.81875.sa2Author response
Essential revisions (for the authors):
The three reviewers recognized the merit of the work and its potential for reconciling previously conflicting results about the role of maternally inherited H3K27me3 on a growth trajectory. However, they also shared strong concerns about the interpretation of the data and in particular, the existence of a placenta-dependent catch-up phase. Additionally, it was not formally demonstrated that H3K27me3 patterns are altered in the placenta of embryos derived from EED-null oocytes, making it difficult to understand direct and indirect gene expression changes.
Please find below four points that require necessary revision. In addition, answer the individual comments of the reviewers in a detailed rebuttal letter and tone down or revise conclusions in your text accordingly.
1. Provide a more in-depth characterization of the placenta phenotype. Please first revise placental efficiency measurements as stated by Reviewer #3. Then, are there some sex-specific differences in placental alterations (considering that the placental transcriptome is differentially affected in males and females from EedKO oocytes)? Is the effect limited to the junctional zone only, and in particular to glycogen cells? The increase in the endocrine compartment may have been emphasized by the chosen methodology. Improved calculation of the different layers, within the same gestational age (E17.5) is required.
2. Whether the catch-up phase may be due to placental compensation prior to birth requires better support, especially in face of the increased gestational length and different post-coitum days of birth between mutants and WT. So far, enhanced growth is reported to happen between E17.5 and P3. The catch-up phase may equally happen prior to birth, in link with a potential placental role, or after birth, in relation to increased feeding, for example. It is really necessary to document when the catch-up phase occurs relative to birth. A measure at E19.5 was done but is only presented in Sup material: there, there is no difference between WT and mutants, which would suggest that extended gestation itself may lead to overgrowth. Please be clearer and report data in post-coitum days rather than postnatal days for WT and mutants, and add additional time points if necessary. Also, there seems to be no former evidence that placentomegaly can enhance fetal growth, and there could equally be an intrinsic fetal contribution to the phenotype.
3. Provide more clarity in the methods section, as descriptions are currently too limited to allow a proper understanding of the work carried out.
4. To claim potential epigenetic programming of placental phenotype and gene expression via Polycomb-dependent mechanisms, H3K27me3 mapping (ChIP-seq or CUT&RUN) would be required. This should be performed on placental samples at the stage when RNA-seq was done. Correlation analyses between gene expression and H3K27me3 changes (if any) are needed. It would be also interesting to compare with H3K27me3 patterns present in the oocyte (for which ChIP-seq data are publicly available).
In response to the Reviewer’s comments, we have significantly rewritten our manuscript and have added substantially more data to the study. These data include new analyses of glucose, amino acid and metabolite levels in fetal and maternal blood samples, more highly resolved fetal growth analyses, and a more detailed study of the hyperplastic placenta including IF analyses of labyrinth and capillary areas and ratios. We have also added analyses of placental DNA methylation state in offspring from oocytes lacking EED, which reveal a range of DNA methylation changes at imprinted and non-imprinted genes in HET-hom offspring compared to HET-het or WT-wt controls. We have also re-titled the manuscript to better reflect the findings of our revised study.
Intergenerational epigenetic inheritance in mammals remains poorly understood, particularly with respect to the contributions of histone modifications. To understand the consequences of disrupting these mechanisms, there remains a significant need for more information on the phenotypic and physiological consequences of perturbed epigenetic inheritance in pre- and post-implantation embryos and postnatal life. Our study extends understanding of intergenerational epigenetic inheritance by providing novel insights into the phenotypic and physiological consequences of failed EED-dependent programming in oocytes. This is particularly important considering that the developmental trajectory experienced by the embryo and fetus can have life-long impacts on health.
