Background: Lysophosphatidic acid (LPA) and sphingosine 1-phosphate (S1P) are ubiquitous lipid messengers found in the blood and most cell types. Both lysophospholipids are ligands of G protein–coupled receptors and mediate important physiological processes. Moreover, lysophospholipids are potential biomarkers for various diseases, including atherosclerosis and cancer. Because existing methodologies are of limited value for systematic evaluations of S1P and LPA in clinical studies, we developed a fast and simple quantification method that uses liquid chromatography–tandem mass spectrometry (LC-MS/MS).
Methods: Sphingoid base 1-phosphates and LPA species were quantified in negative-ion mode with fragments of m/z 79 and 153, respectively. The internal standards LPA 17:0 and [13C2D2]S1P were added before butanol extraction. Application of hydrophilic-interaction chromatography allowed coelution of analytes and internal standards with a short analysis time of 2.5 min.
Results: Comparison of butanol extraction with a frequently used extraction method based on strong acidification of human plasma revealed artificial formation of LPA from lysophosphatidylcholine with the latter method. Validation according to US Food and Drug Administration guidelines showed an overall imprecision (CV) of <12% and a limit of detection <6 nmol/L for all lysophospholipid species. Concentrations of S1P and sphinganine 1-phosphate (SA1P) in EDTA-containing plasma were stable for 24 h at room temperature, whereas LPA concentrations increased substantially over this period.
Conclusions: Our validated LC-MS/MS methodology for quantifying LPA, S1P, and SA1P features simple sample preparation and short analysis times, therefore providing a valuable tool for diagnostic evaluation of these lysophospholipids as biomarkers.
Lysophospholipids, in particular lysophosphatidic acid (LPA)1 and sphingosine 1-phosphate (S1P), are bioactive lipids that elicit a wide range of cellular responses, such as cell survival, differentiation, and migration (1)(2)(3)(4)(5). Blood concentrations of S1P and LPA are derived either from cellular release (S1P by platelets and red blood cells) (5) or from enzymatic conversion of plasma lipids (lysophosphatidylcholine conversion to LPA by autotaxin) (2)(4). Both LPA and S1P act on specific G protein–coupled receptors on cells of the immune, cardiovascular, and nervous systems and thereby regulate the induction of inflammation, atherosclerosis, and cancer (1)(2)(3)(4)(5). Because of their signaling functions and roles in different diseases, LPA and S1P may gain clinical importance as biomarkers; however, the quantitative analysis of LPA and S1P as diagnostic variables in routine testing or in large clinical studies is hindered by the presently laborious and time-consuming methods of analysis.
Liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) has been applied to the measurement of S1P and LPA (6)(7)(8)(9)(10)(11)(12)(13). Several methods lack coelution of the internal standard and the analyte, which is necessary to help compensate for potential matrix effects or for varying ionization efficiency due to mobile-phase gradients (7)(10)(11)(14). Laborious sample preparation (8)(15), large sample volumes (8)(11)(12)(13), time-consuming derivatization procedures (15), and a lack of sufficient validation (8)(9)(10)(11) exclude these methods from routine analysis. Only one methodology exists for combined measurement of LPA and S1P (11). We describe a simple and rapid LC-MS/MS method for quantifying S1P, sphinganine 1-phosphate (SA1P), and LPA species.
We used a hybrid triple-quadrupole linear ion-trap mass spectrometer (API 4000 Q-Trap; Applied Biosystems). The LC equipment consisted of a binary and isocratic pump (Agilent Technologies 1200 series) connected to an HTC Pal autosampler (CTC Analytics). Chromatographic separation was performed on a Waters Atlantis™ hydrophobic-interaction chromatography silica column (50 × 2.1 mm), with a 3-μm particle size and equipped with a guard column (10 × 2.1 mm) containing the same material. The column was maintained at 50 °C. The mobile phase consisted of water containing 2 mL/L formic acid and 50 mmol/L ammonium formate (eluent A) and acetonitrile containing 2 mL/L formic acid (eluent B). A gradient elution was performed with 5% A for 0.7 min, a linear increase to 25% A until 1.5 min, 50% A until 1.7 min, and reequilibration from 1.7–2.5 min with 5% A. A flow rate of 500 μL/min was used except from 0–0.5 min and from 1.7–2.5 min, when the flow rate was increased to 700 μL/min. To minimize contamination of the mass spectrometer, we directed column flow into the mass spectrometer via a divert valve only from 1.3–2.1 min. Otherwise, methanol was delivered into the mass spectrometer at a flow rate of 250 μL/min. The Applied Biosystems TurboIonSpray source was operated in the negative-ionization mode with the following settings: ion spray voltage, −4500 V; ion source heater temperature, 300 °C; source gas 1, 40 psi; source gas 2, 35 psi; curtain gas, 20 psi. Evaluation of the LPA species revealed fragments from the head group (m/z 79, m/z 153) and the fatty acid moiety (m/z 255 for LPA 16:0) (see Fig. 1A in the Data Supplement that accompanies the online version of this Brief Communication at http://www.clinchem.org/content/vol55/issue6 ). To cover all LPA species and to increase the specificity, we chose m/z 153 as the product ion for LPA quantification. Product ion spectra for S1P and SA1P showed only a single intense fragment of m/z 79 (see Fig. 1B in the online Data Supplement), which we then used for analysis (see Table 1 in the online Data Supplement). The collision energy optimums for LPA species and sphingoid base 1-phosphates were 30 eV and 58 eV, respectively.
