Abstract
Aim:
To quantify associations of anxiety and depression during pregnancy with differential cord blood DNA methylation of the glucorticoid receptor (NR3C1).
Materials & methods:
Pregnancy anxiety, trait anxiety and depressive symptoms were collected using the Pregnancy Related Anxiety Scale, State-Trait Anxiety Index and Edinburgh Postnatal Depression Scale, respectively. NR3C1 methylation was determined at four methylation sites.
Results:
DNA methylation of CpG1 in the NR3C1 CpG island shore was higher in infants born to women with high pregnancy anxiety (β 2.54, 95% CI: 0.49–4.58) and trait anxiety (β 1.68, 95% CI: 0.14–3.22). No significant association was found between depressive symptoms and NR3C1 methylation.
Conclusion:
We found that maternal anxiety was associated with increased NR3C1 CpG island shore methylation.
Keywords: : cord blood, developmental epigenetics, DNA methylation, epigenetics, glucocorticoid receptor, maternal anxiety, maternal depression, NR3C1, pregnancy
Maternal anxiety and depression are common during pregnancy, with symptoms of depression estimated to affect 19% of pregnant women and symptoms of anxiety estimated to affect 25% of pregnant women [1,2]. Exposure to maternal depression during pregnancy is a well-known risk factor for preterm birth and low birth weight and contributes to adverse health outcomes throughout childhood, including asthma and an altered long-term stress response [2–6]. While the association between exposure to depression during pregnancy and long-term adverse health outcomes for the child has been widely reported, less is known about in utero exposure to maternal anxiety and its potential effect on fetal programming.
Early life exposure to psychosocial stress has a long-term impact on the developing hypothalamic–pituitary–adrenal axis, which regulates the mammalian stress response [7,8]. Hypermethylation of the promoter region of the glucocorticoid receptor gene, NR3C1, has been associated with exposure to early life psychosocial stressors in animals and humans [6,9]. Cross-fostering studies in rodents have demonstrated a causal effect of early life stress, defined as less licking and grooming of offspring by dams, on DNA methylation of offspring [9]. In humans, infants born to women who experienced depressive symptoms or anxiety during pregnancy demonstrate altered methylation in the NR3C1 promoter, which may lead to long-term dysregulation of an individual’s physiological stress response [6,10–18]. These data suggest that DNA methylation of NR3C1 is sensitive to early life psychosocial stressors and that the pathway is conserved across both rodent and human models.
Trait anxiety, stress and depression have been the primary conditions of studies of maternal psychosocial exposure on offspring DNA methylation [8,19,20]. However, these measures of mental health may not capture the context-specific stressors associated with pregnancy, as they predominantly focus on an individual’s long-term psychosocial state [20,21]. Trait anxiety specifically measures a individual’s overall emotional state during the general course of life [22]. Isolating pregnancy related anxiety as its own variable captures the acuity of maternal stress because it includes concerns specific to the infant, delivery and parenthood within the context of the current pregnancy [20]. Prior studies suggest that pregnancy anxiety is predictive of decreased length of gestation and preterm birth [3,23,24]. The connection between maternal psychosocial state and infant health outcomes in this context suggests that there is a physiological mechanism that may alter the fetal environment and consequently influence fetal programming.
In the present study, we measured the associations of pregnancy anxiety, maternal trait anxiety and depressive symptoms during pregnancy with methylation levels of the NR3C1 gene, using DNA from leukocytes in umbilical cord blood. We hypothesized that infants of mothers with anxiety and depressive symptoms during pregnancy would have higher methylation in the NR3C1 promoter region and the CpG island shore region of NR3C1 compared with infants whose mothers experienced low levels of anxiety and depressive symptoms.
Materials & methods
Participants
We studied mother–infant dyads participating in the Spontaneous Prematurity and Epigenetics of the Cervix (SPEC) study, a prospective cohort study that enrolled pregnant women in Massachusetts, USA from 2013 to 2019. Participants were eligible if they were at least 18 years old and at less than 28 weeks of gestation [25]. While we enrolled 1199 women, the current study analyzes a subset of participants who agreed to analysis of cord blood DNA methylation. A comparison of included and excluded participants is displayed in Supplementary Table 1. Research staff obtained informed consent from all participants and all protocols were approved by the institutional review board at Beth Israel Deaconess Medical Center (Protocol 2013P-000125; MA, USA).
Of the participants enrolled, complete cord blood DNA methylation data and at least one completed psychosocial questionnaire were available for 163. For the four participants who had twin births, we used a cord blood sample from only one infant. The analytic dataset was n = 162 for pregnancy anxiety, n = 163 for trait anxiety and n = 163 for depressive symptoms.