Our study was first submitted to an alternative journal in January 2022, but we were unsatisfied with the Reviewer’s comments and elected to transfer our manuscript to the open process eLife provides. During this time Matoba et al., published a paper which demonstrated that deleting Slc38a4 or C2MC, a micro-RNA cluster embedded within Sfmbt2, was able to rescue placental hyperplasia in offspring generated from oocytes lacking EED (Matoba et al., 2022 Genes and Development). This study also revealed that deletion of both Eed and Xist in oocytes corrected male-biased lethality in HET-hom offspring, but only partially normalised litter size and offspring weight. While the outcomes reported by Matoba et al., are highly informative, the rescue provided by maternal Xist deletion was incomplete and we hypothesise that other mechanisms must contribute to fetal growth and developmental outcomes in offspring from Eed-null oocytes.
In our study we demonstrate that surviving HET-hom offspring were able to attain normal weights by birth without deletion of Xist in the oocyte or correction of H3K27me3-dependent imprinting. Remarkably, the fetal growth delay in HET-hom fetuses was resolved during pregnancy despite reduced placental efficiency, low fetal blood glucose levels and unaltered amino acid and metabolite levels. Together, our observations indicate that HET-hom offspring have an innate ability to ameliorate fetal growth delay without increased placental function and within pregnancies of normal gestational length. Moreover, our comparison of HET-hom and SCNT derived growth and placental data suggest that similar fetal growth recovery occurs in both of these models without a requirement to correct maternally inherited impacts mediated by Xist or H3K27me3-dependent imprinting.
While the reason for this growth and developmental recovery during pregnancy remains unclear, our model demonstrates that these outcomes were initiated by loss of EED in the growing oocyte or during preimplantation development, or both. Given the capacity for fetal growth restriction to be resolved in the absence of enhanced placental function, we speculate that an unknown mechanism may sense growth restricted pregnancies to support adaptation(s) that favour the delivery of fully grown pups. In addition, we show that 60% of pregnancies are extended in length and that this increased gestational time contributes to offspring overgrowth. While it is clear that loss of H3K27me3 at Slc38a4, C2MC or Xist significantly contribute to the placental phenotype, loss of male fetuses and fetal growth delay, our study demonstrates that fetal growth and development are recovered in HET-hom offspring despite loss of H3K27me3 imprinting.
[Editors’ note: what follows is the authors’ response to the second round of review.]
The manuscript has been improved but there are some remaining issues raised by reviewer 2 that need to be addressed, as outlined below:
In particular, the point about normalization of fetal growth in mutants being a consequence of longer gestation period or the proposed intra-uterine catch-up growth has not been adequately addressed. Evidence should be provided that timing of birth is identical between genotypes to make this claim. Without this, conclusions about fetal catch-up should be toned down, removed from abstract and the two hypotheses (gestational length/catch-up) would need to considered alongside in results & discussion, highlighting that future studies are needed to distinguish between the two.
Please also re-write the Results section for clarity, following the suggestions of Reviewer 3 below, and move the interpretations of results to the Discussion section.
We have added a new analysis to demonstrate that the weight deficit in HET-hom offspring is resolved between E14.5 and E19.5 (excluding offspring with extended gestational length). Please refer to our detailed response to Reviewer 3, comment 3. We have also responded to Reviewer 3’s request to edit the Results section.
Given that the new analysis we have provided clearly demonstrates that the weight deficit is resolved in HEThom offspring between E14.5 and E19.5, we have opted not to remove our observation that offspring fetal growth is normalised within a normal gestational length from the abstract. We are confident that our conclusions are correct and our interpretation of these data is more accurately reflected in the current wording of the manuscript.
Lines 350-355 (now 355-359) describe a comparison we performed as part of our experimental work. We extracted SCNT offspring and placental weight data from the supplementary material of Matoba et al., 2018 (26) and compared these data with our own Wt-wt and Het-hom offspring and placental weight data for the oocyte Eed-null model described in this study. This revealed marked similarities between these models. As this is part of our analysis, we prefer to leave this in the Results section, but have now explained this experiment in more detail (please refer to our response to Reviewer 3, Minor issue 2). Lines 355-359 now read: “Given that both Eed HET-hom offspring and SCNT offspring lack H3K27me3 imprinting, we compared E14.5 and E19.5 fetal and placental weight data collected in this study with SCNT offspring data extracted from Matoba et al., 2018 (26). This revealed very similar fetal and placental growth trajectories in the Eed HET-hom and SCNT offspring models (Figure 7).”
