ABSTRACT
Background
Physical activity (PA) prior to and during pregnancy may have intergenerational effects on offspring health through placental epigenetic modifications. We are unaware of epidemiologic studies on longitudinal PA and placental DNA methylation.
Objectives
We evaluated the association between PA before and during pregnancy and placental DNA methylation.
Methods
Placental tissues were obtained at delivery and methylation was measured using HumanMethylation450 Beadchips for participants in the Eunice Kennedy Shriver National Institute of Child Health and Human Development Fetal Growth Studies–Singletons among 298 participants. Using the Pregnancy Physical Activity Questionnaire, women recalled periconception PA (past 12 mo) at 8–13 wk of gestation and PA since last visit at 4 follow-up visits at 16–22, 24–29, 30–33, and 34–37 wk. We conducted linear regression for associations of PA at each visit with methylation controlling for false discovery rate (FDR). Top 100 CpGs were queried for enrichment of functional pathways using Ingenuity Pathway Analysis.
Results
Periconception PA was significantly associated with 1 CpG site. PA since last visit for visits 1–4 was associated with 2, 2, 8, and 0 CpGs (log fold changes ranging from –0.0319 to 0.0080, after controlling for FDR). The largest change in methylation occurred at a site in TIMP2 , which is known to encode a protein critical for vasodilation, placentation, and uterine expansion during pregnancy (log fold change: –0.05; 95% CI: –0.06, –0.03 per metabolic equivalent of task–h/wk at 30–33 wk). Most significantly enriched pathways include cardiac hypertrophy signaling, B-cell receptor signaling, and netrin signaling. Significant CpGs and enriched pathways varied by visit.
Conclusions
Recreational PA in the year prior and during pregnancy was associated with placental DNA methylation. The associated CpG sites varied based on timing of PA. If replicated, the findings may inform the mechanisms underlying the impacts of PA on placenta health. This study was registered at clinicaltrials.gov as NCT00912132.
Keywords: physical activity (PA), maternal exercise, epigenome-wide association study, DNA methylation, fetal programming, epigenetics, intrauterine exposure, pregnancy health, prospective cohort, placenta
Introduction
Recreational physical activity (PA) during pregnancy has been associated with multiple health benefits by improving physiologic, metabolic, and psychological function (1). Specific benefits, for instance, include enhanced maternal cardiorespiratory function (2) and reduced risk of excessive weight gain, gestational diabetes, preeclampsia, and cesarean delivery (3). Maternal PA during pregnancy may thereby be related to favorable fetal, neonatal, and childhood health outcomes in the offspring (4–6), although the underlying molecular mechanisms between maternal PA and offspring health remain unclear. Emerging data from in vivo and animal studies point to a role for epigenetic changes in placenta (7, 8).
The placenta not only mediates all exchanges between the fetus and mother but also controls the hormonal environment of pregnancy and performs essential physiologic functions (9, 10). Despite being a transient organ with the life span of the gestation it supports, the placenta can affect the offspring's life course through fetal programming (11). Placental development and function are controlled partly by gene expression via epigenetic modifications such as DNA methylation (12). Mice studies suggest that PA before pregnancy can facilitate the maintenance of epigenetic modifications in favor of later embryo development in vitro (7). PA before and throughout pregnancy can improve epigenetic regulation of a metabolic master regulator in skeletal muscles for neonatal and 12-mo-old offspring mice exposed to a maternal high-fat diet (13).
Human studies on maternal PA and offspring DNA methylation are scarce. One study using cord blood and another using newborn blood spots identified a handful of methylation differences in imprinting control regions known to be important for fetal development (14, 15) in relation to PA in pregnancy. Both studies mainly assessed PA levels once at enrollment (around the end of the first trimester), capturing different periods before and during early pregnancy, which is an incomplete assessment because women may alter their PA patterns, and time windows beyond the first trimester may be important. Furthermore, neither study measured DNA methylation directly in placental tissue. In the present study, we examined the associations between longitudinal assessments of recreational PA in the year prior and throughout pregnancy and epigenome-wide methylation in placenta among a racially and ethnically diverse cohort.
Methods
Study population
The Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) Fetal Growth Studies–Singletons is a prospective cohort established to develop standards for fetal growth (16). Briefly, low-risk participants with ultrasound-confirmed gestational age (between 8 wk and 0 d and 13 wk and 6 d) were enrolled between July 2009 and January 2013 from 12 clinical sites in the United States (17). At enrollment, participants were randomly allocated to 1 of 4 ultrasonography schedules for longitudinal follow-up at 4 visits during which in-person interviews and anthropometric assessments were conducted (18). All participants were followed to delivery. This analysis includes 298 participants with placental DNA methylation data, no missing covariates, and longitudinal measures of PA (Figure 1). This study was approved by institutional review boards at NICHD (IRB number: 09-CH-N152) and each participating clinical site and coordinating center.
Figure 1.
Flowchart for determining analytical cohort for each visit. 1Details regarding sex and sample identifier mismatch are previously described (25). 2Missing marital status (n = 1), employment (n = 1), or prepregnancy BMI (n = 1). 3Sample size varied due to missing physical activity questionnaire.
Physical activity measures
Using the Pregnancy Physical Activity Questionnaire (PPAQ) (19), women recalled habitual PA in the periconception period and early pregnancy (past 12 mo, inclusive of first trimester), as well as PA during 8–22, 16–29, 24–33, 30–37, and 34–41 wk of gestation (Supplemental Table 1). These data were obtained at study visits that occurred at enrollment (8–13 wk) (baseline visit) and at targeted follow-up visits at 16–22 (visit 1), 24–29 (visit 2), 30–33 (visit 3), and 34–37 (visit 4) wk. This recall pattern ensured that PA at all time periods before and during pregnancy was captured.
Based on the Compendium of Physical Activity (20), we obtained the estimated energy required for each of the 9 recreational activities reported [in metabolic equivalent of task (MET); Supplemental Table 2]. Total recreational MET-h/wk was obtained by incorporating the MET score and amount (h/wk) for each activity and then summing across activities. Total physical activity was determined a priori as our primary exposure. Intensity of activities included light (MET ≥1.5 and <3.0), moderate (MET ≥3.0 and ≤6.0), and vigorous (MET >6.0) activities. The American College of Obstetricians and Gynecologists has issued recommendations regarding the benefits of moderate physical activity and encouraged women habitually engaged in vigorous activities to continue during pregnancy (21); as such, we focus our secondary analyses on activities falling into these intensities. Although some participants performed vigorous activities prior to and during early pregnancy (n = 170), few reported any vigorous activities at visit 4 (n = 64); we therefore combined METs from moderate and vigorous activities as secondary exposure for analyses. We conducted logic checks for extreme values of all PA measures and did not exclude any observations based on activity levels.
Placental DNA methylation and gene expression
Placental samples were collected at delivery following a detailed standardized protocol to ensure specimens were comparable and processed within 1 h. More specifically, after the free membranes were trimmed off the placental disc, the cord was cut from its attachment at the insertion site, and the placental disc was inspected, imaged, and sliced transversely at 1- to 2-cm intervals. This uniformed approach resulted in placental parenchymal biopsy specimens measuring 0.5 × 0.5 × 0.5 cm obtained from the fetal side within 1–2 cm from the cord insertion site. Biopsy samples were placed in RNALater and frozen for molecular analysis, as previously described (22). To profile methylation, extracted DNA was assayed using Infinium HumanMethylation450 Beadchips (Illumina). As our primary outcome, methylation fractions (ratio of methylated and unmethylated signals) were calculated at each CpG site using Genome Studio. To better meet model assumptions (23), methylation fractions (β values) were logit transformed to M values before analysis. Samples that did not pass quality control filters (n = 11) were excluded (24, 25). A total of 409,101 CpGs were available for the analysis (Figure 1).
To quantify RNA in placental samples, RNA was isolated using TRIZOL reagent (Invitrogen) and sequenced using the Illumina HiSeq2000 system. The expression of transcripts was quantified using Salmon, which accounts for experimental attributes and biases (26). Participants with RNA sequencing (RNA-seq) and DNA methylation (n = 80) were included in correlation tests between DNA methylation and nearby gene expression levels.
Covariate measures
Participants self-reported sociodemographic characteristics, reproductive and medical history, and behavioral and lifestyle factors at enrollment. We adjusted for covariates that were related to PA, including maternal age (years), self-identified race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, Asian or Pacific Islander), prepregnancy BMI (in kg/m2), parity (0, 1, 2+), education level (high school or less, some college or associate degree, bachelor or advanced degree), marital status (not married, married or living with partner), employment status (not employed, has at least 1 job, or is a full-time student), prepregnancy alcohol use (never, 3 times a month or less, once a week or more), and periconception sleep duration (6 h or less, 7–8 h, 9 h or more). Prepregnancy BMI was calculated from self-reported weight and height, which were highly correlated with weight and height measured by study personnel during enrollment (27). Three participants missing covariates were excluded, resulting in a total of 298 women included in this study (Figure 1).
Statistical analyses
To test if total recreational PA (MET-h/wk) prior to and during early pregnancy differed by maternal and infant characteristics, we used the Kruskal–Wallis test to compare the differences in medians. Epigenome-wide associations were performed using linear regression to estimate log fold change in DNA methylation for each additional MET-h/wk at each visit. We assessed differential methylation across 409,101 CpGs using the R package limma [i.e., lmFit function for model fitting (least squares) and eBayes function for model summarization (default options; assuming the proportion of genes differentially expressed is 0.01), version 3.48.1] (28). All analyses controlled for false discovery rate (FDR) at 5% at each visit. Analyses controlled for FDR at each visit based on a priori hypothesis that the associations can vary by study visits given the dynamics of placental and fetal development during gestation, which was also one of the reasons that longitudinal data on PA were collected across multiple study visits by design. Analyses were performed for moderate recreational PA and moderate or vigorous recreational PA at each visit (n = 6).
For all analyses described above, we adjusted for a priori covariates, as well as infant sex, 5 methylation plate number (categorical), the first 3 principal components (PCs) from the samples’ percent methylation profile using R package prcomp (continuous) (29), and the first 10 genotype-based PCs (continuous) from multidimensional scaling analysis of a pruned set of uncorrelated genome-wide single-nucleotide polymorphisms generated using PLINK to account for population structure (30). Heterogeneity in cell-type composition was accounted for using surrogate variables, as previously described (25). The optimal number of surrogate variables was identified using the “be” method, a permutation-based approach first applied to expression data in 2007 (31), in R package sva (version 3.40.0). Between 20 and 25 surrogate variables were used in analyses related to visits 0–4. Statistical analyses were conducted in SAS (version 9.4; SAS Institute) and R Studio (version 1.3.1073).