Description of variables
Pregnancy anxiety, trait anxiety and depressive symptoms were the primary exposures in this analysis. Pregnancy anxiety was measured using the seven-item Pregnancy-related Anxiety Scale (PRAS) [26,27], in which participants responded to statements that pertained to the health of their fetus ('You have a lot of fear regarding the health of your baby'; '“'You are worried that the baby could be abnormal') or their own health as a result of this pregnancy ('You are afraid you will be harmed during delivery') [27]. Responses used a 4-point Likert scale (not at all, somewhat, moderately and very much). Responses were then dichotomized for each question, with ‘very much’ given a value of 1 and the other answer choices given a value of 0 and then summed to provide an anxiety score ranging from 0 to 7. Anxiety scores were categorized as low to moderate (scores 0–2) or high anxiety (score ≥3); this cutoff point was used previously [26]. This scale had an internal consistency coefficient ranging from 0.71 to 0.96 [28].
Trait anxiety was measured with the ten-item State Trait Anxiety Index (STAI) [29]. Participants responded to statements regarding how they perceived their overall emotional state ('You lack self-confidence' and 'You feel nervous and restless') on a 4-point Likert scale, with 1 as ‘almost never’, 2 as ‘sometimes’, 3 as ‘often’ and 4 as ‘almost always’. [22] The statements 'You feel satisfied with yourself', 'You feel secure', 'You are a steady person' were reverse-coded. Participants were defined as experiencing high trait anxiety if their summed score was ≥19, which was the median STAI score for all women in the SPEC cohort. The full version of the trait-anxiety scale in the STAI has a Cronbach’s α of 0.92 for females between the ages of 19–39 years [22].
Depressive symptoms were measured using the ten-item Edinburgh Postnatal Depression Scale (EPDS), where participants responded to statements assessing perceived current mental status (“You have been anxious or worried for no good reason”) [30] on a 4-point Likert scale, with 0 as ‘not at all’, 1 as ‘not very often’, 2 as ‘some of the time’ and 3 as ‘most of the time’. These responses were summed, resulting in a total depression score ranging from 0 to 30, with greater scores indicating higher levels of depression. Women were defined as experiencing high levels of depressive symptoms if their summed score on the EPDS was ≥13 [30]. This scale was validated in a cohort of pregnant women [30]. The EPDS demonstrates a standardized α-coefficient of 0.87 [30].
We performed single imputation for the four incomplete psychosocial questionnaires. The present responses were averaged. The averaged value was substituted for the missing values for all unanswered questions within each scale. Summed and dichotomized scores were calculated for all women who had answered at least 50% of each questionnaire. In addition to this imputation, we calculated what the questionnaire score would have been prior to imputation, to determine if the imputed score affected the final categorization for that individual for that score.
Other covariates considered in this analysis were maternal age, race/ethnicity, education, income level, insurance status, prepregnancy BMI, gestational age, method of delivery and antianxiety or antidepressant medications taken during pregnancy. Maternal age and gestational age were used as continuous variables. Due to small numbers, race was categorized as non-Hispanic white, non-Hispanic black and Asian/Hispanic/other. Maternal education was dichotomized as having less than a 4-year college degree or a 4-year college degree or more. Income was dichotomized as less than US$100,000 per year and US$100,000 per year or more. Prepregnancy BMI was dichotomized as 25 kg/m2 or ≥ 25 kg/m2 (overweight and obese). Antianxiety and antidepressant medication use during pregnancy were each treated as a yes/no dichotomous variable. Mode of delivery was categorized as vaginal or cesarean. We performed a secondary analysis using a composite exposure of high pregnancy anxiety or high self-reported depressive symptoms.
Targeted bisulfite sequencing for DNA methylation analysis
Umbilical cord blood samples were collected upon delivery. The buffy coat was separated via centrifugation from each of the samples and immediately stored at -80°C. Genomic DNA was isolated from the leukocytes using the DNeasy kit (Qiagen, Venlo, The Netherlands). DNA yield and purity were quantified on a Nanophotometer Pearl (Implen, Munich, Germany). DNA samples with optical density 260/280 values between 1.8 and 2.2 were submitted to EpigenDx (MA, USA) for NR3C1 methylation pyrosequencing analysis.
Bisulfite conversion was performed using the Zymo Research EZ Methylation kit (cat. no. D5002) [31]. Percent DNA methylation was quantified using PCR and pyrosequencing [32–36]; 2 μl of bisulfite-converted DNA was treated with 3.15 μl HotStar Taq Polymerase and PCR Master Mix with 0.6 μl each of the NR3C1 forward and reverse primers. The assays used for this study were ADS5690-FS (GRCh37/hg19, chr 5, 142780837) and ADS5691-FS (GRCh37/hg19, chr 5, 142780705 – 142770667). These assays have been used in previous studies that have analyzed the NR3C1 promoter region and the CpG island shores [6,37,38]. Pyrosequencing was conducted using the PSQ HS96 Pyrosequencing System (Qiagen Inc., MA, USA) where 6–15 μl of the PCR product was bound to sepharose beads, purified, washed, denatured, washed and annealed to 0.3 μl of the pyrosequencing primer [39]. Percent methylation was calculated by determining the fraction of methylated cytosine of the total cytosines at each CpG site (methylated or unmethylated) [39]. The location of these CpG sites are included in Figure 1. Technical replicates for quality control purposes were run on each sample in triplicate. We used the average value of the triplicates as the final percent methylation.
Figure 1. . NR3C1 promoter region.
