Genetic and non-genetic predictors of LINE-1 methylation in leukocyte DNA

Tajuddin, Salman Muhammad; Amaral, André F.S.; Fernández, Agustín F.; Rodriguez Rodero, Sandra; Rodríguez, Ramón María; Moore, Lee E.; Tardón, Adonina; Carrato, Alfredo; García Closas, Montserrat; Silverman, Debra T.; Jackson, Brian P.; García Closas, Reina; Cook, Ashley L.; Cantor, Kenneth P; Chanock, Stephen J.; Kogevinas, Manolis; Rothman, Nathaniel; Real, Francisco X.; Malats i Riera, Núria

doi:http://dx.doi.org/10.1289/ehp.1206068

Genetic and non-genetic predictors of LINE-1 methylation in leukocyte DNA

Citation

Tajuddin SM, Amaral AFS, Fernandez AF, Rodriguez-Rodero S, Rodriguez RM, Moore LE et al. Genetic and non-genetic predictors of LINE-1 methylation in leukocyte DNA. Environ Health Perspect. 2013;121(6):387-92. DOI: 10.1289/ehp.1206068

Abstract

Background: Altered DNA methylation has been associated with various diseases. Objective: We evaluated the association between levels of methylation in leukocyte DNA at long interspersed nuclear element 1 (LINE-1) and genetic and non-genetic characteristics of 892 control participants from the Spanish Bladder Cancer/EPICURO study. Methods: We determined LINE-1 methylation levels by pyrosequencing. Individual data included demographics, smoking status, nutrient intake, toenail concentrations of 12 trace elements, xenobiotic metabolism gene variants, and 515 polymorphisms among 24 genes in the one-carbon metabolism pathway. To assess the association between LINE-1 methylation levels (percentage of methylated cytosines) and potential determinants, we estimated beta coefficients (βs) by robust linear regression. Results: Women had lower levels of LINE-1 methylation than men (β = –0.7, p = 0.02). Persons who smoked blond tobacco showed lower methylation than nonsmokers (β = –0.7, p = 0.03). Arsenic toenail concentration was inversely associated with LINE-1 methylation (β = –3.6, p = 0.003). By contrast, iron (β = 0.002, p = 0.009) and nickel (β = 0.02, p = 0.004) were positively associated with LINE-1 methylation. Single nucleotide polymorphisms (SNPs) in DNMT3A (rs7581217-per allele, β = 0.3, p = 0.002), TCN2 (rs9606756-GG, β = 1.9, p = 0.008; rs4820887-AA, β = 4.0, p = 4.8 × 10–7; rs9621049-TT, β = 4.2, p = 4.7 × 10–9), AS3MT (rs7085104-GG, β = 0.7, p = 0.001), SLC19A1 (rs914238, TC vs. TT: β = 0.5 and CC vs. TT: β = –0.3, global p = 0.0007) and MTHFS (rs1380642, CT vs. CC: β = 0.3 and TT vs. CC; β = –0.8, global p = 0.05) were associated with LINE-1 methylation. Conclusions: We identified several characteristics, environmental factors, and common genetic variants that predicted DNA methylation among study participants.