In addition, we mapped the top differentially methylated CpG sites to the nearby gene and examined enrichment of molecular pathways using the Ingenuity Pathway Analysis software (IPA; Qiagen). For each visit, the top 100 CpGs (based on FDR-adjusted P values) that could be mapped to a nearby gene were imported into the proprietary IPA Knowledge Base. If multiple CpGs were tied for the top 100, ties were included if the total number of CpGs did not exceed 500. Using the Core Analysis function in IPA, statistically significant overrepresented canonical pathways were determined by Fisher exact test followed by adjustment for multiple testing using the Benjamini–Hochberg method (32). In the subset of women with both RNA-seq and DNA methylation (n = 80), we also estimated the correlation between DNA methylation at significant CpG sites and the mRNA gene expression levels of the annotated genes using Spearman correlations. The analytic codes are available upon request.
Results
Among 298 participants, most (96%) reported performing recreational PA at baseline, with a median of 11.7 (IQR: 4.1, 21.7) MET-h/wk. The typical women reported engaging in 1.5 h/wk of walking slowly, 0.75 h/wk of walking quickly, 0.25 h/wk of walking quickly up hills, and 0.75 h/wk of other activities for fun or exercise. The median amount of recreational PA was 5.2 (IQR: 1.9, 11.7), 5.3 (IQR: 2.4, 10.4), 4.8 (IQR: 2.0, 11.0), and 4.3 (IQR: 1.1, 10.3) MET-h/wk for visits 1–4, respectively. Total recreational PA differed by race/ethnicity, with higher levels reported by non-Hispanic Whites than other races/ethnicities. Greater recreational PA was also associated with having a BMI between 19.0 and 24.9, being nulliparous, having a higher education, being employed, and drinking alcohol more than once a week prepregnancy (Table 1).
Table 1.
Total recreational PA at baseline and its association with maternal/infant characteristics (n = 298)1
| Characteristic | n | PA (MET-h/wk), median (IQR) | P value |
|---|---|---|---|
| Age,2 y | 0.24 | ||
| 18–25 | 95 | 8.7 (3.6, 19.6) | |
| 25–34 | 173 | 12.5 (4.3, 22.7) | |
| 35–40 | 30 | 12.1 (6.0, 23.4) | |
| Race/ethnicity | <0.01 | ||
| Non-Hispanic White | 76 | 17.4 (10.9, 27.8) | |
| Non-Hispanic Black | 71 | 8.3 (3.1, 18.8) | |
| Hispanic | 101 | 9.7 (3.6, 19.2) | |
| Asian/Pacific Islander | 50 | 9.4 (2.4, 16.1) | |
| Prepregnancy BMI,2,3 kg/m2 | 0.04 | ||
| Normal weight (19.0–24.9) | 194 | 12.5 (4.7, 24.1) | |
| Overweight (25.0–29.9) | 75 | 11.0 (3.1, 19.8) | |
| Obese (≥30.0) | 29 | 8.1 (3.4, 11.7) | |
| Parity (number of births) | 0.01 | ||
| 0 | 138 | 13.4 (4.2, 27.0) | |
| 1 | 108 | 11.3 (5.5, 18.4) | |
| ≥2 | 52 | 9.0 (2.5, 15.9) | |
| Education level | <0.01 | ||
| ≤ High school | 91 | 6.3 (2.6, 11.8) | |
| Some college/associate degree | 86 | 11.4 (4.5, 20.9) | |
| ≥ Bachelor degree | 121 | 16.1 (9.6, 26.2) | |
| Marital status | 0.09 | ||
| Other | 73 | 8.7 (3.6, 17.4) | |
| Married/living as married | 225 | 12.5 (4.3, 22.2) | |
| Employment status | 0.02 | ||
| Unemployed | 89 | 8.1 (3.6, 17.1) | |
| Employed/student | 209 | 12.5 (4.7, 23.5) | |
| Alcohol use prepregnancy | <0.01 | ||
| Never | 105 | 5.9 (2.0, 12.6) | |
| <3 times a month | 104 | 11.4 (5.2, 20.9) | |
| ≥ Once a week | 89 | 18.7 (11.7, 29.1) | |
| Periconception sleep duration | 0.12 | ||
| ≤6 h | 55 | 13.3 (5.8, 23.8) | |
| 7–8 h | 196 | 11.5 (3.9, 21.3) | |
| ≥9 h | 47 | 8.6 (2.3, 20.8) | |
| Infant sex | 0.76 | ||
| Male | 150 | 11.6 (4.3, 22.0) | |
| Female | 148 | 11.7 (3.7, 20.9) | |
| Birth weight | 0.15 | ||
| <2500 g | 22 | 7.2 (2.6, 17.7) | |
| ≥2500 and <4000 g | 264 | 11.8 (4.2, 22,1) | |
| ≥4000 g | 12 | 9.3 (3.0, 11.8) | |
| Physical activity | |||
| 8–13 wk of gestation (baseline) | 298 | 11.7 (4.1, 21.7) | |
| 16–22 wk of gestation (visit 1) | 291 | 5.2 (1.9, 11.7) | |
| 24–29 wk of gestation (visit 2) | 292 | 5.3 (2.4, 10.4) | |
| 30–33 wk of gestation (visit 3) | 286 | 4.8 (2.0, 11.0) | |
| 34–37 wk of gestation (visit 4) | 281 | 4.3 (1.1, 10.3) |
P values from Kruskal–Wallis test of differences in median physical activity at visit 0 (8–13 wk of gestation, n = 298). MET, metabolic equivalent of task; PA, physical activity.
Categorized for this table but modeled as continuous variables.
Calculated from self-reported height (meters) and pregravid weight (kilograms). No women were underweight (BMI <18.9 kg/m2).
Total recreational PA was significantly associated with placental DNA methylation for 13 CpG sites (1 at visit 0, 2 at visit 1, 2 at visit 2, 8 at visit 3, and none at visit 4) (Table 2; quantile–quantile plots in Supplemental Figure 1). Except for 1 CpG (cg06265277), all significant sites were associated with decreased methylation. Several CpG sites, including cg14517108 (TIMP2), cg05580512 (LPCAT1), cg21707521 (UBE2F), cg25970266 (CSK), cg05074892 (BRD2), cg10965041 (EYA3), and cg06100552 (MAGI1), were consistently among the top 100 CpG sites at multiple visits in association with PA. The largest decrease in methylation with respect to higher PA occurred at a site located in a CpG island of TIMP2 (log fold change: –0.05; 95% CI: –0.06, –0.03 per MET-h/wk at visit 3). Methylation at one of the sites identified (cg16141228 near CLINT1) was negatively correlated with CLINT1gene expression in placenta (Spearman r = –0.34, P = 0.004, Supplemental Table 3). Methylation at other CpGs was not correlated with gene expression.
Table 2.
Significant CpG sites and nearby annotated genes associated with total recreational physical activity across visits1
| CpG | Gene | Chromosome | Position | Relation to island | logFC2 | 95% CI | P value | FDR-adjusted P value | Visit number3 |
|---|---|---|---|---|---|---|---|---|---|
| cg14122102 | MOB3B | chr9 | 27333402 | OpenSea | –3.19E-02 | –4.15E-02, –2.23E-02 | 3.53E-10 | 1.45E-04* | 0 |
| cg14517108 | TIMP2 | chr17 | 76921504 | Island | –4.58E-02 | –6.00E-02, –3.17E-02 | 9.07E-10 | 3.71E-04* | 3 |
| –3.85E-02 | –5.62E-02, –2.07E-02 | 2.78E-05 | 5.24E-01 | 4 | |||||
| cg05580512 | LPCAT1 | chr5 | 1463045 | N_Shore | –5.31E-03 | –7.92E-03, –2.71E-03 | 7.64E-05 | 3.43E-01 | 0 |
| –1.18E-02 | –1.56E-02, –7.96E-03 | 4.73E-09 | 9.67E-04* | 3 | |||||
| cg21707521 | UBE2F | chr2 | 238875814 | Island | –1.61E-02 | –2.33E-02, –8.80E-03 | 1.98E-05 | 1.42E-01 | 1 |
| –1.57E-02 | –2.32E-02, –8.22E-03 | 5.03E-05 | 3.29E-01 | 2 | |||||
| –2.24E-02 | –3.02E-02, –1.45E-02 | 5.50E-08 | 7.49E-03* | 3 | |||||
| cg25970266 | CSK | chr15 | 75074299 | N_Shore | –1.56E-02 | –2.29E-02, –8.42E-03 | 2.86E-05 | 2.46E-01 | 0 |
| –2.48E-02 | –3.65E-02, –1.30E-02 | 4.74E-05 | 3.29E-01 | 2 | |||||
| –3.36E-02 | –4.59E-02, –2.13E-02 | 1.73E-07 | 1.77E-02* | 3 | |||||
| cg07176514 | C1orf25 | chr1 | 185126191 | Island | –4.09E-02 | –5.54E-02, –2.64E-02 | 8.10E-08 | 2.17E-02* | 2 |
| cg05074892 | BRD2 | chr6 | 32939292 | Island | –1.43E-02 | –2.03E-02, –8.22E-03 | 5.44E-06 | 1.24E-01 | 0 |
| –2.06E-02 | –3.01E-02, –1.11E-02 | 2.80E-05 | 1.67E-01 | 1 | |||||
| –2.66E-02 | –3.62E-02, –1.71E-02 | 1.06E-07 | 2.17E-02* | 2 | |||||
| –2.44E-02 | –3.48E-02, –1.41E-02 | 5.78E-06 | 1.08E-01 | 3 | |||||
| –2.65E-02 | –3.95E-02, –1.34E-02 | 8.36E-05 | 5.61E-01 | 4 | |||||
| cg16141228 | CLINT1 | chr5 | 157286335 | S_Shore | –1.35E-02 | –1.83E-02, –8.77E-03 | 6.26E-08 | 2.56E-02* | 1 |
| cg10965041 | EYA3 | chr1 | 28415294 | Island | –1.88E-02 | –2.59E-02, –1.17E-02 | 4.09E-07 | 3.35E-02* | 3 |
| –1.94E-02 | –2.86E-02, –1.02E-02 | 4.42E-05 | 5.39E-01 | 4 | |||||
| cg06265277 | chr15 | 30260973 | Island | 7.95E-03 | 5.04E-03, 1.09E-02 | 1.74E-07 | 3.56E-02* | 1 | |
| cg15787616 | chr3 | 137482006 | N_Shore | –2.88E-02 | –4.18E-02, –1.58E-02 | 1.83E-05 | 2.27E-01 | 2 | |
| –3.54E-02 | –4.90E-02, –2.19E-02 | 5.70E-07 | 3.89E-02* | 3 | |||||
| cg21087962 | KPNA4 | chr3 | 160283382 | Island | –3.36E-02 | –4.68E-02, –2.05E-02 | 9.09E-07 | 4.69E-02* | 3 |
| cg02984023 | RTN3 | chr11 | 63448764 | Island | –3.25E-02 | –4.52E-02, –1.98E-02 | 9.17E-07 | 4.69E-02* | 3 |
Results from epigenome-wide associations performed using linear regression, adjusting for maternal age (years), self-identified race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, Asian or Pacific Islander), prepregnancy BMI (in kg/m2), parity (0, 1, 2+), education level (high school or less, some college or associate degree, bachelor or advanced degree), marital status (not married, married or living with partner), employment status (not employed, has at least 1 job, or is a full-time student), prepregnancy alcohol use (never, 3 times a month or less, once a week or more), periconception sleep duration (6 h or less, 7–8 h, 9 h or more), infant sex, 5 methylation plate number (categorical), the first 3 principal components (PCs) from the samples’ percent methylation profile, the first 10 genotype-based PCs to account for population structure, and surrogate variables to account for heterogeneity in cell-type composition. FDR-adjusted P value <0.05 for any visit are marked with *, and the corresponding logFC and P values for other visits are also presented if among the top 100 CpGs for that visit. Chr, chromosome; FC, fold change; FDR, false discovery rate; N, north; S, south.
