|
|
||||||||
1 Institut für Klinische Chemie und Laboratoriumsmedizin, Universität Regensburg, D-93042 Regensburg, Germany.
aAuthor for correspondence. Fax 49-941-944-6202; e-mail gerd.schmitz{at}klinik.uni-regensburg.de.
| Abstract |
|---|
|
|
|---|
Methods: Crude lipid extracts of EDTA-plasma samples were used for direct flow injection analysis. LPC 13:0 and LPC 19:0 were added as internal standards, and the ESI-MS/MS was operated in the parent-scan mode for m/z 184. Quantification was achieved by standard addition. Data processing was highly automated by use of the mass spectrometer software and self-programmed Excel macros.
Results: The calibrators LPC 16:0, LPC 18:0, and LPC 22:0 showed a linear response independent of sample dilution and plasma cholesterol concentration for both internal standards. The within-run imprecision (CV) was 3% for the major and 12% for the minor species, whereas the total imprecision was
12% for the major and 25% for the minor species. The detection limit was <1 µmol/L.
Conclusion: The developed ESI-MS/MS methodology with an analysis time of 2 min/sample, simple sample preparation, and automated data analysis allows high-throughput quantification of distinct LPC species from plasma samples, which could be a valuable tool for the evaluation of LPC as diagnostic marker.
| Introduction |
|---|
|
|
|---|
LPC is a relevant component of human plasma originating from lecithin-cholesterol acyltransferase, hepatic secretion (4)(5), or action of phospholipase A2 (6). LPC generated by phospholipase A2 promotes inflammatory effects such as expression of endothelial cell adhesion molecules, growth factors, chemotaxis, and activation of monocytes/macrophages (6). Recently, two G-protein-coupled receptors with a high affinity for LPC were identified: G2A, which is an immunoregulatory receptor (7), and GPR4, which is a receptor for LPC and sphingosylphosphorylcholine (8). LPC has been suggested to play a functional role in the pathogenesis of various diseases, including atherosclerosis (9), diabetes (10), systemic autoimmune diseases (11), and cancer. Moreover, plasma (12) and ascites (13) from ovarian cancer patients had increased LPC concentrations, and plasma samples had altered fatty acid composition compared with healthy controls. Taken together, these data indicate a potential use of LPC as diagnostic marker.
Various methods are used for LPC measurement. The most common involve separation by thin-layer chromatography (TLC) (14)(15) or HPLC (16)(17)(18). Studies analyzing LPC fatty acid composition used either TLC separation followed by gas chromatographic analysis (12)(19) or HPLC coupled to electrospray ionization mass spectrometry (ESI-MS) (20). Recently, a quantification method for different lysophospholipids was published that uses ESI-tandem MS (ESI-MS/MS) with prior TLC separation (21). The application of these methodologies in the routine laboratory, however, is limited because they are complicated, time-consuming procedures that are to some extent insensitive, unselective, and imprecise. Here we describe a methodology based on the highly specific and selective ESI-MS/MS that is applicable for high-throughput routine LPC quantification.
| Materials and Methods |
|---|
|
|
|---|
apparatus
Samples were quantified by direct flow injection analysis. The solvent mixture, 10 mmol/L ammonium acetate in methanolchloroform (3:1 by volume), was provided at a constant flow of 75 µL/min by a Waters Alliance 2790. The triple quadrupole mass spectrometer (Quattro LC; Micromass) was operated with the following settings: capillary voltage, 3.5 kV; cone voltage, 41 V; collision energy, 24 V; collision gas, argon at a pressure of 1.3 10-3 Torr. Quantification was achieved by a precursor ion scan of m/z 184 specific for phosphocholine-containing lipids (22). Mass resolution was above unit resolution. Data analysis was performed with MassLynx software, which included the NeoLynx tool (Micromass) for averaging the scans at half peak height of the total ion count. NeoLynx generates centroid peak data from the continuum spectra and allows selection of the intensities of certain peaks. These NeoLynx results were exported to Excel spreadsheets and further processed by self-programmed Excel macros, which sorted the results, calculated the ratios to the internal standards, generated calibration curves, and calculated quantitative values. Concentrations of the LPC species were calculated with use of the closest related calibration slope. For LPC 14:0, LPC 15:0, LPC 16:1, and LPC 16:0, we used the slope of the LPC 16:0 calibration curve; for LPC 17:0, LPC 18:2, LPC 18:1, and LPC 18:0, we used the slope of the LPC 18:0 calibration curve; and for LPC 20:4, LPC 20:3, LPC 22:6, and LPC 22:5, we used the slope of the LPC 22:0 calibration curve.
