Characterization of Phospholipids by Two-Dimensional Liquid

Article pubs.acs.org/JAFC

Characterization of Phospholipids by Two-Dimensional Liquid Chromatography Coupled to In-line Ozonolysis−Mass Spectrometry Chenxing Sun,* Yuan-Yuan Zhao, and Jonathan M. Curtis Agriculture/Forestry Center 4-10, Department of Agricultural, Food and Nutritional Science, University of Alberta, Edmonton, Alberta T6G 2P5, Canada S Supporting Information *

ABSTRACT: In this study, the characterization of phospholipids (PL) was achieved by using a combination of LC/MS/MS and two-dimensional LC/MS. A HILIC LC column was used for PL class separation, while the further molecular species separation of each PL class was achieved by using online HILIC × C18 LC. The double bond positions along the fatty acyl chains of these PL molecular species were also obtained by using the combination of 2D-LC and in-line ozonolysis−MS analysis. The ozonolysis device is composed of a gas-permeable, liquid-impermeable Teflon tube passing through a glass chamber filled with ozone gas, which is then placed in-line between the 2D-LC and the mass spectrometer. The eluting PL molecules in the LC mobile phase passed through the device where they rapidly reacted with the ozone that penetrated through the tubing wall. The ozonolysis products were then detected by MS in real-time, which allowed the localization of the double bonds along the fatty acyl chains in these PL molecular species. This comprehensive method was successfully applied to an egg yolk PL extract, which revealed the detailed structures of the PL molecules. KEYWORDS: phospholipids, double bond positions, two dimensional liquid chromatography, ozonolysis, mass spectrometry

■

INTRODUCTION Glycerophospholipids (GPs) and sphingolipids (SPs) are the two main classes of phospholipids (PLs) which are important structural and functional components of eukaryotic cell membranes.1 Dietary PLs from animal and plant sources, often with unsaturated fatty acids (FAs) at the sn-2 position, can be delivered to and incorporated into cell membranes.2 Thus, the fatty acid composition of dietary PLs can potentially affect membrane components of certain cells and so ultimately influence cellular function, with implications for the treatment of diseases. For example, in a mouse model dietary intake of sphingomyelins (SMs) has shown a preventive effect on the formation of colon cancer.3 Also, an n-3 polyunsaturated fatty acid (PUFA) rich GP extract from a marine source has shown a blood cholesterol lowering effect in patients suffering from hyperlipidaemia.4 Since PLs are known to be safe and ubiquitous components in food, a recommendation to increase dietary intake of specific PLs for the prevention of disease does not carry the risks associated with an increased consumption of other nonfood compounds. Therefore, a systematic study of the PL structure of foods may help in understanding the role of PLs in studies of nutrition and health. GPs can be divided into different “classes” depending on the polar headgroup at the sn-3 position of the glycerol backbone, such as phosphatidylethanolamine (PE), phosphatidylinositol (PI), phosphatidylcholine (PC), and phosphatidylserine (PS). Within each class, a range of molecular species exists due to different fatty acid substituents at sn-1 and 2 positions. In lipidomics, electrospray ionization−mass spectrometry (ESIMS) and tandem MS (ESI-MS/MS) have become powerful tools for the analysis of PL species.5−7 Since most PL molecules have a characteristic headgroup, lipidomic studies can target these features by using specific scanning modes, such as a © 2015 American Chemical Society

precursor ion scan of m/z 184 for PC and SM, and a neutral loss scan of 260 Da for PI. Such experiments can be used for both identification and quantification of these PL classes.8,9 Since ion suppression can occur in ESI when multiple species coelute and thus compete during the ESI process, an attempt must be made to optimize chromatographic separations of complex lipids mixtures. A long established strategy in PL separation is the use of normal phase liquid chromatography (NPLC) for class separation of PL prior to MS analysis. Chloroform, hexane, methanol, and 2-propanol were used as mobile phases for NPLC separations of PL classes, but in addition, electrospray compatible aqueous buffers such as ammonium hydroxide, formic acid, and ammonium formate were also added in the mobile phase to improve peak shape and ionization response.10−12 It should be pointed out here that the distinction between NPLC including water in the mobile phase and hydrophilic interaction chromatography is not always made in the literature. Therefore, in this work we refer to the terms used by the original authors, but it should be appreciated that in some cases these NPLC separations could be described as hydrophilic interaction chromatography (HILIC) separations. Even though the number of PL species that can be identified increases greatly after NPLC class separation, the lack of separation between species within each PL class means that only the major molecular species can be positively identified by NPLC-MS methods. Thus, two-dimensional LC (2D-LC) using NPLC for PL class separation combined with reverse phase LC (RPLC) for molecular species separation is highly Received: Revised: Accepted: Published: 1442

October 13, 2014 January 8, 2015 January 21, 2015 January 21, 2015 DOI: 10.1021/jf5049595 J. Agric. Food Chem. 2015, 63, 1442−1451

Article

Journal of Agricultural and Food Chemistry

that were used for the direct assignment of double bond positions in unsaturated PC species. In this study, the online 2D-LC method is further developed in order to allow for molecular species separation within the PI and PE classes, in addition to that already described for the PC class.25 In addition, we have also demonstrated for the first time the application of the in-line O3-MS method to the direct assignment of double bond positions in PI, PE, SM, lyso-PE (LPE), and LPC molecules. Overall, this approach can allow the structural characterization of PL molecules, including the definitive determination of each significant molecular species, as demonstrated here for the PLs extracted from chicken egg yolk.