LogFC is the log of fold change (base 2); logFC <0 is associated with hypomethylation and logFC >0 is associated with hypermethylation.
Visits 0–4 are targeted for 8–13, 16–22, 24–29, 30–33, and 34–37 wk of gestation, respectively. The sample size for visits 0–4 is 298, 291, 292, 286, and 281, respectively (Figure 1).
Moderate recreational PA was also significantly associated with methylation at 15 CpG sites (1 at visit 0, 11 at visit 1, 1 at visit 2, 2 at visit 3, and none at visit 4) (Table 3). Only 1 CpG overlapped with the 13 CpGs associated with total recreational PA (i.e., cg16141228 near CLINT1). For cg16141228 (CLINT1), higher amounts of both total and moderate PA were associated with decreased methylation at visit 1, and its methylation level was negatively correlated, as described above.
Table 3.
Significant CpG sites and nearby annotated genes associated with moderate recreational physical activity across visits1
| CpG | Gene | Chromosome | Position | Relation to island | logFC2 | 95% CI | P value | FDR-adjusted P value | Visit number3 |
|---|---|---|---|---|---|---|---|---|---|
| cg07179329 | CDH13 | chr16 | 83171992 | OpenSea | –5.02E-02 | –6.72E-02, –3.32E-02 | 2.03E-08 | 8.32E-03* | 2 |
| cg17462962 | MRPL22 | chr5 | 154347281 | OpenSea | 9.46E-03 | 6.09E-03, 1.28E-02 | 8.28E-08 | 1.85E-02* | 1 |
| cg16141228 | CLINT1 | chr5 | 157286335 | S_Shore | –1.92E-02 | –2.61E-02, –1.23E-02 | 1.02E-07 | 1.85E-02* | 1 |
| cg26169320 | MYCBP2 | chr13 | 77901010 | Island | –3.93E-02 | –5.36E-02, –2.51E-02 | 1.35E-07 | 1.85E-02* | 1 |
| –3.63E-02 | –5.32E-02, –1.95E-02 | 3.07E-05 | 5.06E-01 | 4 | |||||
| cg13179085 | SOCS2 | chr12 | 93963345 | N_Shore | –4.55E-02 | –6.13E-02, –2.96E-02 | 5.14E-08 | 2.10E-02* | 3 |
| cg01106412 | OSBPL3 | chr7 | 25020404 | N_Shore | –4.17E-02 | –5.71E-02, –2.63E-02 | 2.38E-07 | 2.44E-02* | 1 |
| –2.39E-02 | –3.64E-02, –1.14E-02 | 2.17E-04 | 7.32E-01 | 0 | |||||
| cg09575314 | CPNE1 | chr20 | 34220833 | OpenSea | –8.02E-03 | –1.11E-02, –5.00E-03 | 3.91E-07 | 3.20E-02* | 1 |
| cg19596493 | PTP4A2 | chr1 | 32404086 | Island | –6.95E-02 | –9.60E-02, –4.30E-02 | 4.95E-07 | 3.38E-02* | 1 |
| cg16364121 | chr17 | 35291924 | Island | –5.06E-02 | –7.02E-02, –3.11E-02 | 6.79E-07 | 3.47E-02* | 1 | |
| cg11166466 | CDON | chr11 | 125932389 | Island | –5.90E-02 | –8.17E-02, –3.62E-02 | 7.02E-07 | 3.47E-02* | 1 |
| –5.37E-02 | –7.74E-02, –3.00E-02 | 1.27E-05 | 2.48E-01 | 2 | |||||
| cg13392587 | NAGLU | chr17 | 40689561 | S_Shore | –9.44E-03 | –1.31E-02, –5.78E-03 | 7.64E-07 | 3.47E-02* | 1 |
| cg22586884 | PPAP2A | chr5 | 54832140 | S_Shore | 1.63E-02 | 9.97E-03, 2.27E-02 | 8.94E-07 | 3.64E-02* | 1 |
| 1.41E-02 | 7.58E-03, 2.06E-02 | 2.82E-05 | 3.21E-01 | 2 | |||||
| 1.65E-02 | 1.00E-02, 2.30E-02 | 1.02E-06 | 8.37E-02 | 3 | |||||
| cg12425768 | KIF1B | chr1 | 10269828 | N_Shore | –4.62E-02 | –6.44E-02, –2.81E-02 | 9.79E-07 | 3.64E-02* | 1 |
| cg01440514 | NCAPD3 | chr11 | 134094371 | Island | –2.23E-02 | –3.03E-02, –1.43E-02 | 9.27E-08 | 3.79E-02* | 0 |
| cg00384993 | chr20 | 2630663 | N_Shelf | 1.18E-02 | 7.50E-03, 1.62E-02 | 1.95E-07 | 3.99E-02* | 3 |
Results from epigenome-wide associations performed using linear regression, adjusting for maternal age (years), self-identified race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, Asian or Pacific Islander), prepregnancy BMI (in kg/m2), parity (0, 1, 2+), education level (high school or less, some college or associate degree, bachelor or advanced degree), marital status (not married, married or living with partner), employment status (not employed, has at least 1 job, or is a full-time student), prepregnancy alcohol use (never, 3 times a month or less, once a week or more), periconception sleep duration (6 h or less, 7–8 h, 9 h or more), infant sex, 5 methylation plate number (categorical), the first 3 principal components (PCs) from the samples’ percent methylation profile, the first 10 genotype-based PCs to account for population structure, and surrogate variables to account for heterogeneity in cell-type composition. FDR-adjusted P value <0.05 for any visit are marked with *, and corresponding logFC and P values for other visits are also presented if among the top 100 CpGs for that visit. Chr, chromosome; FC, fold change; FDR, false discovery rate; N, north; S, south.
LogFC is the log of fold change (base 2); logFC <0 is associated with hypomethylation and logFC >0 is associated with hypermethylation.
Visits 0–4 are targeted for 8–13, 16–22, 24–29, 30–33, and 34–37 wk of gestation, respectively. The sample size for visits 0–4 is 298, 291, 292, 286, and 281, respectively (Figure 1).
The number of participants performing vigorous activities was small (ranging from 170 at visit 0 to 64 at visit 4). When moderate or vigorous recreational PA was combined, 29 CpGs were identified (1 at visit 0, 20 at visit 1, 2 at visit 2, and 6 at visit 3) (Table 4). Eleven of the 13 CpGs identified with total recreational PA overlapped with these 29 CpGs. Of note, cg16141228 (CLINT1) was identified regardless of recreational PA intensity modeled. In addition to cg16141228 (CLINT1), methylation at cg07424889 near FOXK1 was negatively correlated with FOXK1 expression (r = –0.29, P = 0.015; Supplemental Table 3).
Table 4.
