Background
Infants of mothers with pre-existing types 1, 2 or gestational diabetes have significantly higher rates of perinatal mortality and major congenital anomaly [
1,
2]. Diabetic embryopathy can affect many organ systems including the heart and the neural tube [
3‐
5]. The effects of maternal diabetes mellitus on fetal development have been studied in various animal models. In drug-induced animal models, embryo development was disturbed by maternal hyperglycaemia [
6,
7]. In non-obese diabetic (NOD) mouse models of spontaneous type 1 diabetes mellitus, only 20% of recovered NOD embryos reached the blastocyst stage at 96 hours of superovulation compared with 90% of embryos recovered from nondiabetic animals [
6].
Several reports have shown that teratogenesis may be a direct result of hyperglycemia [
8,
9]. Previous studies found that expressions of genes related to metabolism, development and signal transduction were altered in embryos by the diabetic environment [
10,
11]. Moley
et al. showed that the mRNA and protein expression of the glucose transporters GLUT-1, GLUT-2, and GLUT-3 were decreased in embryos from streptozotocin (STZ)-induced hyperglycaemic mice [
12]. These results partly elucidate how maternal diabetes mellitus causes abnormal embryo development. But it is still not clear how the adverse effects are inherited to the next generations.
Wellen
et al. found that excessive glucose affected histone acetylation via the citrate lyase pathway [
13]. This finding suggests the possibility that the epigenome of the embryo may contribute significantly to abnormal fetal development in diabetic females. DNA methylation, one of the epigenetic modifications on DNA, can regulate relative gene expression, X-chromosome inactivation, as well as genomic imprinting [
14]. Genomic imprinting includes the formation of DNA methylation at specific loci in a parent-of-origin-specific manner [
15]. If DNA methylation on imprinted genes is not acquired/maintained properly, embryonic development and the offspring’s health would be affected [
16,
17].
The DNA methylation pattern of imprinted genes is susceptible to being affected by the environment [
18]. Several reports have shown that pre-implantation culture and manipulation can cause an abnormal methylation status of Differentially Methylated Regions (DMRs) at imprinted loci and these changes may induce abnormal fetal development [
19‐
21]. If maternal nutrients are altered the DNA methylation patterns may also be changed and this will induce abnormalities during fetal development. During gestation, if female rats are fed with choline-deficient diets, the DNA methylation of G9a and Suv39h1 is mis-regulated [
22]. These data indicate that the adverse maternal environment exerts adverse effects on DNA methylation during genomic imprinting establishment and maintenance.
We hypothesized that impaired DNA methylation at imprinted loci may play a key role in causing abnormal embryo development in maternal diabetes mellitus. In our lab, we have found that the DNA methylation patterns in DMRs of imprinted genes
Peg3, Snrpn and
H19 in oocytes was not altered by maternal diabetes at 15 days of injection of STZ [
23], but the embryonic development was affected. This indicated that the uterus environment may have deleterious effects on embryonic development. We examined the methylation patterns of DMRs of
Peg3, Snrpn, and
H19 in day post-coitum (dpc)10.5 placenta and fetus in the STZ-induced mouse model. We found that the expression and methylation levels of the imprinted genes were altered by maternal diabetes mellitus in placentas at 10.5dpc of gestation. Previous studies have shown that if the pre-gestational type 1 diabetes mellitus was cured at pre-pregnancy, the risks of adverse pregnancy outcomes was reduced in women [
24]. In animal models, if diabetic females were treated with insulin, embryonic development was not significantly different from that in non-diabetic females [
6]. Therefore, we also investigated whether the adverse effects caused by maternal diabetes on imprinted genes in placentas could be corrected by embryo transfer.
Methods
Ethics statement
All procedures described were reviewed and approved by the ethical committee of the Institute of Zoology, Chinese Academy of Sciences.
All mice were provided by the Beijing Vital River Experimental Animals Centre and fed in a temperature controlled room with a light cycle of 12 L: 12D (light:dark).
