Maternal hyperglycemia during pregnancy is associated with excess fetal growth and adverse perinatal and developmental outcomes. Placental epigenetic maladaptation may underlie these associations. We performed an epigenome-wide association study (>850,000 CpG sites) of term placentas and prenatal maternal glycemic response 2-h post oral glucose challenge at 24–30 weeks of gestation among 448 mother-infant pairs. Maternal 2-h glycemia postload was strongly associated with lower DNA methylation of four CpG sites (false discovery rate [FDR] q <0.05) within the phosphodiesterase 4B gene (PDE4B). Additionally, three other individual CpG sites were differentially methylated relative to maternal glucose response within the TNFRSF1B, LDLR, and BLM genes (FDR q <0.05). DNA methylation correlated with expression of its respective genes in placental tissue at three out of four independent identified loci: PDE4B (r = 0.31, P < 0.01), TNFRSF1B (r = −0.24, P = 0.013), and LDLR (r = 0.32, P < 0.001). In an independent replication cohort (N = 65–108 samples), results were consistent in direction but not significantly replicated among tested CpG sites in PDE4B and TNFRSF1B. Our study provides evidence that maternal glycemic response during pregnancy is associated with placental DNA methylation of key inflammatory genes whose expression levels are partially under epigenetic control.

Prenatal nutritional, behavioral, and environmental conditions play a key role in fetal development by modulating the intrauterine environment, fetal nutrient availability, and growth. The Pedersen hypothesis states that prenatal maternal glucose crosses the placenta and leads to intrauterine hyperglycemia affecting fetal growth and development (1).

It is now well established that maternal hyperglycemia in pregnancy is associated with adverse maternal and birth outcomes. For example, in the large prospective international multicenter Hyperglycemia and Adverse Pregnancy Outcome (HAPO) study, robust linear associations were found between greater maternal prenatal glucose levels and higher birth weight or cord blood C-peptide levels across the whole spectrum of maternal glycemia, with no lower threshold (2). In the HAPO study, measures of maternal prenatal glycemia, both fasting and post oral glucose load, were linearly associated with adverse outcomes in mothers and in infants at birth (2). Of these, maternal glucose levels 2-h post 75-g glucose load showed slightly stronger associations with neonatal hypoglycemia (3) and with an increased risk of abnormal glucose tolerance in offspring at 7 years of age (4), suggesting that maternal response to glucose loading in pregnancy contributes to offspring metabolic programming at birth and later in life. Indeed, emerging evidence suggests that maternal hyperglycemia could have lasting health consequences due to metabolic programming that occurs during key fetal developmental stages and may act through epigenetic programming and molecular mechanisms (5).

It is now recognized that the placenta is not simply a transferring organ for nutrients from the mother to fetus but participates actively in maternal metabolism and likely contributes to fetal programming. The placenta is the master regulator of the fetal environment and is directly responsible for maternal-fetal nutrient and waste transport as well as hormone synthesis (6). Fetal gluconeogenesis is relatively minimal, and thus the fetus relies heavily on circulating maternal glucose transported into the fetal side by facilitated placental diffusion through members of the glucose transporter (GLUT) family (6). Over 95% of the fetal glucose levels are estimated to be derived from maternal plasma levels by diffusion through the placenta (7,8), and when maternal hyperglycemia occurs, this excess maternal glucose is passed toward the fetus.

The placenta has been shown to adapt to nutrient availability through epigenomic modifications in response to gestational diabetes mellitus (911). Nevertheless, these studies reported epigenetic modifications with cases of gestational diabetes mellitus and thus could not determine whether the observed patterns were consequences of gestational diabetes mellitus (or of its treatments) or actual placental adaptations to hyperglycemia in pregnancy. To address this gap of the role of placental adaptation to maternal glucose levels during fetal development, we conducted an epigenome-wide association study (EWAS) for maternal 2-h glucose post oral glucose tolerance test (OGTT) administered during the second trimester of pregnancy and DNA methylation of full-term placenta tissue. We measured genome-wide DNA methylation of the fetal placenta side among 448 mother-infant pairs and observed strong associations for several inflammatory-relevant genes. Furthermore, we confirmed the functional role of DNA methylation at the discovered CpG loci by quantifying gene expression in placental tissue. Finally, we tested for external replication of our top differentially methylated loci in an independent birth cohort.

Study Population

Samples and study participants were selected from the Genetics of Glucose regulation in Gestation and Growth (Gen3G), a prospective Canadian prebirth cohort. Gen3G was designed to elucidate the biological, environmental, genetic, and epigenetic determinants of glucose regulation during pregnancy and the impact on offspring development (12). Briefly, we recruited expecting mothers during the first trimester of their pregnancy, and we enrolled pregnant women in the study if they were at least 18 years of age or older with a singleton pregnancy and did not have prepregnancy diabetes based on medical history and screening during the first trimester blood sampling. For this study, mother-infant pairs were selected from the larger cohort if they had placental tissue available for DNA isolation as well as >37 weeks of gestation at delivery. Study participants provided written informed consent prior to enrollment in accordance with the Declaration of Helsinki. All study protocols were approved by the ethics review board from the Centre Hospitalier Universitaire de Sherbrooke.

Placental Tissue DNA and RNA Extraction

Trained research personnel collected fetal placenta tissue samples immediately after delivery (<30 min postpartum). A 1 cm3 placenta tissue sample was collected approximately 5 cm from the umbilical cord insertion, from the fetal side of the placenta for each delivery. Placenta samples were collected by trained study staff and stored in RNAlater (Qiagen) at −80°C until DNA or RNA extraction occurred. We purified DNA and RNA from the placenta samples using the AllPrep DNA/RNA/Protein Mini Kit (Qiagen). Purity of extracted DNA was evaluated using a Spectrophotometer (Ultrospec 2000 UV/Visible; Pharmacia Biotech) with an absorbance ratio set at 260–280 nm as recommended (13).

OGTT

During the first trimester, enrolled participants completed a nonfasting 50-g glucose challenge test (GCT), and we measured glucose 1 h after the glucose load (to screen for preexisting diabetes). During the second trimester clinical visit, all women performed a fasting 75-g OGTT, and we measured maternal glucose levels at fasting prior to the oral glucose challenge and at 1 h and 2 h post glucose challenge. Maternal blood glucose concentrations were all measured at the Centre Hospitalier Universitaire de Sherbrooke central laboratory.