Significant CpG sites and nearby annotated genes associated with moderate or vigorous recreational physical activity across visits1
| CpG | Gene | Chromosome | Position | Relation to island | logFC2 | 95% CI | P value | FDR-adjusted P value | Visit number3 |
|---|---|---|---|---|---|---|---|---|---|
| cg14122102 | MOB3B | chr9 | 27333402 | OpenSea | –3.27E-02 | –4.24E-02, –2.30E-02 | 2.12E-10 | 8.69E-05* | 0 |
| cg05580512 | LPCAT1 | chr5 | 1463045 | N_Shore | –1.18E-02 | –1.57E-02, –7.98E-03 | 5.94E-09 | 2.38E-03* | 3 |
| –5.43E-03 | –8.04E-03, –2.82E-03 | 5.73E-05 | 2.42E-01 | 0 | |||||
| cg14517108 | TIMP2 | chr17 | 76921504 | Island | –4.34E-02 | –5.79E-02, –2.90E-02 | 1.16E-08 | 2.38E-03* | 3 |
| –3.69E-02 | –5.50E-02, –1.89E-02 | 7.46E-05 | 5.71E-01 | 4 | |||||
| cg06265277 | chr15 | 30260973 | Island | 8.77E-03 | 5.86E-03, 1.17E-02 | 1.03E-08 | 4.23E-03* | 1 | |
| cg16141228 | CLINT1 | chr5 | 157286335 | S_Shore | –1.40E-02 | –1.88E-02, –9.19E-03 | 2.90E-08 | 5.92E-03* | 1 |
| cg21707521 | UBE2F | chr2 | 238875814 | Island | –2.22E-02 | –3.01E-02, –1.42E-02 | 1.11E-07 | 1.51E-02* | 3 |
| –1.73E-02 | –2.45E-02, –1.00E-02 | 4.47E-06 | 6.76E-02 | 1 | |||||
| –1.58E-02 | –2.34E-02, –8.28E-03 | 4.99E-05 | 3.39E-01 | 2 | |||||
| cg21087962 | KPNA4 | chr3 | 160283382 | Island | –3.59E-02 | –4.91E-02, –2.27E-02 | 2.09E-07 | 1.90E-02* | 3 |
| cg25970266 | CSK | chr15 | 75074299 | N_Shore | –3.36E-02 | –4.60E-02, –2.11E-02 | 2.55E-07 | 1.90E-02* | 3 |
| –1.54E-02 | –2.27E-02, –8.08E-03 | 4.61E-05 | 2.37E-01 | 0 | |||||
| –2.48E-02 | –3.65E-02, –1.31E-02 | 4.47E-05 | 3.32E-01 | 2 | |||||
| cg15787616 | chr3 | 137482006 | N_Shore | –3.68E-02 | –5.04E-02, –2.31E-02 | 2.78E-07 | 1.90E-02* | 3 | |
| –2.94E-02 | –4.24E-02, –1.63E-02 | 1.47E-05 | 2.28E-01 | 2 | |||||
| –1.65E-02 | –2.45E-02, –8.52E-03 | 6.35E-05 | 2.53E-01 | 0 | |||||
| cg05074892 | BRD2 | chr6 | 32939292 | Island | –2.69E-02 | –3.65E-02, –1.72E-02 | 9.73E-08 | 3.10E-02* | 2 |
| –1.47E-02 | –2.08E-02, –8.62E-03 | 3.41E-06 | 1.20E-01 | 0 | |||||
| –2.11E-02 | –3.05E-02, –1.16E-02 | 1.73E-05 | 1.29E-01 | 1 | |||||
| –2.45E-02 | –3.50E-02, –1.39E-02 | 7.45E-06 | 1.34E-01 | 3 | |||||
| –2.74E-02 | –4.06E-02, –1.42E-02 | 5.82E-05 | 5.69E-01 | 4 | |||||
| cg07176514 | C1orf25 | chr1 | 185126191 | Island | –4.10E-02 | –5.59E-02, –2.61E-02 | 1.51E-07 | 3.10E-02* | 2 |
| cg16586192 | APH1B | chr15 | 63569620 | Island | –5.59E-02 | –7.66E-02, –3.51E-02 | 2.60E-07 | 3.55E-02* | 1 |
| –4.53E-02 | –6.77E-02, –2.29E-02 | 8.83E-05 | 3.98E-01 | 2 | |||||
| cg23923070 | ZNF638 | chr2 | 71558899 | Island | –2.97E-02 | –4.13E-02, –1.82E-02 | 7.54E-07 | 4.48E-02* | 1 |
| cg07368146 | NCRNA00219 | chr5 | 111497076 | S_Shore | –2.57E-02 | –3.57E-02, –1.57E-02 | 7.96E-07 | 4.48E-02* | 1 |
| cg14673194 | chr17 | 80132900 | OpenSea | 7.04E-03 | 3.84E-03, 1.02E-02 | 9.39E-07 | 4.48E-02* | 1 | |
| cg15467911 | CHD3 | chr17 | 7791319 | S_Shelf | 8.07E-03 | 4.92E-03, 1.12E-02 | 8.87E-07 | 4.48E-02* | 1 |
| cg22586884 | PPAP2A | chr5 | 54832140 | S_Shore | 1.14E-02 | 6.92E-03, 1.58E-02 | 9.22E-07 | 4.48E-02* | 1 |
| 1.22E-02 | 7.17E-03, 1.72E-02 | 2.89E-06 | 8.61E-02 | 3 | |||||
| 1.04E-02 | 5.63E-03, 1.53E-02 | 2.82E-05 | 2.89E-01 | 2 | |||||
| 1.26E-02 | 6.36E-03, 1.89E-02 | 9.66E-05 | 5.90E-01 | 4 | |||||
| cg06327998 | CDYL | chr6 | 4892662 | OpenSea | 6.82E-03 | 4.15E-03, 9.49E-03 | 9.85E-07 | 4.48E-02* | 1 |
| 5.92E-03 | 3.06E-03, 8.79E-03 | 6.32E-05 | 3.70E-01 | 2 | |||||
| cg10546562 | TNRC18 | chr7 | 5390491 | N_Shore | –1.61E-02 | –2.25E-02, –9.75E-03 | 1.22E-06 | 4.58E-02* | 1 |
| cg19893494 | NSMCE2 | chr8 | 126162903 | OpenSea | –1.03E-02 | –1.44E-02, –6.22E-03 | 1.34E-06 | 4.58E-02* | 1 |
| –1.14E-02 | –1.58E-02, –6.95E-03 | 8.64E-07 | 5.05E-02 | 3 | |||||
| cg04722155 | GAK | chr4 | 851454 | OpenSea | 1.29E-02 | 7.76E-03, 1.80E-02 | 1.40E-06 | 4.58E-02* | 1 |
| cg01920494 | STX18 | chr4 | 4543764 | Island | –2.96E-02 | –4.14E-02, –1.78E-02 | 1.45E-06 | 4.58E-02* | 1 |
| –3.27E-02 | –4.61E-02, –1.94E-02 | 2.35E-06 | 8.02E-02 | 3 | |||||
| –2.95E-02 | –4.18E-02, –1.71E-02 | 4.58E-06 | 1.25E-01 | 2 | |||||
| –1.75E-02 | –2.51E-02, –9.80E-03 | 1.10E-05 | 1.63E-01 | 0 | |||||
| cg07424889 | FOXK1 | chr7 | 4768247 | S_Shelf | –7.68E-03 | –1.08E-02, –4.58E-03 | 1.85E-06 | 4.83E-02* | 1 |
| cg13392587 | NAGLU | chr17 | 40689561 | S_Shore | –6.33E-03 | –8.89E-03, –3.77E-03 | 2.01E-06 | 4.83E-02* | 1 |
| cg08301525 | NXN | chr17 | 867782 | OpenSea | 1.09E-02 | 6.51E-03, 1.54E-02 | 2.11E-06 | 4.83E-02* | 1 |
| cg13631094 | ROR2 | chr9 | 94619684 | OpenSea | –4.09E-03 | –5.75E-03, –2.43E-03 | 2.19E-06 | 4.83E-02* | 1 |
| cg07312357 | ESRRG | chr1 | 217263805 | S_Shore | 1.08E-02 | 6.43E-03, 1.52E-02 | 2.19E-06 | 4.83E-02* | 1 |
| cg08478283 | CD58 | chr1 | 117076803 | N_Shore | –4.28E-03 | –6.02E-03, –2.54E-03 | 2.24E-06 | 4.83E-02* | 1 |
| –3.88E-03 | –5.70E-03, –2.06E-03 | 3.79E-05 | 3.28E-01 | 2 | |||||
| cg13951572 | SORBS3 | chr8 | 22409000 | Island | –4.02E-02 | –5.66E-02, –2.39E-02 | 2.38E-06 | 4.87E-02* | 1 |
Results from epigenome-wide associations performed using linear regression, adjusting for maternal age (years), self-identified race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic, Asian or Pacific Islander), prepregnancy BMI (in kg/m2), parity (0, 1, 2+), education level (high school or less, some college or associate degree, bachelor or advanced degree), marital status (not married, married or living with partner), employment status (not employed, has at least 1 job, or is a full-time student), prepregnancy alcohol use (never, 3 times a month or less, once a week or more), periconception sleep duration (6 h or less, 7–8 h, 9 h or more), infant sex, 5 methylation plate number (categorical), the first 3 principal components (PCs) from the samples’ percent methylation profile, the first 10 genotype-based PCs to account for population structure, and surrogate variables to account for heterogeneity in cell-type composition. FDR-adjusted P value <0.05 for any visit are marked with *, and the corresponding logFC and P values for other visits are also presented if among the top 100 CpGs for that visit. Chr, chromosome; FC, fold change; FDR, false discovery rate; N, north; S, south.
LogFC is the log of fold change (base 2); logFC <0 is associated with hypomethylation and logFC >0 is associated with hypermethylation.
Visits 0–4 are targeted for 8–13, 16–22, 24–29, 30–33, and 34–37 wk of gestation, respectively. The sample size for visits 0–4 is 298, 291, 292, 286, and 281, respectively (Figure 1).
Ten pathways were significantly enriched for genes located near the top 100 CpG sites for total recreational PA (Table 5). All 10 pathways were identified at visit 0 (i.e., habitual PA in the periconception period and early pregnancy). Three pathways were tied for being most significant: cardiac hypertrophy signaling (enhanced), B-cell receptor signaling, and netrin signaling. For moderate recreational PA, no pathways were significantly enriched after adjustment for multiple testing (results not shown). For moderate or vigorous recreational PA, 32 pathways were enriched, with 10 identified at visit 0, none at visit 1, 22 at visit 2, none at visit 3, and none at visit 4 (Table 6). Among these, the aforementioned cardiac hypertrophy signaling (enhanced) was also the most significant pathway.
Table 5.
