Article on Methodology

Rapid separation and identification of phosphatidylethanolamine molecular species

Correspondence to: Lambert M. G. van Golde

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

A novel high performance liquid chromatography method is presented for the separation and identification of intact molecular species of phosphatidylethanolamine (PE). After isocratic separation, detection of species can be achieved by measurement of UV absorbance as well as by the quantitative method of light scattering detection. A mathematical relationship exists between i) the relative retention time of a PE molecular species and ii) the number of carbon atoms and double bonds in the aliphatic groups of the species. This relationship can aid in the identification of the species. Furthermore, the absence of non-volatile components in the solvent allows the use of electrospray mass spectrometry to identify the eluting components and to establish the position of the individual radyl groups at the glycerol backbone. Using this method, samples of bovine heart PE (rich in plasmalogens) and rat liver PE (rich in diacyl species) have been analyzed.—Brouwers, J. F. H. M., E. A. A. M. Vernooij, A. G. M. Tielens, and L. M. G. van Golde. Rapid separation and identification of phosphatidylethanolamine molecular species. J. Lipid Res. 1999. 40: 164–169.

Supplementary key words: ether lipids, phospholipids, plasmalogens, mass spectrometry, HPLC

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

Phosphatidylethanolamine (PE) is a major phospholipid in membranes from bacterial as well as from eukaryotic origin and has received special attention due to several interesting characteristics. Among the best studied are the asymmetric distribution in the plasma membrane of cells and the particular richness of ether-linked aliphatic groups often found in this phospholipid class. Furthermore, PE is rich in polyunsaturated fatty acyl groups. After hydrolysis by phospholipase A₂, released fatty acids play important roles in signal transduction by serving as precursors for biologically active molecules such as prostaglandins and leukotrienes. For a thorough characterization of membrane phospholipids it is therefore necessary to determine not only phospholipid headgroup composition, but also the individual molecular species of the phospholipids.

Analysis of intact molecular species of phospholipids by high performance liquid chromatography (HPLC) is nowadays the method of choice. Without the need of derivatization, the risk for the introduction of artefacts is reduced to a minimum, and the information about the combinations of fatty acids present in the lipid class is preserved. Whereas phospholipid classes can easily be separated and sensitively quantified by a variety of procedures (1) (2) (3), separation of the molecular species within a single phospholipid class without removal of the polar headgroup, is more problematic (reviewed in ref. (4)). Patton, Fasulo, and Robins (5) published a method applicable for a variety of phospholipid classes, including PE. Quantification, however, is difficult using that method, as the response of the UV detector is dependent on the degree of unsaturation of the aliphatic groups, making careful calibration of each component necessary (6). Light scattering has proven to be an excellent choice for the detection of molecular species, as it is relatively inexpensive, very sensitive and, in contrast to a UV detector, the response is independent of the number of double bonds in the molecule (7). Recently, we published a method for the quantitative analysis of intact molecular species of PC using a light scattering detector (8). However, this method is not applicable as such for PE, because the presence of triethylamine in the solvent causes all PE molecular species to elute from the column without retention. We have now developed a method for the separation of intact molecular species of PE that enables the use of a light scattering detector and the isolation of individual species in a pure form.

A PE sample from a biological origin usually contains a large number of fatty acyl combinations bonded via different types of radyl linkage, making identification of each and every peak a laborious task. Mass spectrometry (MS) has been used successfully in the past to aid in the identification of molecular species of PE and its derivatives (9) (10) (11) (12). The electrospray ionization procedure used here poses the same restrictions on the mobile phase as an evaporative light scattering detection (ELSD). Both detectors display best performance when no salts are present and the solvent is volatile. In this paper we describe an HPLC method for the rapid analysis of intact PE molecular species, which is capable of good component separation and quantification. Identification of the individual molecular species and the assignment of the position of the acyl chains at the glycerol backbone was achieved with on-line electrospray mass spectrometry.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

Isolation of PE
Bovine heart PE and rat liver PE were purified by normal phase HPLC according to the method of Letter (3) with some minor modifications as published previously (8).