sample preparation
EDTA-plasma samples were used for analysis. Before lipid extraction, we placed 100 µL of a chloroform solution containing 10 ng/µL each of LPC 13:0 and LPC 19:0 in a glass centrifuge tube and evaporated the solvent. For calibrators, we added known amounts of LPC 16:0, LPC 18:0, and LPC 22:0 and evaporated the solvent. We extracted 20 µL of EDTA plasma according to the procedure described by Bligh and Dyer (23). The separated chloroform phase was dried and dissolved in 10 mmol/L ammonium acetate in methanolchloroform (3:1 by volume), producing a 75-fold dilution corresponding to the initial plasma volume. We injected 20 µL of this solution, and data were acquired over 2 min.
lpc quantification by tlc
Plasma lipid extracts were separated by high-performance TLC as described previously (24).
| Results |
|---|
|
|
|---|
|
plasma species
Analysis of crude plasma lipid extracts by a precursor-ion scan of m/z 184 yielded the main species LPC 16:0, LPC 18:2, LPC 18:1, and LPC 18:0 and the minor species LPC 14:0, LPC 15:0, LPC 16:1, LPC 17:0, LPC 20:4, LPC 20:3, LPC 22:6, and LPC 22:5 (Fig. 2
). The spectra contained no peaks that could not be assigned to distinct LPC species.
|
quantification
For quantitative analysis, we added LPC 13:0 and LPC 19:0, non-naturally occurring species, as internal standards before lipid extraction to compensate for variations in sample preparation and in the ionization efficiency attributable to matrix components or machine instabilities. For all calculations, we used the ratio of analyte to the internal standards rather than the ion counts. To address a possible chain length dependency of species response, we generated calibration curves by adding LPC 16:0, LPC 18:0, and LPC 22:0. Calibration curves calculated with both LPC 13:0 and LPC 19:0 (Fig. 3
and Table 1
) showed a linear response for the added LPC species in the range of concentrations added. The added LPC species displayed a response variation <25% related to the maximum response (Tables 1
and 2
). The linearity of the method was further documented by an experiment in which two samples, one with low and the other with high analyte concentrations, were mixed at different ratios. The results (Fig. 4
) indicated that independent of whether LPC 13:0 or LPC 19:0 was used as internal standard, total LPC concentrations at the different mixture ratios followed a strongly linear relationship. Because ESI is highly affected by matrix components, we tested the influence of dilution of the lipid extract and different plasma cholesterol concentrations on the LPC species response. The calibration curve slopes showed no notable differences at cholesterol concentrations of 2.20 and 6.78 mmol/L or dependence on the dilution of the injected solution at 50- to 400-fold dilutions relative to the extracted plasma volume (Table 2
). A 75-fold dilution of the plasma extract was chosen for further measurements and provided satisfactory ion counts. Concentrations of the LPC species were calculated based on the closest related calibration slope, and the results obtained for both internal standards were similar (Table 1
).
|
|
|
|
assay performance
The within-run imprecision (CV) of the assay was
3% for the total LPC concentration and the major species and
1013% for the minor species (Table 3
). The day-to-day variation was substantially lower when LPC 13:0 was used as the internal standard compared with LPC 19:0,
12% for the major and 25% for the minor species. To assess the detection limit, an analyte-free plasma sample was essential because the effect of the matrix had to be considered. Because such plasma was not available, we determined the limit of detection by calculating the values for the analyte-free regions of m/z 504, m/z 530, and m/z 560, selected from the mass range of the naturally occurring LPC species. The resulting detection limit was 0.60.8 µmol/L for both internal standards, calculated from the mean + 3 SD at the above m/z values measured in 10 aliquots of the same sample.
|
The ESI could be disturbed by contamination in the ion source; we therefore checked the stability of the calibration curve. The calibration curve slope, linearity, and intercept calculated with LPC 13:0 and LPC 19:0 were not influenced by 250 injections of extracted plasma samples showing the stability of the methodology.
sample stability
Because of the possible generation of LPC species by lecithin-cholesterol acyltransferase or phospholipase A2, we assessed the stability of the plasma samples at room temperature (Table 4
). LPC species showed a slow increase at room temperature to a 24% total increase at 4 h. In samples stored 1 day at room temperature, the LPC concentration increased 54%; most the increase was in the saturated species LPC 16:0 and LPC 18:0. In contrast, repeated freeze-thaw cycles had no significant effect on LPC concentration.