Summary of functional pathways enriched for genes near top 100 CpG sites associated with total recreational physical activity at multiple visits1
| Pathways | Molecule | Ratio2 | P value | B-H P value | Top functions and diseases | Visit number3 |
|---|---|---|---|---|---|---|
| Cardiac hypertrophy signaling (enhanced) | CACNA1C, CAMK2B, IL13, JUN, NFATC1, NFATC2, PLCG2, PRKG1, TSC2, WNT10A | 0.019 | 1.02E-04 | 2.63E-02* | Cardiac enlargement; cardiovascular disease; cardiovascular system development and function | 0 |
| B-cell receptor signaling | CAMK2B, CSK, JUN, NFATC1, NFATC2, PLCG2 | 0.031 | 1.86E-04 | 2.63E-02* | Cellular development; cellular growth and proliferation; embryonic development | 0 |
| Netrin signaling | CACNA1C, NFATC1, NFATC2, PRKG1 | 0.056 | 2.57E-04 | 2.63E-02* | Cellular assembly and organization; cellular movement; skeletal and muscular system development and function | 0 |
| PI3K signaling in B lymphocytes | CAMK2B, JUN, NFATC1, NFATC2, PLCG2 | 0.035 | 3.72E-04 | 2.88E-02* | Cellular development; cellular growth and proliferation; embryonic development | 0 |
| B-cell activating factor signaling | JUN, NFATC1, NFATC2 | 0.070 | 8.32E-04 | 2.95E-02* | Cellular development; cellular growth and proliferation; hematologic system development and function | 0 |
| Molecular mechanisms of cancer | ARHGEF4, CAMK2B, JUN, LRP1, NOTCH1, PA2G4, PRKDC, WNT10A | 0.018 | 6.46E-04 | 2.95E-02* | Embryonic development; organismal development; connective tissue development and function | 0 |
| APH1B, HDAC4, JAK1, JUN, PRKCZ, PTCH1 | 0.014 | 1.38E-02 | 2.19E-01 | 2 | ||
| APRIL-mediated signaling | JUN, NFATC1, NFATC2 | 0.071 | 7.76E-04 | 2.95E-02* | Cellular development; cellular growth and proliferation; hematologic system development and function | 0 |
| Role of macrophages, fibroblasts, and endothelial cells in rheumatoid arthritis | CAMK2B, JUN, LRP1, NFATC1, NFATC2, PLCG2, WNT10A | 0.022 | 4.90E-04 | 2.95E-02* | Cellular Development; cellular growth and proliferation; hematologic system development and function | 0 |
| AXIN1, CREB5, FZD6, MAP2K3, WNT10A | 0.015 | 4.57E-02 | 3.12E-01 | 4 | ||
| Th1 and Th2 activation pathway | IL13, JUN, NFATC1, NFATC2, NOTCH1 | 0.029 | 8.71E-04 | 2.95E-02* | Cellular development; cellular growth and proliferation; embryonic development | 0 |
| APH1B, HLA-DOB, JAK1, JUN | 0.023 | 7.08E-03 | 2.12E-01 | 2 | ||
| JAK1, JUN, NOTCH1 | 0.017 | 4.47E-02 | 3.20E-01 | 3 | ||
| Systemic lupus erythematosus in B-cell signaling pathway | CSK, IL13, JUN, NFATC1, NFATC2, PLCG2 | 0.022 | 1.23E-03 | 3.80E-02* | Cellular development; cellular growth and proliferation; embryonic development | 0 |
| CSK, ISG15, JAK1, JUN, PRKCZ | 0.018 | 7.59E-03 | 2.12E-01 | 2 | ||
| CSK, JAK1, JUN, PRKCA | 0.014 | 3.89E-02 | 3.19E-01 | 3 |
Statistically significant overrepresented canonical pathways were determined by Fisher exact test followed by adjustment for multiple testing using the B-H method. BH-adjusted P value <0.05 for any visit are marked with *. Associated molecules and P values for other visits are also presented if nominal P value <0.05. All pathways presented have 2 or more molecules. APRIL, a proliferation-inducing ligand; B-H, Benjamini–Hochberg (adjustment for multiple testing).
Overlap is the number of molecules in our list over the number of molecules in the pathway of interest.
Visits 0–4 are targeted for 8–13, 16–22, 24–29, 30–33, and 34–37 wk of gestation, respectively. The sample size for visits 0–4 is 298, 291, 292, 286, and 281, respectively (Figure 1).
Table 6.
Summary of functional pathways enriched for genes near the top 100 CpG sites associated with moderate or vigorous recreational physical activity at multiple visits1
| Pathways | Molecule | Ratio2 | P value | B-H P value | Top functions & diseases | Visit number3 |
|---|---|---|---|---|---|---|
| Cardiac hypertrophy signaling (enhanced) | CACNA1C, GNB1, GNG4, IL13, JUN, NFATC1, NFATC2, PDE8A, PLCG | 0.020 | 3.31E-05 | 9.77E-03* | Cardiac enlargement; cardiovascular disease; cardiovascular system development and function | 0 |
| Gaq signaling | CSK, GNB1, GNG4, NFATC1, NFATC2, PLCG2 | 0.035 | 1.26E-04 | 1.86E-02* | Posttranslational modification; cellular assembly and organization; lipid metabolism | 0 |
| CSK, GNG7, PRKCA | 0.018 | 4.37E-02 | 3.19E-01 | 3 | ||
| Netrin signaling | CACNA1C, NFATC1, NFATC2, PRKG1 | 0.056 | 3.24E-04 | 3.16E-02* | Cellular assembly and organization; cellular movement; skeletal and muscular system development and function | 0 |
| Glucocorticoid receptor signaling | FOS, GTF2B, HLA-DOB, HSPA6, JAK1, JUN, NDUFB10, NDUFV1 | 0.014 | 1.91E-03 | 3.39E-02* | Gene expression; embryonic development; organismal development | 2 |
| RAR activation | AKR1B1, CSK, FOS, JUN, PRKCZ | 0.026 | 1.00E-03 | 3.39E-02* | Gene expression; cellular development; cellular growth and proliferation | 2 |
| CSK, JUN, PRKCA, RDH13 | 0.020 | 1.26E-02 | 2.10E-01 | 3 | ||
| 14-3-3–mediated signaling | FOS, JUN, PRKCZ, TSC2 | 0.032 | 1.51E-03 | 3.39E-02* | Cell death and survival; organismal injury and abnormalities; cell cycle molecules | 2 |
| JUN, PLCG2, TSC2, TUBA1B | 0.032 | 2.69E-03 | 5.37E-02 | 0 | ||
| PRKCA, YWHAZ | 0.016 | 3.47E-02 | 3.49E-01 | 1 | ||
| Role of PKR in interferon induction and antiviral response | FOS, HSPA6, JAK1, JUN | 0.029 | 1.95E-03 | 3.39E-02* | Cell death and survival; organismal injury and abnormalities; cellular development | 2 |
| IL-3 signaling | FOS, JAK1, JUN, PRKCZ | 0.051 | 2.57E-04 | 3.39E-02* | Cell signaling; cell-to-cell signaling and interaction; cellular development | 2 |
| JAK1, JUN, PRKCA | 0.038 | 5.75E-03 | 1.98E-01 | 3 | ||
| Thrombopoietin signaling | FOS, JUN, PRKCZ | 0.048 | 1.91E-03 | 3.39E-02* | Cell cycle; cancer; organismal injury and abnormalities | 2 |
| JUN, PLCG2 | 0.032 | 3.31E-02 | 1.78E-01 | 0 | ||
| JUN, PRKCA | 0.032 | 3.39E-02 | 3.19E-01 | 3 | ||
| Molecular mechanisms of cancer | APH1B, FOS, HDAC4, JAK1, JUN, PRKCZ, PTCH1 | 0.016 | 1.74E-03 | 3.39E-02* | Embryonic development; organismal development; connective tissue development and function | 2 |
| ARHGEF4, GNB1, GNG4, JUN, NOTCH1, PA2G4, PRKDC | 0.016 | 4.17E-03 | 5.75E-02 | 0 | ||
| APH1B, AXIN1, FZD6, GNG4, HDAC4, ITGB4, PRKCA | 0.016 | 2.24E-03 | 1.80E-01 | 1 | ||
| iNOS signaling | FOS, JAK1, JUN | 0.064 | 8.13E-04 | 3.39E-02* | Cell death and survival; organismal injury and abnormalities; respiratory disease | 2 |
| JAK1, JUN | 0.043 | 1.95E-02 | 2.68E-01 | 3 | ||
| nNOS signaling in neurons | CAPN8, GRIN2A, PRKCZ | 0.064 | 8.13E-04 | 3.39E-02* | nNOS signaling in neurons | 2 |
| UVB-induced MAPK signaling | FOS, JUN, PRKCZ | 0.058 | 1.10E-03 | 3.39E-02* | Cellular development; cellular growth and proliferation; cell cycle | 2 |
| JUN, PRKCA | 0.039 | 2.34E-02 | 2.73E-01 | 3 | ||
| UVC-induced MAPK signaling | FOS, JUN, PRKCZ | 0.059 | 1.02E-03 | 3.39E-02* | Cell death and survival; connective tissue development and function; cancer | 2 |
| JUN, SMPD3 | 0.039 | 2.24E-02 | 1.41E-01 | 0 | ||
| JUN, PRKCA | 0.039 | 2.29E-02 | 2.73E-01 | 3 | ||
| EGF signaling | FOS, JAK1, JUN | 0.055 | 1.29E-03 | 3.39E-02* | Cell cycle; DNA replication, recombination, and repair; cancer | 2 |
| JAK1, JUN, PRKCA | 0.055 | 2.04E-03 | 1.98E-01 | 3 | ||
| IL-2 signaling | FOS, JAK1, JUN | 0.049 | 1.74E-03 | 3.39E-02* | Cellular development; cellular growth and proliferation; embryonic development | 2 |
| JAK1, JUN | 0.033 | 3.16E-02 | 3.19E-01 | 3 | ||
| IGF-1 signaling | FOS, JAK1, JUN, PRKCZ | 0.039 | 7.24E-04 | 3.39E-02* | Cancer; organismal injury and abnormalities; cell death and survival | 2 |
| JAK1, JUN, SOCS2 | 0.029 | 1.20E-02 | 2.10E-01 | 3 | ||
| Th2 pathway | APH1B, HLA-DOB, JAK1, JUN | 0.029 | 2.04E-03 | 3.39E-02* | Hematologic system development and function; tissue morphology; lymphoid tissue structure and development | 2 |
| IL13, JUN, NFATC2, NOTCH1 | 0.029 | 3.55E-03 | 5.50E-02 | 0 | ||
| JAK1, JUN, NOTCH1 | 0.022 | 2.51E-02 | 2.73E-01 | 3 | ||
| Neuroinflammation signaling pathway | APH1B, FOS, GRIN2A, HLA-DOB, JAK1, JUN | 0.019 | 1.45E-03 | 3.39E-02* | Cell-to-cell signaling and interaction; hematologic system development and function; immune cell trafficking | 2 |
| Apelin endothelial signaling pathway | FOS, HDAC4, JUN, PRKCZ | 0.029 | 2.14E-03 | 3.39E-02* | Cardiovascular system development and function; organismal development; embryonic development | 2 |
| GNB1, GNG4, JUN | 0.022 | 2.51E-02 | 1.53E-01 | 0 | ||
| GNG7, JUN, PRKCA | 0.022 | 2.57E-02 | 2.73E-01 | 3 | ||
| Systemic lupus erythematosus in B-cell signaling pathway | CSK, FOS, ISG15, JAK1, JUN, PRKCZ | 0.022 | 7.41E-04 | 3.39E-02* | Cellular development; cellular growth and proliferation; embryonic development | 2 |
| CSK, IL13, JUN, NFATC1, NFATC2, PLCG2 | 0.022 | 1.66E-03 | 4.90E-02* | 0 | ||
| CSK, JAK1, JUN, PRKCA | 0.014 | 3.89E-02 | 3.19E-01 | 3 | ||
| MSP-RON signaling in macrophages pathway | FOS, HLA-DOB, JUN, PRKCZ | 0.034 | 1.12E-03 | 3.39E-02* | Cellular movement; hematologic system development and function; immune cell trafficking | 2 |
| B-cell activating factor signaling | JUN, NFATC1, NFATC2 | 0.070 | 9.77E-04 | 3.72E-02* | Cellular development; cellular growth and proliferation; hematologic system development and function | 0 |
| FOS, JUN | 0.047 | 1.23E-02 | 7.94E-02 | 2 | ||
| G protein signaling mediated by tubby | GNB1, GNG4, PLCG2 | 0.068 | 1.05E-03 | 3.72E-02* | Cancer; hematologic disease; immunologic disease | 0 |
| APRIL-mediated signaling | JUN, NFATC1, NFATC2 | 0.071 | 9.12E-04 | 3.72E-02* | Cellular development; cellular growth and proliferation; hematologic system development and function | 0 |
| FOS, JUN | 0.048 | 1.15E-02 | 7.94E-02 | 2 | ||
| Sperm motility | CSK, GNB1, GNG4, NTRK3, PLCG2, PRKG1 | 0.024 | 1.07E-03 | 3.72E-02* | Cellular function and maintenance; molecular transport; amino acid metabolism | 0 |
| CSK, GNG7, JAK1, PRKCA, TEC | 0.020 | 6.31E-03 | 1.98E–01 | 3 | ||
| Estrogen receptor signaling | CACNA1C, CTBP2, GNB1, GNG4, JUN, NOTCH1, PLCG2, PRKDC | 0.020 | 5.13E-04 | 3.72E-02* | Gene expression; embryonic development; organismal development | 0 |