desirable for the detailed analysis of a complex PL mixture. Offline NPLC × RPLC/MS was first used for the study of PL molecular species. In general, the PL extract was first separated by NPLC, then each fraction was collected off-line and injected onto a RPLC column separately for further molecular species elucidation.13,14 In order to solve the possible immiscibility problem between the mobile phases from NPLC and RPLC, a solvent-evaporating interface was used for online NPLC × RPLC/MS analysis.15 In the lipid profiling of a biological sample, Liu et al. used a vacuum pump at the interface of the NPLC and RPLC; in this way, the mobile phase from the NPLC was evaporated; then the mobile phase from RPLC passed through the interface and transferred the fraction into the second dimension C18 column for further molecular species separation.16 Later, HILIC was used as the first dimension LC for online 2D-LC separation of PLs.18,25 HILIC is a versatile tool for the separation of polar compounds; in addition, the aqueous−organic (usually acetonitrile) mobile phase used during HILIC separation is miscible with the second dimension RPLC mobile phase.17 For the analysis of a milk PL extract, 23 fractions from the first dimension HILIC column were collected and injected onto a C18 column for species separation via a 100 μL sample loop under the stop-flow mode.18 Further improvement of the stop-flow 2D-LC method was made by using an intermediate column to trap the components eluting from the HILIC column; these components were then eluted from the trap using a makeup flow.19 In general, 2D-LC/MS and -MS/MS analyses have revealed a large diversity in both PL classes and molecular species within each class. However, the identification of an specific molecular species may rely on a combination of LC/MS analysis with the separate identification of the overall fatty acid profile, as determined by gas chromatography coupled to a flame ionization detector (GC-FID)20,21 or from the literature.18,22,23 A necessary part of the full identification of fatty acids is the specific assignment of double bond positions. In GC-FID, this can often be achieved via the retention times of fatty acid methyl ester (FAME) derivatives. In other cases, the fatty acids from hydrolysis of PLs have been converted into dimethyloxazolines (DMOX) derivatives. Under electron ionization mass spectrometry, DMOX derivatives of fatty acids undergo charge remote fragmentations, and these fragmentation patterns can reveal double bond locations.11,24 However, in these GC/MS analyses of DMOX or FAME derivatives, the complete PL molecular structure is lost, and the derivatization procedures add additional complexity to the analysis. Recently, we successfully coupled online 2D-LC with in-line ozonolysis-MS (O3-MS) to allow the detailed structure determination of PC molecular species, as demonstrated for the identification of PC molecules in the PL extract from rat liver.25 The fraction containing PC species eluted from the first dimension HILIC column at between 8.2 and 9.0 min was directly injected onto the second dimension C18 column through a 10-port 2-position switching valve. The PC molecular species, eluting from the HILIC × C18 LC separation, then passed through the ozonolysis device that was connected between the C18 column and the ESI source. The ozonolysis device is a length of gas-permeable Teflon tube that passes through a chamber filled with ozone vapor, as described in detail by our previous study.26 The ozonolysis reaction occurred to all double bonds in the PC molecules, resulting in the formation of diagnostic ozonolysis product aldehydes

■

MATERIALS AND METHODS

Material. HPLC grade water, tetrahydrofuran (THF), chloroform, methanol, acetonitrile (ACN), and isopropanol (IPA) were purchased from Fisher Scientific Company (Ottawa, ON, Canada). HPLC grade ammonium formate and formic acid were obtained from Sigma (St. Louis, MO, USA). All of the standards PI(18:0/20:4 (n-6,9,12,15)), PE(18:0/20:4 (n-6,9,12,15)), SM(d18:1/18:1(n-9)) (“d” means dihydroxy base), LPE(18:1(n-9)), and LPC(18:1(n-9)) were purchased from Avanti polar lipids, Inc. (Alabaster, AL, USA). Each standard solution was prepared in methanol at a concentration of 200 μg/mL. The Teflon AF-2400 tubing (0.020″ OD, 0.010″ ID) was purchased from Biogeneral Inc. (San Diego, CA, USA). Nomenclature. We adopted as much as possible the nomenclature and abbreviation conventions from Liebisch et al.27 In this study, it is advantageous to locate the double bond by counting from the terminal methyl group, thus the “n-” terminology is used. For example, 1octadecanoyl-2-(9Z,12Z-octadecadienoyl)-sn-glycero-3-phosphocholine is expressed as PC(18:0/18:2 (n-6,9)). If the sn-position of the fatty acids is not known, the separator “_” is used, for example, PC(18:0_18:2 (n-6,9)). Extraction of PLs from Egg Yolk. Fresh eggs were purchased from local markets. The egg yolk was separated from the egg white, and was rolled on the filter membrane in order to completely remove the egg white. Six egg yolks were mixed well together and used for PL extraction according to a modified Bligh and Dyer method.23 In brief, 100 mg of egg yolk was mixed with 2 mL of extraction solvent (chloroform/methanol/water, 1:2:0.8) and homogenized at 10,000 rpm for 5 min on a Polytron PT1300D homogenizer (Kinematica AG, Switzerland) and then centrifuged at 3000 rpm for 5 min. The extraction procedure was repeated three times, and all of the supernatants were combined and made up to 10 mL using methanol. The extract was further diluted 10-fold with methanol prior to analysis. LC and MS Instrument. An Agilent 1200 series HPLC system (Agilent Technologies Inc., Palo Alto, CA, USA) coupled to a hybrid quadrupole time-of-flight mass spectrometer (QSTAR Elite, Applied Biosystems/MDS Sciex, Concord, ON, Canada) with ESI source was used for all of the LC-MS analysis. The ion source temperature was kept at 375 °C and the ionspray voltage was at 5500 V for ESI (+) and −4500 V for ESI (−). Source region nitrogen gas flows in arbitrary units assigned by the data system were as follows: curtain gas 25; auxiliary gas 10, and nebulizing gas 50. The declustering potential (DP), focus potential (FP), and DP2 were 40 V, 150 V, 10 V, and −40 V, and −150 V and −10 V for ESI (+) and ESI (−), respectively. Analyst QS 2.0 software was used for data acquisition and analysis. HILIC × C18 LC/MS Analysis. The 2D-LC configuration was the same as that in our previous study25 using a 10-port 2-position switching valve (Rheodyne, Rohnert Park, CA) as the interface between the HILIC column and the C18 column (see Figure S1, Supporting Information). Separation conditions were modified for the shorter analysis time. The Ascentis Express HILIC column (2.1 mm i.d. × 150 mm, 2.7 μm particles) (Sigma, St. Louis, MO) was used as the first dimension LC for PL class separation. Mobile phase A was ACN, and B was 10 mM ammonium formate in water at pH 3.0, and the flow rate was at 0.2 mL/min. The gradient was as follows: 0−10 min, from 8% to 30% B; 10.0−10.1 min, from 30% to 95% B; 10.1− 1443

DOI: 10.1021/jf5049595 J. Agric. Food Chem. 2015, 63, 1442−1451

Article


Figure 1. In-line O3-MS spectrum of PL standard (a) LPE(18:1(n-9)) (+ve); (b) LPC(18:1(n-9)) (+ve); (c) PE(18:0/20:4 (n-6,9,12,15))(+ve); (d) SM(d18:1/18:1(n-9)) (+ve); and (e) PI(18:0/20:4 (n-6,9,12,15)) (−ve). 15.0 min, held at 95% B; 15.0−15.1 min, from 95% to 8% B; and 15.1−20.0 min, held at 8% B. The Ascentis Express C18 column (2.1 mm i.d. × 75 mm, 2.7 μm particles, Supelco, Bellefonte, PA USA) was used as the second dimension RPLC. The mobile phase at 0.475 mL/ min was composed of THF/ACN/IPA/water with 10 mM ammonium formate at pH 3.0 (5:20:16:9, v/v/v) at 45 °C. The injection of the eluent from the HILIC column to the C18 column was achieved by switching the valve from the A to B position and back to the A position (Figure S1, Supporting Information). For the PI class, the fraction between 4.0 and 4.8 min from the first dimension HILIC column was transferred to the second dimension C18 column and detected under ESI (−). For PE and PC, the fractions between 5.6 and 6.4 min and 8.2−9.0 min were transferred onto the C18 column and analyzed under ESI (+). In this study, LPE, LPC, and SM were analyzed by one dimension HILIC/MS instead of 2D-LC/MS, as described below. In-line O3-MS Analysis. The method development of the in-line O3-MS approach for double bond position determination has been described in detail in our previous study.26 In this study, a 20 cm length of gas permeable and liquid impermeable Teflon tube passed through a 0.5 L glass chamber filled with oxygen and ozone gas (ozone concentration 56.2 g/m3) at room temperature. This ozonolysis device was placed in-line between 2D-LC and ESI source of the mass spectrometer for the analysis of PI, PE, and PC molecular species. For