The NR3C1 gene promoter region contains nine noncoding exons numbered 1A–1J (omitting G). Exons 1D–1H span across approximately 3 kb and comprise the proximal promoter region of the NR3C1 gene. Shaded regions represent the CpG island shore, noncoding regions 2 kb upstream and downstream of the CpG island. CpG1 is indicated as a red diamond.
Statistical analyses
We performed chi-squared bivariate analyses of the covariates with each exposure to describe our sample. After visual inspection revealed a normal distribution of DNA methylation at each of the 18 CpG sites, we calculated the mean and standard deviation of the % methylation and used analysis of variance to assess associations with levels of pregnancy anxiety, trait anxiety and depressive symptoms. We performed technical replicates of each site and only included CpG sites in which the coefficient of variation among replicates was less than 0.1. We conducted independent covariate-adjusted linear regression analyses with maternal pregnancy anxiety, depression and trait anxiety as the independent variables and DNA methylation at each CpG site as the dependent variables. The adjusted models included maternal age, race, education, prepregnancy BMI, antianxiety or antidepressant medications taken during pregnancy, mode of delivery and gestational age at delivery. All covariates aside from gestational age at delivery were entered into the model as categorical variables. A secondary analysis was conducted in which a composite exposure of either pregnancy anxiety or high depressive symptoms was the independent variable of interest.
Results
In the overall sample, 24 women (15%) reported high pregnancy anxiety, 51 (31%) reported high trait anxiety during pregnancy and 22 (13%) reported high levels of depressive symptoms during pregnancy; 38 women (23%) reported experiencing the composite exposure of either high pregnancy anxiety, high levels of depressive symptoms during pregnancy. Participants who reported experiencing depressive symptoms were younger, had lower educational attainment and had lower income levels than those who did not report depressive symptoms. Those who reported high pregnancy anxiety, high trait anxiety and depressive symptoms were less likely to have private insurance. Preterm birth was more common in participants who reported pregnancy anxiety and depression. Complete demographic information of the participants is included in Table 1.
Table 1. . Characteristics of the Spontaneous Prematurity and Epigenetics of the Cervix study participants (n = 163), by levels of anxiety and depressive symptoms.
| Characteristic | High pregnancy-related anxiety | High trait anxiety | High level of depressive symptoms | |||
|---|---|---|---|---|---|---|
| Yes (n = 24) n (%) |
No (n = 138) n (%) |
Yes (n = 51) n (%) |
No (n = 112) n (%) |
Yes (n = 22) n (%) |
No (n = 141) n (%) |
|
| Maternal age (years) | ||||||
| <30 | 5 (20.8) | 31 (22.5) | 13 (25.5) | 23 (20.5) | 9 (40.9) | 27 (19.1) |
| 30–<35 | 8 (33.3) | 51 (37.0) | 18 (35.3) | 42 (37.5) | 4 (18.2) | 56 (39.7) |
| ≥35 | 11 (45.8) | 56 (40.6) | 20 (39.2) | 47 (42.0) | 9 (40.9) | 58 (41.1) |
| Maternal race/ethnicity | ||||||
| Non-Hispanic White | 15 (62.5) | 86 (62.3) | 32 (62.7) | 70 (62.5) | 13 (59.1) | 89 (63.1) |
| Non-Hispanic Black | 4 (16.7) | 20 (14.5) | 9 (17.6) | 15 (13.4) | 6 (27.3) | 18 (12.8) |
| Asian, Hispanic, Other | 5 (20.8) | 32 (23.2) | 10 (19.6) | 27 (24.1) | 3 (13.6) | 34 (24.1) |
| Education | ||||||
| < College degree | 9 (37.5) | 41 (29.7) | 18 (35.3) | 32 (28.6) | 12 (54.5) | 38 (27.0) |
| College graduate | 15 (62.5) | 97 (70.3) | 33 (64.7) | 80 (71.4) | 10 (45.5) | 103 (73.0) |
| Income † | ||||||
| < US$100,000 | 12 (50.0) | 68 (49.3) | 30 (58.8) | 50 (44.6) | 16 (72.7) | 64 (45.4) |
| ≥ US$100,000 | 12 (50.0) | 67 (48.6) | 20 (39.2) | 60 (53.6) | 6 (27.3) | 74 (52.5) |
| Insurance | ||||||
| Private | 15 (62.5) | 111 (80.4) | 30 (58.8) | 97 (86.6) | 11 (50.0) | 116 (82.3) |
| Public, self-pay, other | 9 (37.5) | 27 (19.6) | 21 (41.2) | 15 (13.4) | 11 (50.0) | 25 (17.7) |
| Prepregnancy BMI (kg/m2) | ||||||
| <25 | 13 (54.2) | 71 (51.4) | 26 (51.0) | 59 (52.7) | 11 (50.0) | 74 (52.5) |