|
comparison with tlc
To compare the developed methodology with the existing TLC methodology, we quantified a selected plasma sample by both ESI-MS/MS and TLC. The total LPC concentration obtained by ESI-MS/MS was 184 µmol/L, compared with 213 µmol/L for the TLC measurement.
| Discussion |
|---|
|
|
|---|
Our proposed methodology, based on the highly sensitive and selective ESI-MS/MS, requires only a one-step lipid extraction for sample preparation, which is monitored by addition of two non-naturally occurring LPC species as internal standards. The resulting crude lipid extracts were dried and analyzed by direct flow injection in an optimized solvent mixture. These internal standards also were required to compensate for the various matrix effects on the ESI attributable to changing lipid composition. To address the possible influence of chain length on the ionization response, different LPC species have to be analyzed. To achieve a calibration with a reasonable effort, we chose three species, LPC 16:0, LPC 18:0, and LPC 22:0, which cover the chain length range of naturally occurring LPC species, for calibration by standard addition. In addition, not all naturally occurring LPC species were commercially available. Calibration curves constructed with LPC 13:0 and LPC 19:0 as internal standards showed a linear response in the measured concentration range (Table 1
and Fig. 3
), and the chain length-dependent response displayed a variation within 25% related to the maximum. These similar responses for LPC 16:0, LPC 18:0, and LPC 22:0 were not in accordance with a study by Xiao et al. (21), who found a 75% decrease in the response for the long-chain species LPC 22:0 and LPC 24:0 from the maximum for LPC 16:0 and LPC 18:0. However, this may be related to differences in the mass spectrometer settings as well as in the ionization matrix. The small response variation observed in our protocol should permit calculation of quantitative values of noncalibrated LPC species by use of the closest related calibration curve slope. Although this calculation might not provide exact values, it should at least allow a close quantitative approximation. Moreover, neither sample dilution nor different cholesterol concentrations as surrogate markers of total lipid content influenced LPC quantification (Table 2
), and the responses were stable during multiple injections. Taken together, the internal standards LPC 13:0 and LPC 19:0 allow sufficient compensation of different ionization matrices and provide a constant, similar linear response for LPC 16:0, LPC 18:0, and LPC 22:0, permitting calculation of quantitative values.
Our quantitative results showed excellent precision for the major species, with a within-run CV of
3%, whereas the lower precision for the minor species may be attributable to a concentration close to the detection limit. One reason for the lower precision between runs may be different machine conditions. These variations were possibly compensated more accurately by LPC 13:0, which had a lower total variation compared with LPC 19:0. However, for the evaluation of minor species, the ratio to other naturally occurring species may be relevant (12), and this ratio showed a much better day-to-day precision (e.g., the ratio of LPC 22:6 to LPC 18:0 had a CV of 11%). Taken together, the application of LPC 13:0 for the calculation of quantitative results and the use of ratios to other naturally occurring species for relevant minor species provide a assay precision well suited for routine measurements. Additionally, LPC 19:0 added in a constant ratio to LPC 13:0 could be used as an internal quality control to confirm the proper ionization response of each sample.
Direct comparison of the LPC concentration determined by ESI-MS/MS with a TLC method showed a good correlation. Moreover, a study by Croset et al. (19), in which the authors analyzed the fatty acid composition of LDL, HDL, and albumin-bound LPC by TLC and gas chromatography, showed a good correlation of the LPC concentration and species with our data. They found a total LPC concentration in these fractions of 150 µmol/L, of which 40% was unsaturated fatty acids (Table 3
; total, 181 µmol/L and 34% unsaturated).
All peaks in the spectra of crude plasma lipid extracts could be assigned to distinct LPC species, demonstrating the high selectivity of the parent scan at m/z 184. The high sensitivity is demonstrated by the small plasma volume (20 µL) used for lipid extraction and that an injection volume corresponding to 0.3 µL of plasma was sufficient for LPC quantification. In contrast to existing protocols, this assay offers advantages regarding analysis time (2 min/sample), the stability of the measurement, the simple sample preparation, and the detailed information obtained for different LPC species. Additionally, self-programmed Excel macros coupled to the commercial mass spectrometer software enable highly automated data processing. The precision of this methodology and its high-throughput capability thus make it suitable for routine application.
