Optimization of ultra-high pressure liquid chromatography - tandem mass spectrometry determination in plasma and red blood cells of four sphingolipids and their evaluation as biomarker candidates of Gaucher's disease.

While important advances have been recently achieved in the optimization of lipid classes' separation, information on the specific determination of medium polarity lipids such as sphingolipids (SLs) in highly complex matrices remains fragmentary. In human, disorders of SL metabolism known as sphingolipidoses are a heterogeneous group of inherited disorders affecting primarily the central nervous. Early diagnosis of these conditions is of importance notably when a corrective therapy is available. The diagnosis is generally based on the determination of specific SLs in plasma and red blood cells (RBCs). For instance, glucosylceramide (GL1), glucosylsphingosine (Lyso-GL1), sphingosine (Sph), and sphingosine-1-phosphate (S1P) are proposed as relevant biomarkers for Gaucher disease (GD). Our main objective was to evaluate these biomarker candidates in a cohort of GD patients. However, most of current methods of GL1, Lyso-GL1, Sph, and S1P determination in plasma of GD patients require at least two liquid chromatographic runs. On the other hand, except for GL1 nothing is known concerning the RBC sphingolipid content. Yet, several reversed phase LC-MS methods of SLs separation and/or determination in various media with different sample preparation approaches have been proposed since 2010. Here we focused on stationary phase selection and mobile phase composition as well as on the sample preparation step to optimize and validate an UHPLC-MS/MS method for the simultaneous quantification of the four sphingolipids in both plasma and RBCs. A comparison between seven stationary phases including two RP18, two polar embedded RP18, and three HILIC phases shows that under our conditions polar embedded RP18 phases are the most appropriate for the separation of the four SLs, in terms of efficiency, peak symmetry, and separation time. In the same way, a comparison between a single step extraction with methanol and a liquid-liquid extraction with a mixture of methanol/methyl tert-butyl ether, shows that the latter mixture is the most appropriate for the extraction of SLs in terms of recovery and absence of matrix effect. After validation, this method was applied to the evaluation of the targeted SLs in a cohort of 15 known GD patients. The obtained results show that Lyso-GL1 is the only relevant biomarker in both plasma and RBCs for GD diagnosis. As the proposed method is applicable to the determination in such a highly complex matrices of four SLs with a large difference in polarity, and as the sample preparation procedure is freedom of matrix effects, this method can be easily adapted to a large diversity of samples.