Selenium-associated DNA methylation modifications in placenta and neurobehavioral development of newborns: An epigenome-wide study of two U.S. birth cohorts

Abstract

Background/aim: Selenium (Se) levels in pregnancy have been linked to neurobehavioral development of the offspring. DNA methylation is a potential mechanism underlying the impacts of environmental exposures on fetal development; however, very few studies have been done elucidating the role of DNA methylation linking prenatal Se and child neurobehavior. We aimed to investigate the associations between placental Se concentration and epigenome-wide DNA methylation in two U.S. cohorts, and to assess the association between Se-related DNA methylation modifications and newborns' neurobehavior.

Methods: We measured placental Se concentrations in 343 newborns enrolled in the New Hampshire Birth Cohort Study and in 141 newborns in the Rhode Island Child Health Study. Genome-wide placental DNA methylation was measured by HumanMethylation450 BeadChip, and newborn neurobehavioral development was assessed by the NICU Network Neurobehavioral Scales (NNNS). We meta-analyzed the associations between placental Se concentration and DNA methylation in each cohort, adjusting for covariates. We also fit multiple linear regression and ordinal logistic regression for methylation and newborn NNNS summary scores.

Results: We identified five Se-related differentially methylated CpG sites. Among them was cg09674502 (GFI1), where selenium concentration was positively associated with methylation (β-coefficient = 1.11, FDR-adjusted p-value = 0.045), and where we observed that a one percent methylation level increase was associated with a 15% reduced odds of higher muscle tone in the arms, legs and trunk of newborns, (OR [95% Confidence Interval, CI] = 0.85 [0.77, 0.95]). We also observed for each interquartile range (IQR) increase in selenium concentration in the placenta, there was 1.76 times greater odds of higher hypotonicity (OR [95% CI] = 1.76 [1.12, 2.82]).

Conclusions: Placental selenium concentration was inversely associated with muscle tone of newborns, and hypermethylation of GFI1 could be a potential mechanism underlying this association.

Keywords: DNA methylation; Muscle tone; NNNS; Neurobehavior; Placenta; Selenium.

Figure 1.

Manhattan plot of meta-analysis. The red line represents the Bonferroni corrected threshold (p <= 1.55 × 10⁻⁷). The blue line represents the false discovery rate (FDR) corrected 5% threshold.

Figure 2.

The scatterplot shows the correlations between IQR-scaled selenium concentrations and the differentially methylated CpGs (FDR < 0.05) that are consistent between two cohorts, using the pooled data of two cohorts. The regression lines were fitted with robust linear regression models that were corrected for maternal age, education, pre-pregnant BMI, smoking status, gestational age, child’s sex, child’s race/ethnicity, estimated tissue heterogeneity (the first four components for NHBCS and all the four components for RICHS), and the cohort’s indicator. The correlation coefficients were estimated using Spearman partial correlation analysis, adjusted for the same covariates corrected in the robust linear regression models.

Figure 3.

The distribution of cg09674502 DNA methylation on each level of hypertonicity. The black line represents the median of DNA methylation on each level. The blue square represents the mean of DNA methylation on each level.

Figure 4.

The distribution of selenium concentration on each level of hypotonicity. The black line represents the median of DNA methylation on each level. The blue square represents the mean of DNA methylation on each level.