the analysis of LPE, LPC, and SM species, the ozonolysis device was coupled in-line with the HILIC column. In this way, each PL species eluting from either the second dimension C18 column or the HILIC column passed through the semipermeable tube, ozone vapor immediately reacted with carbon−carbon double bonds existing in the molecules, and the ozonolysis products would be directly detected by ESI-MS.

■

RESULTS AND DISCUSSION

In this study, ESI-MS, MS/MS, and in-line O3-MS analyses were performed on PE, SM, PI, LPE, and LPC standards in order to demonstrate how these can yield similar structural information as was achieved for the analysis of PC in our previous study.25 Then we further developed the 2D-LC and 2D-LC/O3-MS methods for the direct identification of the PE, PI, SM, LPC, and LPE molecular species. Finally, these methods were applied to the characterization of the PL extracted from egg yolk. ESI-MS and MS/MS Analysis of PE, SM, LPE, LPC, and PI Standards. Flow injection analysis (FIA) was performed on PI(18:0/20:4 (n-6,9,12,15)), PE(18:0/20:4 (n-6,9,12,15)), SM(d18:1/18:1(n-9)), LPE(18:1(n-9)), and LPC(18:1(n-9)) 1444


Article


n-9 double bond in LPE. The mass loss of 110 Da after ozonolysis has also been observed in other lipids with double bonds located at the n-9 position such as methyl oleate26 and PC(16:0/18:1(n-9)).25 The ion at m/z 352 was probably due to the neutral loss of one H2O molecule from the ion at m/z 370. For the LPC(18:1(n-9)) standard, a similar mass loss of 110 Da from the intact molecular ion at m/z 522 was observed resulting in the ion at m/z 412 after in-line ozonolysis (Figure 1b). In the O3-MS spectrum of the PE(18:0/20:4 (n-6,9,12,15)) standard (Figure 1c), the ozonolysis product aldehyde ions at m/z 700, 68 Da mass lower than that of intact [M + H]+ ion at m/z 768, were generated by the ozonolysis of the double bond at the n-6 position. The product ions at m/z 660, 620, and 580 corresponded to the further oxidative cleavage of successive methylene interrupted double bonds. In addition, the neutral loss of one H2O molecule from the aldehyde ion at m/z 580 was observed at m/z 562. The O3-MS spectrum of the SM(d18:1/18:1(n-9)) standard (Figure 1d) contained the intact [M + H]+ ions at m/z 729 and the high intensity ozonolysis product ions at m/z 619, which resulted from ozonolysis of the double bond at the n-9 position in the N-18:1 chain. The low intensity ions at m/z 549, 180 Da lower mass than that of the [M + H]+ ions at m/z 729, were formed by ozonolysis of the 4,5-trans double bond located along the sphingosine (d18:1) chain. The ozonolysis of both double bonds in SM(d18:1/18:1(n-9)) resulted in the ions at m/z 439. The ozonolysis product aldehydes formed by oxidative cleavage of double bonds located along the N-acyl chain (m/z 619 in this case) can be used for the direct assignment of the double bond positions within the N-acyl chain. The in-line ozonolysis products of the PI(18:0/20:4 (n6,9,12,15)) standard were detected in negative ion mode, since the PI can be observed as [M − H]− ions. In the resulting O3MS spectrum (Figure 1e), intact [M − H]− ions at m/z 885 were still observed along with the ozonolysis product aldehydes ions at m/z 817, 777, 737, and 697, due to the oxidative cleavage of double bonds at the n-6,-9,-12,-15 positions. The same group of ozonolysis product aldehyde ions were also reported from the OzID-MS analysis of PI(18:0/20:4 (n6,9,12,15)) under ESI (−).30 Along with these characteristic ozonolysis aldehyde ions, another set of ions at m/z 851, 811, 771, and 731 were also observed. These ions have 34 Da higher mass than the corresponding deprotonated aldehydes ions at m/z 817, 777, 737, and 697. These ions might be attributed to the further reaction of “Criegee intermediates” that are zwitterions formed during the oxidative cleavage of each double bond. They are probably formed by the addition of a H2O molecule to the Criegee intermediates, resulting in the formation of hydroperoxide groups. As an example, a possible structure of the product ion at m/z 731 is shown in the inset of Figure 1e. The Criegee intermediates themselves were observed during the OzID-MS analysis of PI under ESI (−) as the carbonyl oxide ions at 16 Da are higher in mass than their corresponding aldehyde ions.30 Whether the Criegee intermediate adducts observed in the present study were generated during the in-line ozonolysis reaction or during the negative electrospray ionization is not clear, but they do not diminish the use of the ozonolysis product aldehydes for the definitive assignment of double bond positions in PI molecules. In summary, for LPE, LPC, PE and SM, both the intact molecules and the ozonolysis product aldehydes that are