| FOS, JAK1, JUN, NDUFB10, NDUFV1, PRKCZ | 0.015 | 4.90E-03 | 5.01E-02 | 2 | ||
| GNG7, JAK1, JUN, NOTCH1, PRKCA | 0.012 | 3.80E-02 | 3.19E-01 | 3 | ||
| Th1 and Th2 activation pathway | IL13, JUN, NFATC1, NFATC2, NOTCH1 | 0.029 | 1.12E-03 | 3.72E-02* | Cellular development; cellular growth and proliferation; embryonic development | 0 |
| APH1B, HLA-DOB, JAK1, JUN | 0.023 | 4.57E-03 | 5.01E-02 | 2 | ||
| JAK1, JUN, NOTCH1 | 0.017 | 4.47E-02 | 3.20E-01 | 3 | ||
| IL-10 signaling | FOS, JAK1, JUN | 0.042 | 2.82E-03 | 3.98E-02* | Cell signaling; cell-to-cell signaling and interaction; organismal development | 2 |
| JAK1, JUN | 0.028 | 4.27E-02 | 3.19E-01 | 3 | ||
| Corticotropin releasing hormone signaling | FOS, JUN, PRKCZ, PTCH1 | 0.027 | 2.75E-03 | 3.98E-02* | DNA replication, recombination, and repair; cellular development; cellular growth and proliferation | 2 |
| CACNA1C, JUN, PLCG2 | 0.020 | 3.02E-02 | 1.69E-01 | 0 | ||
| Hypoxia signaling in the cardiovascular system | JUN, P4HB, UBE2F | 0.041 | 3.02E-03 | 4.17E-02* | Cell cycle; carbohydrate metabolism; cell death and survival | 2 |
| JUN, UBE2F, UBE2J2 | 0.041 | 4.79E-03 | 1.98E-01 | 3 | ||
| JUN, UBE2O | 0.027 | 4.47E-02 | 2.12E-01 | 0 |
Statistically significant overrepresented canonical pathways were determined by Fisher exact test followed by adjustment for multiple testing using the B-H method. Pathways with BH-adjusted P value <0.05 for any visit are marked with *. Associated molecules and P values for other visits are also presented if nominal P value <0.05. All pathways presented have 2 or more molecules. APRIL, a proliferation-inducing ligand; B-H, Benjamini–Hochberg (adjustment for multiple testing); EGF, epidermal growth factor; IGF-1, insulin-like growth factor 1; iNOS, Inducible nitric oxide synthase; MAPK, mitogen-activated protein kinase; MSP, macrophage stimulating protein; RON, recepteur d’origine nantais; nNOS, neuronal nitric oxide synthases; PKR, protein kinase receptor; RAR, retinoic acid receptor.
Overlap is the number of molecules in our list over the number of molecules in the pathway of interest.
Visits 0–4 are targeted for 8–13, 16–22, 24–29, 30–33, and 34–37 wk of gestation, respectively. The sample size for visits 0–4 is 298, 291, 292, 286, and 281, respectively (Figure 1).
Discussion
In a cohort of women from diverse race/ethnicity groups with longitudinal PA measures from periconception through pregnancy, recreational PA before and during pregnancy was significantly associated with several DNA methylation sites in placenta. Importantly, these associations differed by timing of PA. Furthermore, pathway analysis suggested the impact of PA on placental DNA methylation likely occurs through pathways related to cardiovascular system development and function, cellular development, and skeletal and muscular system development and function.
To the best of our knowledge, we are unaware of studies on PA and placental epigenetic modifications. Biological links have been suggested by previous studies examining epigenetics using blood samples [i.e., cord blood (14) or offspring blood spots (15) at delivery] based on candidate gene approaches. For example, total maternal PA in early pregnancy has been associated with decreased methylation at a differentially methylated region of PLAGL1, which is a key regulator of placental and fetal growth (14, 33). No studies have examined methylation in placental tissues using an epigenome-wide association study approach, nor have prior studies assessed PA beyond enrollment (in the first half of pregnancy) (14, 15).
Although the precise underlying mechanisms of the association between PA and placental DNA methylation remain elusive, several lines of evidence suggest the observed associations are biologically plausible. A recent study found that maternal PA could induce the secretion of a placenta-derived protein (i.e., superoxide dismutase 3) that, in turn, regulated glucose metabolic genes in mice fetal liver (8). Maternal PA has also been linked to placentation (34, 35), placental development and growth (36, 37), and placental blood flow to the fetus (38, 39). Given these prior findings, we anticipated maternal PA would influence genes associated with placental function. In our study, the methylation site with the largest fold change was cg14517108, located on a CpG island of TIMP2, which is previously known to encode a protein critical for vasodilation, placentation, and uterine expansion during normal pregnancy (40). However, based on experimental data available in the Human Protein Atlas (www.proteinatlas.org, version 21.0) (41), protein expression of TIMP2 was not detected in decidual or trophoblastic cells (Supplemental Table 4).
Other genes significantly associated with total recreational PA were related to cell growth and differentiation, as well as cell assembly and organization. MOB3B plays a pivotal role in organ size control and tumor suppression by restricting proliferation and promoting apoptosis (42). LPCAT1 exhibits acyltransferase activity and may synthesize phosphatidylcholine in pulmonary surfactant, thereby playing a role in respiratory physiology (43, 44). UBE2F interacts with the E3 ubiquitin ligase and is involved in protein neddylation (45). CSK plays an important role in the regulation of cell growth, differentiation, migration, and immune response (46, 47). A meta-analysis shows that CSK polymorphism is associated with higher systolic and diastolic blood pressure in Asian populations (48). C1orf25 may play a role in motor coordination and exploratory behavior and may be involved in postnatal neuronal functions (49). BRD2 may play a role in spermatogenesis or folliculogenesis and nucleosome assembly, as well as regulate the transcription of other genes (50). CLINT1 binds to membranes and may have a role in transporting vesicles from the trans-Golgi network to endosomes (51). EYA3 promotes efficient DNA repair and regulates transcription during organogenesis and may be involved in the development of the eye (52, 53). KPNA4 functions in nuclear protein import as an adapter protein for nuclear receptor and may be involved in cataract formation (54). RTN3 induces the formation of endoplasmic reticulum tubules (55–57). MAGI1 may play a role as scaffolding protein at cell–cell junctions with tumor suppressive and vascular functions (58).
The most significant pathway for the link between both total and moderate or vigorous recreational PA and placental epigenetic modifications was the cardiac hypertrophy signaling pathway at the periconception and early pregnancy periods. Cardiac remodeling is an important physiologic adaption induced by pregnancy, and the placenta can mediate adaptions required for pregnancy (59). Other highly significant pathways associated with total recreational PA include the B-cell receptor signaling pathway and the netrin signaling pathway, which are important for humoral immune response (60–62), as well as involved in nervous system signaling and the development of the skeletal and muscular system (63–65). Moderate or vigorous PA is also strongly associated with the netrin signaling pathway, as well as Gaq signaling, which plays a role in intracellular and second-messenger signaling and lipid metabolism (66–68).
We also observed the association of PA and DNA methylation varied by timing of PA. Although we cannot compare our findings with other studies due to limited longitudinal PA assessment in the literature, we suspect timing of PA is important because temporal modulation of different pathways may influence placental bed blood flow, arterial blood sugar, and oxygen content (37). In addition, a recent study from our group found that timing of moderate or vigorous PA was related to maternal glucose metabolism (69), which has been implicated in the modulation of placenta function. In the present study, our results suggest that timing of PA may be associated with activation of cellular pathways important for fetal programming in specific temporal windows. Importantly, the periconception to early period of PA identified the majority of pathways, despite few individual CpGs becoming FDR significant. These associations continue to emphasize the importance of having information prior to and early in pregnancy, as an important window when methylation patterns are being established.
Strengths of this study include having longitudinal assessment of PA capturing from periconception through pregnancy and implementing a standard protocol for obtaining placental tissues at delivery. We also considered a comprehensive list of covariates that were related to PA. Limitations included that PA was self-reported, which may not necessarily reflect physiologic changes such as changes in oxygen intake and heart rate. However, the PPAQ is a validated questionnaire specifically designed for pregnant women (19). We also have insufficient power to examine sex differences and limited gene expression data to corroborate methylation impact, as well as limited sample size to compare light compared with moderate and vigorous activities separately. In addition, pathways identified are not specific to the placenta, and our interpretation of results is limited by our understanding of how specific pathways influence placental development and function. Last, we may have uncontrolled confounding because of the lack of information on unobserved factors such as paternal lifestyle. We also did not adjust for smoking status because only 1 woman reported smoking in the past 6 mo.
In conclusion, recreational PA both before and during pregnancy was associated with placental DNA methylation with associations varying by activity timing. The impact of PA is potentially related to pathways with functions such as cellular development. These novel findings provide further evidence supporting the significance of pre- and periconception lifestyle on epigenetic changes in the placenta.