| ≥25 | 11 (45.8) | 67 (48.6) | 25 (49.0) | 53 (47.3) | 11 (50.0) | 67 (47.5) |
| Antianxiety medication | ||||||
| Yes | 1 (4.2) | 1 (0.7) | 1 (2.0) | 1 (0.9) | 1 (4.5) | 1 (0.7) |
| No | 23 (95.8) | 137 (99.3) | 50 (98.0) | 111 (99.1) | 21 (95.5) | 140 (99.3) |
| Antidepressant medication | ||||||
| Yes | 0 | 5 (3.6) | 3 (5.9) | 2 (1.8) | 2 (9.1) | 3 (2.1) |
| No | 24 (100.0) | 133 (96.4) | 48 (94.1) | 110 (98.2) | 20 (90.9) | 138 (97.9) |
| Mode of delivery | ||||||
| Cesarean | 6 (25.0) | 50 (36.2) | 17 (33.3) | 39 (34.8) | 5 (22.7) | 51 (36.2) |
| Vaginal | 18 (75.0) | 88 (63.8) | 34 (66.7) | 73 (65.2) | 17 (77.3) | 90 (63.8) |
| Gestational age | ||||||
| <37 weeks | 5 (20.8) | 13 (9.4) | 5 (9.8) | 13 (11.6) | 5 (22.7) | 13 (9.2) |
| ≥37 weeks | 19 (79.2) | 125 (90.6) | 46 (90.2) | 99 (88.4) | 17 (77.3) | 128 (90.8) |
| Birth weight | ||||||
| <2500 g | 2 (8.3) | 11 (7.8) | 2 (3.9) | 11 (9.8) | 2 (9.1) | 11 (7.8) |
| ≥2500 g | 22 (91.7) | 130 (92.2) | 49 (96.1) | 101 (90.2) | 20 (90.9) | 130 (92.2) |
Maternal income missing for three participants.
A total of 18 methylation sites in the NR3C1 gene were initially analyzed and 14 of the 18 methylation sites were located in the 1F exon in the promoter region of the NR3C1 gene; these were the same sites analyzed by Oberlander et al. [6] Each of these methylation sites exhibited a mean percent methylation level of < 2% and analysis of technical replicates demonstrated a coefficient of variation ≥0.1 (range: 0.1–1.7). Thus, these 14 CpG sites were not included as dependent variables for the statistical analyses. The four methylation sites located in the CpG island shore of the NR3C1 promoter region performed well. Two of the methylation sites (CpG3 and CpG4) were highly methylated (>90%) with little variation; Cpg3 ranged from 92.4–92.8% and CpG4 ranged from 94.0–94.9%. The remaining methylation sites (CpG1 and CpG2) showed variable methylation levels ranging from 64.0–76.7% (Table 2). Supplementary Table 2 shows demographics and DNA methylation of four CpGs of the CpG island shore of the glucocorticoid receptor NR3C1.
Table 2. . Mean percent DNA methylation at four sites within the CpG island shore of exon 1F in the promoter region of NR3C1 by exposure to anxiety and depressive symptoms (n = 164).
| Symptom | CpG 1 mean (SD) |
p-value† | CpG 2 mean (SD) |
p-value† | CpG 3 mean (SD) |
p-value† | CpG 4 mean (SD) |
p-value† |
|---|---|---|---|---|---|---|---|---|
| High pregnancy-related anxiety | ||||||||
| No | 64.2 (4.8) | 0.071 | 76.5 (5.6) | 0.919 | 92.7 (3.9) | 0.735 | 94.9 (3.0) | 0.191 |
| Yes | 66.1 (4.3) | 76.7 (4.3) | 92.4 (4.0) | 94.0 (2.8) | ||||
| High trait anxiety | ||||||||
| No | 64.0 (4.9) | 0.046 | 76.7 (5.2) | 0.773 | 92.8 (3.8) | 0.685 | 94.7 (3.0) | 0.666 |
| Yes | 65.6 (4.1) | 76.4 (6.0) | 92.5 (4.1) | 94.9 (3.0) | ||||
| High level of depressive symptoms | ||||||||
| No | 64.3 (4.9) | 0.185 | 76.6 (4.9) | 0.654 | 92.7 (3.9) | 0.839 | 94.8 (3.0) | 0.947 |
| Yes | 65.7 (3.7) | 76.1 (8.1) | 92.5 (4.1) | 94.7 (3.2) | ||||
t-test
ANOVA: Analysis of variance; CpG 1: chr5:143401272; CpG 2: chr5:143401140; CpG 3: chr5:143401129CpG 4: chr5:143401102; SD: Standard deviation.
Exposure to high pregnancy anxiety was associated with 2.07% higher methylation of CpG1 compared with infants born to women with low to moderate levels of pregnancy anxiety (95% CI: -0.02–4.17; p = 0.05). The association persisted after multivariable adjustment (β = 2.54%, 95% CI: 0.49–4.58; Table 3).
Table 3. . Unadjusted and adjusted percent methylation of CpG1 of NR3C1 (chr5:142780837) with high anxiety and depressive symptoms (n = 163†) .
| Symptom | Unadjusted β (95% CI) | Adjusted‡β (95% CI) |
|---|---|---|
| High pregnancy-related anxiety (vs low–moderate) | 2.07 (-0.02–4.17) | 2.54 (0.49–4.58) |
| High trait anxiety (vs low) | 1.60 (0.03–3.17) | 1.68 (0.14–3.22) |
| High depressive symptoms (vs low) | 1.45 (-0.70–3.59) | 1.34 (-0.85–3.53) |
| Composite high pregnancy-related anxiety or high depressive symptoms | 1.68 (-0.06–3.41) | 1.88 (0.16–3.61) |
Sample sizes for models: pregnancy-related anxiety (n = 162); trait anxiety (n = 163); depressive symptoms (n = 163).