standards in positive ion mode using a mobile phase of ACN/ water with 10 mM ammonium formate (85/15, v/v). The PE, SM, LPE, and LPC standards were observed as [M + H]+ ions with high abundance, whereas the PI [M + H]+ ion at m/z 887 appeared at very low intensity, and instead, the ion at m/z 627 generated from the neutral loss of 260 Da was the dominant ion. For PE and SM, mass measurements of the [M + H]+ ions were used to deduce their elemental compositions. However, the ESI (+)-MS/MS analyses of PE and SM [M + H]+ ions failed to provide fragment ions that could be used for the determination of fatty acyl chain composition. The dominant product ions in these MS/MS spectra arise due to the fragmentation of the headgroup, such as phosphocholine ions at m/z 184 for SM or the product ions from neutral loss of phosphoethanolamine (141 Da) from [M + H]+ ions of PE. This has also been observed by other studies.8,23,28 ESI(−)/MS analysis of PE(18:0/20:4 (n-6,9,12,15)) resulted in [M − H]− ions at m/z 766. MS/MS analysis of the ion at m/ z 766 generated carboxylate anions at m/z 283 and 303 that are characteristic of 18:0 and 20:4 fatty acyl chains (Figure S2a, Supporting Information). For SM(d18:1/18:1(n-9)), [M + HCOO]− ions at m/z 773 were observed, and the following MS/MS analysis of these ions (Figure S2b, Supporting Information) resulted in the demethylated molecular ion [M − CH3]− at m/z 713 as the dominant product ions. However, fragment ions directly corresponding to the N-acyl chain did not appear in the MS/MS spectrum. Instead, the ion at m/z 449 was observed, which was formed by the neutral loss of the N-acyl chain (18:1 ketene) from the demethylated molecular ion at m/z 713. Thus, the mass difference between the ion [M − CH3 − R1CO]− (m/z 449 in this case) and the demethylated molecular ion [M − CH3]− (m/z 713 in this case) can be used to assign the N-acyl chain. A similar fragmentation pattern to the SM [M + HCOO]− ion was also observed by another study and used for molecular species assignment of SM.9 In contrast to the weak [M + H]+ ions seen under positive ion mode, the PI(18:0/20:4 (n-6,9,12,15)) deprotonated molecule [M − H]− at m/z 885 was seen with high intensity under ESI (−). Thus, with the PI class, mass measurements of [M − H]− ions can be used for the determination of elemental compositions. The MS/MS spectrum of the ions at m/z 885 is presented in Figure S2c (Supporting Information). The [RCOO]− ions at m/z 283 and 303 correspond to the 18:0 and 20:4 fatty acyl chains, and in addition, the inositol phosphate ion at m/z 241 is present with high intensity. Other product ions at m/z 599 ([lysostearoyl PI−H]−) and 581 [lysostearoyl PI − H − H2O]−) are also observed, probably formed by the loss of the 20:4 acyl chain and the further loss of H2O. The product ion at m/z 419 is due to the further loss of inositol from the [lysostearoyl PI − H]− ion at m/z 599. A similar fragmentation pattern for PI under ESI(−)-MS/MS was also observed by other studies.10,29 In-line O3-MS Analysis of LPE, LPC, PE, SM, and PI Standards. For LPE and LPC, even though the chain length and number of double bonds can already be determined by the mass measurement of [M + H]+ ions, the position of double bonds along the chain cannot be assigned. Hence, in-line O3MS analyses of LPE(18:1(n-9)) and LPC(18:1(n-9)) standards were performed under ESI (+) (Figure 1a and b). The ozonolysis product of LPE(18:1(n-9)) was observed at m/z 370 as well as the intact [M + H]+ ion at m/z 480 (Figure 1a). The ion at m/z 370 corresponds to the characteristic ozonolysis product aldehyde that was formed by oxidative cleavage of the 1445


Article


Figure 2. TIC of HILIC/MS analysis of egg yolk PL extract: (a) detected under ESI (+ve); (b) detected under ESI (−ve).

H]+ ions. Under ESI(−) (Figure 2b), the PI class was also observed as [M − H]− ions at elution times between 4.0 and 5.0 min. The PE, LPE, PC, SM, and LPC classes could still be observed under negative mode but with about half of the intensity of that seen under ESI (+). Hence, the mass measurements of PE, LPE, PC, SM, and LPC [M + H]+ ions and PI [M − H]− ions were used to determine their elemental compositions. From the mass measurements of LPE and LPC [M + H]+ ions, the fatty acyl chain compositions could be directly assigned. In order to obtain information on the fatty acyl chain composition for PI, PE, PC, and SM, information dependent acquisition (IDA) mode was used during HILIC/ESI(−) analyses. During IDA, which is used to automatically trigger MS/MS scans, CID with CE of −35 and −42 eV were applied to ions with intensity higher than 10 cps. This results in the formation of fragment ions that are the deprotonated fatty acid residues [R1COO]− and [R2COO]−. Thus, the MS/MS spectra of [M − H]− ions of PE, PI, and the [M + HCOO]− ions of PC were used for the determination of their fatty acyl chain compositions. For SM, the mass difference between the fragment ions [M − CH3]− and [M − CH3 − R1CO]− in the MS/MS spectrum of [M + HCOO]− was used to determine the composition of the N-acyl chain. Following this, the composition of the sphingosine group in SM could be deduced. In this way, the fatty acid residue chain lengths and numbers of double bonds in PI, PE, PC, and SM were assigned even for isomeric species such as PE(18:0_18:3) and PE(18:1_18:2). These are listed in Table 1. 2D-LC/MS and 2D-LC/O3-MS Analysis of PI, PE, and PC Classes. In our previous study, HILIC × C18 LC, achieved by using a 10-port 2-position switching valve as the interface, was developed for online PC molecular species separation of a rat liver extract. In the present study, a C18 column with the same

characteristic of double bond positions are detected as [M + H]+ ions during O3 -ESI(+)/MS analysis. For PI, the deprotonated molecule can only be observed as [M − H]− ions under negative ion mode. The in-line ozonolysis products of PI are also detected under ESI (−) from which the aldehydes can still be used for the unambiguous assignment of double bond locations. Characterization of Egg Yolk PL Extract. Egg yolk is a good source of dietary PL, and recently, PL extracted from egg yolk has been introduced as a novel food in the EU.3 In addition, it is known that the egg yolk fatty acid profile can be modified through the hen’s diet. For example, the abundance of n-3 PUFA in egg yolk can be increased by enriching seal blubber oil in the diet of layers.10 The main classes of PLs in chicken egg yolk that have been reported include PI, PE, LPE, PC, SM, and LPC.10,23,31 In the following sections, HILIC/MS and HILIC/MS/MS analyses are used for PL class separation and for the determination of elemental and fatty acyl chain composition. Additionally, online 2D-LC/MS analyses were carried out for the separation of molecular species within each PI, PE, and PC class, and 2D-LC/O3-MS analysis was also performed to assign the double bond positions directly. Finally, HILIC was coupled to O3-MS (HILIC/O3-MS) in order to determine the double bond positions in LPE, LPC, and SM classes since the molecular species within these three classes are relatively simple and can be adequately separated on a HILIC column. HILIC/MS and HILIC/MS/MS Analysis. We have previously demonstrated the separation of PL classes using a HILIC column with gradient elution.10,23,31 Here, the egg yolk PL extract was analyzed by HILIC/ESI-MS under both positive and negative ion modes (Figure 2). As can be seen in the total ion current chromatogram (TIC) of HILIC/ESI(+)-MS (Figure 2a), the egg yolk PLs were well separated into PE, LPE, PC, SM, and LPC classes which formed abundant [M + 1446


Article

Journal of Agricultural and Food Chemistry Table 1. PL Species Identified in Egg Yolk PL Extract observed ions (m/z)

a

formula

mass accuracy (ppm)