Generation of the diabetic mouse model
Female CD-1® (strain code; 022) mice, aged 6–7 weeks, received a single intraperitoneal injection of streptozotocin (STZ) at a dose of 230 mg/kg [
25]. Four days later, blood glucose levels were checked using a glucometer, Blood Testing Equipment, Accu-CHEK Active (Roche Diagnostic, Germany). If glucose levels were higher than 17.0 mmol/l, the mice were selected and used as the diabetic model (diabetic mice n = 53). Mice of similar age injected with buffer were selected as control (non-diabetic mice n = 49).
Collecting placentas and fetuses at mid-gestation
Diabetic/nondiabetic mice were mated naturally with normal male mice within 15 days of STZ/buffer injection and were determined to be pregnant when the vaginal plug was examined at 0.5d. At 10.5d of gestation, placentas (trophoblast population) and whole fetuses were collected. Samples were frozen immediately in liquid nitrogen and stored at -80°C until used.
DNA purification and quantitative analysis of methylation by PCR (qAMP)
DNA was purified from one half of the bisected fetus or placentas using DNA Tissue Kit (Tiangen, China) following the manufacturer’s directions. To investigate the methylation conditions, qAMP was conducted as described by Oakes
et al. and Lopes
et al.[
26,
27]. Briefly, fetal or placental DNA (~1.5 μg/enzyme) was digested with methylation-sensitive (
HhaI
, New England Biolabs, Beijing, China;
HpaII, New England Biolabs, Beijing, China) and methylation-dependent (
McrBc, New England Biolabs, Beijing, China) restriction enzymes, respectively. The resulting product was used as template for real-time PCR reaction using the SYBR green kit (Kangwei Inc., China). The qAMP was performed employing a Rotor-Gene Q Real-Time PCR instrument (Qiagen). The percentage of methylation for cytosine-phosphate-guanine (CpG) sites was based on ⊿Ct (cycle threshold values between digested and undigested sham aliquots). For
HhaI, the percent methylation is 100 (e
-0.7(∆Ct)); for
McrBC, the percent methylation is 100 (1- e
-0.7(∆Ct)). The primers described in Table
1 were designed to span ~ 180 bps of the DMRs of the maternally methylated sequences:
Peg3 and
Snrpn, as well as the paternally methylated
H19.
Table 1
Oligonucleotides and annealing temperature utilized for qAMP and qRT-PCR of imprinted genes
qAMP
|
H19
| 5′-AGCCGTTGTGAGTGGAAAGA-3′ | 5′-CATAGCGGCTTCGGACATT-3′ |
Snrpn
| 5′-CTCCTCAGAACCAAGCGTCT-3′ | 5′-ATTCCGGTCAGAGGGACAGA-3′ |
Peg3
| 5′-GGTGTCCCGCAGCCCTTG-3′ | 5′-CGGAGCACAGCACTCTACGC-3′ |
chr9:106724005-106724149 | 5′-GATCTATTCCTTCCTTTACTTT-3′ | 5′-TCCTGGGAAATGAAGTTT-3′ |
Peg3
|
qRT-PCR
| 5′-TTGGACTGGACAGAGATGATGACA-3′ | 5′-ATTCTGGTATGACTCGGCATCCT-3′ |
Snrpn
| 5′-AGGCCCATCCCAGCAGGTCAT-3′ | 5′-GCGGGTACTGGGTTGGGGCTC-3′ |
H19
| 5′- CTTGTCGTAGAAGCCGTCTGTTC-3′ | 5′- GTAGCACCATTTCTTTCATCTTGAGG-3′ |
Ppia
| | 5′- CGCGTCTCCTTCGAGCTGTTTG-3′ | 5′- TGTAAAGTCACCACCCTGGCACAT-3′ |
RNA isolation and mRNA expression assays
RNA was extracted from the other half of the bisected fetus and placentas using RNA Tissue kit (Tiangen, China) according to the manufacturer’s instructions. cDNA first-strand synthesis was performed using Superscript II (Invitrogen). The first-strand cDNA was used as template and the expressions of
Peg3, Snrpn and
H19 were evaluated by quantitative real-time PCR (qRT-PCR) using Rotor-gen Q (Qiagen, Germany). Triple reactions were analyzed for each sample and the threshold cycle value was normalized to the housekeeping gene of peptidylprolyl isomerase A (
Ppia). The primers are shown in Table
1. The expression levels were calculated using the 2
-△△Ct method.