Adjusted for maternal age, race/ethnicity, education, prepregnancy body mass index, antianxiety medication, antidepressants, mode of delivery and gestational age.
CpG 1: chr5:143401272.
Infants born to women who reported high trait anxiety exhibited significantly higher mean methylation levels of CpG1 in both the unadjusted (1.60%, 95% CI: 0.03–3.17) and adjusted (1.68%, 95% CI: 0.14–3.22) models.
Infants born to women who reported high levels of depressive symptoms, compared with low or moderate levels, exhibited non-significantly higher methylation of CpG1 in unadjusted (1.45%, 95% CI: -0.70, 3.59) and adjusted (1.34%, 95% CI: -0.85, 3.53) models.
There were no significant associations of depressive symptoms or anxiety with the remaining three CpG sites (Table 2). Multivariable adjustment did not reveal any significant associations with these sites (data not shown). With respect to the secondary analysis using a composite exposure of either high pregnancy anxiety or high depressive symptoms during pregnancy, demonstrated that infants born to women with this exposure had significantly higher DNA methylation in CpG 1 (β 1.88, 95% CI: 0.16–3.61).
Discussion
We found that a single CpG site in the CpG island shore of the promoter region of the NR3C1 gene was more highly methylated in infants born to women with high pregnancy-related and trait anxiety during the perinatal period. Notably, the increase in average methylation of 2.54% among infants exposed to pregnancy anxiety and of 1.68% for infants exposed to trait anxiety in utero are at a degree that has been shown to contribute to hypothalamic–pituitary–adrenal axis sensitization; these findings support those of prior epigenetic studies of this gene [40,41]. Though not statistically significant, we observed a positive association with methylation levels in infants and maternal reports of depressive symptoms during pregnancy. Our findings are consistent with prior studies demonstrating a relationship between psychosocial stress exposure and DNA methylation in the same methylation sites in the NR3C1 gene [37,38].
This is the first published study to assess the association between in utero maternal anxiety and depressive symptoms and methylation of sites in the CpG island shore promoter region of the NR3CI gene in infant cord blood, rather than solely in the CpG island itself. Epigenetic studies of early life stressors have traditionally focused on DNA methylation within the 1F promoter region of the NR3C1 gene, as this region is thought to be the most functionally important in glucocorticoid receptor gene expression based on prior studies involving rodent models [9]. Yet, methylation in the regions 2 kb around the CpG islands, deemed CpG island shores, have been shown to be equally as important in predicting downstream physiological function [42]. Prior research identifies functionally relevant methylation sites in the CpG island shores of the promoter region in the NR3C1 gene that are also vulnerable to environmental stressors [37,38,43]. Shields et al. identified that sites within the CpG island shore of the NR3C1 gene are more highly methylated in African–American women who had experienced abuse in childhood compared with women without history of abuse [43]. Another study by Bollati et al. showed that hypermethylation in this same region is related to years of exposure to stressful working conditions in a population of Italian men [37]. These two examples support our finding that this methylation site may be responsive to psychosocial exposures and may serve as a potential candidate for future study of perinatal environmental stressors.
To be consistent with the original study by Oberlander et al., we analyzed the methylation sites of the NR3C1 gene in cord blood DNA that were previously shown to be associated with with early life exposure to depression. However, we were unable to replicate those findings due to low methylation at those sites, which indicates that these sites may be uninformative due to lack of precision in technical replicates (Supplementary file 1). The original study did not report on the performance of technical replicates [6]. Prior literature also has reported that pyrosequencing values lower than 10% may be difficult to interpret; our findings agree with this assertion [34,44]. Variation in methylation values observed across studies of the same CpG sites of the NR3C1 gene suggest that more than one region may be sensitive to psychosocial exposures, which merits further study [41].
In addition to studying CpG island shores, which broadens the inquiry into the NR3C1 gene, our study has several strengths, including a larger sample size than prior studies, as well as detailed covariate data collected prospectively that allows for control of many potential confounding variables [6,18]. Yet, there are several limitations that should be noted. This sample is fairly homogenous, as the majority of participants included were highly educated and non-Hispanic white women. However, the psychosocial measures were relatively representative, as the percentages of women who reported anxiety and depressive symptoms in our study are similar to those in other studies of psychosocial state during pregnancy [2,45]. The genomic DNA was extracted from leukocytes with no information regarding the subtype of leukocytes present in the sample mixture. Therefore, we did not adjust for cell type in our analyses. However, prior research suggests that there is no substantial alteration in associations with gene-specific DNA methylation after adjustment for cell type [46,47]. Additionally, further research is needed to evaluate whether methylation in leukocytes parallels methylation in neural and endocrine tissue more directly involved in modulating the stress response. Finally, we did not have RNA or protein available for analysis to determine the functional significance of our findings. Nonetheless, our findings are biologically plausible given they are largely consistent with prior literature on the association of psychosocial stressors and higher DNA methylation of NR3C1.