Epigenome-Wide DNA Methylation Measurements

Epigenome-wide DNA methylation measurements were performed on DNA from placenta samples using bisulfite conversion followed by quantification using the Infinium MethylationEPIC BeadChip (Illumina, San Diego, CA) that measures over 850,000 CpG sites at a single nucleotide resolution. Samples were randomly allocated to different plates and chips to minimize confounding. Methylation data were imported into R for preprocessing using minfi (14). We performed quality control at the sample level, excluding samples that failed (n = 8), mismatch on genotype (n = 12) or sex (n = 1), and technical duplicates (n = 10). A total of 448 high-quality samples were retained for subsequent analyses. We performed quality control on individual probes by computing a detection P value and excluded 2,003 probes with nonsignificant detection (P > 0.05) for 5% or more of the samples. We also excluded 19,129 probes annotated to sex chromosomes, 2,836 non-CpG probes, 5,552 single nucleotide polymorphism–associated probes at the single base extension with a minor allele frequency of ≥5%, and 4,453 probes with a single nucleotide polymorphism at the target CpG site with a minor allele frequency of ≥5%. Finally, we excluded 40,448 cross-reactive probes previously identified (15). A total of 791,131 CpG sites were included in the final analyses. We processed our data using functional normalization with the default of two principal components from control probes (16). We also adjusted for probe-type bias using RCP, a regression method approach that uses genomic proximity to adjust the distribution of type 2 probes (17). Last, we used the ComBat function from the sva package to adjust for sample plate (18). We visualized the data using density distributions at all processing steps and performed principal component analyses to examine the association of both technical and biological variables. We logit transformed the β-values to M-values for statistical analyses (19). However, we report effect estimated and summary statistics on the β-value scale to ease interpretability.

RNA Quantification of Top Differentially Methylated Loci

We selected a simple random sample of 104 participants with DNA methylation data for RNA quantification. RNA concentrations and RNA Integrity Number (RIN) were assessed using Agilent 2100 Bioanalyzer and the Agilent RNA 6000 Nano Kit (Agilent Technologies). Complementary DNA of placental RNA was generated using a random primer hexamer (High Capacity cDNA RT; Applied Biosystems). Amplicons were generated in duplicate in 20 μL with 10 μL of TaqMan Universial PCR Master Mix (Applied Biosystems). RNA expression for genes annotated to the differentially methylated loci significantly associated with 2-h maternal glucose (PDE4B: Hs00277080_m1, TNFRSF1B: Hs00961748_m1, BLM: Hs00172060_m1, LDLR: Hs00181192_m1; Applied Biosystems) were measured with quantitative real-time PCR (qRT-PCR) using 7500 Real Time PCR system (Thermo Fisher Scientific). Expression levels are reported as a Ct ratio with respect to the reference gene, the tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein ζ (YWHAZ: Hs01122445_g1; Applied Biosystems) previously shown to be stable in human placenta (20). We report 1/ratio values (to ease interpretation as a higher 1/ratio value reflect higher expression) for correlation analyses with DNA methylation.

Replication Cohort and Pyrosequencing: ECO-21

The independent replication cohort (ECO-21) consisted of mother-infant pairs recruited in the first trimester of pregnancy, from a French Canadian population from the Saguenay-Lac-Saint-Jean region of Québec (Saguenay city area, Québec, Canada) (21). Pregnant women were excluded if they were <18 or >40 years of age, had a history of alcohol/drug abuse in pregnancy, or had diagnosed familial hypercholesterolemia, pregestational diabetes, or other prepregnancy disorders impairing glucose homeostasis. For this analysis, we excluded women who were treated for gestational diabetes mellitus, to avoid confounding by treatment. The Chicoutimi Hospital ethics committee reviewed and approved this project. Women provided written informed consent prior to enrollment in the study in accordance with the Declaration of Helsinki.

In the replication cohort, DNA methylation levels of the top CpG site annotated to the TNFRSF1B gene (cg26189983) as well as three other CpG sites (cg07734160, cg13866577, and cg03442467) within the PDE4B region were measured using pyrosequencing (PyroMark Q24; Qiagen). Although we also intended to pyrosequence our other top loci, our primers failed. To pyrosequence TNFRSF1B and PDE4B CpG sites, DNA underwent a sodium bisulfite (NaBis) treatment (EpiTect Bisulfite Kits; Qiagen). Target NaBis-DNA loci were PCR-amplified with specific primers designed using the PyroMark Assay Design software (version 2.0.1.15; Qiagen) and were then pyrosequenced. Pyrosequencing runs performed included a negative PCR and sodium bisulfite conversion controls. Additionally, pyrosequencing quality control was assessed for each sample, as recommended by the manufacturer, using PyroMark Q24 Analysis Software (v1.0.10.134).

Statistical Analyses

For the 448 participants eligible for analyses, we report our sample characteristics using means, SD, or proportions. We performed CpG-by-CpG analyses by fitting robust linear regression models for each site adjusted for covariates with DNA methylation as the response variable on the M-value scale using maternal 2-h glucose levels post OGTT as the main predictor. Robust linear regression was used to protect against potential heteroskedasticity (22). To control for cell-type heterogeneity, we used the top 10 components from ReFACTor, a reference-free method that adjusts for cell-type mixture in heterogenous tissues (23). Robust linear regression models were adjusted for maternal age in years, BMI, parity, smoking during pregnancy, gestational age at birth, sex, and the first 10 principal components estimated from ReFACTor as proxy for placenta cellular heterogeneity. CpG-by-CpG analyses were controlled for the false discovery rate (FDR) at 5% (q <0.05). Quantile-quantile plots of P values were used to inspect genomic inflation. In epigenome-wide association analyses, bioinformatic adjustment reduced the genomic inflation (λ) from 1.39 to 1.01 (Supplementary Fig. 1). Regional and genome-wide Manhattan plots were used to report results from epigenome-wide association analysis.

We calculated Pearson correlation coefficients to estimate the association between DNA methylation and gene expression among a randomly selected sample of 104 participants for top differentially methylated placenta genes associated with maternal glucose response (FDR q <0.05). Significant statistical correlations between DNA methylation and gene expression were considered with P < 0.05. In additional analyses, we evaluated associations adjusting for the same covariates among top differentially methylated CpG sites and maternal 1-h glucose levels post 50-g GCT during the first trimester as well as baseline fasting glucose levels and 1-h glucose levels post 75-g OGTT at second trimester per SD of each glycemic trait to allow comparability between traits. We performed these additional analyses with other measures of glycemia taken in the first trimester or during OGTT to ensure robustness of results and to assess the specificity of our associations with regard to postchallenge glucose response.

