Epigenome-wide association study for the identification of tobacco smoke effects on gingival methylation as a risk factor of periodontitis
Final Report Abstract
Periodontitis (PD) is regarded as one of the most common inflammatory diseases worldwide and has an estimated heritability of 50%. Epigenetic factors contribute strongly to the susceptibility of complex diseases. In order to better understand the molecular genetic aetiology of PD, it is important to understand how noncoding methylation variation affect gene expression, and how methylation and gene expression differ in the inflamed and healthy gingiva of the same individuals. Epigenome-wide association studies (EWAS) combined with whole transcriptome RNA sequencing (RNA-Seq) enable the unbiased and systematic identification of biologically active differences in the methylation patterns of promoters and enhancers between individuals. Despite the prominent function of the masticatory oral mucosa in maintaining barrier integrity at the gateway to the gastrointestinal and respiratory tracts, no comprehensive data of the methylome of healthy and inflamed oral masticatory mucosa existed. Likewise, no well-powered EWAS on inflammatory disease phenotypes of the healthy and inflamed oral mucosa have been performed with tissue samples of PD patients. To address the lack of reference data of the methylome of the healthy gingiva and to establish the prerequisite for a well-powered EWAS, we created a large collection of ex-vivo tissue samples from the uninflamed and inflamed masticatory mucosa of PD patients (N=140) and of uninflamed PD-free controls (N=130). We further performed an EWAS to analyse the methylation pattern of the healthy masticatory mucosa of smokers (n = 18) and non-smokers (n = 24). 155 CpG sites showed a genome-wide significant association of changes in methylation values in a regression analysis. The most significant association located to the gene CYP1B1 (cytochrome P450 family 1 subfamily B member 1), with p = 3.2 x 10E-12. We generated well-powered transcriptome-wide gene expression data of the oral masticatory healthy mucosa of healthy non-smokers and smokers by performing mRNA- Sequencing (RNA-Seq with a coverage of 16 Mio 50 bp single-end reads per sample) in 41 samples that passed quality control (QC) for RNA-Seq., and validated the results from the EWAS on the effects of smoking on the gene CYP1B1, which showed a highly significant differential expression between smokers and non-smokers (p = 5.8 x 10E-8). In conclusion, this study gives statistical evidence that smoking has a biological functional effect on heritable characteristics of the DNA and identified CYP1B1 as having a strong, epigenetically altered activity in smokers compared to non-smokers. This gene has been related to various forms of cancer. In addition, this study set the requisites to perform a well-powered EWAS of PD, to construct of a methylome map of the healthy and inflamed masticatory mucosa, to construct an expression quantitative trait locus (eQTL) as well as a methylation quantitative trait locus (meQTL) map for healthy and inflamed masticatory mucosa, and to identify putative risk variants of PD in genes that are differentially expressed in the healthy and inflamed masticatory mucosa. These studies have strong potential to provide novel insights into the molecular etiology of oral inflammation and thus will contribute to the development of novel therapies and disease prevention.
