Prenatal exposure to triclosan assessed through repeated urine samples and placental DNA methylation in the SEPAGES cohort

Background: Previously we showed associations between elevated maternal urinary concentrations of triclosan in French pregnant women and increased placental DNA methylation in male fetuses, an effect potentially mediated by placental cell composition. Using a more recent cohort with improved exposure assessment, we aimed to replicate these findings and perform an exploratory study. Methods: We assessed triclosan in two pools of 21 urine samples each, collected during the second and third trimester among 395 pregnant women from the SEPAGES cohort. We used Infinium MethylationEPIC arrays to measure DNA methylation in placenta collected at delivery. We used adjusted linear regressions on 500 previously identified candidate CpGs and investigated differentially methylated regions (DMRs) within CpGs overlapping with our previous study. Additionally, we performed an exploratory epigenome-wide association study and DMR analysis on all CpGs available in the assay. Results: In the sex-stratified replication study, we found 75 triclosan-associated genes (three for males and 72 for females including imprinted FBRSL1, KCNQ1, RHOBTB3, and SMOC1). The effect estimates showed a similar pattern to our previous study for females but not males (correlation coefficient rho = 0.75 and 0.07, respectively), with most associations being positive but weaker. When both sexes were considered, 18 triclosan-associated genes were identified. The correlation with effect estimates of our previous study was 0.6. As previously, we showed that placental heterogeneity could mediate the associations between triclosan concentrations and DNA methylation. In the exploratory analysis, we identified only a few isolated sex-specific associations. Conclusions: We observed several positive associations between pregnancy triclosan and placental DNA methylation similarly to our previous study, however most of the associations were identified for females but not males. Our results suggest that pregnancy exposure to triclosan could affect placenta through DNA methylation and that female placentas may be particularly vulnerable. Keywords: placenta, methylation, pregnancy triclosan, pooled biospecimens