Metabolic profile in childhood intra-thoracic tuberculosis
Background: We hypothesized that the serum metabolic profile of children with intra-thoracic tuberculosis is different from children with no TB, and therefore with the use of untargeted metabolomics approach we aimed to identify TB specific biomarkers for the diagnosis of childhood intra-thoracic TB.
Methods and materials: Serum samples obtained from a total of 36 subjects i.e., from 23 children with intra-thoracic tuberculosis confirmed for Mycobacterium tuberculosis by liquid culture (MGIT 960), eight healthy controls (children with no apparent signs of any disease), and 5 disease controls (children with typical symptoms of respiratory tract infections not consistent with active TB) were analyzed by 1H Nuclear Magnetic Resonance (NMR) spectroscopy at 700 MHz field strength and one dimensional NMR spectra were acquired using Car-Purcell-Meiboom-Gill (CPMG) pulse sequence. The differences in serum metabolic profiles of these subjects were explored by multivariate analysis using partial least square discriminatory analysis (PLS-DA) and Orthogonal PLS-DA (OPLS-DA) models and significant metabolites separating the groups were further tested for statistical significance by univariate analysis done by Mann Whitney U test. Deregulated pathways with p < 0.05 and impact value > 0.0 were selected as significantly altered metabolic pathways.
Results: Eight significant discriminatory metabolites with VIP score >1 and statistical significance with p-value <0.05 were found to be the most specific biomarkers differentiating TB from non-TB controls. Compared to healthy and disease controls, sera of TB patients were characterized by significantly increased levels of NAAL, leucine, PUFA and phenylalanine and decreased levels of asparagine and choline. The levels of methionine and glutamine were significantly decreased in TB patients compared to healthy controls, but not lower than the disease controls. Receiver operating characteristics (ROC) curve analysis revealed that sensitivity, specificity and area under the curve (AUC) for this biomarker panel were 77%, 96% and 0.976 respectively. Pathway analysis revealed alterations in amino acid and lipid metabolisms in TB children.
Conclusion: The biomarkers identified in our study (NAAL, leucine, methionine, glutamine, asparagine, phenylalanine, choline and PUFA) show a diagnostic potential to differentiate children with TB from non-TB controls efficiently which on further validation in larger cohort may form the basis of a new diagnostic test for active TB.
Authors and affiliations
1. Nupur Sharma, Ph.D., Department of Microbiology, AIIMS, New Delhi
2. Dr. Urvashi B. Singh, Professor, Department of Microbiology, AIIMS, New Delhi
3. Dr. Uma Sharma, Additional Professor, Department of NMR, AIIMS, New Delhi
4. Dr. Rakesh Lodha, Professor, Department of Pediatrics, AIIMS, New Delhi
5. Dr. S. K. Kabra, Professor, Department of Pediatrics, AIIMS, New Delhi
6. Dr. Bimal K. Das, Professor, Department of Microbiology, AIIMS, New Delhi
7. Dr. Arti Kapil, Professor, Department of Microbiology, AIIMS, New Delhi
8. Dr. Hitender Gautam, Associate Professor, Department of Microbiology, AIIMS, New Delhi