Gaseous air pollutants and DNA methylation in a methylome-wide association study of an ethnically and environmentally diverse population of U.S. adults

Katelyn M Holliday*, Rahul Gondalia, Antoine Baldassari, Anne E Justice, James D Stewart, Duanping Liao, Jeff D Yanosky, Kristina M Jordahl, Parveen Bhatti, Themistocles L Assimes, James S Pankow, Weihua Guan, Myriam Fornage, Jan Bressler, Kari E North, Karen N Conneely, Yun Li, Lifang Hou, Pantel S Vokonas, Cavin K Ward-CavinessRory Wilson, Kathrin Wolf, Melanie Waldenberger, Josef Cyrys, Annette Peters, H Marike Boezen, Judith M Vonk, Sergi Sayols-Baixeras, Mikyeong Lee, Andrea A Baccarelli, Eric A Whitsel

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

5 Downloads (Pure)

Abstract

Epigenetic mechanisms may underlie air pollution-health outcome associations. We estimated gaseous air pollutant-DNA methylation (DNAm) associations using twelve subpopulations within Women's Health Initiative (WHI) and Atherosclerosis Risk in Communities (ARIC) cohorts (n = 8397; mean age 61.3 years; 83% female; 46% African-American, 46% European-American, 8% Hispanic/Latino). We used geocoded participant address-specific mean ambient carbon monoxide (CO), nitrogen oxides (NO2; NOx), ozone (O3), and sulfur dioxide (SO2) concentrations estimated over the 2-, 7-, 28-, and 365-day periods before collection of blood samples used to generate Illumina 450 k array leukocyte DNAm measurements. We estimated methylome-wide, subpopulation- and race/ethnicity-stratified pollutant-DNAm associations in multi-level, linear mixed-effects models adjusted for sociodemographic, behavioral, meteorological, and technical covariates. We combined stratum-specific estimates in inverse variance-weighted meta-analyses and characterized significant associations (false discovery rate; FDR<0.05) at Cytosine-phosphate-Guanine (CpG) sites without among-strata heterogeneity (PCochran's Q > 0.05). We attempted replication in the Cooperative Health Research in Region of Augsburg (KORA) study and Normative Aging Study (NAS). We observed a -0.3 (95% CI: -0.4, -0.2) unit decrease in percent DNAm per interquartile range (IQR, 7.3 ppb) increase in 28-day mean NO2 concentration at cg01885635 (chromosome 3; regulatory region 290 bp upstream from ZNF621; FDR = 0.03). At intragenic sites cg21849932 (chromosome 20; LIME1; intron 3) and cg05353869 (chromosome 11; KLHL35; exon 2), we observed a -0.3 (95% CI: -0.4, -0.2) unit decrease (FDR = 0.04) and a 1.2 (95% CI: 0.7, 1.7) unit increase (FDR = 0.04), respectively, in percent DNAm per IQR (17.6 ppb) increase in 7-day mean ozone concentration. Results were not fully replicated in KORA and NAS. We identified three CpG sites potentially susceptible to gaseous air pollution-induced DNAm changes near genes relevant for cardiovascular and lung disease. Further harmonized investigations with a range of gaseous pollutants and averaging durations are needed to determine the effect of gaseous air pollutants on DNA methylation and ultimately gene expression.

Original languageEnglish
Article number113360
Number of pages9
JournalEnvironmental Research
Volume212
Issue numberPt C
Early online date30-Apr-2022
DOIs
Publication statusPublished - Sep-2022

Cite this