833.52 857.52 859.53 861.55 863.57 885.55 887.57

C43H78O13P C45H78O13P C45H80O13P C45H82O13P C45H84O13P C47H82O13P C47H84O13P

−0.50 −0.50 −4.10 4.40 3.40 −1.50 0.20

716.52 718.54 740.52

C39H75NO8P C39H77NO8P C41H75NO8P

−0.30 −3.50 0.80

742.54

C41H77NO8P

−0.60

744.55 746.57 764.53 768.56 792.55

C41H79NO8P C41H81NO8P C43H75NO8P C43H79NO8P C45H79NO8P

−0.30 1.20 4.80 1.80 −0.50

732.55 734.57 756.56 758.57 760.59 762.60 780.55 782.57 784.58

C40H79NO8P C40H81NO8P C42H79NO8P C42H81NO8P C42H83NO8P C42H85NO8P C44H79NO8P C44H81NO8P C44H83NO8P

−0.10 −3.00 1.60 −1.50 1.20 −1.20 −0.40 −0.10 −2.40

786.60

C44H85NO8P

−2.50

788.62 806.57 808.58


3.90 0.20 −2.60

810.60 812.62 834.60


0.20 −1.60 −2.00

454.29

C21H45NO7P

0.40

466.33 480.31 482.32


0.60 −1.20 −0.40

703.57 705.59 731.61 813.68

C39H80N2O6P C39H82N2O6P C41H84N2O6P C47H94N2O6P

−1.70 3.10 2.10 0.20

496.34

C24H51NO7P

0.40

522.35

C26H53NO7P

−0.50

524.37

C26H55NO7P

−1.10

molecular species

ECN

PI [M − H]− 34:2 16:0_18:2 30 36:4 16:0_20:4 28 36:3 16:0_20:3 30 36:2 18:0_18:2 32 36:1 18:0_18:1 34 38:4 18:0_20:4 30 38:3 18:0_20:3 32 PE [M + H]+ 34:2 16:0_18:2 30 34:1 16:0_18:1 32 36:4 16:0_20:4 28 18:2_18:2 28 36:3 18:0_18:3 30 18:1_18:2 30 36:2 18:0_18:2 32 36:1 18:0_18:1 34 38:6 16:0_22:6 26 38:4 18:0_20:4 30 40:6 18:0_22:6 28 PC [M + H]+ 32:1 16:0_16:1 30 32:0 16:0_16:0 32 34:3 16:0_18:3 28 34:2 16:0_18:2 30 34:1 16:0_18:1 32 34:0 16:0_18:0 34 36:5 16:0_20:5 26 36:4 16:0_20:4 28 36:3 16:0_20:3 30 18:1_18:2 30 36:2 18:1_18:1 32 18:0_18:2 32 36:1 18:0_18:1 34 38:6 16:0_22:6 26 38:5 18:0_20:5 28 18:1_20:4 28 38:4 18:0_20:4 30 38:3 18:0_20:3 32 40:6 18:0_22:6 28 LPE [M + H]+ sn-2−16:0 sn-1−16:0 18:1-eLPE 18:1 sn-2−18:0 sn-1−18:0 SM [M + H]+ d18:1/16:0 d18:0/16:0 d18:1/18:0 d18:1/24:1 LPC [M + H]+ sn-2−16:0 sn-1−16:0 sn-2−18:1 sn-1−18:1 sn-2−18:0 sn-1−18:0

tR (min)

double bond position n-

ozonolysis product aldehyde ions (m/z)

7.37 7.16 7.60 8.30 9.43 8.11 8.60

n-6,9 N/A n-6,9,12 n-6,9 n-9 n-6,9,12,15 n-6,9,12

765,725 N/Aa 791,751,711 793,753 753 817,777,737,697 819,779,739

10.14 11.36 9.91 9.39 10.53 10.21 11.73 13.58 9.56 11.38 10.89

n-6,9 n-9 n-6,9,12,15 N/A n-6,9,12 N/A n-6,9 n-9 n-3,6,9,12,15,18 n-6,9,12,15 n-3,6,9,12,15,18

648,608 608 672,632,592,552 N/Aa 674,634,594 N/Aa 676,636 636 738,698,658,618,578,538 700,660,620,580 766,726,686,646,606,566

11.86 12.86 11.43 12.02 12.93 14.52 10.95 11.84 12.27 12.07 13.01 13.26 14.68 11.57 12.21 11.88 12.98 13.67 12.59

n-9 saturated n-3,6,9 n-6,9 n-9 saturated N/A n-6,9,12,15 n-6,9,12 n-9/n-9,12 n-9/n-9 n-6,9 n-9 n-3,6,9,12,15,18 N/A N/A n-6,9,12,15 n-6,9,12 n-3,6,9,12,15,18

622 saturated 730,690,650 690,650 650 saturated N/Aa 714,674,634,594 636,676,716 674/606,566 676/566 718,678 678 780,740,700,660,620,580 N/Aa N/Aa 742,702,662,622 744,704,664 808,768,728,688,648,608

7.72 7.87 7.47 7.78 7.57 7.73

saturated saturated n-9 n-9 saturated saturated

saturated saturated 356 370 saturated saturated

9.84 9.73 9.72 9.43

saturated saturated saturated n-9

saturated saturated saturated 703

10.54 10.72 10.40 10.60 10.32 10.50

saturated saturated n-9 n-9 saturated saturated

saturated saturated 412 412 saturated saturated

The intensity of ozonolysis product aldehyde ions is too low to observe. 1447


Article


ions at m/z 793 and 753 corresponded to ozonolysis product aldehydes from the oxidative cleavage of double bonds at n-6 and −9 positions along the 18:2 acyl chain. For other PI molecular species in this sample, the ozonolysis product aldehydes along with the products at 34 Da or higher are also observed and used for the determination of all double bond positions (Table 1). For the molecular species separation of the PE class, the fraction eluting at between 5.6 and 6.4 min from the HILIC column (Figure 2a) was transferred onto the C18 column. Although a small degree of PE species separation was achieved on the HILIC column (Figure 2a), the diversity of the PE molecular species was far better resolved after 2D-LC separation (Figure 4a). Since PE [M + H]+ ions under ESI

stationary phase but only half the length (75 mm) was used along with a modified gradient (see Materials and Methods) in order to shorten the second dimension separation time from 30 to 20 min, with only minor effect on the efficiency of separation. Using this arrangement, 2D-LC/MS analysis was applied to the molecular species separation of PI, PE, and PC classes from egg yolk PL extract. 2D-LC/MS analysis coupled to in-line ozonolysis was also performed for more detailed structural investigation of each eluting molecule within these 3 classes. In the 2D-LC separation of PI molecular species, the PI fraction eluting from the HILIC column was transferred onto the C18 column by switching the valve to position B at 4.0 min and back to position A at 4.8 min (see Materials and Methods and Figure S1, Supporting Information). Compared to the first dimension HILIC separation (Figure 2b), the molecular species of PI were much better separated on the second dimension C18 column (Figure 3a). The observed molecular species and

Figure 4. (a) TIC of 2D-LC/MS analysis of the PE class in egg yolk PL extract in positive ion mode; (b) O3 -MS spectrum of PE(18:0_22:6) in the egg yolk sample. Figure 3. (a) TIC of 2D-LC/MS analysis of the PI class in egg yolk PL extract in negative ion mode; (b) O3-MS spectrum of PI(18:0_18:2) in the egg yolk sample.