Participant Characteristics

A total of 448 mother-infant pairs were eligible for analyses with complete DNA methylation and 2-h glucose levels post OGTT. All mothers were Caucasian, 49.3% were primiparous, with a mean ± SD age of 28.2 ± 4.3 years and BMI of 25.45 ± 5.7 kg/m2, and 90.2% reported to be nonsmokers during pregnancy. Mean ± SD gestational age at birth was 39.5 ± 1.0 weeks, mean birth weight was 3,448 ± 428 g, and 52.7% of births were males. At enrollment and by design of the study, none of the pregnant women had pregestational diabetes evaluated by first trimester glycemic testing (GCT and HbA1c). At the second trimester 75-g OGTT, mean ± SD fasting glucose level was 4.20 ± 0.37 mmol/L, 1-h glucose was 7.11 ± 1.61 mmol/L, and 2-h glucose was 5.80 ± 1.33 mmol/L. Additional participants characteristics are summarized in Table 1. Concentrations for 2-h glucose were approximately normally distributed (Supplementary Fig. 2).

Table 1

Participant characteristics from the Gen3G prospective cohort (N = 448)

Maternal age (years)
 
28.2 ± 4.3
 
BMI (kg/m2)
 
25.45 ± 5.7
 
Parity
 

 
 Primiparous (%)
 
221 (49.3)
 
Ethnicity
 

 
 Caucasian
 
448 (100)
 
Maternal smoking during pregnancy
 

 
 No
 
404 (90.2)
 
 Yes
 
39 (8.7)
 
 Unknown
 
5 (1.1)
 
First trimester GCT
 

 
 1-h glucose (mmol/L)
 
5.55 ± 1.41
 
Second trimester OGTT
 

 
 Fasting glucose (mmol/L)
 
4.20 ± 0.38
 
 1-h glucose (mmol/L)
 
7.11 ± 1.61
 
 2-h glucose (mmol/L)
 
5.80 ± 1.33
 
Child sex
 

 
 Male (%)
 
236 (52.7)
 
Gestational age at birth (weeks)
 
39.5 ± 1.0
 
Birth weight (g) 3,448 ± 428 
Maternal age (years)
 
28.2 ± 4.3
 
BMI (kg/m2)
 
25.45 ± 5.7
 
Parity
 

 
 Primiparous (%)
 
221 (49.3)
 
Ethnicity
 

 
 Caucasian
 
448 (100)
 
Maternal smoking during pregnancy
 

 
 No
 
404 (90.2)
 
 Yes
 
39 (8.7)
 
 Unknown
 
5 (1.1)
 
First trimester GCT
 

 
 1-h glucose (mmol/L)
 
5.55 ± 1.41
 
Second trimester OGTT
 

 
 Fasting glucose (mmol/L)
 
4.20 ± 0.38
 
 1-h glucose (mmol/L)
 
7.11 ± 1.61
 
 2-h glucose (mmol/L)
 
5.80 ± 1.33
 
Child sex
 

 
 Male (%)
 
236 (52.7)
 
Gestational age at birth (weeks)
 
39.5 ± 1.0
 
Birth weight (g) 3,448 ± 428 

Data are mean ± SD or n (%).

Association of Maternal 2-Hour Glucose Levels With DNA Methylation of Placenta

In adjusted CpG-by-CpG analyses, we found seven CpG sites of placental DNA significantly associated (FDR q <0.05) with maternal 2-h glucose levels post 75-g OGTT (Table 2). Epigenome-wide results are shown in Fig. 1. We observed lower placental DNA methylation of four CpG sites annotated to PDE4B (phosphodiesterase 4B gene) in response to higher maternal 2-h glucose levels. Specifically, a 1 mmol/L greater 2-h glucose level was associated with a 1.16%, 0.88%, 1.86%, and 0.58% lower DNA methylation at cg07734160 (P = 1.20 × 10−9), cg13866577 (P = 1.11 × 10−7), cg03442467 (P = 2.84 × 10−7), and cg13349623 (P = 2.06 × 10−9), respectively. The four CpG sites were annotated within the transcription start site (1,500–200 bps), first exon, or body of PDE4B and surrounded by another four sites just below our FDR statistical threshold (Fig. 2). Scatter plots for the association between DNA methylation and 2-h glucose levels are shown in Fig. 3. In addition, a 1 mmol/L greater maternal 2-h glucose level was associated with 1.22% (P = 1.70 × 10−7) higher DNA methylation of cg26189983 annotated to TNFRSF1B/TNFR2 (tumor necrosis factor receptor superfamily member 1B) and located within the gene body. We also observed an inverse association for cg20254265 annotated to the exon boundary of BLM (bloom syndrome RecQ like helicase gene) with a 1 mmol/L greater 2-h glucose level associated with a 0.63% lower DNA methylation (P = 7.58 × 10−8). Last, we observed 0.27% lower DNA methylation at cg08483713 per 1 mmol/L greater 2-h glucose level, and this CpG site was annotated to the gene body of LDLR (low density lipoprotein receptor).

Table 2

Adjusted differences in DNA methylation associated with a 1 mmol/L increase in prenatal 2-h glucose levels post OGTT

CpGDNA methylation (%), mean ± SDChromosomePositionGeneGene groupPercent difference in DNA methylation95% CIP
cg26189983
 
49.7 ± 8.2
 
chr1
 
12227700
 
TNFRSF1B
 
Body
 
1.22
 
(0.80, 1.7)
 
1.70 × 10−7
 
cg07734160
 
13.0 ± 7.5
 
chr1
 
66797378
 
PDE4B
 
TSS1500, body
 
−1.16
 
(−1.5, −0.8)
 
1.20 × 10−9
 
cg13866577
 
9.4 ± 6.6
 
chr1
 
66797481
 
PDE4B
 
Body, TSS1500
 
−0.88
 
(−1.2, −0.6)
 
1.11 × 10−7
 
cg03442467
 
24.6 ± 13.1
 
chr1
 
66797701
 
PDE4B
 
Body, TSS200
 
−1.86
 
(−2.6, −1.2)
 
2.84 × 10−7
 
cg13349623
 
5.8 ± 3.6
 
chr1
 
66798221
 
PDE4B
 
First exon, body
 
−0.58
 
(−0.8, −0.4)
 
2.06 × 10−9
 
cg20254265
 
79.6 ± 9.1
 
chr15
 
91306178
 
BLM
 
ExonBnd, body
 
−0.63
 
(−0.9, −0.4)
 
7.58 × 10−8
 
cg08483713 5.86 ± 2.9 chr19 11241669 LDLR Body −0.27 (−0.4, −0.2) 1.39 × 10−6 
CpGDNA methylation (%), mean ± SDChromosomePositionGeneGene groupPercent difference in DNA methylation95% CIP
cg26189983
 
49.7 ± 8.2
 
chr1
 
12227700
 
TNFRSF1B
 
Body
 
1.22
 
(0.80, 1.7)
 
1.70 × 10−7
 
cg07734160
 
13.0 ± 7.5
 
chr1
 
66797378
 
PDE4B
 
TSS1500, body
 
−1.16
 
(−1.5, −0.8)
 