To obtain the naturally occurring lysophospholipid species, we analyzed plasma lipid extracts with a precursor ion scan of 79 m/z and 153 m/z and obtained the following main species: S1P and SA1P containing a dihydroxy-C18:1 base, LPA 16:0, LPA 18:0, LPA 18:1, LPA 18:2, and LPA 20:4. This pattern is in good accord with the literature (7)(9)(11)(16). Quantification was achieved by the addition of 5 different concentrations of standards for S1P, SA1P, LPA 16:0, LPA 18:0, LPA 18:1, LPA 18:2, and LPA 20:4 to samples of human plasma (Table 1⇓ ).
Lysophospholipids such as LPA and S1P show poor extraction efficiencies with such organic solvents as chloroform at neutral pH. Therefore, as a first step, we used a modified extraction procedure according to Bligh and Dyer with 6 mol/L HCl (8)(12). Although both LPA and S1P were extracted with efficiencies of 85%–95%, this extraction method had poor reproducibility for plasma samples. Because lysophosphatidylcholine (LPC) can be degraded to LPA under highly acidic conditions (11), we added nonnaturally occurring LPC 19:0 to plasma samples. Acidic chloroform extraction of spiked plasma converted about 2% of the LPC 19:0 to LPA 19:0 (see Table 2 in the online Data Supplement). This artificial formation of LPA from LPC required the presence of plasma, because acidic extraction of LPC 19:0 without plasma did not produce any LPA 19:0 (data not shown). For an LPC plasma concentration of 250 μmol/L (17), strong acidic extraction could potentially produce an approximately 7-fold increase in the LPA concentration. To circumvent these problems, we used the butanolic extraction procedure described by Baker et al. (16) and observed no LPC-to-LPA conversion (data not shown). In brief, 75 μL EDTA-containing plasma was mixed with 20 ng each of LPA 17:0 and S1P labeled with a stable isotope ([13C2D2]S1P; Toronto Research) and 400 μL of a buffer containing 30 mmol/L citric acid and 40 mmol/L disodium hydrogen phosphate (pH 4.0). Extraction was performed with 1 mL of 1-butanol and 500 μL of water-saturated 1-butanol. The recovered butanol phase was evaporated to dryness under reduced pressure. The residue was redissolved in 200 μL ethanol, and 10 μL was injected into the LC system.
Most of the LC-MS/MS methods that have been reported for LPA and S1P analysis use reversed-phase chromatography (6)(7)(10)(11)(12)(15)(17) resulting in a separation of analytes and internal standards. In contrast, we established a separation based on hydrophilic-interaction chromatography, which coelutes analytes with their respective internal standards (Fig. 1⇓ ). The retention times of LPA and S1P were 1.74 min and 1.84 min, respectively. Because of their coelution, SA1P, S1P, and LPA species exhibited isotope overlap. To avoid overestimating species concentrations, we corrected peak areas according to principles previously described (18) (see Table 3 in the online Data Supplement).
Our established LC-MS/MS method was validated according to the US Food and Drug Administration Guidelines for Bioanalytical Methods (19). First, we evaluated ion suppression of the analyte due to matrix effects. Because no analyte-free EDTA-containing plasma was available, we analyzed a mixture of internal standards with or without plasma extract (spiked analyte concentrations of 0.5, 1, and 2 μmol/L). The presence of plasma extract caused mean signal reductions (SD) of 25.0% (1.8%) and 2.5% (1.4%) for [13C2D2]S1P and LPA 17:0, respectively. Moreover, coelution of analytes and internal standards may suppress the MS response for multicomponent analytes. Although we found signal suppression by analyte coelution, the ratios of the internal standard to the analyte did not change substantially (see Table 4 in the online Data Supplement).