(+) have higher intensity than the [M − H]− ions seen under ESI (−), 2D-LC/MS analysis was carried out in positive ion mode. The first eluting species was PE(16:0_22:6) (ECN = 26) at 9.56 min, and the rest of the PE molecular species eluted generally according to their ECN, ranging from 26 to 34. However, one exception was observed for PE(18:0_22:6) with an ECN of 28 eluting at 10.89 min, which was retained longer than PE(16:0_18:2) (tR 10.14 min) with an ECN of 30. Similar retention order exceptions for PLs containing a PUFA chain were also observed during 2D-LC separation of PC in rat liver in our previous study.25 Even though the PE species containing PUFA such as PE(16:0_22:6), PE(18:0_22:6), PE(16:0_20:4), and PE(18:0_20:4) can be relatively well separated from other species, no more detailed structure information can be derived from just the 2D-LC/MS analysis to identify them as n-3 or n-6 PUFA. Therefore, 2D-LC/O3-MS analysis was also performed on these PE species. As an example, the O3-MS spectrum of

their corresponding tR are listed in Table 1. The elution order of these PI species was consistent with their equivalent carbon number (ECN), meaning the PI molecules with higher ECN were retained longer on the C18 column. In order to assign the specific fatty acyl chains in these PI species, 2D-LC was coupled to O3-MS in the negative ion mode. As an example of the results, the O3-MS spectrum of PI(18:0_18:2) in this egg yolk PL extract is shown in Figure 3b. In addition to the intact [M − H]− ions at m/z 861, ozonolysis product ions appeared at m/z 753, 787, 793, and 827. The ions at m/z 787 and 827, that are 34 Da higher in mass than the ions at m/z 753 and 793, were generated by H2O addition to the corresponding Criegee intermediates, consistent with results from the example of inline ozonolysis of a PI(18:0/20:4) standard (Figure 1e). The 1448


Article


5b), two peaks appear at 13.01 and 13.26 min, indicating that they are PC(18:1_18:1) and PC(18:0_18:2) isomers. In the O3-MS spectrum of the peak at 13.01 min (Figure 5c), ozonolysis product ions were observed at m/z 676 along with the intact molecular ion at m/z 786, which directly assigned the double bond at n-9 in one of the 18:1 acyl chains. The other ozonolysis product ion at m/z 566 was generated from the further ozonolysis of the double bond in the other chain, and the mass loss of 110 Da from m/z 676 also located the second double bond at the n-9 position in the second 18:1 acyl chain. Also seen in the O3-MS spectrum of the peak eluting at 13.01 min (Figure 5c) is an abundant [M + H]+ ion at m/z 760 from PC(16:0_18:1), along with its ozonolyis product ion at m/z 650. However, the partial coelution of this species does not affect the assignment of double bond positions in PC(18:1_18:1) at 13.01 min. In Figure 5d, the O3-MS spectrum of the isomeric species at m/z 786 which elutes at 13.26 min is shown. From this spectrum, the ozonolysis product ions at m/z 718 and 678 allow assignment of the double bond positions at n-6 and 9 in PC(18:0_18:2). HILIC/MS Analysis of LPE, LPC, and SM Class. Initially, similar to the 2D-LC/MS analysis of PI, PE, and PC classes, LPE eluting between 7.2 and 8.0 min and LPC eluting between 10.2 and 11.0 min from the HILIC column were also switched onto the C18 column and eluted using the same isocratic conditions. However, all of the species in both the LPE and LPC classes had very little retention and almost eluted as one single peak on the C18 column. The low retention of the LPE and LPC classes on the C18 column was also observed in the off-line 2D-LC study, in which LPE and LPC classes from the HILIC column were collected and injected onto a C18 column separately for molecular species separation.32 However, it can be seen by the insets of Figure 2a that significant molecular species separation of LPE and LPC has already been achieved on the HILIC column. The accurate mass measurement of the ion at m/z 454 identified it as LPE(16:0). In the XIC of the ion at m/z 454 from Figure 2a (not shown), two peaks at 7.73 and 7.87 min are present, representing regio-isomers with the 16:0 acyl chain at either sn-1 or sn-2 position. The regio-isomers of LPE(18:0) at m/z 482 were also observed at 7.57 and 7.73 min. Although unambiguous assignment of regio-isomers cannot be achieved by our study, it was shown earlier that sn-1 LPE and LPC were retained longer than the corresponding sn-2 isomers on a HILIC column.32 Thus, the assignment of sn-isomers presented is solely based on the retention time on the HILIC column in our study (Table 1). We also observed low intensity ions at m/z 466 from LPE species eluting at a tR of 7.47 min. The accurate mass measurement at m/z 466.3295 assigned the elemental composition of these LPE molecules as C23H49NO6P+, which excluded the possibility that this ion represented the isobaric LPE(17:1). Therefore, this LPE species was tentatively identified as either eLPE(18:1) which has the long alkyl chain bound as a 1-O-alkyl ether, or as pLPE(18:0) with the chain bound as 1-O-alk-1′-enyl (note that plasmenyl- or p(18:0) refers to the vinyl ether form, which is derived from the saturated 18:0 but includes a double bond at C-1 and hence is isobaric with eLPE(18:1)). However, the exact structure of this LPE species could not be determined because the MS/MS spectra in both positive and negative ionizations failed to provide product ions that could be used for this specific assignment. In order to differentiate between these two ether lipids, other work has used acidic hydrolysis performed on the ether PL before MS analysis since the double bond in 1-O-

PE(18:0_22:6) in the sample is shown in Figure 4b. The mass difference of 26 Da between the intact PE(18:0_22:6) [M + H]+ ion at m/z 792 and the ozonolysis product ion at m/z 766 directly assigns the first double bond at the n-3 position. The successive ozonolysis product ions at m/z 726, 686, 646, 606, and 566, with 40 Da mass differences between each other, demonstrate that as expected, all of the double bonds are separated only by a methylene group. Thus, PE(18:0_22:6) in egg yolk contains the n-3 docosahexaenoic acid chain. The PC class eluting at between 8.2 and 9.0 min from the HILIC column was also directed onto the C18 column for the further separation of molecular species. The major PC species are indicated in Figure 5a; other minor PC species are listed in

Figure 5. (a) TIC of 2D-LC/MS analysis of PC class in egg yolk PL extract in positive ion mode; (b) XIC of m/z 786 from (a); (c) O3-MS spectrum of the peak at 13.01 min; (d) O3-MS spectrum of the peak at 13.26 min.

