Tuberculosis diagnostic markers based on the whole gene transcriptome

Early diagnosis is essential for the timely treatment of tuberculosis patients. At present, there is a lack of ideal clinical diagnostic methods for this disease. Therefore, rapid and simple diagnostic methods are urgently needed. This study analyzed expression of the whole genome transcript of nine tuberculosis patients before treatment, six cases of latent tuberculosis infection, and six healthy controls. The expression profiles of tuberculosis patients were followed before and after treatment. The results showed that, compared with healthy controls and patients infected with latent tuberculosis, 77 genes were highly expressed in tuberculosis patients and 50 genes were expressed at low levels. In response to different periods of anti-tuberculosis therapy (0, 3, and 6 months), the expression of 12 genes was reduced. Gene Ontology enrichment and Kyoto Encyclopedia of Genes and Genomes biological pathways analyses identified nine biological processes related to the inflammatory response, chemotaxis, and the cellular immune response. In addition, one signaling pathway was associated with a cytokine-cytokine receptor interaction. Chemokines and cytokines were involved in each signaling pathway. Our results were verified in additional subjects using the quantitative real-time polymerase chain reaction. Specifically, most measured mRNA levels were consistent with the microarray data. In the eight genes with the best consistency, two ( HBB and TNFRSF10C ) were ideal biomarkers for diagnosing tuberculosis. The receiver operating characteristic curve and the area under the curve were above 0.9, and the sensitivity and specificity were over 95% for these genes. In summary, two differentially expressed genes were identified as novel biological markers for tuberculosis. They provide a new basis for the establishment of diagnostic methods, efficacy monitoring, and revealing the pathogenesis for this disease.