Abstract 1200Untargeted metabolomics denotes the profiling of several thousand low molecular weight biochemicals, including lipids, hormones, saccharides, nucleotides, organic acids, and amino acids. The analytes are measured using mass spectrometry (MS) as molecule mass peaks without specific “targeting” of specific metabolites or analytes. We undertook an untargeted metabolomics investigation to define the effect of warfarin on plasma metabolite profiles in a crossover study of 17 venous thrombosis patients (9 male, 8 female) (age 58.1 ± 18.6 yr old; range 27–83 yr old). Blood was drawn at 4–6 weeks after initiation of warfarin (patients with an average INR 2.50 ± 0.65) and again at ∼ 10 days after discontinuation of the clinically determined course of warfarin. Plasmas were methanol-extracted and applied to a capillary reverse phase HPLC column, interfaced to an ESI-TOF mass spectrometer. Data were converted and aligned in the time domain using the program XCMS, with 3500–4500 distinct MS features found. Statistically significant differences were ranked using a paired t-test. In addition to warfarin metabolites, other changes were discovered. Several unique lipid metabolites, lyso-phosphatidylcholines (lyso-PC) (16:0 and 18:2), palmitoyl-ethanolamide (N-palmitoylethanolamine)(PEA) and stearoyl-ethanolamide (N-palmitoylethanolamine)(SEA), were identified as significantly decreased during warfarin therapy (p < 0.03)(See Table). PEA and SEA are naturally occurring saturated N-acylethanolamines (ethanolamides), both of which are structurally related to arachidonoyl (20:4) ethanolamide which binds to the cannabinoid receptor. PEA and SEA are devoid of affinity for cannabinoid receptors and physiological roles for PEA and SEA are unknown. When tested for effects on coagulation, exogenously added PEA (16:0) dose-dependently enhanced thrombin generation in plasma induced by tissue factor or calcium ions. SEA (18:0) and other lipid ethanolamides, e.g., arachidonoyl ethanolamide (20:4), also enhanced recalcification-induced thrombin generation in plasma. However, two analogs of PEA, palmitoyl N-isopropylamide and N-palmitoyl taurine which are lacking hydroxyl molecule in the head group, showed little enhancement of thrombin generation in plasma. These results suggest that the free hydroxyl group in the head group appears to be a key component for the observed procoagulant activity in ethanolamide family. How is it that warfarin might affect the metabolism of these lipids? The R-racemer of warfarin interacts with the pregnane X receptor (PNR) and increases CYP3A4 and CYP2C9 mRNAs levels in cultured human cells. Alternations in the catabolic activity of CYP3A4 might alter levels of ethanolamides and lyso-PC in patients. In summary, we employed the innovative untargeted metabolomics methods to identify endogenous biomarkers in warfarin-treated patients to help reveal biochemical changes triggered by this widely used drug which led to the discovery of a new family of plasma procoagulant molecules, namely ethanolamides. The downregulation of procoagulant ethanolamides by warfarin and its anticoagulant consequent effect might be minor, but may yet be an important additional anticoagulant property of warfarin. This study has implications for coagulation and plasma lipid relationships with potential clinical and pharmacologic applications.TablePlasma metabolites changed by warfarin treatmentNameMolecular formulaMH+ calculatedm/z observedFold changeP valuePEAC18H37NO2300.2897300.29082.8 ↓0.026SEAC20H41NO2328.3210328.31933.5 ↓0.030Lyso-PC(18:2)C26H50NO7P520.3397520.34243.5 ↓0.029Lyso-PC(16:0)C24H50NO7P496.3397496.33721.6 ↓0.006 Disclosures:No relevant conflicts of interest to declare.
Read full abstract