Author:

Keywords:

Science & Technology, Life Sciences & Biomedicine, Environmental Sciences, Public, Environmental & Occupational Health, Environmental Sciences & Ecology, Maternal green space, Epigenomics, DNA methylation, Epigenome-wide, Differentially methylated regions, Gene expression, CARCINOMA-CELLS, MODEL CHAIN, HEALTH, ASSOCIATION, SUPPRESSES, MORTALITY, Pregnancy, Female, Humans, Infant, Newborn, Maternal Exposure, Epigenome, DNA Methylation, Fetal Blood, Parks, Recreational, Prenatal Exposure Delayed Effects, Particulate Matter, Kruppel-Like Transcription Factors, Receptors, Progesterone, 03 Chemical Sciences, 05 Environmental Sciences, 06 Biological Sciences, Toxicology, 31 Biological sciences, 34 Chemical sciences, 41 Environmental sciences

Abstract:

BACKGROUND: DNA methylation programming is sensitive to prenatal life environmental influences, but the impact of maternal exposure to green space on newborns DNA methylation has not been studied yet. METHODS: We conducted a meta-epigenome-wide association study (EWAS) of maternal exposure to green space during gestation with cord blood DNA methylation in two subsets of the ENVIRONAGE cohort (N = 538). Cord blood DNA methylation was measured by Illumina HumanMethylation 450K in one subset (N = 189) and EPICarray in another (N = 349). High (vegetation height>3 m (m)), low (vegetation height<3 m) and total (including both) high-resolution green space exposures during pregnancy were estimated within 100 m and 1000 m distance around maternal residence. In each subset, we sought cytosine-phosphate-guanine (CpG) sites via linear mixed models adjusted on newborns' sex, ethnicity, gestational age, season at delivery, sampling day, maternal parity, age, smoking, education, and estimated blood cell proportions. EWASs results were meta-analysed via fixed-effects meta-analyses. Differentially methylated regions (DMRs) were identified via ENmix-combp and DMRcate algorithms. Sensitivity analyses were additionally adjusted on PM2.5, distance to major roads, urbanicity and neighborhood income. In the 450K subset, cord blood expression of differentially methylated genes was measured by Agilent microarrays and associated with green space. RESULTS: 147 DMRs were identified, 85 of which were still significant upon adjustment for PM2.5, distance to major roads, urbanicity and neighborhood income, including HLA-DRB5, RPTOR, KCNQ1DN, A1BG-AS1, HTR2A, ZNF274, COL11A1 and PRSS36 DMRs. One CpG reached genome-wide significance, while 54 CpGs were suggestive significant (p-values<1e-05). Among them, a CpG, hypermethylated with 100 m buffer total green space, was annotated to PAQR9, whose expression decreased with 1000 m buffer low green space (p-value = 1.45e-05). CONCLUSIONS: Our results demonstrate that maternal exposure to green space during pregnancy is associated with cord blood DNA methylation, mainly at loci organized in regions, in genes playing important roles in neurological development (e.g., HTR2A).