Conclusion & future perspective
We found that perinatal exposure to maternal anxiety was associated with increased methylation in the NR3C1 gene in cord blood. This is the first study to identify an association with maternal anxiety and methylation of CpG sites in the island shore of the infant glucocorticoid receptor gene. Further studies to understand the impact of maternal psychologic state on fetal programming of later disease, including mental health disorders, are warranted.
Summary points.
Animal and human studies have linked maternal stress to differential fetal and offspring DNA methylation in the promoter region of the NR3C1 gene.
The connection between maternal psychosocial state and infant health outcomes in this context suggests that there is a physiological mechanism that may alter the fetal environment and consequently influence fetal programming.
Self-reported levels of pregnancy anxiety, trait anxiety and depressive symptoms were collected using the Pregnancy Related Anxiety Scale, State-Trait Anxiety Index and Edinburgh Postnatal Depression Scale, respectively.
NR3C1 methylation was analyzed at four methylation sites.
At CpG 1, we found higher methylation in infants of women who reported both high pregnancy anxiety (β: 2.54, 95% CI: 0.49–4.58) and high trait anxiety (β: 1.68, 95% CI: 0.14–3.22).
No significant association was found between depressive symptoms and CpG 1 methylation (β: 1.34, 95% CI: -0.85–3.53).
A composite exposure of either high pregnancy anxiety or depression symptoms was significantly associated with higher DNA methylation at CpG 1 (β: 1.88, 95% CI: 0.16–3.61).
We found that a single CpG site in the CpG island shore of the promoter region of the NR3C1 gene was more highly methylated in infants born to women with high pregnancy-related and trait anxiety during the perinatal period.
We were unable to replicate previous study findings at an additional 14 CpG sites due to low methylation at those sites, which indicates that these sites may be uninformative.
We found that perinatal exposure to maternal anxiety was associated with higher methylation in the NR3C1 gene in cord blood.
Acknowledgments
The authors would like to acknowledge the SPEC study staff and the participants for their contributions to this work.
Footnotes
Supplementary data
To view the supplementary data that accompany this paper please visit the journal website at: www.futuremedicine.com/doi/suppl/10.2217/epi-2020-0022
Financial & competing interests disclosure
The authors would like to acknowledge the Charles H Hood Foundation, National Institutes of Health/K23ES022242 and Harvard Catalyst, The Harvard Clinical and Translational Science Center (National Center for Advancing Translational Sciences, National Institutes of Health Award UL 1TR002541). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
No writing assistance was utilized in the production of this manuscript.
Ethical conduct of research
The authors state that they have obtained appropriate institutional review board approval or have followed the principles outlined in the Declaration of Helsinki for all human or animal experimental investigations. In addition, for investigations involving human subjects, informed consent has been obtained from the participants involved.
References
Papers of special note have been highlighted as: • of interest; •• of considerable interest
- 1.Field T. Prenatal anxiety effects: a review. Infant Behav. Dev. 49, 120–128 (2017). [DOI] [PubMed] [Google Scholar]
- 2.Ashley JM, Harper BD, Arms-Chavez CJ, Lobello SG. Estimated prevalence of antenatal depression in the US population. Arch. Womens Ment. Health 19(2), 395–400 (2016). [DOI] [PubMed] [Google Scholar]
- 3.Dole N, Savitz DA, Hertz-Picciotto I, Siega-Riz AM, Mcmahon MJ, Buekens P. Maternal stress and preterm birth. Am. J. Epidemiol. 157(1), 14–24 (2003). [DOI] [PubMed] [Google Scholar]
- 4.Ciesielski TH, Marsit CJ, Williams SM. Maternal psychiatric disease and epigenetic evidence suggest a common biology for poor fetal growth. BMC Pregnancy Childbirth 15, 192 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lee A, Mathilda Chiu Y-H, Rosa MJ et al. Prenatal and postnatal stress and asthma in children: temporal- and sex-specific associations. J. Allergy Clin. Immunol. 138(3), 740–747.e3 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Oberlander TF, Weinberg J, Papsdorf M, Grunau R, Misri S, Devlin AM. Prenatal exposure to maternal depression, neonatal methylation of human glucocorticoid receptor gene (NR3C1) and infant cortisol stress responses. Epigenetics 3(2), 97–106 (2008). [DOI] [PubMed] [Google Scholar]; •• First identifies associations between maternal depression during pregnancy and associated DNA methylation in neonatal cord blood, within the promotor region of the NR3C1 gene.