1.20 × 10−9
 
cg13866577
 
9.4 ± 6.6
 
chr1
 
66797481
 
PDE4B
 
Body, TSS1500
 
−0.88
 
(−1.2, −0.6)
 
1.11 × 10−7
 
cg03442467
 
24.6 ± 13.1
 
chr1
 
66797701
 
PDE4B
 
Body, TSS200
 
−1.86
 
(−2.6, −1.2)
 
2.84 × 10−7
 
cg13349623
 
5.8 ± 3.6
 
chr1
 
66798221
 
PDE4B
 
First exon, body
 
−0.58
 
(−0.8, −0.4)
 
2.06 × 10−9
 
cg20254265
 
79.6 ± 9.1
 
chr15
 
91306178
 
BLM
 
ExonBnd, body
 
−0.63
 
(−0.9, −0.4)
 
7.58 × 10−8
 
cg08483713 5.86 ± 2.9 chr19 11241669 LDLR Body −0.27 (−0.4, −0.2) 1.39 × 10−6 

Body, between the ATG and stop codon, irrespective of the presence of introns, exons, TSS, or promoters; ExonBnd, within 20 bases of an exon boundary, i.e., the start or end of an exon; TSS200, 0–200 bases upstream of the transcriptional start site (TSS); TSS1500, 200–1,500 bases upstream of the TSS.

Figure 1

Manhattan plot for the EWAS of maternal 2-h glucose levels post OGTT with DNA methylation in placenta (solid line: Bonferroni threshold; dashed line: FDR q <0.05).

Figure 1

Manhattan plot for the EWAS of maternal 2-h glucose levels post OGTT with DNA methylation in placenta (solid line: Bonferroni threshold; dashed line: FDR q <0.05).

Close modal
Figure 2

Regional Manhattan plot and correlation heat map for CpG sites near the differentially methylated loci in the PDE4B gene: associations with maternal 2-h glucose levels post OGTT.

Figure 2

Regional Manhattan plot and correlation heat map for CpG sites near the differentially methylated loci in the PDE4B gene: associations with maternal 2-h glucose levels post OGTT.

Close modal
Figure 3

Scatter plots for the associations between DNA methylation of placenta CpG sites and 2-h glucose levels post OGTT.

Figure 3

Scatter plots for the associations between DNA methylation of placenta CpG sites and 2-h glucose levels post OGTT.

Close modal

Association of Additional Maternal Glucose Measures With CpG Sites Associated With 2-Hour Glucose

As additional analyses, we tested associations with 1-h glucose levels measured post 50-g GCT performed at first trimester as well as fasting glucose and 1-h glucose levels post 75-g OGTT measured during the second trimester per SD increase for each glycemic trait to allow comparability (Table 3). Greater postload maternal glucose at either first (1-h glucose level post 50-g GCT) or second trimester (1-h glucose level post 75-g OGTT) was associated with lower DNA methylation at previously discovered CpG sites in PDE4B, BLM, and LDLR genes (P < 0.05), consistent in the direction of association found in discovery associations with 2-h glucose levels at second trimester but with relatively smaller effect sizes. Associations between maternal fasting glucose at second trimester and DNA methylation at CpG sites in PDE4B were in the same direction as initial findings in the 2-h glucose analyses but were not significant. Associations among other maternal glucose measures and the TNFRSF1B CpG site were less consistent.

Table 3

Adjusted differences in DNA methylation (per SD change for each glycemic trait) among CpG sites discovered in the EWAS of maternal 2-h glucose levels post OGTT

GeneCpG1-h glucose post 50-g GCT*
Baseline glucose prior to 75-g OGTT
1-h glucose post 75-g OGTT
2-h glucose post 75-g OGTT
Percent difference in DNA methylation (95% CI)PPercent difference in DNA methylation (95% CI)PPercent difference in DNA methylation (95% CI)PPercent difference in DNA methylation (95% CI)P
TNFRSF1B cg26189983 0.13 (−0.6, 0.8) 0.72 −0.06 (−0.6, 0.5) 0.84 1.00 (0.2, 1.8) 1.22 × 10−2 1.62 (1.0, 2.2) 1.70 × 10−7 
PDE4B cg07734160 −0.91 (−1.6, −0.2) 7.08 × 10−3 −0.62 (−1.2, 0.01) 0.05 −1.08 (−1.6, −0.5) 1.42 × 10−4 −1.55 (−2.1, −1.1) 1.20 × 10−9 
PDE4B cg13866577 −0.59 (−1.1, −0.1) 3.07 × 10−2 −0.50 (−1.0, −0.01) 0.04 −0.95 (−1.4, −0.5) 4.58 × 10−5 −1.17 (−1.6, −0.7) 1.11 × 10−7 
PDE4B cg03442467 −1.46 (−2.08, −0.1) 2.95 × 10−2 −0.78 (−1.9, 0.4) 0.18 −1.57 (−2.7, −0.5) 4.56 × 10−3 −2.48 (−3.4, −1.5) 2.84 × 10−7 
PDE4B cg13349623 −0.49 (−0.8, −0.2) 4.30 × 10−3 −0.26 (−0.6, 0.1) 0.10 −0.52 (−0.8. −0.2) 5.22 × 10−4 −0.77 (−1.0, −0.5) 2.06 × 10−9 
BLM cg20254265 −0.50 (−0.9, −0.1) 1.85 × 10−2 0.03 (−0.3, 0.4) 0.86 −0.65 (−1.0, −0.3) 1.45 × 10−4 −0.84 (−1.2, −0.5) 7.58 × 10−7 
LDLR cg08483713 −0.22 (−0.4, −0.1) 5.75 × 10−3 −0.02 (−0.2, 0.1) 0.80 −0.17 (−0.3, −0.02) 2.79 × 10−2 −0.37 (−0.5, −0.2) 1.39 × 10−6 
GeneCpG1-h glucose post 50-g GCT*
Baseline glucose prior to 75-g OGTT
1-h glucose post 75-g OGTT
2-h glucose post 75-g OGTT
Percent difference in DNA methylation (95% CI)PPercent difference in DNA methylation (95% CI)PPercent difference in DNA methylation (95% CI)PPercent difference in DNA methylation (95% CI)P
TNFRSF1B cg26189983 0.13 (−0.6, 0.8) 0.72 −0.06 (−0.6, 0.5) 0.84 1.00 (0.2, 1.8) 1.22 × 10−2 1.62 (1.0, 2.2) 1.70 × 10−7 
PDE4B cg07734160 −0.91 (−1.6, −0.2) 7.08 × 10−3 −0.62 (−1.2, 0.01) 0.05 −1.08 (−1.6, −0.5) 1.42 × 10−4 −1.55 (−2.1, −1.1) 1.20 × 10−9 
PDE4B cg13866577 −0.59 (−1.1, −0.1) 3.07 × 10−2 −0.50 (−1.0, −0.01) 0.04 −0.95 (−1.4, −0.5) 4.58 × 10−5 −1.17 (−1.6, −0.7) 1.11 × 10−7 
PDE4B cg03442467 −1.46 (−2.08, −0.1) 2.95 × 10−2 −0.78 (−1.9, 0.4) 0.18 −1.57 (−2.7, −0.5) 4.56 × 10−3 −2.48 (−3.4, −1.5) 2.84 × 10−7 
PDE4B cg13349623 −0.49 (−0.8, −0.2) 4.30 × 10−3 −0.26 (−0.6, 0.1) 0.10 −0.52 (−0.8. −0.2) 5.22 × 10−4 −0.77 (−1.0, −0.5) 2.06 × 10−9 
BLM cg20254265 −0.50 (−0.9, −0.1) 1.85 × 10−2 0.03 (−0.3, 0.4) 0.86 −0.65 (−1.0, −0.3) 1.45 × 10−4 −0.84 (−1.2, −0.5) 7.58 × 10−7 
LDLR cg08483713 −0.22 (−0.4, −0.1) 5.75 × 10−3 −0.02 (−0.2, 0.1) 0.80 −0.17 (−0.3, −0.02) 2.79 × 10−2 −0.37 (−0.5, −0.2) 1.39 × 10−6 