The EDTA plasma used for LPC quantification showed a slow but continuous increase in LPC concentrations primarily because of increases in the saturated species LPC 16:0 and LPC 18:0, which may be a result of lecithin-cholesterol acyltransferase (26) or phospholipase A2 activity (27), both of which could hydrolyze phosphatidylcholine mostly at the unsaturated sn-2-acyl group to generate saturated 1-acyl-LPC.
In summary, LPC has been shown to function as a signaling molecule that binds to the G-protein-coupled receptors G2A (7) and GPR4 (8). This may be the functional basis for the role of LPC in various diseases (9)(10)(11)(12)(13). A potential diagnostic use has been implicated for patients with ovarian cancer, who exhibit increased plasma LPC concentrations attributable to increases in the saturated species LPC 16:0 and LPC 18:0 (12). In contrast, unsaturated LPC 18:1 and especially LPC 18:2 are decreased in the plasma of ovarian cancer patients, which produces an increased LPC 16:0-to-LPC 18:2 ratio, whereas patients with other cancers, such as leukemia, have decreased plasma LPC. Additionally, increased lysophospholipid concentrations, including LPC, have been detected in ascites of these patients (13). LPC may also play an important role during acute phases because acute-phase HDL contained more LPC compared with control samples (20). Additional support for the importance of the distinct LPC species comes from a study in which the plasma of patients with dementia and cognitive impairment was analyzed: only patients who were cognitively impaired but did not have dementia had decreased LPC 22:6 concentrations (28). These data indicate a potential use of LPC as a diagnostic marker, although this must be evaluated in further studies involving systematic screening of patient groups with conditions associated with a possible pathophysiologic role of LPC. The described methodology, based on ESI-MS/MS, provides an ideal tool for these investigations.
| Acknowledgments |
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
The following articles in journals at HighWire Press have cited this article:
![]() |
M. Scherer, G. Schmitz, and G. Liebisch High-Throughput Analysis of Sphingosine 1-Phosphate, Sphinganine 1-Phosphate, and Lysophosphatidic Acid in Plasma Samples by Liquid Chromatography-Tandem Mass Spectrometry Clin. Chem., June 1, 2009; 55(6): 1218 - 1222. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Wiesner, K. Leidl, A. Boettcher, G. Schmitz, and G. Liebisch Lipid profiling of FPLC-separated lipoprotein fractions by electrospray ionization tandem mass spectrometry J. Lipid Res., March 1, 2009; 50(3): 574 - 585. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Yu, Q. Li, T. M. Badger, and N. Fang Quantitative Analysis of Polar Lipids in the Nanoliter Level of Rat Serum by Liquid Chromatography/Mass Spectrometry/Mass Spectrometry Exp Biol Med, February 1, 2009; 234(2): 157 - 163. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. Hou, H. Zhou, F. Elisma, S. A. L. Bennett, and D. Figeys Technological developments in lipidomics Briefings in Functional Genomics, September 19, 2008; (2008) eln042v1. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Xiao, T. Watanabe, Y. Zhang, B. Trigatti, L. Szeto, P. W. Connelly, S. Marcovina, T. Vaisar, J. W. Heinecke, and G. F. Lewis Enhanced Cellular Uptake of Remnant High-Density Lipoprotein Particles: A Mechanism for High-Density Lipoprotein Lowering in Insulin Resistance and Hypertriglyceridemia Circ. Res., July 18, 2008; 103(2): 159 - 166. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Binder, G. Liebisch, T. Langmann, and G. Schmitz Metabolic Profiling of Glycerophospholipid Synthesis in Fibroblasts Loaded with Free Cholesterol and Modified Low Density Lipoproteins J. Biol. Chem., August 4, 2006; 281(31): 21869 - 21877. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. M. Ham, R. B. Cole, and J. T. Jacob Identification and comparison of the polar phospholipids in normal and dry eye rabbit tears by maldi-tof mass spectrometry. Invest. Ophthalmol. Vis. Sci., August 1, 2006; 47(8): 3330 - 3338. [Abstract] [Full Text] [PDF] |
||||
![]() |
W. Drobnik, G. Liebisch, F.-X. Audebert, D. Frohlich, T. Gluck, P. Vogel, G. Rothe, and G. Schmitz Plasma ceramide and lysophosphatidylcholine inversely correlate with mortality in sepsis patients J. Lipid Res., April 1, 2003; 44(4): 754 - 761. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |