Methylation of serotonin regulating genes in cord blood cells: association with maternal metabolic parameters and correlation with methylation in peripheral blood cells during childhood and adolescence

Background Serotonin (5-hydroxytryptamine, 5-HT) signaling is involved in neurodevelopment, mood regulation, energy metabolism, and other physiological processes. DNA methylation plays a significant role in modulating the expression of genes responsible for maintaining 5-HT balance, such as 5-HT transporter (SLC6A4), monoamine oxidase A (MAOA), and 5-HT receptor type 2A (HTR2A). Maternal metabolic health can influence long-term outcomes in offspring, with DNA methylation mediating these effects. We investigated associations between maternal metabolic parameters—pre-pregnancy body mass index (pBMI), gestational weight gain (GWG), and glucose tolerance status (GTS), i.e., gestational diabetes mellitus (GDM) versus normal glucose tolerance (NGT)—and cord blood methylation of SLC6A4, MAOA, and HTR2A in participants from our PlaNS birth cohort. CpG sites (15, 9, and 2 in each gene, respectively) were selected based on literature and in silico data. Methylation levels were quantified by bisulfite pyrosequencing. We also examined the stability of methylation patterns in these genes in circulating blood cells from birth to adolescence using longitudinal DNA methylation data from the ARIES database. Results None of the 203 PlaNS mothers included in this study had preexisting diabetes, 99 were diagnosed with GDM, and 104 had NGT; all neonates were born at full term by planned Cesarean section. Methylation at most CpG sites differed between male and female newborns. SLC6A4 methylation correlated inversely with maternal pBMI and GWG, while methylation at HTR2A site -1665 correlated positively with GWG. None of the maternal metabolic parameters statistically associated with MAOA methylation. DNA methylation data in cord blood and peripheral blood at ages 7 and 15 years were available for 808 participants from the ARIES database; 4 CpG sites (2 in SLC6A4 and 2 in HTR2A) overlapped between the PlaNS and ARIES cohorts. A positive correlation between methylation levels in cord blood and peripheral blood at 7 and 15 years of age was observed for both SLC6A4 and HTR2A CpG sites. Conclusions Methylation of 5-HT regulating genes in cord blood cells is influenced by neonatal sex, with maternal metabolism playing an additional role. Inter-individual variations present in circulating blood cells at birth are still pronounced in childhood and adolescence. Supplementary Information The online version contains supplementary material available at 10.1186/s13148-023-01610-w.

a Women who reported never having smoked or having quit smoking at least 6 months before pregnancy were categorized as non-smokers, while women who reported having smoked throughout pregnancy or having quit smoking during pregnancy were categorized as smokers; unclear cases (5 NGT, 6 GDM) were treated as missing data.p-values for the difference between the NGT and GDM groups were determined using b Mann-Whitney test, c Student's ttest, or d Fisher's exact test.Statistically significant differences are in bold.

Figure S2 .Figure S3 .Figure S4 .
Figure S2.Association of newborn sex and maternal pre-pregnancy body weight status with methylation of SLC6A4 and HTR2A CpG sites in cord blood monocytes, as determined in GSE212174 dataset [7].Methylation in GSE212174 dataset was quantified using Infinium MethylationEPIC BeadChip (Illumina).Shown are methylation data (normalized β values) for CpG sites that overlapped with those analyzed in the PlaNS cohort: CpG sites A. #10 (cg03363743) and B. #14 (cg22584138) in SLC6A4, and C. -1665 (cg0207079) in HTR2A; methylation data for other CpG sites analyzed in the PlaNS cohort were not available.Each dot represents a single participant (N=29), with empty and filled dots indicating lean and obese mothers, respectively.Horizontal lines indicate median and interquartile range.** p<0.01

Figure S5 .
Figure S5.Temporal mapping of methylation quantitative trait loci (mQTLs) for Illumina 450K CpG sites in A. SLC6A4 and B. HTR2A genes in the ARIES cohort.Annotations of CpG sites correspond to Illumina Infinium 450K identifiers (IDs).CpG sites studied in the PlaNS cohort are indicated by red rectangles.Only CpG sites associated with at least one mQTL are shown (5 of 16 for SLC6A4 and 21 of 27 for HTR2A).The length of the bars represents the number of mQTLs associated with each of the CpG sites.Each bar is subdivided by the time points at which mQTLs were detected, as indicated in the legend.Data are from the mQTL database [9].