*First trimester, nonfasting.

†Second trimester, fasting.

DNA Methylation and Gene Expression in Placenta

We investigated potential functional expression adaptations of genes identified by our epigenetic investigations using placenta samples from 104 randomly selected participants. We observed that greater DNA methylation at cg03442467 (gene body and TSS200) within PDE4B was significantly correlated with higher PDE4B expression in placental tissue (r = 0.31; P = 1.81 × 10−3) (Fig. 4). The other CpG sites in PDE4B were positively correlated but not statically significant (Supplementary Fig. 3). Greater DNA methylation of TNFRSF1B at cg26189983 (gene body) was negatively correlated with expression (r = –0.24; P = 0.013), whereas greater DNA methylation at the LDLR cg08483713 (gene body) site was associated with greater expression (r = 0.32, P = 9.62 × 10−4) in placenta tissue. No association was observed between BLM DNA methylation and expression at the site (Fig. 4). There was a strong positive correlation between placental PDE4B and TNFRSF1B expression (r = 0.82, P < 2.20 × 10−16), and the expression levels of all other genes were also positively associated (Supplementary Fig. 4).

Figure 4

Pearson correlation coefficients and fitted scatter plot lines for the association between placental DNA methylation and gene expression among top loci associated with prenatal maternal 2-h glucose levels post OGTT (N = 104).

Figure 4

Pearson correlation coefficients and fitted scatter plot lines for the association between placental DNA methylation and gene expression among top loci associated with prenatal maternal 2-h glucose levels post OGTT (N = 104).

Close modal

Replication of Results in an Independent Study Cohort: ECO-21

In the external independent birth cohort with a smaller sample size (N = 65–108), we sought to replicate of our top differentially methylated site (cg26189983) at the TNFRSF1B gene and three CpG sites (cg07734160, cg13866577, cg03442467) within the PDE4B region. Study characteristics for the replication cohort are described in Supplementary Table 1. In this external replication cohort, estimated adjusted associations were consistent in direction of association but did not reach statistical significance (Supplementary Table 2). However, we had relatively low statistical power and would have required at least 200 samples to achieve adequate power for our strongest signal and main finding at PDE4B (Supplementary Table 3).

In this prepregnancy birth cohort, we conducted the largest EWAS of placenta for maternal prenatal glucose response during pregnancy thus far. We found CpG sites for which DNA methylation levels were associated with maternal glucose response 2-h post 75 g glucose loading, performed during the second trimester of pregnancy. Notably, there was evidence that multiple CpG sites in a genomic region of PDE4B were strongly and inversely associated with higher maternal 2-h glucose levels. Furthermore, greater methylation levels at identified CpG sites in the PDE4B locus correlated with higher gene expression of PDE4B in placental tissue, supporting the functional role of these placental epigenetic markers influenced by maternal glucose response in pregnancy. Other discovered epigenomic loci mapped to three additional genes—TNFRSF1B, BLM, and LDLR—observed to be associated with maternal 2-h glucose, of which DNA methylation at TNFRSF1B and LDLR loci correlated with expression of the respective gene in placental tissue. Our study provides evidence that maternal glucose response postchallenge in midpregnancy is associated with differential methylation of genes within the placenta at birth and that these loci are partially under epigenetic control for gene expression. Our results highlight the ability of the placenta to epigenetically adapt to the maternal nutritional environment that might play a functional role in metabolic programming of the offspring.

PDE4B is a member of the cyclic nucleotide phosphodiesterases family responsible for the hydrolysis of cyclic AMP and GMP (24). The PDE4 family of enzymes catalyzes the hydrolysis of second messenger cyclic AMP, a key signaling molecule for immune response regulation (25). Inhibition of PDE4 decreases secretion of tumor necrosis factor-α (TNF-α) (26), a potent proinflammatory cytokine (27). Specifically, the PDE4B isoform has been shown to predominantly mediate TNF-α release (26). Indeed, PDE4 inhibition in mice has been shown to block intrauterine inflammation, decrease cytokine production, and delay preterm birth (28). For instance, in an experimental mouse study, intrauterine injection with Escherichia coli LPS increased expression of PDE4B, eliciting an inflammatory response and triggering preterm delivery (29): in this mouse study, investigators also showed that PDE4B inhibition blocked intrauterine inflammatory response and prevented preterm delivery. Remarkably, this inflammatory response was localized in glycogen trophoblast cells of the placenta from the fetal compartment, suggesting a direct role of the placenta (29). The relationship between PDE4 inhibition and the proinflammatory cytokine TNF-α is of high interest, as TNF-α levels have been observed to be increased in adipose and placental tissue of obese pregnant women compared with nonobese pregnant women (30). Circulating levels of TNF-α are also a strong independent predictor of insulin resistance in pregnancy, with placenta tissue being a primary contributor to maternal TNF-α levels (31). The association between gestational insulin resistance and TNF-α levels in pregnancy has also been previously demonstrated in our cohort (32).

There is emerging evidence that suggests that PDE4B plays an important role in adiposity and metabolic function. For example, PDE4B-null mice have been shown to be leaner, with lower fat pad weights, smaller adipocytes, and decreased serum leptin levels compared with wild-type littermates (33). Treatment with a PDE4 inhibitor reduced the body weight of mice fed a Western-type diet mediated by an increase in energy expenditure and PDE4B mRNA in white adipose tissue (34). Furthermore, chronic treatment with PDE4 inhibitors was shown to delay the progression of diabetes in an experimental animal model for obesity, diabetes, and metabolic syndrome (db/db mice) (35). In addition, a randomized controlled trial of newly diagnosed patients with type 2 diabetes demonstrated that treatment with a PDE4 inhibitor (roflumilast) successfully lowered glucose levels (36). Therefore, PDE inhibitors have been proposed as a potential therapeutic agent for diabetes and metabolic syndrome (37). Our results add to the growing body of evidence suggesting that PDE4B plays a functional role in metabolic programming.

Higher DNA methylation of TNFRSF1B loci was associated with greater 2-h glucose levels, and a weaker positive association was also observed with 1-h glucose levels postload during the second trimester of pregnancy. TNFRSF1B encodes a high-affinity receptor for TNF-α. TNF-α is linked to metabolism and insulin sensitivity in human tissues as well as in experimental studies and genetically linked to hyperlipidemia (38,39). Expression of TNFRSF1B, also known as TNFR2, has been detected in placental trophoblasts and distributed across the cytosol with multiple functions such as apoptosis, inhibition of trophoblast cell fusion, and invasion or epithelial shedding (40). In an animal model, pharmacological attenuation of TNF-α signaling with soluble TNFR2-IgG (etanercept) protected the placenta from deformities due to infection. In obese adult individuals, prolonged treatment with etanercept improved fasting glucose and adiponectin levels (41). Furthermore, among individuals with normal glucose levels, plasma soluble TNFR2 (sTNFR2) was negatively associated with insulin sensitivity (42), while higher sTNFR2 has been observed in offspring of subjects with type 2 diabetes (43). In line with these findings, plasma sTNFR2 concentration has been proposed to serve as a marker of TNF-α–related insulin resistance (44), but it is still unclear whether circulating sTNFR2 acts as a buffer in response to higher TNFa or also participates in the inflammatory process. In our findings, higher maternal 2-h glucose is associated with greater DNA methylation of TNFRSF1B, which in turn is associated with lower expression levels, suggesting a possible maladaptation of the sTNFR2 buffering system and allowing greater inflammation within the placenta. Epigenomic modifications at the TNFRSF1B and PDE4B genes and our findings that both genes correlated highly in placental expression suggests that they act in common pathways in response to maternal glucose and point toward a likely role of TNF-α, a proinflammatory cytokine amply associated with insulin sensitivity and metabolic dysregulation (45,46).

Another CpG site associated with prenatal maternal glucose response was annotated to the body of LDLR. LDLR encodes a lipoprotein receptor that mediates endocytosis of LDL particles into the cell and it is known to be expressed in the placenta (47). Human studies have shown that intrauterine growth restriction is associated with changes in placental LDLR expression compared with normal pregnancies (48,49). Additionally, an increase in placenta LDLR expression has been documented in placentas of full-term pregnancies with gestational diabetes mellitus and has been suggested to be attributed to maternal inflammation (47). Supporting this hypothesis, in vitro studies have shown that inflammatory cytokines such as TNF-α regulate cholesterol-mediated LDL receptor regulation (50).

Lower DNA methylation at a CpG site annotated to the body of BLM was associated with greater maternal glucose levels post oral glucose load. BLM codes for an enzyme that restores replication breaks in DNA and is associated with genome stability and maintenance. Mutations of this gene are associated with an autosomal recessive syndrome, Bloom syndrome (51). However, its role in glucose homeostasis or placental functions is unknown. Given the limited literature on BLM and glycemic and metabolic traits or potential role in placenta biology, this finding must be interpreted with caution.

We did not find overlap between our differentially methylated CpG sites and those previously reported in placentas in pregnancies with gestational diabetes mellitus (911). Additionally, among the discovered CpG sites, we tested for associations in paired cord blood samples, but we did not observe consistent associations, suggesting that results are placenta specific (Supplementary Table 4).

Our study has strengths and limitations. Our strengths include our prospective design, our relatively large sample size, careful placenta collection for methylation and expression studies, and the use of the most recent technology covering >850,000 CpG sites across the genome. Although results were not directly replicated in a smaller independent cohort of pregnant women without diabetes, estimated relationships were consistent in direction but nonsignificant, possibly due to sample size and low statistical power. Despite our attempts, we were not successful in finding another cohort with a larger sample size that had appropriate phenotypes and tissue samples for replication.

Given the prospective design, our findings suggest that exposure to maternal hyperglycemia gives rise to DNA methylation alterations as part of the placental adaptations reflected in placental DNA collected at birth. Alternatively, the observed associations might be part of the pathophysiology of impaired glucose response during pregnancy and, therefore, DNA methylation shifts could be seen as a biomarker of this physiological process or even contributing to maternal glycemic regulation. Our prospective design minimizes the latter but does not rule out the possibility of reverse causality. In addition, even with bioinformatic adjustment for cell-type composition, there is currently no gold standard for cell-type adjustment for placental DNA methylation. However, the ReFACTor method has been shown to perform well when compared with reference-based methods in blood (52). Yet, adjustment with reference-free methods or reference-based methods does not guarantee that associations do not originate from cell-type differences, and results could reflect cellular lineage commitment and differentiation. The correlations observed between DNA methylation and expression could be driven by or reflect chromatin configuration. Last, our sample is composed of women from European descent and may not be generalizable to other ethnicities.

In this prospective study of healthy expecting mothers and term births, we observed robust associations between maternal glucose response and DNA methylation of the placenta at several genes implicated in inflammatory processes. Methylation levels at the discovered loci correlated with functional changes in gene expression, potentially reflecting placental adaptions to maternal impaired glucose response that could underlie fetal metabolic programming.

Funding. This work was supported by American Diabetes Association accelerator award 1-15-ACE-26 (M.-F.H.), Fonds de Recherche du Québec - Santé 20697 (M.-F.H.), Canadian Institutes of Health Research MOP 115071 (M.-F.H.), and Diabetes Québec grants (P.P. and L.B.).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. A.C. is the lead author of the study and carried out all epigenome-wide analyses with the guidance of M.-F.H., who conceived the original study design and analyses plan with the help of P.P. and L.B. V.G.-O. performed the gene expression analyses and replication of results for ECO-21 with the guidance of D.B. and L.B. C.A. helped with the data analyses. All authors helped write the manuscript and approved the final version. A.C. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented in abstract form at the 77th Scientific Sessions of the American Diabetes Association, San Diego, CA, 9–13 June 2017.

