Early prediction of cerebral malaria by 1H NMR based metabolomics. (December 2016)

• Record Type: Journal Article
• Authors: Ghosh, Soumita; Sengupta, Arjun; Sharma, Shobhona; Sonawat, Haripalsingh
• Is Part Of: Malaria journal. Volume 15:Number 1(2016)
• Page Start: 1
• Page End: 10
• Publication Date: 2016-12
• Subjects: Cerebral malaria -- NMR -- Prediction -- Metabolomics -- OPLS-DA -- Lipids; Malaria -- Periodicals; 616.9362
• DOI: 10.1186/s12936-016-1256-z
• Languages: English
• ISSNs: 1475-2875

Abstract

Abstract Background Cerebral malaria (CM) is a life-threatening disease, caused mainly byPlasmodium falciparum in humans. In adults only 1–2 % ofP. falciparum -infected hosts transit to the cerebral form of the disease while most exhibit non-cerebral malaria (NCM). The perturbed metabolic pathways of CM and NCM have been reported. Early marker(s) of CM is(are) not known and by the time a patient exhibits the pathological symptoms of CM, the disease has progressed. Murine CM, like the human disease, is difficult to assign to specific animals at early stage and hence the challenge to treat CM at pre-clinical stage of the disease. This is the first report of prediction of CM in mice using a novel strategy based on1 H nuclear magnetic resonance (NMR)-based metabolomics. Methods Mice were infected with malarial parasites, and serum was collected from all the animals (CM/NCM) before CM symptoms were apparent. The assignment of mice as NCM/CM at an early time point is based on their symptoms at days 8–9 post-infection (pi). The serum samples were subjected to1 H NMR-based metabolomics.1 H NMR spectra of the serum samples, collected at various time points (pi) in multiple sets of experiments, were subjected to multivariate analyses. Results The results from orthogonal partial least square discriminant analyses (OPLS-DA) suggest that the animals with CM start to diverge out in metabolic profile and were distinct on day 4 pi, although by physical observation they were indistinguishable from the NCM. The metabolites that appeared to contribute to this distinction were serum lipids and lipoproteins, and 14–19 % enhancement was observed in mice afflicted with CM. A cut-off of 14 % change of total lipoproteins in serum predicts 54–71 % CM in different experiments at day 4 pi. Conclusion This study clearly demonstrates the possibility of differentiating and identifying animals with CM at an early, pre-clinical stage. The strategy, based on metabolite profile of serum, tested with different batches of animals in both the sex and across different times of the year, is found to be robust. This is the first such study of pre-clinical prognosis of CM.