High performance liquid chromatography
All solvents were from Labscan, Dublin, Ireland and of HPLC grade. Molecular species of approximately 50 nmol of PE were separated on two 250 x 4 mm Lichrospher RP-18 endcapped columns in series (Merck, Darmstadt, Germany). Isocratic elution was performed with a solvent consisting of acetonitrile–methanol 3:7 (v/v) containing 5 µM ethanolamine, or a solvent consisting of acetonitrile–methanol 2:3 (v/v) containing 2 µM ethanolamine at a flow rate of 1.25 ml · min^-1. The column effluent was monitored at 206 nm using an LKB 2251 Uvicord (Pharmacia, Upsala, Sweden) and subsequent ELSD was performed using a Varex MKIII obtained from Alltech, (Deerfield, IL) operating at a gas flow rate of 1.9 l · min^-1 and a drift tube temperature of 100°C. The detector was calibrated by triplicate injections of 1, 2.5, 5, 7.5, 10, 15, 20, 25, and 30 µg (approx. 1– 40 nmol) dioleoyl phosphatidylethanolamine. For on-line mass spectrometry, the flow rate of the HPLC was adjusted to 1 ml · min^-1 and a flow splitter was inserted between column and mass spectrometer. The flow into the mass spectrometer was set to approximately 0.1 ml · min^-1.

Mass spectrometry
A Fisons VG Platform II single quadrupole mass spectrometer (Micromass, Manchester, UK), fitted with an electrospray ion source operated at atmospheric pressure, was used for the identification of components eluting from the column. Nitrogen was used as nebulizer gas (0.2 l · min^-1) and as curtain gas (7 l · min^-1). The capillary voltage was set to 3.0 kV and the cone voltage to 100 V.

Data analysis
Data obtained from the ELSD were recorded and analyzed using EZChrom chromatography data system (Scientific Software, San Ramon, CA). MS data were recorded and analyzed using Masslynx software (Micromass, Manchester, UK).

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

High performance liquid chromatography
Non-specific interaction is known to occur between the phospholipid headgroup and column material and to interfere with an optimal separation. In the past, this has been prevented by the addition of either choline chloride or triethylamine to the solvent system (5) (8). Here, we were able to reduce the interaction between the PE headgroup and the column material by the addition of low concentrations of ethanolamine (2–5 µM) and frequent regeneration of the column with acetone, hexane, and ethanol. Despite the fact that two columns in series are used for the separation of molecular species, the total time required for analysis is only 75 min at a flow rate of 1.25 ml · min^-1, which compares favorably with the well-known method published by Patton et al. (5), requiring 150 min at a flow rate of 2 ml · min^-1. Figure 1 shows the separation of intact PE molecular species from bovine heart as detected by UV absorbance and subsequent ELSD. Although both detectors yield comparable chromatograms, the increased response of the UV detector to highly unsaturated species (e.g., (16:0–20:5)PlasEth, peak 4) relative to more saturated species (e.g., (18:0–18:2)PtdEth, peak 16) is evident. For the quantification of species based on the absorbance at 206 nm, response factors for the individual molecular species are required, thereby excluding the analysis of samples with an unknown composition or fully saturated species (6). Light scattering detection, on the other hand, permits quantification at the nanomole level and is unaffected by the degree of unsaturation of lipids, but careful calibration of the detector is necessary as the response of the detector is at low concentrations nonlinear with the amount of sample (1) (8).

View larger version (19K): [in this window] [in a new window]   Figure 1. The separation of intact PE molecular species from bovine heart. Molecular species were resolved on two Lichrosphere RP-18 columns in series using a mobile phase consisting of 5 µM ethanolamine in methanol–acetonitrile 7:3 (v/v). Detection was performed by measurement of the absorbance at 206 nm (UV) and subsequent light scattering detection (ELSD). Peak numbers correspond to species listed in Table 2.

After calibration of the light scattering detector in the range of 1 to 40 nmol with (18:1–18:1)PtdEth, the molecular species of PE and their abundance in rat liver and bovine heart were determined as listed in Table 1 and Table 2. The species composition of PE isolated from rat liver has previously been described (5) (13) (14), and although the exact amounts of individual molecular species vary somewhat between the different studies, the main characteristics do not. Typical is the low amount of ether lipids in rat liver found in the literature (5) (13) (14) as well as in the present study; all common species are of the diacyl subclass. The molecular species (16:0–20:4) PtdEth, (16:0–22:6) PtdEth, (18:0–20:4) PtdEth, and (18:0–22:6) PtdEth are invariably reported as the most abundant species. These are also major components according to our analysis (Table 1). Earlier analysis of membranes from (canine) heart tissue by reverse phase (RP)-HPLC and fast atom bombardment mass–spectrometry (FAB-MS) has revealed that the PE fraction of these membranes is rich in plasmalogens and that arachidonic acid is the predominant fatty acyl at the sn -2 position (15) (16) (17). The (bovine) heart PE fraction analyzed here (Table 2), is also rich in plasmalogens (53% of total PE), and contains arachidonic acid as the most common fatty acyl at the sn -2 position (40%).

Identification of species
In the past, the high sensitivity of electrospray ionization mass spectrometry (ESI-MS) has successfully been used for the analysis of phospholipids and their molecular species by direct injection of chloroform extracts (18) (19) (20) (21). Although direct injection does allow detection of subpicomole amounts of phospholipids (18) (21), individual species cannot be isolated for further use or analysis. Furthermore, complex mixtures of lipids require tandem mass spectrometry (MS/MS), making their analysis time-consuming and difficult to automate. Here, we achieved on-line identification of individual species eluting from the column using a single quadrupole mass spectrometer fitted with an electrospray ion source operating in the negative ion mode. Spectra of eluting plasmalogen species displayed the fatty acid carboxylate anion derived from the sn -2 position, the molecular ion [M–H]^-, and two ions resulting from fragmentation at the ester bond ( Figure 2). Fragmentation of the vinyl-ether bond at the sn-1 position did not occur. These (single MS) spectra thereby showed large resemblance to tandem MS spectra obtained after collision-induced dissociation of PE molecular ions (21) (22).

View larger version (15K): [in this window] [in a new window]   Figure 2. Negative ion mass spectrum recorded during the elution of component 20 in Figure 1 ((18:0–18:2)PlasEth). Observed ions correspond to [M–H]- (m/z 726.6) and ions resulting from fragmentation at the sites indicated in the structure formula of the [M–H]- anion.

Eluting diacyl species also had fragmentation patterns that were comparable to previously published data from more complex tandem mass spectrometry ( Figure 3, refs. 19, 21, 22). Characteristic is the presence of the molecular ion [M–H]^- and both the fatty acid carboxylate anions from the sn -1 (m/z 283) and sn -2 (m/z 303) position, the sn -2 carboxylate anion being more abundant than the sn -1 carboxylate anion, thereby enabling determination of the position of the fatty acyl groups along the glycerol backbone. A recently published HPLC/ESI-MS method for the determination of acyl chain position in PC molecular species is based on the relative abundance of the sn-1 and sn-2 lyso-PC fragment ions (23). Similarly, further confirmation of acyl position in PE species could be obtained from the abundance of the sn -1 and sn -2 lyso-PE ions in the mass spectrum (see inset, Figure 3). Loss of the acyl group from the sn -2 position as R₂ = C = O (leading to the fragment ion at m/z 480) was considerable. From the sn -1 position, loss of R₁ = C = O was hardly observed (m/z 500). From both the sn -1 and sn -2 position the radyl group was lost as intact fatty acid (leading to m/z 482 and m/z 462, respectively).

View larger version (19K): [in this window] [in a new window]   Figure 3. Negative ion mass spectrum during the elution of (18:0–20:4) PtdEth in a sample of rat liver PE. Elution was performed as in Figure 1. Observed ions correspond to [M–H]- (m/z 766.8) and ions resulting from fragmentation at the sites indicated in the structure formula of the [M–H]- anion.

