Novel insight into the epigenomic signature of regular smoking using serum cotinine

Owing to the adverse health effects of nicotine and tobacco, understanding the consequences and mechanisms involved in nicotine exposure are crucial. Exposure to tobacco smoke associates extensively with DNA methylation... [ view full abstract ]

Owing to the adverse health effects of nicotine and tobacco, understanding the consequences and mechanisms involved in nicotine exposure are crucial. Exposure to tobacco smoke associates extensively with DNA methylation alterations, providing a plausible link between smoking and adverse health effects. As self-reported smoking may be erroneous due to misreporting (recall bias or under reporting), we utilized cotinine, the primary metabolite of nicotine and a reliable biomarker of nicotine exposure, to identify DNA methylation alterations associated with smoking. We performed an epigenome-wide association study of serum cotinine levels among regular (current daily) smokers in the Finnish Twin Cohort (N=310) and identified DNA methylation at 50 CpG sites significantly (FDR P-value<0.05) associated with cotinine levels. Seventeen of these associations were novel and reside in smoking-related genes such as THSD4, LSM6, and CACNA2D4. As cotinine levels are influenced not only by nicotine intake but also by the rate of cotinine formation and removal, primarily mediated by CYP2A6, we performed secondary analyses accounting for such influence with a genetic risk score for CYP2A6, and identified five additional novel associations. We further investigated the role of genetic variants in the highlighted genes. We observed 124 cis and 3898 trans methylation quantitative trait loci (meQTLs), among which 19 meQTLs also directly associated with cotinine levels. Based on these findings we performed causal inference test at the 19 meQTLs and observed a trend (P-value <0.05) for mediation by DNA methylation at seven CpG sites. This suggests that methylation at these CpG sites may be on the causal pathway between underlying genotype and nicotine exposure (intake) and not directly altered due to nicotine exposure. In conclusion, using a reliable biomarker of nicotine exposure, we replicated and identified novel epigenetic associations in nicotine exposure pertinent genes. We also demonstrate evidence of an interplay between the epigenome and the genome, and propose that DNA methylation alteration at a small proportion of the highlighted CpG sites may be a molecular mediator for the effects of underlying genotype and not a consequence of smoking.