Genome-wide DNA methylation profile of peripheral blood lymphocytes from subjects with nonsteroidal anti-inflammatory drug-induced respiratory diseases

Abstract

Background Significant changes in CpG methylation have been identified in nasal polyps, which are the main targets of nonsteroidal anti-inflammatory drug-exacerbated respiratory disease (NERD); however, these polyps are composed of various cellular components. In the present study, whole-genome CpG methylation in peripheral blood lymphocytes (PBLs) was analyzed to define the epigenetic changes in lymphocytes, which are the primary immune cells involved in NERD. Materials and methods Genomic DNA from peripheral blood mononuclear cells from 27 NERD and 24 aspirin-tolerant asthma (ATA) was subjected to bisulfate conversion and a methylation array. Quantitative CpG methylation, the beta-values as a quantitative measure of DNA methylation, in lymphocytes were calculated after adjustments for cellular composition. Results Fifty-six hypermethylated and three hypomethylated differentially methylated CpGs (DMCs) in PBLs in the NERD compared with ATA. The top 10 CpG loci predicted the methylation risk score, with a positive predictive value of 91.3%, a negative predictive value of 81.5% and an accuracy of 84.3%. As demonstrated in the nasal polyps, 30 DMCs were predicted to bind to the following 10 transcription factors, ranked in descending order: AP-2alphaA, TFII-1, STAT4, FOXP3, GR, c-Est-1, E2F-1, XBP1, ENKTF-1 and NF-1. Gene ontology analysis identified 13 categories such as regulation of T-helper 17 cell differentiation, including SMAD7 and NFKBIZ. PBLs in NERD contained no DMCs in genes associated with the prostaglandin and leukotriene pathways, which were found in ATA. Conclusion PBLs in NERD form a unique pattern of DNA CpG methylation, and the combined analysis may provide predictive values for NERD.