- 7.Weinstock M. The potential influence of maternal stress hormones on development and mental health of the offspring. Brain Behav. Immun. 19(4), 296–308 (2005). [DOI] [PubMed] [Google Scholar]
- 8.Seckl JR, Meaney MJ. Glucocorticoid programming. Ann. NY Acad. Sci. 1032, 63–84 (2004). [DOI] [PubMed] [Google Scholar]
- 9.Weaver IC, Cervoni N, Champagne FA et al. Epigenetic programming by maternal behavior. Nat. Neurosci. 7(8), 847–854 (2004). [DOI] [PubMed] [Google Scholar]
- 10.Conradt E, Lester BM, Appleton AA, Armstrong DA, Marsit CJ. The roles of DNA methylation of NR3C1 and 11β-HSD2 and exposure to maternal mood disorder in utero on newborn neurobehavior. Epigenetics 8(12), 1321–1329 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Monk C, Feng T, Lee S, Krupska I, Champagne FA, Tycko B. Distress during pregnancy: epigenetic regulation of placenta glucocorticoid-related genes and fetal neurobehavior. Am. J. Psychiatry 173(7), 705–713 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Capron LE, Ramchandani PG, Glover V. Maternal prenatal stress and placental gene expression of NR3C1 and HSD11B2: the effects of maternal ethnicity. Psychoneuroendocrinology 87, 166–172 (2018). [DOI] [PubMed] [Google Scholar]
- 13.Mansell T, Vuillermin P, Ponsonby A-L, Collier F, Saffery R, Ryan J. Maternal mental well-being during pregnancy and glucocorticoid receptor gene promoter methylation in the neonate. Dev. Psychopathol. 28(4pt2), 1421–1430 (2016). [DOI] [PubMed] [Google Scholar]
- 14.Braithwaite EC, Kundakovic M, Ramchandani PG, Murphy SE, Champagne FA. Maternal prenatal depressive symptoms predict infant NR3C1 1F and BDNF IV DNA methylation. Epigenetics 10(5), 408–417 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Perroud N, Rutembesa E, Paoloni-Giacobino A et al. The Tutsi genocide and transgenerational transmission of maternal stress: epigenetics and biology of the HPA axis. World J. Biol. Psychiatry 15(4), 334–345 (2014). [DOI] [PubMed] [Google Scholar]
- 16.Bleker LS, Milgrom J, Sexton-Oates A et al. Exploring the effect of antenatal depression treatment on children's epigenetic profiles: findings from a pilot randomized controlled trial. Clin. Epigenetics 11(1), 18–18 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Stonawski V, Frey S, Golub Y et al. Associations of prenatal depressive symptoms with DNA methylation of HPA axis-related genes and diurnal cortisol profiles in primary school-aged children. Dev. Psychopathol. 31(2), 419–431 (2019). [DOI] [PubMed] [Google Scholar]
- 18.Hompes T, Izzi B, Gellens E et al. Investigating the influence of maternal cortisol and emotional state during pregnancy on the DNA methylation status of the glucocorticoid receptor gene (NR3C1) promoter region in cord blood. J. Psychiatr. Res. 47(7), 880–891 (2013). [DOI] [PubMed] [Google Scholar]
- 19.Ryan J, Mansell T, Fransquet P, Saffery R. Does maternal mental well-being in pregnancy impact the early human epigenome? Epigenomics 9(3), 313–332 (2017). [DOI] [PubMed] [Google Scholar]
- 20.Dunkel Schetter C, Tanner L. Anxiety, depression and stress in pregnancy: implications for mothers, children, research and practice. Curr. Opin. Psychiatry 25(2), 141–148 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Van Den Bergh BRH, Van Den Heuvel MI, Lahti M et al. Prenatal developmental origins of behavior and mental health: the influence of maternal stress in pregnancy. Neurosci. Biobehav. Rev. 117, 26–64 (2017). [DOI] [PubMed] [Google Scholar]
- 22.Spielberger CD, Gorsuch RL, Lushene R, Vagg PR, Jacobs GA. Manual for the State – Trait Anxiety Inventory. Consulting Psychologists Press, CA, USA: (1983). [Google Scholar]
- 23.Kramer MS, Lydon J, Seguin L et al. Stress pathways to spontaneous preterm birth: the role of stressors, psychological distress and stress hormones. Am. J. Epidemiol. 169(11), 1319–1326 (2009). [DOI] [PubMed] [Google Scholar]
- 24.Orr ST, Reiter JP, Blazer DG, James SA. Maternal prenatal pregnancy-related anxiety and spontaneous preterm birth in Baltimore, Maryland. Psychosom. Med. 69(6), 566–570 (2007). [DOI] [PubMed] [Google Scholar]
- 25.Burris HH, Just AC, Haviland MJ et al. Long noncoding RNA expression in the cervix mid-pregnancy is associated with the length of gestation at delivery. Epigenetics 13(7), 742–750 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Cardenas A, Faleschini S, Cortes Hidalgo A et al. Prenatal maternal antidepressants, anxiety and depression and offspring DNA methylation: epigenome-wide associations at birth and persistence into early childhood. Clin. Epigenetics 11(1), 56 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Rini CK, Dunkel-Schetter C, Wadhwa PD, Sandman CA. Psychological adaptation and birth outcomes: the role of personal resources, stress and sociocultural context in pregnancy. Health Psychol. 18(4), 333–345 (1999). [DOI] [PubMed] [Google Scholar]