Relation between retention time and carbon number
The isocratic elution in this system had another advantage in that a mathematical correlation existed between the relative retention time (R'_t) and the equivalent carbon number (ECN) of a molecular species ( Figure 4). We found that the ECN of a given species can be calculated from the number of carbon atoms (CN) and double bonds in the aliphatic groups (n) of the species by the empirically obtained formulas ECN = CN - 1.505n (diacyl species) and ECN = CN - 1.538n (plasmalogen species). The formulas relating R'_t to ECN are illustrated in Figure 4 for plasmalogen and diacyl species of the analyzed samples of bovine heart and rat liver PE, respectively, and can assist in the identification of unknown PE species when a mass spectrometer is not available.

View larger version (14K): [in this window] [in a new window]   Figure 4. Exponential relationship between the relative retention time of molecular species and their equivalent carbon numbers. A: Plasmalogen species of bovine heart PE eluted with methanol–acetonitrile 7:3 (v/v) from two endcapped Lichrosphere RP18 columns in series in the presence of 5 µM ethanolamine. Numbers refer to the species listed in Table 2. B: Diacyl species of rat liver PE eluted with methanol–acetonitrile 3:2 (v/v) in the presence of 2 µM ethanolamine. Numbers refer to the species listed in Table 1.

Method performance
In conclusion, the HPLC method described here separates a wide variety of molecular species of PE in an intact form using solvents that enable UV detection as well as ELSD. The sensitivity of ELSD for mass rather than for degree of unsaturation enables the sensitive and quantitative analysis of PE samples of unknown composition. Furthermore, unequivocal species identification, including the assignment of acyl chain position, can be achieved by on-line electrospray MS. The observed relationship between ECN and retention time can aid in the identification of unknown species. The method can be applied for the isolation of individual PE species in order to detect their biological activities.

	ACKNOWLEDGMENTS

This work was supported by the Netherlands Foundation for Chemical Research (SON) and the Life Science Foundation (SLW) of The Netherlands Organization for Scientific Research (NWO).

Manuscript received April 23, 1998; and in revised form September 1, 1998.

Abbreviations: ECN, equivalent carbon number; ELSD, evaporative light scattering detection; ESI, electrospray ionization; FAB, fast atom bombardment; (RP-) HPLC, (reverse phase-) high performance liquid chromatography; MS, mass spectrometry; PC, phosphatidylcholine; PE, phosphatidylethanolamine (referring collectively to all diradyl ethanolamine glycerophospholipid species); PtdEth, diacyl phosphatidylethanolamine; PlasEth, plasmalogen PE species (1-alk-1'-enyl 2-acyl ethanolamine glycerophospholipids); R'_t, relative retention time

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

1. Christie, W. W. 1985. Rapid separation and quantification of lipid classes by high performance liquid chromatography and mass (light-scattering) detection. J. Lipid Res. 26:507-512[Abstract].

2. Bunger, H., Pison, U. 1995. Quantitative analysis of pulmonary surfactant phospholipids by high-performance liquid chromatography and light-scattering detection. J. Chromatogr. B. Biomed. Appl. 672:25-31[Medline].

3. Letter, W. S. 1992. A rapid method for phospholipid class separation by HPLC using an evaporative light scattering detector. J. Liq. Chromatogr. 15:253-266.

4. Olsson, N. U., Salem, N. J. 1997. Molecular species analysis of phospholipids. J. Chromatogr. B. Biomed. Appl. 692:245-256.

5. Patton, G. M., Fasulo, J. M., Robins, S. J. 1982. Separation of phospholipids and individual molecular species of phospholipids by high-performance liquid chromatography. J. Lipid Res. 23:190-196[Abstract].

6. McHowat, J., Jones, J. H., Creer, M. H. 1996. Quantitation of individual phospholipid molecular species by UV absorption measurements. J. Lipid Res. 37:2450-2460[Abstract].

7. Guiochon, G., Moysan, A., Holley, C. 1988. Influence of various parameters on the response factors of the evaporative light scattering detector for a number of non-volatile compounds. J. Liq. Chromatogr. 11:2547-2570.

