Determination of Sphingosine-1-Phosphate in Human Plasma Using Liquid Chromatography Coupled with Q-Tof Mass Spectrometry

Abstract

Evidence suggests that high-density lipoprotein (HDL) components distinct from cholesterol, such as sphingosine-1-phosphate (S1P), may account for the anti-atherothrombotic effects attributed to this lipoprotein. The current method for the determination of plasma levels of S1P as well as levels associated with HDL particles is still cumbersome an assay method to be worldwide practical. Recently, a simplified protocol based on liquid chromatography-tandem mass spectrometry (LC-MS/MS) for the sensitive and specific quantification of plasma levels of S1P with good accuracy has been reported. This work utilized a triple quadrupole (QqQ)-based LC-MS/MS system. Here we adapt that method for the determination of plasma levels of S1P using a quadrupole time of flight (Q-Tof) based LC-MS system. Calibration curves were linear in the range of 0.05 to 2 µM. The lower limit of quantification (LOQ) was 0.05 µM. The concentration of S1P in human plasma was determined to be 1 ± 0.09 µM (n = 6). The average accuracy over the stated range of the method was found to be 100 ± 5.9% with precision at the LOQ better than 10% when predicting the calibration standards. The concentration of plasma S1P in the prepared samples was stable for 24 h at room temperature. We have demonstrated the quantification of plasma S1P using Q-Tof based LC-MS with very good sensitivity, accuracy, and precision that can used for future studies in this field.

Keywords: high-density lipoprotein; high-density lipoprotein (HDL); liquid chromatography-mass spectrometry (LC-MS); liquid chromatography-tandem mass spectrometry (LC-MS/MS); quadrupole time of flight (Q-Tof); sphingosine-1-phosphate (S1P).

Figure 1

Representative extracted ion chromatogram (EIC) at m/z 380.2560 corresponding to the [M + H]⁺ ion of sphingosine-1-phosphate (S1P). S1P elution time and peak shape are illustrated. The retention time corresponds to that determined for the standards.

Figure 2

Comparison between the Waters X-Select CSH column, as used previously [32] (Red), and a Phenomenex Kinetex EVO C18 used here (Green).

Figure 3

(A) When only the ions of interest are extracted [M+H]⁺, both mass accuracy and isotope pattern are consistent with the presence of S1P. The black lines are the spectrum overlaid with predicted isotope pattern; (B) Representative raw spectrum extracted under the S1P peak with background subtracted. The evidence of blood proteins (probably serum albumin) co-eluting is shown by the peaks in the charge envelope between ~700 and 1500 m/z.

Figure 4

1/x weighted calibration curve recommended to be used. Black dots: calibration standards; blue triangles: QC samples; pink arrow: denotes human plasma. Black dots represent calibration standards; blue triangles are QC samples and pink arrow is a human plasma sample.

Figure 5

Non-weighted linear regression calibration curve of three injection replicates at seven calibration levels (4 to 10) (0.05 to 4.5 µM) with very good correlation and linearity over the entire range. The heteroscedasticity of the data is clear with absolute variance increasing with concentration. Black dots represent calibration standards; blue triangles are QC samples.

Figure 6

The data is heteroscedastic and would be better modelled using a weighted linear regression. Plot of the positive and negative standard deviation against calibration levels show increased absolute variation with increasing concentration.

Figure 7

Response/concentration ratio is constant from 1.5 to 4.5 µM with a decline noticeable at 0.05 µM.

Figure 8

The consistency of the signal over greater than 12 h and 50 injections demonstrates the stability of the instrument and the samples. (Quality control and calibration injections were performed between the above replicates).