Plasmodium vivax malaria is more prevalent than P. falciparum malaria in most parts of India . Infection of P. vivax is historically believed to be a benign form of malaria. However recent reports suggest that P. vivax can also result in several complications including multiple organ failure, acute respiratory distress syndrome and CNS related problems [21–23]. It is, therefore, becoming even more important to understand the host-parasite interaction for vivax malaria, specifically how the host metabolism responds to the parasite infection. In this paper, a 1H NMR approach is described to understand this process. Urine metabolic profiles of P. vivax-infected individuals, healthy individuals, and patients with non-malarial fever were compared in order to get the malaria specific signatures.
Metabolic syndromes like lactic acidosis due to enhanced glycolysis are known complications during severe malaria [15, 24]. However, enhancement in excretion of lactic acid was not observed in vivax-infected patients. This might be due to low parasitaemia (ranging from 0.01%-0.1%), resulting in insignificant perturbation of glucose metabolism. However, a range of other metabolites were found, which differentiate between the malarial patients and the two control categories.
The urine of malaria patients contained significantly increased amounts of valerylglycine and phenylpyruvic acid compared to that of healthy individuals (Figure 7A, E). However, the level of valerylglycine was not significantly different between the malarial patients and non-malarial fever patients (Figure 9G). Valerylglycine is a minor metabolite in fatty acid break down. The elevation of the urinary acylglycines is an indication of impaired β-oxidation of fatty acid . Therefore, it may be concluded that the impairment in fatty acid β-oxidation is associated with febrile condition.
Data presented here also indicates a disturbance in the phenylalanine metabolism. Phenylalanine is known to be metabolized by phenylalaninehydroxylase and tetrahydrobiopterin (THBP) to tyrosine. Alteration of this pathway leads to the formation of phenylpyruvate. Usually phenylpyruvate is a minor metabolite of phenylalanine. However, in patients with phenylketonuria, the phenylalanine-tyrosine pathway is blocked leading to excess formation of phenylpyruvate . A significantly enhanced excretion of phenylpyruvic acid in the malarial patients compared to healthy individuals was observed in our experiments (Figure 7E). This 'phenylketonuric condition' is probably arising out of an impaired phenylalanine metabolism. The data also suggests a possible increase in the biopterin concentration in the urine of malaria patients (Figure 7F). Some atypical form of phenylketonuria has been characterized by an increase in the urinary biopterin, which is caused by malfunction of dihydropteridinereductase (DHPR) . Hyperphenylalaninemia is associated with severe P. falciparum malaria in African children . However, this does not seem to be the case in the present study as the phenylpyruvic acid excretion is not significantly different between the malaria and non-malarial fever patients (Figure 9E). Although, the overlapped peaks of phenylalanine and hippurate showed a significant increase in the urine of malaria patients compared to that of non-malarial fever patients (Figure 8A), this cannot be taken to an indication of hyperphenylalaninemia. The hyperphenylalaninemic condition may be an associated complication of fever. However, there seems to be no study that deals with the metabolite profiling during fever in humans, therefore it remains an interesting aspect to be investigated.
Ornithine is a part of the urea cycle. In the first step of the cycle, ornithinecarbamoyltransferase (OCT) acts on ornithine and carbamoylphosphate to begin the natural detoxification of ammonia. Earlier reports suggested that elevation of serum value of OCT can be a potential marker of malaria and this was associated with the damage of the liver cells in the liver stage of the parasite [27, 28].
Perturbation of OCT levels are likely to result in an impaired urea cycle. Urea cycle is the major pathway for ammonia detoxification in mammals. Earlier reports indicated ammonia toxicity during malaria in mice model . Observation of significant increase in the urinary level of both ornithine (Figure 8D) and N-acetylornithine (Figure 9A) in malaria patients (compared to non-malarial fever patients) and an increased level of N-aceylornithine (compared to healthy individuals) (Figure 7B) also points to possible hepatic injury and impaired urea cycle resulting in ammonia toxicity during P. vivax infection. In the light of results presented here, earlier reports of statistically significant decrease of ornithine in the plasma of patients with clinical malaria  may be interpreted as an enhanced excretion of ornithine in malaria patients.
Pipecolic acid is a minor metabolite of lysine catabolism. In humans, pipecolic acid is found to be associated with chronic liver disease, pyridoxine dependent epilepsy and Zellweger syndrome [30–32]. Recent studies in the murine malaria model show elevated pipecolic acid level in the urine of NMRI mice . Liver dysfunction is associated with malaria. However, the cause for the elevated pipecolic acid during malaria remains unclear. Although statistically not very significant, this metabolite was present in larger quantities in the urine of malarial patients compared with non-malarial fever patients (Figure 9D). Earlier studies showed hyperlysinaemia in the sera of P. berghei ANKA infected mice . Hyperlysinaemia is often related to activation of minor metabolic pathway hence an enhanced pipecolic acid excretion . From the data reported here and earlier data from murine malaria model it seems likely that lysine metabolism is perturbed during the course of malarial infection. However, the underlying biochemical reason remains to be addressed.
The non-malarial fever category serves as a control for general fever related changes that are not specific to P. vivax infection. Febrile illnesses are often likely to be misdiagnosed as malaria. Lundqvist and co-workers showed that borreliosis can be misdiagnosed as malaria due to the similarity in the symptoms . Poorly executed microscopy also sometimes adds on to this, leading to malaria overdiagnosis and overuse of anti-malarial in malaria endemic regions [35, 36]. In this report, it is shown that the non-malarial fevers can be separated from vivax patients on the basis of the urinary metabolite profile of the patients. The urine metabolomic profile seems to be significantly different in these two categories (Figures 3 and 4). This is due to the difference in the levels of Phenylalanine, hippurate, glucose, glutamine, alanine, and ornithine which are increased in the urine of the P. vivax-infected individuals and n-butyrate and acetate which are decreased in these patients (Figure 8, Table 4). Pipecolic acid also showed a higher trend in the urine of malaria patients compared to non-malarial fever patients (Figure 9D). As is evident from a very high Q2(cum) which is equal to 0.89 from the OPLS-DA model made from the urinary metabolite profile of the non-malarial fever patients and P. vivax-infected individuals, urinary metabolic profiles have a potential to be of diagnostic use for detecting malaria and to differentiate malaria from other non-malarial fevers.