Table 1 due to their partial coelution with the major species. Compared to PI and PE, the PC class has an even higher level of species diversity, with ECNs that ranged from 26 to 34. PC molecular species also elute based on their ECN, with the exception of the PC molecule with PUFA such as PC(16:0_22:6) and PC(18:0_22:6), which have greater retention than some PC species with higher ECNs. In-line ozonolysis was also performed on PC species separated by 2D-LC, and the diagnostic ozonolysis product aldehydes were detected as [M + H]+ ions. The isomers of PC 36:2 in the egg yolk is used as an example to illustrate how the double bond positions were assigned using in-line ozonolysis. In the extracted ion chromatogram (XIC) of m/z 786 (Figure 1449


Article

Journal of Agricultural and Food Chemistry alk-1′-enyl linkage is unstable to acid treatment.33 It was also previously reported that product ions from the loss of R1OH can only be observed from the MS/MS analysis of the pLPC [M+Li]+ ions and not from the MS/MS analysis of eLPC [M + Li]+ ions.34 For the LPC class, regio-isomers of LPC(16:0), LPC(18:0), and LPC(18:1) were also observed in their XIC traces of m/z 496, 524, and 522, respectively. As with LPE, the identification of regio-isomers only relied on their retention time on the HILIC column (Table 1). For the SM class, even though there are two substituents in the molecules besides the polar headgroup, the SM fraction eluting at between 9.2 and 10.0 min from the HILIC column had poor retention and resolution under the current C18 separation. However, the SM molecular species were separated well on the HILIC column (the inset of Figure 2a). SM(d18:1/ 24:1) eluted at 9.43 min and was separated from SM(d18:1/ 16:0) (tR 9.84 min) and SM(d18:1/18:0) (tR 9.72 min); only SM(d18:0/16:0) coeluted with SM(d18:1/18:0) at 9.73 min. HILIC/O3-MS Analysis of LPE, LPC, and SM Class. While molecular species separation of LPE, LPC, and SM classes was achieved on the HILIC column, the double bond positions in unsaturated species such as LPE(18:1), LPC(18:1), and SM(d18:1/24:1) could still not be assigned. Therefore, the in-line ozonolysis device was coupled with the HILIC column (HILIC/O3-MS). In this way, the ozonolysis products from each eluting LPE, LPC, and SM molecule was detected by ESI (+)-MS. For example, in Figure S3a (Supporting Information) an ion at m/z 370 appeared in the O3-MS spectrum of LPE(18:1) along with the intact [M + H]+ ions at m/z 480. This mass loss of 110 Da allows the assignment of the only double bond in this LPE to the n-9 position. Also seen in Figure S3a (Supporting Information) is an ion at m/z 454, which is due to the coelution of LPE(16:0). However, this does not compromise the value of the diagnostic ozonolysis product ion for the determination of double bond position in LPE(18:1). In the O3-MS spectrum averaged at between 7.45 and 7.50 min (see Figure 2a and Table 1), the [M + H]+ ion at m/z 466 was still observed but was accompanied by an ozonolysis product ion at m/z 356. The latter ion locates the double bond to the n9 position, which in turn confirms that the ion at m/z 466 corresponds to the [M + H]+ ion of eLPE(18:1(n-9)) rather than to pLPE(18:0). Hence, in this study the differentiation between possible eLPE and pLPE isomers was achieved using O3-MS analysis, although further experiments would be needed in order to validate this as a general approach. Figure S3b (Supporting Information) is the O3-MS spectrum of the SM (d18:1/24:1) at m/z 813 from the egg yolk sample. The double bond along the N-24:1 chain was assigned to the n9 position based on the observation of the ozonolysis product ion at m/z 703, 110 Da below the [M + H]+ ion at m/z 813. The ion observed at m/z 523 is the ozonolysis product aldehyde generated by the subsequent oxidative cleavage of the 4,5-trans double bond located along the 18:1 sphingosine group. An analogous pattern of ozonolysis product ions was seen in the O3-MS spectrum of the SM(d18:1/18:1(n-9)) standard (Figure 1d). In the above discussions, it has been demonstrated that for the egg yolk PL extract, all of the molecular species in PI, PE, and PC can be unambiguously identified using HILIC/MS and MS/MS analysis followed by 2D-LC/O3-MS. It has also been shown that LPE, LPC, and SM molecular species can be accurately assigned using HILIC/MS and HILIC/O3-MS analyses. Our findings have revealed even more of the diversity

of PL species that are present in egg yolk, compared to that in previous reports.10,23 Furthermore, the methods that we have used have allowed for the definitive identification of each PL species with the position of the double bond directly assigned. In summary, we have developed a comprehensive PL characterization method which involves using a combination of HILIC/MS, HILIC/MS/MS, 2D-LC/MS, and O3-MS analyses. The use of in-line ozonolysis coupled with a HILIC column and 2D-LC is necessary for the elucidation of the structure of the chains within the molecules. The experimental setup is very easily adapted to a range of LC and MS configurations, which is especially helpful for the characterization of PL, a group of chemically diverse compounds. While the use of shotgun lipidomics can reveal a great diversity of lipid structures, the minor species and most isomers can only be observed with chromatographic separation prior to MS detection, and other analytical techniques are often required to confirm species identification. The comprehensive method demonstrated here has shown great success, as exemplified by the detailed structure elucidation of egg yolk PL molecules. The additional level of structural detail for lipid analyses that can be generated by this approach will be complementary to other experimental methods used in lipidomics.

