Translational toxicoepigenetic Meta-Analyses identify homologous gene DNA methylation reprogramming following developmental phthalate and lead exposure in mouse and human offspring

Abstract

Although toxicology uses animal models to represent real-world human health scenarios, a critical translational gap between laboratory-based studies and epidemiology remains. In this study, we aimed to understand the toxicoepigenetic effects on DNA methylation after developmental exposure to two common toxicants, the phthalate di(2-ethylhexyl) phthalate (DEHP) and the metal lead (Pb), using a translational paradigm that selected candidate genes from a mouse study and assessed them in four human birth cohorts. Data from mouse offspring developmentally exposed to DEHP, Pb, or control were used to identify genes with sex-specific sites with differential DNA methylation at postnatal day 21. Associations of human infant DNA methylation in homologous mouse genes with prenatal DEHP or Pb were examined with a meta-analysis. Differential methylation was observed on 6 cytosines (adjusted-p < 0.05) and 90 regions (adjusted-p < 0.001). This translational approach offers a unique method that can detect conserved epigenetic differences that are developmentally susceptible to environmental toxicants.

Keywords: DEHP; DNA methylation; Developmental exposures; Epigenetics; Lead (Pb); Translational toxicology.

Fig. 1.

DNA Methylation Reprogramming after Prenatal DEHP Exposure in Three Human Cohorts. A) Proportions of CpG sites from the EPIC array and included in the meta-analyses for females (lighter color) and males (darker color) in various genes annotations (31,471 and 37,302 total sites included in the female and male analyses, respectively). B) Proportions of CpG sites from the EPIC array and included in the meta-analyses for females (lighter color) and males (darker color) in relation to CpG islands. C) coMET plot (Martin et al., 2015) of the ZYMND8 DMR from female meta-analysis. Each CpG in the DMR is shown by p-value and genomic location. Gene region details are displayed below the plot. A methylation correlation matrix of raw CpG data is shown at the bottom of the panel. D) coMET plot of the SLC12A7 DMR from male meta-analysis, presented as described for panel C. Abbreviations: DMCs – differentially methylated cytosine; DMR – differentially methylated region; TSS200 – within 200 bp of the transcriptional start site; TSS1500 within 1500 bp of the transcriptional start site; UTR – untranslated region.

Fig. 2.

DNA Methylation Reprogramming after Prenatal Lead (Pb) Exposure in Two Human Cohorts. A) Proportions of CpG sites from the EPIC array and included in the meta-analyses for females (lighter color) and males (darker color) in various gene annotations (35,928 and 39,390 CpG sites included in the female and male analyses, respectively). B) Proportions of CpG sites from the EPIC array and included in the meta-analyses for females (lighter color) and males (darker color) in relation to CpG islands. C) Differentially methylated cytosines (DMCs) from female meta-analysis. All CpGs that had a p_adj < 0.05 are shown in colored triangles. Labeled triangles denote mapped to CpGs with a p_adj < 0.05 and with an I² < 50 %. Open circles denote CpGs with a p_adj > 0.05. D) coMET plot of the RGMA DMR from female meta-analysis. Each CpG in the DMR is shown by p-value and genomic location. Gene region details are displayed below the plot. A methylation correlation matrix of raw CpG data is shown at the bottom of the panel. E) DMCs from male meta-analysis, presented as described for panel C. F) coMET plot of the TNXB DMR from male meta-analysis, presented as described for panel D. Abbreviations: DMCs – differentially methylated cytosine; DMR – differentially methylated region; TSS200 – within 200 bp of the transcriptional start site; TSS1500 within 1500 bp of the transcriptional start site; UTR – untranslated region.