Western blot and immunohistochemistry analysis
Placental proteins (n = 6 from 4 litters) were extracted from cell lysis and quantified for western blot analysis as previously described [
28]. Briefly, proteins were heated for 5 min at 100°C. Then proteins were separated by SDS-PAGE and electrically transferred to polyvinylidene fluoride membranes. After that, the membranes were blocked in TBST containing 5% skimmed milk for 2 h, followed by incubation overnight at 4°C with rabbit anti-peg3 (1:500, Bioss, China, code: bs-1870R). After washed with TBST, membranes were incubated with goat anti-rabbit IgG (1:1000) for 1 h at 37°C. Beta-actin was used as loading control.
For immunohistochemistry analysis, fresh placentas (n = 6 from 4 litters) were fixed in 4% paraformaldehyde overnight at 4°C and washed in 50% ethanol for 1 hour at room temperature. After that, samples were stored in 70% ethanol at 4°C until used. The fixed samples were dehydrated in a graded ethanol series, cleared in xylene, and embedded in paraffin wax. The embedded placentas were sectioned at 8 μm and incubated with rabbit anti-Peg3 (1:500, Bioss, China, code: bs-1870R) overnight at 4°C. The samples were then incubated with the biotin-labeled secondary antibody for 30 minutes. Staining was carried out using the Vectastain ABC kit and DAB peroxidase substrate kit (Vector Laboratories, Burlingame, CA).
Embryo transfer
To obtain pronuclear stage embryos from diabetic and non-diabetic mice, the estrous female diabetic/ non-diabetic mice were naturally mated with normal males, respectively. Vaginal plug was examined the next morning. The mice with a vaginal plug were killed and the pronuclear embryos were collected from their oviductal ampullae.
To transfer embryos to pseudopregnant mice which were produced by mating the normal estrous female mice with males with vasoligation, the pesudopregnant mice were anesthetized by intraperitoneal injection of sodium pentobarbital (40 mg/kg, Sigma). Then eight embryos were transferred into the oviduct of a pseudopregnant mouse. The mice were fed until gestational 10.5dpc and then killed. Embryos and placentas were collected.
Statistical analysis
Data are presented as means ± SD. The methylation level of DMRs and expression levels of genes from different groups were compared by independent-sample t-test. A probability level of P < 0.05 was considered significant.
Discussion
Proper establishment and maintenance of DNA methylation patterns are crucial for embryo development and survival. In the present study, we investigated whether maternal diabetes mellitus could perturb the acquisition/maintenance of DNA methylation during embryo development. In our study, we found that the expressions of Peg3 and H19 were altered in placentas at 10.5dpc; we also identified evident changes in the methylation level in the DMRs of them. Strikingly, these effects were not observed when diabetic pronuclear embryos were transferred into non-diabetic recipients.