While lines 415-417 (now lines 417-421) could be moved to discussion, they are included in results as an example of a non-imprinted gene that may be functionally important in the placenta. We included the sentence “An interesting example is NTN1 (Netrin1), which is reduced in the placentas of women with fetal growth restriction and potentially influences placental size by increasing the viability of placental microvascular endothelial cells (48, 49)” to explain why NTN DNA methylation status was highlighted in comparison to its placental expression, including reference to these data in Supplementary Table 8. We have edited this sentence and prefer to leave this as now described, though we will move the sentence to discussion if required.
To capture the context here, the text referred to on lines 414-421 now reads:
“The latter included five DMRs within or nearby eight non imprinted genes (Ntn1, Unc45b, Fam83h, Tiam1, Pcdhga1, Pcdhga5, Pcdhgb5, and Pcdhgb7) that were differentially expressed between HET-het and HET-hom placentas (Figure 6I; Supplementary File 1G). Of these, NTN1 (Netrin1) is of particular interest as it is reduced in the placentas of women with fetal growth restriction and potentially influences placental size by increasing the viability of placental microvascular endothelial cells (48, 49). While Ntn1 is not imprinted, it had reduced DNA methylation, consistent with its increased transcription in HET-hom compared to HET-het placentas (Supplementary File 1H).”
Reviewer #3 (Recommendations for the authors):
Main points about the claims of the paper:
1) Offspring of mouse oocytes lacking EED are initially developmentally delayed and growth restricted.
Well founded and supported by the data, though this was previously established.
We appreciate Reviewers comments. While previous studies have noted developmental delay, there has been no attempt to study this in detail or to explain how overgrowth could be observed in one study (our work in Prokopuk et al., 2018) while developmental delay was observed in another (Inoue et al., 2018). In this study we have taken great care to characterise the fetal growth in the context of the placental phenotype and have provided significantly more detail than previous studies. This includes determining when growth delay is initiated in preimplantation embryos and its eventual resolution during the later stages of pregnancy.
2) Placentas from EED-deleted oocytes are dysmorphic with increased junctional zone and glycogen cell volume.
Histologically well demonstrated, but the molecular evaluation is incomplete and muddled. There are considerable differences in gene expression between EED-mutant and control placentas. This would be expected since the cell proportions are very different. The transcriptional data does not provide much additional explanatory power, at least as currently presented.
We assume that by molecular evaluation of the placenta, reviewer 2 refers to the RNA sequencing and DNA methylation analyses. In our revised manuscript, as well as the DNA methylation analysis in placenta, we added analyses of metabolic state and circulating glucose levels in fetal and maternal blood samples and a detailed analysis of placental vasculature content using IF. These additional data substantially added to our initial analyses and have a bearing on the potential worth of additional molecular analyses in HET-hom placentas.
In HET-hom offspring we found substantially larger placentas (hyperplasia) and smaller fetuses, which are indicative of lower placental function assessed by fetal / placenta weight ratio. However, we found no difference in metabolic state in the serum of fetal offspring or in their mothers, strongly suggesting that the placenta does not have a major impact on metabolic state in the blood of the fetus or the mother. Glucose levels were also similar in fetuses and mothers at E17.5, although the average level was lower at E18.5, an observation most likely explained by the delay in parturition in around 60% of litters. Together these data indicate that despite the clear placental hyperplasia, metabolic state was not altered in HET-hom offspring blood samples and it is therefore difficult to attribute changes in late mid-gestational fetal growth to altered placental function at the metabolic level.
We appreciate that the global transcriptional analyses of the placenta lack spatial power and changes in cell proportions in the placenta will also contribute to overall changes in transcription. The way to resolve this would be by single cell or spatial RNAseq. However, if we performed single cell or spatial analyses, any cell specific transcriptional changes identified must still be interpreted in the light of the metabolic analyses. Given that we did not observe changes (either up or down) in the metabolic state of HET-hom offspring or their mothers, it is difficult to see how more detailed transcriptional analyses of the placental data could then be used to explain how the developmental delay and weight deficit that was observed at E14.5 in HET-hom offspring was resolved by birth (further discussed in our response to Reviewer 3, comment 3).
Single cell and spatial analyses are very expensive and time consuming. Moreover, given our findings, such an approach would have to include detailed analyses of fetal tissues as well as the placenta as, given the placental data, the fetal growth recovery we observed is likely to have at least some origin in the fetus. This would take substantial time and resources and would be unlikely to increase our ability to explain the central outcomes in this model. We therefore consider this to be outside the scope of the current study.
3) Offspring undergo accelerated development and growth in late gestation.
Evidence for this claim is incomplete. I remain unconvinced that the normalisation in size between Het-hom and control offspring on the day of birth does not reflect a difference in the length of gestation. All embryo weights up to e18.5 show growth restriction in the mutants (Figure 2B-F). It is only in post-natal animals that normalisation is observed.
Our data do not support the conclusion that growth normalisation was only observed in post-natal animals. The data we provided must be taken in context of the whole developmental and growth curve of these offspring, from early embryonic stages until birth. We observed very marked developmental delay and weight deficit of HET-hom offspring at E9.5, E12.5 and at E14.5 (Figure 2A-D). At E17.5 HET-hom fetuses were still underweight compared to controls (Figure 2E). However, the difference at E17.5 was reduced compared to E14.5 and, while there remained a small difference in fetal weight between HET-hom offspring and controls at E18.5 (Figure 2F), the difference was modest. Moreover, while HET-hom embryos were very obviously morphologically delayed at E9.5 (Figure 2A), E12.5 (Figure 2B) and E14.5 (Figure 2 supplement 1B), with the exception of weight, it was no longer possible to morphologically distinguish HEThom offspring at the gross morphological level at E18.5. Finally, the weight of HET-hom offspring born on E19.5 was the same as controls (Figure 2j-K).
We have now illustrated these data by providing ratios of HET-hom offspring weight vs WT-wt, WT-het and HET-het control weight at E12.5, E14.5, E17.5, E18.5 and E19.5. Importantly, E19.5 data included only litters born on E19.5. (New Figure 2L). This clearly demonstrates that compared to all control genotypes, HET-hom weight was normalised between E14.5 and E19.5.
To reflect this analysis, we have added the following text and edits to lines 184-194 of the Results section: “However, ratios of HET-hom fetal weight over WT-wt, WT-het or HET-het fetal weight at E12.5, E14.5, E17.5, E18.5 and E19.5 (excluding litters of extended gestational length) revealed that the weight deficit in HET-hom was resolved between E14.5 and birth on E19.5 (Figure 2L). Moreover, with the exception of the modest weight deficit, E18.5 HET-hom offspring were indistinguishable from controls at the gross morphological level. Together, these data demonstrate that the gross morphological delay and the fetal weight deficit in HET-hom offspring was resolved by E19.5 and that this occurred independently of litter size. In addition, by P3 HET-hom offspring were heavier than controls. As the overgrown HET-hom offspring in the P3 weight cohort (Figure 2H) included litters born at E19.5 E20.5 and E21.5, both post-natal offspring growth and extended gestation presumably contributed to the overgrowth phenotype.”
Our observations clearly show that the developmental delay and weight deficit are resolved in HET-hom offspring between E14.5 and birth on E19.5. Our data do not support the conclusion that resolution of developmental delay and weight deficit occurred as a result of the extended gestational length in HET-hom offspring. However, it is clear that extended gestational length contributed to the overgrowth phenotype we observed in this study, and that was observed in our previous study (Prokopuk et al., 2018).
Previous studies have not examined the apparently discordant observations that offspring from Eed null oocytes are delayed during fetal development (Inoue et al., 2018, Matoba et al., 2022), but are overgrown post-natally (Prokopuk et al., 2018). Here we now demonstrate that these observations are concordant and provide clear evidence that developmental delay and the deficit in fetal weights were resolved during gestation and that offspring are ultimately overgrown partly as a result of extended gestation.
Moreover, by discovering that deletion of Eed in oocytes results in extended gestational length in some litters we reveal a fascinating aspect of this model that had previously remained unknown.
As far as I can tell from the methods, the estimation of gestation length is based on the observation of a copulation plug on the morning following a mating. It is not clear if animals were separated in between mating days and what time they were set up. In addition, the time of birth was not recorded. Therefore, gestational length could reflect an interval of up to 24 hours. This is unlikely since mice usually mate at the beginning of the dark cycle and give birth at the end of it. However, without monitoring the time of birth there could be a significant delay in parturition in the mutant pregnancies that could still account for the small amount of 'growth recovery' observed in this study. Without a more thorough study of birth timing, I don't feel that the authors can hang the major conclusions of their paper on the single finding that the P0 weights are similar between genotypes.
We have performed studies using standard time-mating protocol for more than 20 years and have a great deal of experience in staging in embryos and fetuses. We used the same protocol to generate the litters examined in this study and to test whether gestational length was affected in the Eed model. Our protocol included checking for copulation plugs early on the morning following mating and immediately separating all females from males when a plug was discovered. To ensure we did not miss the day of birth, all females were visually monitored for litters without disturbance twice a day – first thing in the morning and last thing in the afternoon.
We apologise for omitting the important detail that females were separated when a plug was detected in timed mates, but have now added this in Materials and methods.
The methods now read (Lines 651-656) “Mice were time mated for two-four nights, with females plug checked daily for copulation plugs. Positive plugs were noted as day E0.5 and all females for which a plug was discovered were immediately separated from the male. Gestational length was measured in days post copulation by recording the morning of a copulation plug was detected E0.5 and visually monitoring females twice daily (morning and afternoon) for births from late gestation (E18.5) until pups were delivered (E19.5E21.5, depending on oocyte genotype).”
The data produced for gestational length are very clear. Extended gestation was only detected in litters from Eed-null oocytes (i.e. HET-hom offspring). In these animals 40% of litters were born on E19.5, 30% were born on E20.5 and 30% were born on E21.5. All litters (100%) from wild type or heterozygous oocytes were born on E19.5. Moreover, given that we monitored births in the morning and evening there could be a maximum of a half day difference for the 40% of litters born on E19.5 and for most the difference is less than that.
The data we have presented for E12.5-E19.5 offspring weights, including the new analysis we provide in Figure 2L, clearly demonstrate that the weight deficit of HET-hom offspring was resolved between E14.5 and E19.5. To conclude that the HET-hom weight deficit was resolved in the very last few hours of the pregnancy is not consistent with our data, especially given that this period also involved delivery.
To our knowledge, no other study has monitored gestational length in offspring from Eed-null oocytes or in studies of H3K27me3 imprinting in SCNT offspring. Our study is the first and in doing so provides important insight into the physiological responses that may help Het-hom offspring survive.
For all genotypes (experimental and control) there is some variation in plugging time, fertilisation time and birth time in mice and all of these factors influence exact gestational length. Measuring exact fertilisation time would require IVF followed by embryos transfer, rather than natural mates. However, IVF followed by embryo transfers may also affect gestational length as the procedure, timing of implantations and number of implantations are likely to vary between recipient females. In our opinion using IVF would be inferior to using natural mating to produce offspring in this model.
The most effective way to properly observe the exact hour of birth would be to place cameras in cages. This is impractical and would be hampered by cage contents such as tissues, wood shavings, nesting material etc, which enrich the female’s birthing environment. Such an approach would require repetition of all the term deliveries performed in this study. Moreover, it would be very expensive and time consuming, questionable from practical and ethical viewpoints and is unlikely to enhance the interpretation of the data that we have already provided.