Plasma metabolomics reveals membrane lipids, aspartate/asparagine and nucleotide metabolism pathway differences associated with chloroquine resistance in Plasmodium vivax malaria

Abstract

BackgroundChloroquine (CQ) is the main anti-schizontocidal drug used in the treatment of uncomplicated malaria caused by Plasmodium vivax. Chloroquine resistant P. vivax (PvCR) malaria in the Western Pacific region, Asia and in the Americas indicates a need for biomarkers of resistance to improve therapy and enhance understanding of the mechanisms associated with PvCR. In this study, we compared plasma metabolic profiles of P. vivax malaria patients with PvCR and chloroquine sensitive parasites before treatment to identify potential molecular markers of chloroquine resistance.MethodsAn untargeted high-resolution metabolomics analysis was performed on plasma samples collected in a malaria clinic in Manaus, Brazil. Male and female patients with Plasmodium vivax were included (n = 46); samples were collected before CQ treatment and followed for 28 days to determine PvCR, defined as the recurrence of parasitemia with detectable plasma concentrations of CQ ≥100 ng/dL. Differentially expressed metabolic features between CQ-Resistant (CQ-R) and CQ-Sensitive (CQ-S) patients were identified using partial least squares discriminant analysis and linear regression after adjusting for covariates and multiple testing correction. Pathway enrichment analysis was performed using Mummichog.ResultsLinear regression and PLS-DA methods yielded 69 discriminatory features between CQ-R and CQ-S groups, with 10-fold cross-validation classification accuracy of 89.6% using a SVM classifier. Pathway enrichment analysis showed significant enrichment (p<0.05) of glycerophospholipid metabolism, glycosphingolipid metabolism, aspartate and asparagine metabolism, purine and pyrimidine metabolism, and xenobiotics metabolism. Glycerophosphocholines levels were significantly lower in the CQ-R group as compared to CQ-S patients and also to independent control samples.ConclusionsThe results show differences in lipid, amino acids, and nucleotide metabolism pathways in the plasma of CQ-R versus CQ-S patients prior to antimalarial treatment. Metabolomics phenotyping of P. vivax samples from patients with well-defined clinical CQ-resistance is promising for the development of new tools to understand the biological process and to identify potential biomarkers of PvCR.

Metabolic perturbations in classic galactosemia beyond the Leloir pathway: Insights from an untargeted metabolomic study

“…While we successfully identified some metabolites by comparing accurate mass m/z and RT to our confirmed reference library, we also sought to assign high‐confidence annotations to the remaining discriminatory features. To do this, we used xMSannotator to first map each discriminatory feature to candidate matches from the HMDB on the basis of accurate mass m/z and then scored the strength of each candidate match, as explained in Materials and Methods. These annotations served to supplement the initial annotation results from Mummichog.…”

“…We used a combination of univariate and multivariate statistical methods to identify features with differential abundance between cases and controls . First, for each individual feature, we fit a linear regression model of log 2 ‐transformed intensities that included CG status, the potential confounders identified by tests described above (gender, age, and storage time), and batch number as predictors.…”

“…We considered pathways to differ between groups when P < 0.05. We further annotated discriminatory features using xMSannotator version 1.3.2, first identifying accurate‐mass matches to known metabolites in the Human Metabolome Database (HMDB) on the basis of monoisotopic mass m/z with a mass error threshold of ±10 ppm (Level 5 identification by the criteria of . We considered multiple adducts of each metabolite (eg, M + H, M + 2H, M + H + NH 4 , M + ACN + 2H, M + 2ACN + 2H, M + NH 4 , M + Na, M + ACN + H, M + ACN + Na, M + 2ACN + H, 2 M + H, 2 M + Na, 2 M + ACN + H, M + 2Na‐H, M + H‐H 2 O, M + H‐2H 2 O) when first searching HMDB prior to confidence assignment.…”

Predicting human health from biofluid-based metabolomics using machine learning

“…Cancer represented the largest category, with 14 studies 11,12,25,29,[31][32][33][34][35][36][37][38][39] ; cardiovascular 28 , rheumatologic 26 , ‡ , renal ‡ , and respiratory 16 diseases were represented by only one or two studies. Other health states with a small to medium number of studies included neuro/neuropsychiatric 18,19 , endocrine 21,24,40,41 , and infectious disease 14,15,27,30,42,43 , as well as a general 'other' category 17,[44][45][46] .…”

“…These significant features are then used by algorithms like partial least squares discriminant analysis (PLS-DA) for both additional dimensionality reduction and health-state classification. Additional classification methods employed include random forests [24][25][26] and support vector machines 27 . Other analyses are strictly based on univariate or multivariate receiver operating characteristic-area under the curve (ROC-AUC) 26,28 .…”