8. Brouwers, J. F. H. M., Gadella, B. M., van Golde, L. M. G., Tielens, A. G. M. 1998. Quantitative analysis of phosphatidylcholine molecular species using HPLC and light scattering detection. J. Lipid Res. 39:344-353[Abstract/Free Full Text].

9. Hullin, F., Kim, H. Y., Salem, N. 1989. Analysis of aminophospholipid molecular species by high performance liquid chromatography. J. Lipid Res. 30:1963-1975[Abstract].

10. Menguy, L., Christon, R., van Dorsselaer, A., Leger, C. L. 1992. Apparent relative retention of the phosphatidylethanolamine molecular species 18:0–20:5(n–3), 16:0–22:6(n–3) and the sum 16:0–20:4(n–6) plus 16:0–20:3(n–9) in the liver microsomes of pig on an essential fatty acid-deficient diet. Biochim. Biophys. Acta. 1123:41-50[Medline].

11. Kim, H. Y., Wang, T. C., Ma, Y. C. 1994. Liquid chromatography/mass spectrometry of phospholipids using electrospray ionization. Anal. Chem. 66:3977-3982[Medline].

12. Jungalwala, F. B., Evans, J. E., McCluer, R. H. 1984. Compositional and molecular species analysis of phospholipids by high performance liquid chromatography coupled with chemical ionization mass spectrometry. J. Lipid Res. 25:738-749[Abstract].

13. Sugiyama, K., Yamakawa, A. 1996. Dietary eritadenine-induced alteration of molecular species composition of phospholipids in rats. Lipids. 31:399-404[Medline].

14. Samborski, R. W., Ridgway, N. D., Vance, D. E. 1990. Evidence that only newly made phosphatidylethanolamine is methylated to phosphatidylcholine and that phosphatidylethanolamine is not significantly deacylated–reacylated in rat hepatocytes. J. Biol. Chem. 265:18322-18329[Abstract/Free Full Text].

15. Gross, R. W. 1984. High plasmalogen and arachidonic acid content of canine myocardial sarcolemma: a fast atom bombardment mass spectroscopic and gas chromatography–mass spectroscopic characterization. Biochemistry. 23:158-165[Medline].

16. Gross, R. W. 1985. Identification of plasmalogen as the major phospholipid constituent of cardiac sarcoplasmic reticulum. Biochemistry. 24:1662-1668[Medline].

17. Schulz, R., Strynadka, K. D., Panas, D. L., Olley, P. M., Lopaschuk, G. D. 1993. Analysis of myocardial plasmalogen and diacyl phospholipids and their arachidonic acid content using high-performance liquid chromatography. Anal. Biochem. 213:140-146[Medline].

18. Han, X., Gross, R. W. 1994. Electrospray ionization mass spectroscopic analysis of human erythrocyte plasma membrane phospholipids. Proc. Natl. Acad. Sci. USA. 91:10635-10639[Abstract/Free Full Text].

19. Smith, P. B., Snyder, A. P., Harden, C. S. 1995. Characterization of bacterial phospholipids by electrospray ionization tandem mass spectrometry. Anal. Chem. 67:1824-1830[Medline].

20. Kerwin, J. L., Tuininga, A. R., Ericsson, L. H. 1994. Identification of molecular species of glycerophospholipids and sphingomyelin using electrospray mass spectrometry. J. Lipid Res. 35:1102-1114[Abstract].

21. Han, X., Gross, R. W. 1995. Structural determination of picomole amounts of phospholipids via electrospray ionization tandem mass spectrometry. J. Am. Soc. Mass Spectrom. 6:1202-1210.

22. Murphy, R. C. 1993. Mass spectrometry of lipids. In Handbook of Lipid Research. Vol. 7. F. Snyder, editor. Plenum Press, New York.

23. Vernooij, E. A. A. M., Kettenes-van den Bosch, J. J., Crommelin, D. J. A. 1998. Rapid determination of acyl chain position in egg phosphatidylcholine by high performance liquid chromatography electrospray mass spectrometry. Rapid Commun. Mass Spectrom. 12:83-86.