Calibration by standard addition to plasma samples showed linearity and recovery of the analytes in the interval of spiked concentrations (Table 1⇓ ). For accurate quantification given the different responses of species (e.g., only a 30% response of LPA 20:4 compared with other LPA species; Table 1⇓ ), we used multiple calibration curves. Furthermore, we tested the influence of the plasma lipid content on the response for the analytes by varying cholesterol concentrations as a surrogate marker. The slopes of the calibration curves showed no notable differences at cholesterol concentrations of 2.97, 4.52, and 7.49 mmol/L (see Table 5 in the online Data Supplement).
The intraday imprecision values (CVs) were <9%, and interday CVs were <11% (see Table 6 in the online Data Supplement). The limit of detection, defined as a signal-to-noise ratio of 3, was approximately 6 nmol/L for S1P and SA1P and <2 nmol/L for LPA (see Table 1 in the online Data Supplement).
To investigate sample stability, we tested EDTA-containing plasma samples that had been stored immediately at −80 °C or frozen after 1, 4, 8, and 24 h of incubation at ambient temperature. For immediately stored samples, we observed no changes in concentration compared with freshly analyzed plasma samples (data not shown). S1P and SA1P were stable in EDTA-containing plasma and serum up to 24 h at room temperature. In contrast, LPA concentrations increased up to 8-fold when samples were stored at room temperature for 24 h (see Table 7 in the online Data Supplement). This increase is most likely due to LPC conversion to LPA by the lysophospholipase D activity of autotaxin (2). In samples of whole blood, the concentrations of sphingoid base 1-phosphates rapidly increased at room temperature (see Table 9 in the online Data Supplement). Consequently, immediate plasma separation and storage at −80 °C is of paramount importance to achieve reliable results.
Finally, we measured LPA, S1P, and SA1P in samples of EDTA-containing plasma prepared from blood samples freshly drawn from healthy human volunteers (n = 10; see Table 10 in the online Data Supplement). Approval for this study was obtained from the University of Regensburg ethics committee. The mean S1P and SA1P concentrations were 0.59 μmol/L and 0.19 μmol/L, respectively. The mean concentration of total LPA was 0.7 μmol/L, with the 2 predominant species [LPA 20:4 (approximately 55%) and LPA 18:2 (approximately 25%)] contributing 80% of the overall measured LPA concentration. These values are in good agreement with previously reported concentrations (7)(11)(12)(14)(15)(16). In contrast, the concentrations of total LPA measured after highly acidic extraction were much higher (approximately 5 μmol/L), most likely because of LPC degradation (8).
Compared with previous reports based on LC-MS analyses (6)(7)(11)(12)(15)(16), our method appears to have major advantages: small sample volumes, easy sample preparation, and short run times (2.5 min) (see Table 11 in the online Data Supplement). Most importantly, hydrophilic-interaction chromatography permits coelution of analytes and internal standards, an important factor for optimal compensation of matrix effects and varying ionization efficiencies.
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.
Authors’ Disclosures of Potential Conflicts of Interest: Upon manuscript submission, all authors completed the Disclosures of Potential Conflict of Interest form. Potential conflicts of interest:
Employment or Leadership: None declared.
Consultant or Advisory Role: None declared.
Stock Ownership: None declared.
Honoraria: None declared.
Research Funding: This work was supported by Deutsche Forschungsgemeinschaft (Li 923/2–1/2 and SFB-TR 13/A3) and by the Seventh Framework Program of the EU-funded “LipidomicNet” (proposal number 202272) and related to “The European Lipidomics Initiative; Shaping the Life Sciences,” a specific support action subsidized by the EC 2006–2007 (proposal number LSSG-CT-2004-013032).
Expert Testimony: None declared.
Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.
Acknowledgments: We thank Jolante Aiwanger, Doreen Müller, and Simone Peschel for excellent technical assistance; Martin Hermansson (University of Helsinki) for technical advice concerning phosphate determination; and Thomas Langmann (Institute of Human Genetics, University of Regensburg) for helpful discussions and critical reading.
↵1 Nonstandard abbreviations: LPA, lysophosphatidic acid; S1P, sphingosine 1-phosphate; LC-MS/MS, liquid chromatography–tandem mass spectrometry; SA1P, sphinganine 1-phosphate; LPC, lysophosphatidylcholine.
- © 2009 The American Association for Clinical Chemistry