Supplementary Material
Acknowledgements
We thank the study participants of the NICHD Fetal Growth Studies; research teams at all participating clinical centers (Christina Care Health Systems; Columbia University; Fountain Valley Hospital, California; Long Beach Memorial Medical Center; New York Hospital, Queens; Northwestern University; University of Alabama at Birmingham; University of California, Irvine; Medical University of South Carolina; Saint Peters University Hospital; Tufts University; and Women and Infants Hospital of Rhode Island); and the Wadsworth Center, C-TASC, and The EMMES Corporations in providing data and imaging support. This work used the computational resources of the NIH HPC Biowulf cluster (http://hpc.nih.gov) and benefited from editing from the NIH Fellows Editorial Board.
The authors’ responsibilities were as follows—EHY and CZ: conceptualized, designed, and oversaw the study; KLG, FT-A, and CZ: contributed to data acquisition; SKZ, MO, SNH, and JW: analyzed the data; SKZ: drafted the manuscript; EHY, MO, SNH, KLG, SDM, JW, DRS, SC, FT-A, and CZ: contributed to data interpretation and critical revision of the manuscript for important intellectual content; and all authors: read and approved the final version of the manuscript. The authors report no conflicts of interest.
Notes
Supported by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, including American Recovery and Reinvestment Act funding via contract numbers HHSN275200800013C, HHSN275200800002I, HHSN27500006, HHSN275200800003IC, HHSN275200800014C, HHSN275200800012C, HHSN275200800028C, HHSN275201000009C, and HHSN27500008. Additional support was obtained from the NIH Office of the Director, the National Institute on Minority Health and Health Disparities, and the National Institute of Diabetes and Digestive and Kidney Diseases.
The placental genome-wide DNA methylation, gene expression, and genotype data are available through dbGaP with accession number phs001717.v1.p1. All other relevant data are available within the manuscript and its Supporting Information files.
Supplemental Tables 1–4 and Supplemental Figure 1 are available from the “Supplementary data” link in the online posting of the article and from the same link in the online table of contents at https://academic.oup.com/ajcn/.
Abbreviations used: FDR, false discovery rate; MET, metabolic equivalent of task; NICHD, National Institute of Child Health and Human Development; PA, physical activity; PC, principal component; PPAQ, Pregnancy Physical Activity Questionnaire; RNA-seq, RNA sequencing.
Contributor Information
Sifang Kathy Zhao, Epidemiology Branch, Division of Population Health Research, Division of Intramural Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA.
Edwina H Yeung, Epidemiology Branch, Division of Population Health Research, Division of Intramural Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA.
Marion Ouidir, Epidemiology Branch, Division of Population Health Research, Division of Intramural Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA.
Stefanie N Hinkle, Epidemiology Branch, Division of Population Health Research, Division of Intramural Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA; Department of Biostatistics, Epidemiology and Informatics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA.
Katherine L Grantz, Epidemiology Branch, Division of Population Health Research, Division of Intramural Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA.
Susanna D Mitro, Epidemiology Branch, Division of Population Health Research, Division of Intramural Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA.
Jing Wu, Glotech, Inc, Rockville, MD, USA.
Danielle R Stevens, Epidemiology Branch, Division of Population Health Research, Division of Intramural Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA.
Suvo Chatterjee, Epidemiology Branch, Division of Population Health Research, Division of Intramural Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA.
Fasil Tekola-Ayele, Epidemiology Branch, Division of Population Health Research, Division of Intramural Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA.
Cuilin Zhang, Epidemiology Branch, Division of Population Health Research, Division of Intramural Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA; Global Center for Asian Women's Health, Bia-Echo Asia Centre for Reproductive Longevity & Equality (ACRLE), Yong Loo Lin School of Medicine, National University of Singapore, Singapore; Department of Obstetrics and Gynaecology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore.
References
- 1. Wolfe LA, Weissgerber TL. Clinical physiology of exercise in pregnancy: a literature review. J Obstet Gynaecol Can. 2003;25(6):473–83. [DOI] [PubMed] [Google Scholar]
- 2. Gavard JA, Artal R. Effect of exercise on pregnancy outcome. Clin Obstet Gynecol. 2008;51(2):467–80. [DOI] [PubMed] [Google Scholar]
- 3. Physical Activity Guidelines Advisory Committee . Physical Activity Guidelines Advisory Committee scientific report. Washington (DC): US Department of Health and Human Services; 2018. [Google Scholar]
- 4. da Silva SG, Ricardo LI, Evenson KR, Hallal PC. Leisure-time physical activity in pregnancy and maternal-child health: a systematic review and meta-analysis of randomized controlled trials and cohort studies. Sports Med. 2017;47(2):295–317. [DOI] [PubMed] [Google Scholar]
- 5. Boulet SL, Alexander GR, Salihu HM, Pass M. Macrosomic births in the united states: determinants, outcomes, and proposed grades of risk. Am J Obstet Gynecol. 2003;188(5):1372–8. [DOI] [PubMed] [Google Scholar]
- 6. Boney CM, Verma A, Tucker R, Vohr BR. Metabolic syndrome in childhood: association with birth weight, maternal obesity, and gestational diabetes mellitus. Pediatrics. 2005;115(3):e290–6. [DOI] [PubMed] [Google Scholar]
- 7. Xu WH, Wu H, Xia WL, Lan H, Wang Y, Zhang Y, Hua S. Physical exercise before pregnancy helps the development of mouse embryos produced in vitro. Mitochondrion. 2017;34:36–42. [DOI] [PubMed] [Google Scholar]
- 8. Kusuyama J, Alves-Wagner AB, Conlin RH, Makarewicz NS, Albertson BG, Prince NB, Kobayashi S, Kozuka C, Møller M, Bjerre Met al. Placental superoxide dismutase 3 mediates benefits of maternal exercise on offspring health. Cell Metab. 2021;33(5):939–56..e8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Wilson ML. Fertility and pregnancy: an epidemiologic perspective by Allen J Wilcox. Am J Epidemiol. 2011;173(6):712–3. [Google Scholar]
- 10. Griffiths SK, Campbell JP. Placental structure, function and drug transfer. Contin Educ Anaesth Crit Care Pain. 2015;15(2):84–9. [Google Scholar]
- 11. Longtine MS, Nelson DM. Placental dysfunction and fetal programming: the importance of placental size, shape, histopathology, and molecular composition. Semin Reprod Med. 2011;29(3):187–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Campbell NA, Mitchel LG. Biology (5th edition). San Francisco,Benjamin-Cummings; 1999. [Google Scholar]
- 13. Laker RC, Lillard TS, Okutsu M, Zhang M, Hoehn KL, Connelly JJ, Yan Z. Exercise prevents maternal high-fat diet-induced hypermethylation of the Pgc-1α gene and age-dependent metabolic dysfunction in the offspring. Diabetes. 2014;63(5):1605–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. McCullough LE, Mendez MA, Miller EE, Murtha AP, Murphy SK, Hoyo C. Associations between prenatal physical activity, birth weight, and DNA methylation at genomically imprinted domains in a multiethnic newborn cohort. Epigenetics. 2015;10(7):597–606. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Marshall MR, Paneth N, Gerlach JA, Mudd LM, Biery L, Ferguson DP, Pivarnik JM. Differential methylation of insulin-like growth factor 2 in offspring of physically active pregnant women. J Dev Orig Health Dis. 2018;9(3):299–306. [DOI] [PubMed] [Google Scholar]
- 16. Hediger ML, Fuchs KM, Grantz KL, Grewal J, Kim S, Gore-Langton RE, Buck Louis GM, D'Alton ME, Albert PS. Ultrasound quality assurance for singletons in the National Institute of Child Health and Human Development fetal growth studies. J Ultrasound Med. 2016;35(8):1725–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Grewal J, Grantz KL, Zhang C, Sciscione A, Wing DA, Grobman WA, Newman RB, Wapner R, D'Alton ME, Skupski Det al. Cohort profile: NICHD fetal growth studies–singletons and twins. Int J Epidemiol. 2018;47(1):25–25l. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Buck Louis GM, Grewal J, Albert PS, Sciscione A, Wing DA, Grobman WA, Newman RB, Wapner R, D'Alton ME, Skupski Det al. Racial/ethnic standards for fetal growth: the NICHD fetal growth studies. Am J Obstet Gynecol. 2015;213(4):449.e1–e41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Chasan-Taber L, Schmidt MD, Roberts DE, Hosmer D, Markenson G, Freedson PS. Development and validation of a pregnancy physical activity questionnaire. Med Sci Sports Exerc. 2004;36(10):1750–60. [DOI] [PubMed] [Google Scholar]
- 20. Ainsworth BE, Haskell WL, Leon AS, Jacobs JD, Montoye HJ, Sallis JF, Paffenbarger JR. Compendium of physical activities: classification of energy costs of human physical activities. Med Sci Sports Exerc. 1993;25(1):71–80. [DOI] [PubMed] [Google Scholar]
- 21. Physical activity and exercise during pregnancy and the postpartum period: ACOG Committee Opinion, Number 804. Obstet Gynecol. 2020;135(4):e178–88. [DOI] [PubMed] [Google Scholar]
