Tuberculosis (TB) is an ancient infectious disease that remains one of the major health threats in humans worldwide. Biosignatures can play a significant role in the development of novel intervention measures against TB and blood transcriptional profiling is increasingly exploited for their rational design. We have compared whole blood gene expression in TB patients, as well as in healthy infected and uninfected individuals in a cohort in The Gambia, West Africa and validated previously identified signatures showing high similarities of expression profiles among different cohorts. In this study, we applied a unique combination of classical gene expression analysis with pathway and functional association analysis integrated with intra-individual expression correlations. These gene signature analyses were employed for pathognomonic associations, identifying a network of Fc gamma receptor 1 signaling with correlating transcriptional activity as hallmark of gene expression in TB. Remarkable similarities to characteristic signatures in the autoimmune disease systemic lupus erythematosus (SLE) were observed, and functional gene clusters of immunoregulatory interactions provide detailed insights into the dysregulation of critical immune processes in TB. Transcriptomics thus (i) provides a robust system for identification and validation of biosignatures for TB diagnosis and (ii) application of integrated analysis tools yields novel insights into functional networks underlying TB pathogenesis.
Overall design
A total of 46 sputum smear and chest x-ray positive TB patients (TB), 25 latently infected healthy donors (LTBI, TB case contacts with Mantoux test induration >= 10mm) and 37 uninfected donors (NID, Mantoux test induration = 0mm) were included from a cohort in The Gambia. All subjects were HIV– and samples from TB patients were taken prior to chemotherapy. Whole blood RNA was extracted using the PAXgene Blood RNA Kit (PreAnalytix) and labeled with Agilents Fluorescent Linear Amplification Kit. Quantity and labeling efficiency were verified before hybridization of the samples to whole-genome 4×44k human expression arrays and scanned at 5 μm using an Agilent scanner.
Less...
More...