Differential gene expression patterns between smokers and non-smokers: cause or consequence?

The molecular mechanisms causing smoking-induced health decline are largely unknown. To elucidate the molecular pathways involved in cause and consequences of smoking behavior, we conducted a genome-wide gene expression study in peripheral blood samples targeting 18 238 genes. Data of 743 smokers, 1686 never smokers and 890 ex-smokers were available from two population-based cohorts from the Netherlands. In addition, data of 56 monozygotic twin pairs discordant for ever smoking were used. One hundred thirty-two genes were differentially expressed between current smokers and never smokers (P< 1.2x10(-6), Bonferroni correction). The most significant genes were G protein-coupled receptor 15 (P< 1x10(-15)0) and leucine-rich repeat neuronal 3 (P< 1x10(-44)). The smoking-related genes were enriched for immune system, blood coagulation, natural killer cell and cancer pathways. By taking the data of ex-smokers into account, expression of these 132 genes was classified into reversible (94 genes), slowly reversible (31 genes), irreversible (6 genes) or inconclusive (1 gene). Expression of 6 of the 132 genes (three reversible and three slowly reversible) was confirmed to be reactive to smoking as they were differentially expressed in monozygotic pairs discordant for smoking. Cis-expression quantitative trait loci for GPR56 and RARRES3 (downregulated in smokers) were associated with increased number of cigarettes smoked per day in a large genome-wide association meta-analysis, suggesting a causative effect of GPR56 and RARRES3 expression on smoking behavior. In conclusion, differential gene expression patterns in smokers are extensive and cluster in several underlying disease pathways. Gene expression differences seem mainly direct consequences of smoking, and largely reversible after smoking cessation. However, we also identified DNA variants that may influence smoking behavior via the mediating gene expression.