- 28.Wadhwa PD, Sandman CA, Porto M, Dunkel-Schetter C, Garite TJ. The association between prenatal stress and infant birth weight and gestational age at birth: a prospective investigation. Am. J. Obstet. Gynecol. 169(4), 858–865 (1993). [DOI] [PubMed] [Google Scholar]
- 29.Wu S, Gennings C, Wright RJ et al. Prenatal stress, methylation in inflammation-related genes and adiposity measures in early childhood: the Programming Research in Obesity, Growth Environment and Social Stress Cohort study. Psychosom. Med. 80(1), 34–41 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Cox JL, Holden JM, Sagovsky R. Detection of postnatal depression: development of the 10-item Edinburgh Postnatal Depression Scale. Br. J. Psychiatry 150(6), 782–786 (1987). [DOI] [PubMed] [Google Scholar]
- 31.Kile ML, Baccarelli A, Hoffman E et al. Prenatal arsenic exposure and dna methylation in maternal and umbilical cord blood leukocytes. Environ. Health Perspect. 120(7), 1061–1066 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Uhlmann K, Brinckmann A, Toliat MR, Ritter H, Nürnberg P. Evaluation of a potential epigenetic biomarker by quantitative methyl-single nucleotide polymorphism analysis. Electrophoresis 23(24), 4072–4079 (2002). [DOI] [PubMed] [Google Scholar]
- 33.Colella S, Shen L, Baggerly KA, Issa J-PJ, Krahe R. Sensitive and quantitative universal Pyrosequencing™ methylation analysis of CpG sites. BioTechniques 35(1), 146–150 (2003). [DOI] [PubMed] [Google Scholar]
- 34.Tost J, Dunker J, Gut IG. Analysis and quantification of multiple methylation variable positions in CpG islands by Pyrosequencing. BioTechniques 35(1), 152–156 (2003). [DOI] [PubMed] [Google Scholar]
- 35.Dupont J-M, Tost J, Jammes H, Gut IG. De novo quantitative bisulfite sequencing using the pyrosequencing technology. Anal. Biochem. 333(1), 119–127 (2004). [DOI] [PubMed] [Google Scholar]
- 36.Yang AS, EstéCio MRH, Doshi K, Kondo Y, Tajara EH, Issa JPJ. A simple method for estimating global DNA methylation using bisulfite PCR of repetitive DNA elements. Nucleic Acids Res. 32(3), e38–e38 (2004). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Bollati V, Baccarelli A, Sartori S et al. Epigenetic effects of shiftwork on blood DNA methylation. Chronobiol. Int. 27(5), 1093–1104 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Shields AE, Wise LA, Ruiz-Narvaez EA et al. Childhood abuse, promoter methylation of leukocyte NR3C1 and the potential modifying effect of emotional support. Epigenomics 8(11), 1507–1517 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]; •• First to look at CpG island shores rather than CpG islands in association with maternal psychosocial distress.
- 39.Bollati V, Baccarelli A, Hou L et al. Changes in DNA methylation patterns in subjects exposed to low-dose benzene. Cancer Res. 67(3), 876–880 (2007). [DOI] [PubMed] [Google Scholar]
- 40.Tyrka AR, Price LH, Marsit C, Walters OC, Carpenter LL. Childhood adversity and epigenetic modulation of the leukocyte glucocorticoid receptor: preliminary findings in healthy adults. PLoS ONE 7(1), e30148 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Palma-Gudiel H, Cordova-Palomera A, Leza JC, Fananas L. Glucocorticoid receptor gene (NR3C1) methylation processes as mediators of early adversity in stress-related disorders causality: a critical review. Neurosci. Biobehav. Rev. 55, 520–535 (2015). [DOI] [PubMed] [Google Scholar]
- 42.Irizarry RA, Ladd-Acosta C, Wen B et al. The human colon cancer methylome shows similar hypo- and hypermethylation at conserved tissue-specific CpG island shores. Nat. Genet. 41(2), 178–186 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Lepeule J, Bind MA, Baccarelli AA et al. Epigenetic influences on associations between air pollutants and lung function in elderly men: the normative aging study. Environ. Health Perspect. 122(6), 566–572 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Dejeux E, Audard V, Cavard C, Gut IG, Terris B, Tost J. Rapid identification of promoter hypermethylation in hepatocellular carcinoma by pyrosequencing of etiologically homogeneous sample pools. J. Mol. Diagn. 9(4), 510–520 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Bennett HA, Einarson A, Taddio A, Koren G, Einarson TR. Prevalence of depression during pregnancy: systematic review. Obstet. Gynecol. 103(4), 698–709 (2004). [DOI] [PubMed] [Google Scholar]; • A systematic review of studies looking at maternal psychosocial state during pregnancy.
- 46.Burris HH, Baccarelli AA, Motta V et al. Association between length of gestation and cervical DNA methylation of PTGER2 and LINE 1-HS. Epigenetics 9(8), 1083–1091 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Burris HH, Braun JM, Byun HM et al. Association between birth weight and DNA methylation of IGF2, glucocorticoid receptor and repetitive elements LINE-1 and Alu. Epigenomics 5(3), 271–281 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]