Abstract
Background Polycythemia vera (PV) is a malignant clonal hematological disease of hematopoietic stem cells characterized by proliferation of peripheral blood cells, and JAK2 mutation is one of the main causes of PV peripheral blood cell proliferation. Abnormal cell metabolism is a new feature of malignant proliferation of tumor cells, but the role of metabolism in the pathogenesis and prognosis of PV remains unclear.
Methods Metabolic profiling of peripheral blood sera from 32 patients with PV and 20 HCs was performed using liquid chromatography-mass spectrometry (LC-MS). Partial least-squares with latent structure discriminant analysis (PLS-DA) was used for data analysis.
Results To assess whether there were significant changes in the metabolic profiles of PV patients and healthy controls, we performed a multifactorial analysis of the metabolic profiles of the two groups. Figure 1 shows the score plots of the PLS-DA in positive and negative modes. Both data from positive and negative modes suggested that the serum metabolites of PV patients were significantly changed. Subsequently, we identified endogenous metabolites between PV patients and HCs. These potential biomarkers were eventually identified as fatty acids, acylcarnitine, sphingolipids, amino acids, etc. To clarify the abnormal metabolic pathways in PV, we performed enrichment analysis of the metabolic pathways of the biomarkers. The results showed there were 9 important metabolic pathways associated with PV, namely: D-glutamine and D-glutamate metabolism, sphingolipid metabolism, pyruvate metabolism, alanine, aspartate and glutamate metabolism, arginine and proline metabolism, tryptophan metabolism, lysine degradation, arginine biosynthesis and glycolysis/gluconeogenesis (Figure 2). Therefore, fatty acid metabolism, glucose metabolism, sphingolipid metabolism and amino acid metabolism were significantly altered in PV.
To explore the effect of JAK2-associated metabolic abnormalities on cell proliferation in PV patients, we screened for endogenous metabolites associated with JAK2, and found 7 metabolites that differed significantly between the JAK2 mutated and unmutated groups. The levels of Cer(d18:2/22:6-2OH(7S, 17S)), SM(d18:0/PGF1α), CerP(d18:1/16:0), glutamic acid, lactic acid, melibiose and xylose were significantly higher in the JAK2-mutated group than in the JAK2-unmutated group. Among them, Cer(d18:2/22:6-2OH(7S, 17S)) (r=0.817; P<0.001) and SM(d18:0/PGF1α) (r=0.463; P=0.021) levels were positively correlated with JAK2 mutational burden. Then we evaluated the relationship between JAK2 mutation and blood cell count in PV patients. PLT and HCT levels were significantly higher in JAK2 mutated PV patients than in JAK2 unmutated PV patients (P=0.006; P=0.044) and found a significant positive correlation between JAK2 mutational burden and WBC and PLT levels (r=0.62, P=0.001; r=0.55, P=0.007). Finally, we found that Cer(d18:2/22:6-2OH(7S, 17S)) levels were positively correlated with WBC (r=0.650, P<0.001), and SM(d18:0/PGF1α) levels were positively correlated with WBC and PLT (r=0.444, P=0.011 and r=0.623; P<0.001, respectively). Therefore, we speculate that Cer(d18:2/22:6-2OH(7S, 17S)) and SM(d18:0/PGF1α) are closely associated with JAK2 mutations and may contribute to the proliferation of peripheral blood cells in PV patients.
According to the prognostic point system, PV patients can be divided into low-risk group (n=11) and intermediate/high-risk group (n=21). Compared to the low-risk group, the levels of 8 endogenous metabolites in the intermediate/high-risk group were significantly elevated. We next used binary logistic regression to analyze the combined biomarkers and then determine the ROC curve. A set of potential biomarkers containing four endogenous metabolites with an AUC > 0.77 was found to provide an AUC of 0.952, with a sensitivity of 90.905% and specificity of 90.909% at the optimal cutoff point. Therefore, elevated levels of Cer(d18:2/22:6-2OH(7S, 17S)), SM(d18:0/PGF1α), CerP(d18:1/16:0) and octadec-13-enoylcarnitine could be used as potential metabolic biomarkers of poor prognosis in patients with PV.
Conclusion Metabonomics provides an important tool for the study of the pathogenesis of PV and the relationship between JAK2 gene mutation. Furthermore, the potential biomarkers of this study may provide a reference for the prognosis of PV.
Disclosures
No relevant conflicts of interest to declare.
