Re-analysis of DNA methylation data from tuberculosis-infected individuals with a focus on target genes

Motivation and Aim: Along with the structural variability of the genome, epigenetic modifications can be significant for susceptibility to various, including infectious diseases. The aim of this study was to re-analyze the differential methylation of genes (DMG), for which associations with tuberculosis and its pathogenetically significant features were previously established, as well as for those of interest from the point of view of studying the phenomenon of reverse comorbidity of asthma and tuberculosis. Methods and Algorithms: Re-analysis of two DMG datasets presented in the public repository of functional genomic data Gene Expression Omnibus [1] was carried out. The GSE72338 dataset The study was performed on peripheral blood monocytes (PBMC) and neutrophils of patients with active tuberculosis (TB, n = 8) and clinically healthy individuals, with household contact with active TB patients (LTBI, n = 8)). The patients lived in an area with a high incidence TB in Cape Town, South Africa The second dataset, GSE118469 [3], was obtained from a study of TB in Taiwanese residents. The study included patients with TB (n = 12) without concomitant infections (including HIV) and malignant neoplasms. The control group comprised of healthy individuals (n = 6). Methylation analysis was performed on an Infinium HumanMethylation450K BeadChip v1.2 chip [3]. In the GSE72338 and GSE118469 datasets we were primarily interested in genes that, according to our previously data were associated with infectious diseases, in particular thirteen genes IFNG, SOCS5, TNFB, TNFRSF1B, PIAS3, PIASY, CXCL10, ATM, NBN, MRE11, MLH1, PMS2, and TP53BP1. Results: Only the SOCS5 gene (out of 13 genes) included in the Top-50 DMG list, topped the DMG list when comparing neutrophils from TB and LTBI patients (p = 0.00000226), but adjusting for multiple comparisons makes these differences not statistically significant (adj.p = 0.595). It should be noted that, taking into account the FDR correction no significant differences were obtained in any of the comparisons. Further, we carried out the analysis, focusing on the initial level of significance achieved, without introducing any corrections. In monocytes, 13 sites in seven target genes (NBN, MRE11A, MLH1, IFNG, LTA, TNFRSF1B, CXCL10) were differentially methylated in patients with active TB compared with LTBI. In neutrophils, 7 sites in regions of genes ATM, MLH1, PMS2, SOCS5, TNFRSF1B, CXCL10 were differentially methylated in patients with active TB compared with LTBI. 23 sites in ATM, PMS2, TP53BP1, SOCS5, TNFRSF1B, CXCL10 genes were differentially methylated in PBMC when comparing between TB patients and healthy individuals.