Integrated bioinformatics analysis and validation revealed potential immune-regulatory miR-892b, miR-199b-5p and miR-582-5p as diagnostic biomarkers in active tuberculosis.

Tuberculosis (TB) is one of the most prevalent pulmonary diseases caused by Mycobacterium tuberculosis (Mtb). MiRNAs (miRNAs) participate in TB progression by modulating the host-pathogen interaction. Bioinformatics advancements provide basis for exploring novel immunoregulatory miRNAs and their performance as diagnostic biomarkers. Gene and miRNA expression datasets, GSE29190 and GSE54992, were downloaded from Gene Expression Omnibus (GEO) database. Based on fold changes and statistical significance, a total of 7463 differentially expressed mRNAs (DE-mRNAs) and 38 differentially expressed miRNAs (DE-miRNAs) were screened. Function annotation and protein-protein interaction (PPI) network were constructed to reveal underlying mechanisms of TB pathogenesis. Functional annotation identified the MAPK signalling pathway and leukocyte migration as the top enriched processes. The PPI and MGIP networks indicated that chemokine ligands like CXCL1/CXCL2 and receptors, like CCR7 were important down-regulated genes, implying that a protective mechanism against overdue inflammation induced cell death. MiRNA-gene-immune processes (MGIP) network enriched 7 deregulated miRNAs, and their expression was further examined with quantitative real-time PCR (qRT-PCR), in PBMC samples of 20 active TB patients and 20 healthy donors. The diagnostic performance was evaluated with ROC curves. MiR-892b; miR-199b-5p and miR-582-5p were significantly deregulated in TB patients, compared with healthy participants. The best overall performance was from miR-892b, with an area under curve (AUC) of 0.77, 55% sensitivity and 90% specificity. AUC of miR-199b-5p and miR-582-5p were 0.71 and 0.70, respectively. MiR-892b, miR-199b-5p and miR-582-5p could be considered promising novel diagnostic biomarkers for active tuberculosis.