- 22. Delahaye F, Do C, Kong Y, Ashkar R, Salas M, Tycko B, Wapner R, Hughes F. Genetic variants influence on the placenta regulatory landscape. PLos Genet. 2018;14(11):e1007785–e. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Du P, Zhang X, Huang CC, Jafari N, Kibbe WA, Hou L, Lin SM. Comparison of Beta-value and M-value methods for quantifying methylation levels by microarray analysis. BMC Bioinf. 2010;11(1):587. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Tekola-Ayele F, Workalemahu T, Gorfu G, Shrestha D, Tycko B, Wapner R, Zhang C, Louis GMB. Sex differences in the associations of placental epigenetic aging with fetal growth. Aging. 2019;11(15):5412–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Shrestha D, Ouidir M, Workalemahu T, Zeng X, Tekola-Ayele F. Placental DNA methylation changes associated with maternal prepregnancy BMI and gestational weight gain. Int J Obes. 2020;44(6):1406–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Patro R, Duggal G, Love MI, Irizarry RA, Kingsford C. Salmon provides fast and bias-aware quantification of transcript expression. Nat Methods. 2017;14(4):417–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Zhang C, Hediger ML, Albert PS, Grewal J, Sciscione A, Grobman WA, Wing DA, Newman RB, Wapner R, D'Alton MEet al. Association of maternal obesity with longitudinal ultrasonographic measures of fetal growth: findings from the NICHD Fetal Growth Studies–Singletons. JAMA Pediatr. 2018;172(1):24–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, Smyth GK. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Akalin A, Kormaksson M, Li S, Garrett-Bakelman FE, Figueroa ME, Melnick A, Mason CE. methylKit: a comprehensive R package for the analysis of genome-wide DNA methylation profiles. Genome Biol. 2012;13(10):R87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Barfield RT, Almli LM, Kilaru V, Smith AK, Mercer KB, Duncan R, Klengel T, Mehta D, Binder EB, Epstein MPet al. Accounting for population stratification in DNA methylation studies. Genet Epidemiol. 2014;38(3):231–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Leek JT, Storey JD. Capturing heterogeneity in gene expression studies by surrogate variable analysis. PLos Genet. 2007;3(9):e161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Benjamini Y, Yekutieli D. The control of the false discovery rate in multiple testing under dependency. Ann Stat. 2001;29(4):1165–88. [Google Scholar]
- 33. Azzi S, Sas TC, Koudou Y, Le Bouc Y, Souberbielle JC, Dargent-Molina P, Netchine I, Charles MA. Degree of methylation of ZAC1 (PLAGL1) is associated with prenatal and post-natal growth in healthy infants of the EDEN mother child cohort. Epigenetics. 2014;9(3):338–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Bergmann A, Zygmunt M, Clapp JF III. Running throughout pregnancy: effect on placental villous vascular volume and cell proliferation. Placenta. 2004;25(8–9):694–8. [DOI] [PubMed] [Google Scholar]
- 35. Gilbert JS, Banek CT, Bauer AJ, Gingery A, Needham K. Exercise training attenuates placental ischemia-induced hypertension and angiogenic imbalance in the rat. Hypertension. 2012;60(6):1545–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Gilbert JS, Banek CT, Bauer AJ, Gingery A, Dreyer HC. Placental and vascular adaptations to exercise training before and during pregnancy in the rat. Am J Physiol Regul Integr Comp Physiol. 2012;303(5):R520–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Clapp JF. The effects of maternal exercise on fetal oxygenation and feto-placental growth. Eur J Obstet Gynecol Reprod Biol. 2003;110(Suppl 1):S80–5. [DOI] [PubMed] [Google Scholar]
- 38. Lashley CJ, Supik DA, Atkinson JT, Murphy RJ, O'Hagan KP. Effect of pregnancy on the uterine vasoconstrictor response to exercise in rats. Physiol Rep. 2015;3(3):e12337. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. O'Hagan KP, Alberts JA. Uterine artery blood flow and renal sympathetic nerve activity during exercise in rabbit pregnancy. Am J Physiol Regul Integr Comp Physiol. 2003;285(5):R1135–44. [DOI] [PubMed] [Google Scholar]
- 40. Chen J, Khalil RA. Matrix metalloproteinases in normal pregnancy and preeclampsia. Prog Mol Biol Transl Sci. 2017;148:87–165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Uhlén M, Fagerberg L, Hallström BM, Lindskog C, Oksvold P, Mardinoglu A, Sivertsson Å, Kampf C, Sjöstedt E, Asplund Aet al. Tissue-based map of the human proteome. Science. 2015;347(6220):1260419. [DOI] [PubMed] [Google Scholar]
- 42. Joung J, Engreitz JM, Konermann S, Abudayyeh OO, Verdine VK, Aguet F, Gootenberg JS, Sanjana NE, Wright JB, Fulco CPet al. Genome-scale activation screen identifies a lncRNA locus regulating a gene neighbourhood. Nature. 2017;548(7667):343–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Moessinger C, Kuerschner L, Spandl J, Shevchenko A, Thiele C. Human lysophosphatidylcholine acyltransferases 1 and 2 are located in lipid droplets where they catalyze the formation of phosphatidylcholine. J Biol Chem. 2011;286(24):21330–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Soupene E, Fyrst H, Kuypers FA. Mammalian acyl-CoA:lysophosphatidylcholine acyltransferase enzymes. Proc Natl Acad Sci U S A. 2008;105(1):88–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Huang DT, Ayrault O, Hunt HW, Taherbhoy AM, Duda DM, Scott DC, Borg LA, Neale G, Murray PJ, Roussel MFet al. E2-RING expansion of the NEDD8 cascade confers specificity to cullin modification. Mol Cell. 2009;33(4):483–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Bergman M, Mustelin T, Oetken C, Partanen J, Flint NA, Amrein KE, Autero M, Burn P, Alitalo K. The human p50csk tyrosine kinase phosphorylates p56lck at Tyr-505 and down regulates its catalytic activity. EMBO J. 1992;11(8):2919–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Sun G, Budde RJ. Expression, purification, and initial characterization of human Yes protein tyrosine kinase from a bacterial expression system. Arch Biochem Biophys. 1997;345(1):135–42. [DOI] [PubMed] [Google Scholar]
- 48. Qian X, Li Y, Liu X, Tu R, Tian Z, Zhang H, Zhang X, Zhou W, Jiang J, Wang Yet al. C-src tyrosine kinase gene rs1378942 polymorphism and hypertension in Asians: review and meta-analysis. Clin Chim Acta. 2018;487:202–9. [DOI] [PubMed] [Google Scholar]
- 49. Vauti F, Goller T, Beine R, Becker L, Klopstock T, Hölter SM, Wurst W, Fuchs H, Gailus-Durner V, de Angelis MHet al. The mouse Trm1-like gene is expressed in neural tissues and plays a role in motor coordination and exploratory behaviour. Gene. 2007;389(2):174–85. [DOI] [PubMed] [Google Scholar]
- 50. LeRoy G, Rickards B, Flint SJ. The double bromodomain proteins Brd2 and Brd3 couple histone acetylation to transcription. Mol Cell. 2008;30(1):51–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Kalthoff C, Groos S, Kohl R, Mahrhold S, Ungewickell EJ. Clint: a novel clathrin-binding ENTH-domain protein at the Golgi. Mol Biol Cell. 2002;13(11):4060–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52. Cook PJ, Ju BG, Telese F, Wang X, Glass CK, Rosenfeld MG. Tyrosine dephosphorylation of H2AX modulates apoptosis and survival decisions. Nature. 2009;458(7238):591–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53. Krishnan N, Jeong DG, Jung SK, Ryu SE, Xiao A, Allis CD, Kim SJ, Tonks NK. Dephosphorylation of the C-terminal tyrosyl residue of the DNA damage-related histone H2A.X is mediated by the protein phosphatase eyes absent. J Biol Chem. 2009;284(24):16066–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54. Ping X, Cheng Y, Bao J, Shi K, Zou J, Shentu X. KPNA4 is involved in cataract formation via the nuclear import of p53. Gene. 2021;786:145621. [DOI] [PubMed] [Google Scholar]
- 55. Siddiqi S, Zhelyabovska O, Siddiqi SA. Reticulon 3 regulates very low density lipoprotein secretion by controlling very low density lipoprotein transport vesicle biogenesis. Can J Physiol Pharmacol. 2018;96(7):668–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56. Grumati P, Morozzi G, Hölper S, Mari M, Harwardt MI, Yan R, Müller S, Reggiori F, Heilemann M, Dikic I. Full length RTN3 regulates turnover of tubular endoplasmic reticulum via selective autophagy. Elife. 2017;6:e25555. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57. Caldieri G, Barbieri E, Nappo G, Raimondi A, Bonora M, Conte A, Verhoef L, Confalonieri S, Malabarba MG, Bianchi Fet al. Reticulon 3-dependent ER-PM contact sites control EGFR nonclathrin endocytosis. Science. 2017;356(6338):617–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58. Wörthmüller J, Rüegg C. MAGI1, a scaffold protein with tumor suppressive and vascular functions. Cells. 2021;10(6):1494. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59. Yagel S, Cohen SM, Goldman-Wohl D. An integrated model of preeclampsia: a multifaceted syndrome of the maternal cardiovascular-placental-fetal array. Am J Obstet Gynecol. 2022;226(2S):S963–72. [DOI] [PubMed] [Google Scholar]
- 60. Valentine MA, Czernik AJ, Rachie N, Hidaka H, Fisher CL, Cambier JC, Bomsztyk K. Anti-immunoglobulin M activates nuclear calcium/calmodulin-dependent protein kinase II in human B lymphocytes. J Exp Med. 1995;182(6):1943–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 61. Katagiri T, Ogimoto M, Hasegawa K, Mizuno K, Yakura H. Selective regulation of Lyn tyrosine kinase by CD45 in immature B cells. J Biol Chem. 1995;270(47):27987–90. [DOI] [PubMed] [Google Scholar]
- 62. Kincaid RL, Takayama H, Billingsley ML, Sitkovsky MV. Differential expression of calmodulin-binding proteins in B, T lymphocytes and thymocytes. Nature. 1987;330(6144):176–8. [DOI] [PubMed] [Google Scholar]
- 63. Thuerauf DJ, Arnold ND, Zechner D, Hanford DS, DeMartin KM, McDonough PM, Prywes R, Glembotski CC. p38 Mitogen-activated protein kinase mediates the transcriptional induction of the atrial natriuretic factor gene through a serum response element: a potential role for the transcription factor ATF6. J Biol Chem. 1998;273(32):20636–43. [DOI] [PubMed] [Google Scholar]
- 64. Sugden PH. Signaling in myocardial hypertrophy: life after calcineurin?. Circ Res. 1999;84(6):633–46. [DOI] [PubMed] [Google Scholar]
- 65. Ross RS, Borg TK. Integrins and the myocardium. Circ Res. 2001;88(11):1112–9. [DOI] [PubMed] [Google Scholar]
- 66. Avila MA, Otero G, Cansado J, Dritschilo A, Velasco JA, Notario V. Activation of phospholipase D participates in signal transduction pathways responsive to gamma-radiation. Cancer Res. 1993;53(19):4474–6. [PubMed] [Google Scholar]
- 67. Van Aelst L, D'Souza-Schorey C. Rho GTPases and signaling networks. Genes Dev. 1997;11(18):2295–322. [DOI] [PubMed] [Google Scholar]
- 68. Rizzo MA, Shome K, Vasudevan C, Stolz DB, Sung TC, Frohman MA, Watkins SC, Romero G. Phospholipase D and its product, phosphatidic acid, mediate agonist-dependent raf-1 translocation to the plasma membrane and the activation of the mitogen-activated protein kinase pathway. J Biol Chem. 1999;274(2):1131–9. [DOI] [PubMed] [Google Scholar]
- 69. McDonald SM, May LE, Hinkle SN, Grantz KL, Zhang C. Maternal moderate-to-vigorous physical activity before and during pregnancy and maternal glucose tolerance: does timing matter?. Med Sci Sports Exerc. 2021;53(12):2520–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.