The placenta is not only the specialized organ used by the developing embryo and fetus to obtain nutrients and oxygen from the mother, but it also acts as a key selectively permeable barrier between the mother and the fetus [
29]. In pregnancies complicated by pre-gestational diabetes (type 1 and type 2), the
de novo synthesis of blood vessels in the placenta shows abnormalities [
30]. Several studies show that the expressions of leptin, leptin receptors, androgen receptor and peroxisome proliferator-activated receptors (PPARs) are altered by maternal diabetes mellitus in placentas [
31,
32]. In our study, we also observed that the mRNA levels of
H19 and
Peg3 were significantly altered in placentas from diabetic females compared to controls. These imprinted genes are expressed monoallelically in a parent-of-origin specific manner. They are generally located in clusters, epigenetically marked by DNA methylation on differentially methylated regions (DMRs) which regulate relative genes’ expression [
33‐
35]. Interestingly, there were significant changes of average methylation levels in DMRs of
H19 and
Peg3 in placentas from diabetic mice. It is well known that DNA methylation is a mechanism regulating expression of imprinted genes. Our results indicate that abnormal methylation levels may be a reason for the changed expressions of these genes. However, these changes might be induced by different distribution of
Peg3 in different placental cell populations. Hiby et al. found that
Peg3 was expressed in the trophoblast of the developing placenta at 9 days post-coitum (d.p.c.) and the trophoblast populations of the well-developed placenta at 15 dpc [
36]. In our study, the placenta was separated and used to carry out all the relative assays. By immunohistochemistry analysis, we observed that all the trophoblast populations were stained. The Western blot analysis also indicated that the protein level of
Peg3 in diabetic placentas was obviously lower. These findings might partly elucidate how maternal diabetes mellitus induces abnormal embryo development, because proper DNA methylation patterns are important for placenta and fetus development.
Previous studies found that the diabetic condition is detrimental to pre- and post-implantation embryo development. Congenital malformations are approximately 3–4 times more frequent in infants from diabetic mothers than from non-diabetic mothers [
37,
38]. In both chemically induced and spontaneous diabetic models, significant abnormalities in pre-implantation embryo development have been observed [
6,
39,
40]. But the DNA methylation patterns in fetus from diabetic females are still unclear. In our study, we examined the expression and average methylation levels of some imprinted genes in the placenta and embryo at 10.5dpc. We found that the expression of relative imprinted genes was not altered in diabetic dpc10.5 fetus compared with controls. The expression of
H19 was consistent with that reported by Shao
et al.[
41]. The average methylation levels in DMRs of them were also similar to controls. Although another study found that the methylation level of
H19-Igf2 imprint control region was increased in the embryonic day 14 fetus from diabetic mice [
41], we found that the methylation level of
H19 DMR was not affected by maternal diabetes in the 10.5dpc embryo. Different methods employed in these studies may be a reason for the difference in results. Another reason may relate to the fact that we selected different fragments as targets. Because severe loss of DNA methylation in imprinted genes causes embryo absorption and miscarriage during early development, it was not possible to identify significant loss of DNA methylation in DMRs of imprinted genes in live fetus from diabetic females. Another reason may be the protective function of the placenta which serves as filter between mother and fetus. In our study, we found no changes of imprinted genes in fetus from mothers with diabetes mellitus, but the expression and methylation levels were significantly affected by maternal diabetes in placentas. These findings suggest that maternal diabetes may affect fetus development by altering placental functions.
Studies have shown that if maternal diabetes mellitus is well controlled, the deleterious effects on embryo development can be partially corrected. Studies also have shown that when the diabetic females are treated with insulin, embryo development is not significantly different from non-diabetic females [
6]. These results indicate that the diabetic uterus environment may play a key role in abnormal embryonic development. So we transferred pronuclear embryos of normal females to normal/diabetic pseudopregnant females and found that the DNA methylation patterns and expression of
Peg3 in dpc10.5 placentas were significantly changed by the diabetic uterus environment. However, while pronuclear embryos of diabetic mothers were transferred to normal pseudopregnant females (DN), the expression and methylation levels of
H19 and
Peg3 in placentas were similar between DN and NN. Another study has demonstrated that these genes’ methylation status in oocytes of diabetic females was not affected [
23] at 15 days of injection of STZ. So we propose that the altered expression and methylation levels of imprinted genes may primarily be caused by the adverse uterus environment of diabetic females and it could be corrected by embryo transfer.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
ZJG carried out the epigenetic analysis, designed the assay and wrote the manuscript. QXL carried out the epigenetic analysis. SML, YCW and ZJG carried out PCR amplification and epigenetic analysis. HZM participated in analyzing the data. HS participated in revising the manuscript. QYS and CLZ participated in the design of the assay and revising the manuscript. All authors read and approved the final manuscript.