Investigation into the Role of Epigenetic Modifications of DNA Methylation in the Relationship between Mode of Exposure to Arsenic and Manifestation of Arsenical Skin Lesions in Thailand

Abstract

Background: Chronic exposure to arsenic results in a variety of types of abnormal skin pigmentation and keratinization and is associated with altered levels of DNA methylation. The relationships between type of lesion, level of DNA methylation and mode of exposure are still unclear. This study aimed to examine these relationships. Methods: The study was set mostly in Ronphibun sub-district of Nakhon Sri Thammarat Province, Southern Thailand, which has a long history of environmental arsenic contamination as a result of the release of arseno-pyrite from tin- mining activities. From June to October 2014, cases with arsenical skin lesions (n=131) and controls free of such lesions (n=163) were recruited from among residents of this sub-district, as well as a group of external controls (n=105) resident in a 20km-distant arsenic-free sub-district. The levels of global DNA methylation (LINE-1) and of p16 methylation in saliva specimens were estimated using pyrosequencing technology. History of exposure was classified as occupational if the subject had a history of working in tin-mining, and/or domestic if residence was in an area known to have high well-water arsenic and had used well-water for drinking. Recent exposure was represented by current toe-nail arsenic concentration. Associations between a) case status and exposure mode/ DNA methylation level, b) lesion type/severity and DNA methylation levels, and c) DNA methylation level and exposure mode were explored using binary logistic and proportional odds regression models. Bias in the estimated associations was minimized by conditioning on minimal sets of covariates identified by a previously constructed directed acyclic graph. Results: Overall, compared with subjects living in the non- contaminated sub-district, those living in the arsenic-contaminated area, both symptomatic and non-symptomatic, had lower mean levels of LINE-1 methylation (mean [95% CI]: 74.65% [74.10, 75.21] and 74.61% [74.11, 75.11] vs 76.05% [75.43, 76.67], P<0.001 for each) and higher concentrations of toenail arsenic (geometric mean [95% CI]: 1.32 [1.14, 1.53] and 1.18 [1.04, 1.53] vs 0.54 [0.46, 0.63], P<0.001 for each). Among subjects living in the contaminated area, those with arsenical skin lesions were more likely to have a history of domestic exposure to arsenic (ORtotal [95% CI]: 1.76 [1.00, 3.11] and 2.22 [1.22, 4.03], Ptrend-0.005 for exposure <36 and ≥36 years respectively), and this relationship persisted after adjustment for global DNA methylation level. Among the case group, greater severity of spotty hyperpigmentation was significantly related to higher levels of LINE-1 methylation (cumulative OR [95% CI]: 1.70 [0.82, 3.53] and 4.40 [1.47, 13.12], Ptrend=0.006) for middle and top tertiles of LINE-1), respectively). Global DNA methylation level was positively associated with toenail arsenic only among non-symptomatic exposed subjects (OR [95% CI]: 1.31 [1.06, 1.64], P=0.014). P16 methylation levels showed no relationship with group, skin lesion or exposure status. Conclusion: The results support the concept that exposure to an arsenic- contaminated environment results in global DNA hypomethylation overall. However, among symptomatic subjects, increasing LINE-1 levels were evident in subjects with increasing severity of the spotty hyperpigmentation type of arsenical skin lesion.