1.
Pedersen J. The Pregnant Diabetic and Her Newborn: Problems and Management. Copenhagen, Munksgaard, 1977
2.
HAPO Study Cooperative Group, Metzger BE, Persson B, et al. Hyperglycemia and adverse pregnancy outcome study: neonatal glycemia. Pediatrics 2010;126:e1545–e1552
3.
Metzger
BE
,
Persson
B
,
Lowe
LP
, et al.
Hyperglycemia and adverse pregnancy outcome study: neonatal glycemia. Pediatrics
2010
;
126
:
e1545
e1552
4.
Tam
WH
,
Ma
RCW
,
Ozaki
R
, et al
.
In utero exposure to maternal hyperglycemia increases childhood cardiometabolic risk in offspring
.
Diabetes Care
2017
;
40
:
679
686
[PubMed]
5.
Symonds
ME
,
Sebert
SP
,
Hyatt
MA
,
Budge
H
.
Nutritional programming of the metabolic syndrome
.
Nat Rev Endocrinol
2009
;
5
:
604
610
[PubMed]
6.
Brett
KE
,
Ferraro
ZM
,
Yockell-Lelievre
J
,
Gruslin
A
,
Adamo
KB
.
Maternal-fetal nutrient transport in pregnancy pathologies: the role of the placenta
.
Int J Mol Sci
2014
;
15
:
16153
16185
[PubMed]
7.
Staat
BC
,
Galan
HL
,
Harwood
JE
, et al
.
Transplacental supply of mannose and inositol in uncomplicated pregnancies using stable isotopes
.
J Clin Endocrinol Metab
2012
;
97
:
2497
2502
[PubMed]
8.
Illsley
NP
.
Glucose transporters in the human placenta
.
Placenta
2000
;
21
:
14
22
[PubMed]
9.
Ruchat
S-M
,
Houde
A-A
,
Voisin
G
, et al
.
Gestational diabetes mellitus epigenetically affects genes predominantly involved in metabolic diseases
.
Epigenetics
2013
;
8
:
935
943
[PubMed]
10.
Finer
S
,
Mathews
C
,
Lowe
R
, et al
.
Maternal gestational diabetes is associated with genome-wide DNA methylation variation in placenta and cord blood of exposed offspring
.
Hum Mol Genet
2015
;
24
:
3021
3029
[PubMed]
11.
Rong
C
,
Cui
X
,
Chen
J
,
Qian
Y
,
Jia
R
,
Hu
Y
.
DNA methylation profiles in placenta and its association with gestational diabetes mellitus
.
Exp Clin Endocrinol Diabetes
2015
;
123
:
282
288
[PubMed]
12.
Guillemette
L
,
Allard
C
,
Lacroix
M
, et al
.
Genetics of Glucose regulation in Gestation and Growth (Gen3G): a prospective prebirth cohort of mother-child pairs in Sherbrooke, Canada
.
BMJ Open
2016
;
6
:
e010031
[PubMed]
13.
Taylor
S
,
Wakem
M
,
Dijkman
G
,
Alsarraj
M
,
Nguyen
M
.
A practical approach to RT-qPCR-Publishing data that conform to the MIQE guidelines
.
Methods
2010
;
50
:
S1
S5
[PubMed]
14.
Aryee
MJ
,
Jaffe
AE
,
Corrada-Bravo
H
, et al
.
Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays
.
Bioinformatics
2014
;
30
:
1363
1369
[PubMed]
15.
Pidsley
R
,
Zotenko
E
,
Peters
TJ
, et al
.
Critical evaluation of the Illumina MethylationEPIC BeadChip microarray for whole-genome DNA methylation profiling
.
Genome Biol
2016
;
17
:
208
[PubMed]
16.
Fortin
J-P
,
Labbe
A
,
Lemire
M
, et al
.
Functional normalization of 450k methylation array data improves replication in large cancer studies
.
Genome Biol
2014
;
15
:
503
[PubMed]
17.
Xu
Z
,
Langie
SA
,
De Boever
P
,
Taylor
JA
,
Niu
L
.
RELIC: a novel dye-bias correction method for Illumina Methylation BeadChip
.
BMC Genomics
2017
;
18
:
4
[PubMed]
18.
Johnson
WE
,
Li
C
,
Rabinovic
A
.
Adjusting batch effects in microarray expression data using empirical Bayes methods
.
Biostatistics
2007
;
8
:
118
127
[PubMed]
19.
Du
P
,
Zhang
X
,
Huang
C-C
, et al
.
Comparison of Beta-value and M-value methods for quantifying methylation levels by microarray analysis
.
BMC Bioinformatics
2010
;
11
:
587
[PubMed]
20.
Meller
M
,
Vadachkoria
S
,
Luthy
DA
,
Williams
MA
.
Evaluation of housekeeping genes in placental comparative expression studies
.
Placenta
2005
;
26
:
601
607
[PubMed]
21.
Desgagné V, Hivert MF, St-Pierre J, et al. Epigenetic dysregulation of the IGF system in placenta of newborns exposed to maternal impaired glucose tolerance. Epigenomics 2014;6:193–207
22.
Fox
J
,
Weisberg
S
.
An R Companion to Applied Regression: Robust Regression in R
.
Thousand Oaks, CA
,
Sage Publications
,
2011
23.
Rahmani
E
,
Zaitlen
N
,
Baran
Y
, et al
.
Sparse PCA corrects for cell type heterogeneity in epigenome-wide association studies
.
Nat Methods
2016
;
13
:
443
445
[PubMed]
24.
Torphy
TJ
.
Phosphodiesterase isozymes: molecular targets for novel antiasthma agents
.
Am J Respir Crit Care Med
1998
;
157
:
351
370
[PubMed]
25.
Serezani
CH
,
Ballinger
MN
,
Aronoff
DM
,
Peters-Golden
M
.
Cyclic AMP: master regulator of innate immune cell function
.
Am J Respir Cell Mol Biol
2008
;
39
:
127
132
[PubMed]
26.
Jin
S-LC
,
Conti
M
.
Induction of the cyclic nucleotide phosphodiesterase PDE4B is essential for LPS-activated TNF-α responses
.
Proc Natl Acad Sci U S A
2002
;
99
:
7628
7633
[PubMed]
27.
Horiuchi
T
,
Mitoma
H
,
Harashima
S
,
Tsukamoto
H
,
Shimoda
T
.
Transmembrane TNF-α: structure, function and interaction with anti-TNF agents
.
Rheumatology (Oxford)
2010
;
49
:
1215
1228
[PubMed]
28.
Méhats
C
,
Schmitz
T
,
Oger
S
,
Hervé
R
,
Cabrol
D
,
Leroy
M-J
.
PDE4 as a target in preterm labour
.
BMC Pregnancy Childbirth
2007
;
7
(
Suppl. 1
):
S12
[PubMed]
29.
Schmitz
T
,
Souil
E
,
Hervé
R
, et al
.
PDE4 inhibition prevents preterm delivery induced by an intrauterine inflammation
.
J Immunol
2007
;
178
:
1115
1121
[PubMed]
30.
Basu
S
,
Haghiac
M
,