Table S1C .
Association between maternal glucose tolerance status and other sample characteristics.
NGT, normal glucose tolerance; GDM, gestational diabetes mellitus; BMI, body mass index.Continuous data are reported as median [interquartile range].Categorical data are reported as number of subjects (N) and percentage (%).

Table S2 .
CpG sites targeted by methylation analyses in PlaNS cohort.

Table S3 .
Methylation levels (%) at individual CpG sites in PlaNS participants.
aData are reported as median [interquartile range].N, number of subjects.p-values for the difference between female and male newborns were determined using a Mann-Whitney test or b Student's t-test.Statistically significant differences are in bold.

Table S4 .
Significance levels (p-values) for bivariate associations of maternal and neonatal characteristics with methylation of serotonin regulating genes in cord blood cells of PlaNS participants.
Shown are p-values for the association of maternal pre-pregnancy body mass index (pBMI), gestational weight gain (GWG), glucose tolerance status (GTS; normoglycemia vs. gestational diabetes mellitus), parity (PAR; primiparous vs. multiparous), maternal age at childbirth (MAD), maternal smoking during pregnancy (TOB), neonatal sex (SEX), gestational age (GA), birth weight (BW) and ponderal index (PI) with methylation of SLC6A4 (average of CpG sites #1 to #15), MAOA (average of CpG sites #1 to 9), and HTR2A (sites -1665 and -1224) in cord blood cells.Number of participants was 203, except for TOB, where it was 192 due to missing values.p-values were determined by Student's t test, Mann-Whitney test, Pearson correlation, or Spearman correlation, as appropriate.Statistically significant associations are in bold.In analyses separated by neonatal sex, no significant associations were found between MAOA methylation and any of maternal or neonatal characteristics.

Table S5 .
SignificanceShown are p-values for correlation of lymphocyte, monocyte, neutrophil, eosinophil and basophil counts in cord blood with methylation of SLC6A4 (average of CpG sites #1 to #15), MAOA (average of CpG sites #1 to #9), and HTR2A (sites -1665 and -1224) in cord blood cells of subjects with available hematogram data (N=122).p-values were determined by Pearson or Spearman correlation, as appropriate.In analyses separated by neonatal sex, no significant associations were found between cell type counts and methylation levels.Hierarchical linear regression analysis of maternal metabolic parameters as predictors of methylation at individual CpG sites in SLC6A4 gene in cord blood cells of PlaNS participants.
levels (p-values) for correlations between cord blood cell type count and methylation of serotonin regulating genes in cord blood cells in a subset of PlaNS participants (N=122).

Table S6A .
Maternal metabolic parameters as predictors of methylation at CpG site #1 in SLC6A4 gene.
B, unstandardized beta coefficient; SE (B), standard error of B; p (B), significance for each of the predictors; R 2 , proportion of variance explained by predictors; p (R 2 ), significance of R 2; ΔR 2 , increase in R 2 resulting from the addition of predictor/s; p (ΔR 2 ), significance of ΔR 2 ; pBMI, pre-pregnancy body mass index; GWG, gestational weight gain; GTS, glucose tolerance status (normal glucose tolerance (NGT) vs. gestational diabetes mellitus, with NGT as a reference).Significant p values are in bold.

Table S6B .
Maternal metabolic parameters as predictors of methylation at CpG site #12 in SLC6A4 gene.
See footnotes to TableS6A.

Table S6C .
Maternal metabolic parameters as predictors of methylation at CpG site #13 in SLC6A4 gene.
See footnotes to TableS6A.

Table S6B .
Maternal metabolic parameters as predictors of methylation at CpG site #14 in SLC6A4 gene.

Table S6E .
Maternal metabolic parameters as predictors of methylation at CpG site #15 in SLC6A4 gene.

Table S8 .
[1]relation between SLC6A4 methylation and expression in cord blood cells.Data were obtained from publicly available datasets of participants in the ENVIRONAGE cohort (N=150)[1].Expression levels were expressed as normalized signal intensities of SLC6A4 relative to normalized signal intensities of ACTB.

Table S9 .
Primers used in DNA methylation analyses by bisulfite pyrosequencing in PlaNS cohort.