Background: Tuberculosis meningitis (TBM) is the most severe form of tuberculosis, and currently lacks efficient diagnostic approaches. Metabolomics has the potential to differentiate patients with TBM from those with other forms of meningitis and meningitis-negative individuals. However, no systemic metabolomics research has compared the cerebrospinal fluid (CSF) of these patients.
Methods: 1H nuclear magnetic resonance (NMR) was used for CSF metabolic profiling. Principal component analysis and orthogonal signal correction-partial least squares-discriminant analysis (OPLS-DA) were used to screen for important variables. The Human Metabolome Database was used to identify metabolites, and MetaboAnalyst 4.0 was used for pathway analysis and over-representation analysis.
Results: OPLS-DA modeling could distinguish TBM from other forms of meningitis, and several significantly changed metabolites were identified. Additionally, 23, 6, and 21 metabolites were able to differentiate TBM from viral meningitis, bacterial meningitis, and meningitis-negative groups, respectively. Pathway analysis indicated that these metabolites were mainly involved in carbohydrate and amino acid metabolism, and over-representation analysis indicated that some of these pathways were over-represented.
Conclusions: The metabolites identified have the potential to serve as biomarkers for TBM diagnosis, and carbohydrate and amino acid metabolism are perturbed in the CSF of patents with TBM. Metabolomics is a valuable approach for screening TBM biomarkers. With further investigation, the metabolites identified in this study could aid in TBM diagnosis.
Keywords: Bacterial meningitis (BM); Cerebrospinal fluid (CSF); Metabolic profiling; Metabolomics; Tuberculosis meningitis (TBM); Viral meningitis (VM).
Copyright © 2019 Elsevier B.V. All rights reserved.