Surace
P
, et al
.
Pregravid obesity associates with increased maternal endotoxemia and metabolic inflammation
.
Obesity (Silver Spring)
2011
;
19
:
476
482
[PubMed]
31.
Kirwan
JP
,
Hauguel-De Mouzon
S
,
Lepercq
J
, et al
.
TNF-α is a predictor of insulin resistance in human pregnancy
.
Diabetes
2002
;
51
:
2207
2213
[PubMed]
32.
Guillemette
L
,
Lacroix
M
,
Battista
M-C
, et al
.
TNFα dynamics during the oral glucose tolerance test vary according to the level of insulin resistance in pregnant women
.
J Clin Endocrinol Metab
2014
;
99
:
1862
1869
[PubMed]
33.
Zhang
R
,
Maratos-Flier
E
,
Flier
JS
.
Reduced adiposity and high-fat diet-induced adipose inflammation in mice deficient for phosphodiesterase 4B
.
Endocrinology
2009
;
150
:
3076
3082
[PubMed]
34.
Möllmann
J
,
Kahles
F
,
Lebherz
C
, et al
.
The PDE4 inhibitor roflumilast reduces weight gain by increasing energy expenditure and leads to improved glucose metabolism
.
Diabetes Obes Metab
2017
;
19
:
496
508
[PubMed]
35.
Vollert
S
,
Kaessner
N
,
Heuser
A
, et al
.
The glucose-lowering effects of the PDE4 inhibitors roflumilast and roflumilast-N-oxide in db/db mice
.
Diabetologia
2012
;
55
:
2779
2788
[PubMed]
36.
Wouters
EF
,
Bredenbröker
D
,
Teichmann
P
, et al
.
Effect of the phosphodiesterase 4 inhibitor roflumilast on glucose metabolism in patients with treatment-naive, newly diagnosed type 2 diabetes mellitus
.
J Clin Endocrinol Metab
2012
;
97
:
E1720
E1725
[PubMed]
37.
Plock
N
,
Vollert
S
,
Mayer
M
,
Hanauer
G
,
Lahu
G
.
Pharmacokinetic/pharmacodynamic modeling of the PDE4 inhibitor TAK-648 in type 2 diabetes: early translational approaches for human dose prediction
.
Clin Transl Sci
2017
;
10
:
185
193
[PubMed]
38.
van Greevenbroek
MM
,
van der Kallen
CJ
,
Geurts
JM
,
Janssen
RG
,
Buurman
WA
,
de Bruin
TW
.
Soluble receptors for tumor necrosis factor-α (TNF-R p55 and TNF-R p75) in familial combined hyperlipidemia
.
Atherosclerosis
2000
;
153
:
1
8
[PubMed]
39.
Uysal
KT
,
Wiesbrock
SM
,
Marino
MW
,
Hotamisligil
GS
.
Protection from obesity-induced insulin resistance in mice lacking TNF-α function
.
Nature
1997
;
389
:
610
614
[PubMed]
40.
Haider
S
,
Knöfler
M
.
Human tumour necrosis factor: physiological and pathological roles in placenta and endometrium
.
Placenta
2009
;
30
:
111
123
[PubMed]
41.
Stanley
TL
,
Zanni
MV
,
Johnsen
S
, et al
.
TNF-α antagonism with etanercept decreases glucose and increases the proportion of high molecular weight adiponectin in obese subjects with features of the metabolic syndrome
.
J Clin Endocrinol Metab
2011
;
96
:
E146
E150
[PubMed]
42.
Hivert
M
,
Sullivan
L
,
Shrader
P
, et al
.
The association of TNFr2 and TNFα with insulin resistance and the influence of adipose tissue biomarkers in humans
.
Metabolism
2010
;
59
:
540
[PubMed]
43.
Straczkowski
M
,
Kowalska
I
,
Stepien
A
,
Dzienis-Straczkowska
S
,
Szelachowska
M
,
Kinalska
I
.
Increased plasma-soluble tumor necrosis factor-α receptor 2 level in lean nondiabetic offspring of type 2 diabetic subjects
.
Diabetes Care
2002
;
25
:
1824
1828
[PubMed]
44.
Dzienis-Straczkowska
S
,
Straczkowski
M
,
Szelachowska
M
,
Stepien
A
,
Kowalska
I
,
Kinalska
I
.
Soluble tumor necrosis factor-α receptors in young obese subjects with normal and impaired glucose tolerance
.
Diabetes Care
2003
;
26
:
875
880
[PubMed]
45.
Katsuki
A
,
Sumida
Y
,
Murashima
S
, et al
.
Serum levels of tumor necrosis factor-α are increased in obese patients with noninsulin-dependent diabetes mellitus
.
J Clin Endocrinol Metab
1998
;
83
:
859
862
[PubMed]
46.
Winkler
G
,
Salamon
F
,
Salamon
D
,
Speer
G
,
Simon
K
,
Cseh
K
.
Elevated serum tumour necrosis factor-alpha levels can contribute to the insulin resistance in type II (non-insulin-dependent) diabetes and in obesity
.
Diabetologia
1998
;
41
:
860
861
[PubMed]
47.
Dubé
E
,
Ethier-Chiasson
M
,
Lafond
J
.
Modulation of cholesterol transport by insulin-treated gestational diabetes mellitus in human full-term placenta
.
Biol Reprod
2013
;
88
:
16
[PubMed]
48.
Wadsack
C
,
Tabano
S
,
Maier
A
, et al
.
Intrauterine growth restriction is associated with alterations in placental lipoprotein receptors and maternal lipoprotein composition
.
Am J Physiol Endocrinol Metab
2007
;
292
:
E476
E484
[PubMed]
49.
Stepan
H
,
Faber
R
,
Walther
T
.
Expression of low density lipoprotein receptor messenger ribonuclein acid in placentas from pregnancies with preterm delivery and pregnancies with intrauterine growth retardation
.
BJOG
1999
;
106
:
1221
1222
[PubMed]
50.
Conover
CA
,
Harrington
SC
,
Bale
LK
.
Differential regulation of pregnancy associated plasma protein-A in human coronary artery endothelial cells and smooth muscle cells
.
Growth Horm IGF Res
2008
;
18
:
213
220
[PubMed]
51.
Kaneko
H
,
Fukao
T
,
Kondo
N
.
The function of RecQ helicase gene family (especially BLM) in DNA recombination and joining
.
Adv Biophys
2004
;
38
:
45
64
52.
Rahmani
E
,
Zaitlen
N
,
Baran
Y
, et al
.
Correcting for cell-type heterogeneity in DNA methylation: a comprehensive evaluation
.
Nat Methods
2017
;
14
:
218
219
[PubMed]
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.

Supplementary data