■

ASSOCIATED CONTENT

S Supporting Information *

HILIC×C18 LC/O3-MS configuration; ESI (−)-MS/MS analysis of PL standard; and O3-MS spectrum of LPE(18:1) and SM(d18:1/24:1). This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Clinical Science Building 10-102, Department of Laboratory Medicine & Pathology, University of Alberta, Edmonton, Alberta T6G 2G3, Canada. Tel: 780-492-5096. E-mail: [email protected] Funding

This work was supported by an NSERC Discovery grant 197273 awarded to J.M.C. and by a Quality Food for Health grant to J.M.C. and coinvestigators jointly funded by Alberta Livestock and Meat Agency, Alberta Innovates Biosolutions, and the Alberta Egg Producers. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Vance, D. E.; Vance, J. E. Biochemistry of Lipids, Lipoproteins and Membranes; Elsevier Science B.V.: Amsterdam, The Netherlands, 2008. (2) Amate, L.; Gil, A.; Ramirez, M. Feeding infant piglets formula with long-chain polyunsaturated fatty acids as triacylglycerols or phospholipids influences the distribution of these fatty acids in plasma lipoprotein fractions. J. Nutr. 2001, 131, 1250−1255. (3) Küllenberg, D.; Taylor, L. A.; Schneider, M.; Massing, U. Health effects of dietary phospholipids. Lipids Health Dis. 2012, 11, 1−16. (4) Bunea, R.; El, F. K.; Deutsch, L. Evaluation of the effects of Neptune Krill Oil on the clinical course of hyperlipidemia. Altern. Med. Rev. 2004, 9, 420−428. (5) Han, X.; Gross, R. W. Structural determination of picomole amounts of phospholipids via electrospray ionizaion tandem mass spectrometry. J. Am. Soc. Mass Spectrom. 1995, 6, 1202−1210. (6) Smith, M.; Jungalwala, F. B. Reversed-phase high performance liquid chromatography of phosphatidylcholine: a simple method for 1450


Article

Journal of Agricultural and Food Chemistry determining relative hydrophobic interaction of various molecular species. J. Lipid Res. 1981, 22, 697−704. (7) Pulfer, M.; Murphy, R. C. Electrospay mass spectrometry of phospholipids. Mass Spectrom. Rev. 2003, 22, 332−364. (8) Ekroos, K.; Chernushevich, I. V.; Simons, K.; Shevchenko, A. Quantitative profiling of phospholipids by multiple precursor ion scanning on a hybrid quadrupole time-of-flight mass spectrometer. Anal. Chem. 2002, 74, 941−949. (9) Houjou, T.; Yamatani, K.; Nakanishi, H.; Imagawa, M.; Shimizu, T.; Taguchi, R. Rapid and selective identification of molecular species in phosphatidylcholine and sphingomyelin by conditional neutral loss scanning and MS3. Rapid Commun. Mass Spectrom. 2004, 18, 3123− 3130. (10) Pacetti, D.; Boselli, E.; Hulan, H. W.; Frega, N. G. High performance liquid chromatography-tandem mass spectrometry of phospholipid molecular species in eggs from hens fed diets enriched in seal blubber oil. J. Chromatogr., A 2005, 1097, 66−73. (11) Berdeaux, O.; Juaneda, P.; Martine, L.; Cabaret, S.; Bretillon, L.; Acar, N. Identification and quantification of phosphatidylcholines containing very-long-chain polyunsaturated fatty acid in bovine and human retina using liquid chromatography/tandem mass spectrometry. J. Chromatogr., A 2010, 1217, 7738−7748. (12) Brouwers, J. F. Liquid chromatographic−mass spectrometric analysis of phospholipids. Chromatography, ionization and quantification. Biochim. Biophys. Acta 2011, 1811, 763−775. (13) Houjou, T.; Yamatani, K.; Imagawa, M.; Shimizu, T.; Taguchi, R. A shotgun tandem mass spectrometric analysis of phospholipids with normal-phase and/or reverse-phase liquid chromatography/ electrospray ionization mass spectrometry. Rapid Commun. Mass Spectrom. 2005, 19, 654−666. (14) Sato, Y.; Nakamura, T.; Aoshima, K.; Oda, Y. Quantitative and wide-ranging profiling of phospholipids in human plasma by twodimensional liquid chromatography/mass spectrometry. Anal. Chem. 2010, 82, 9858−9864. (15) Tian, H.; Xu, J.; Xu, Y.; Guan, Y. Multidimensional liquid chromatography system with an innovative solvent evaporation interface. J. Chromatogr., A 2006, 1137, 42−48. (16) Nie, H.; Liu, R.; Yang, Y.; Bai, Y.; Guan, Y.; Qian, D.; Wang, T.; Liu, H. Lipid profi ling of rat peritoneal surface layers by online normal- and reversed-phase 2D LC QToF-MS. J. Lipid Res. 2010, 51, 2833−2844. (17) Jandera, P. Stationary and mobile phases in hydrophilic interaction chromatography: a review. Anal. Chim. Acta 2011, 692, 1−25. (18) Dugo, P.; Fawzy, N.; Cichello, F.; Cacciola, F.; Donato, P.; Mondello, L. Stop-flow comprehensive two-dimensional liquid chromatography combined with mass spectrometric detection for phospholipid analysis. J. Chromatogr., A 2013, 1278, 46−53. (19) Shi, X.; Qiao, L.; Lu, X.; Xu, G. A novel stop-flow twodimensional liquid chromatography−massspectrometry method for lipid analysis. J. Chromatogr., A 2013, 1321, 65−72. (20) Grandois, J. L.; Marchiono, E.; Zhao, M.; Giuffrida, F.; Ennahar, S.; Bindler, F. Investigation of natural phosphatidylcholine sources: Separation and identification by liquid chromatography-electrospray ionization-tandem mass spectrometry (LC -ESI-MS2) of molecular species. J. Agric. Food Chem. 2009, 57, 6014−6020. (21) Montealegre, C.; Verardo, V.; Gómez-Caravaca, A. M.; GarcíaRuiz, C.; Marina, M. L.; Caboni, M. F. Molecular characterization of phospholipids by high-performance liquid chromatography combined with an evaporative light scattering detector, high-performance liquid chromatography combined with mass spectrometry, and gas chromatography combined with a flame ionization detector in different oat varieties. J. Agric. Food Chem. 2012, 60, 10963−10969. (22) Shen, Q.; Wang, Y.; Gong, L.; Guo, R.; Dong, W.; Cheung, H. Shotgun lipidomics strategy for fast analysis of phospholipids in fisheries waste and its potential in species differentiation. J. Agric. Food Chem. 2012, 60, 9384−9393. (23) Zhao, Y.; Xiong, Y.; Curtis, J. M. Measurement of phospholipids by hydrophilic interaction liquid chromatography coupled to tandem

mass spectrometry: The determination of choline containing compounds in foods. J. Chromatogr., A 2011, 1218, 5470−5479. (24) Spitzer, V. Structure analysis of fatty acids by gas chromatography-low resolution electron impact mass spectrometry of their 4,4-dimethyloxazoline derivatives - a review. Prog. Lipid Res. 1997, 35, 387−408. (25) Sun, C.; Zhao, Y.; Curtis, J. M. Elucidation of phosphatidylcholine isomers using two dimensional liquid chromatography coupled inline with ozonolysis mass spectrometry. J. Chromatogr., A 2014, 1351, 37−45. (26) Sun, C.; Zhao, Y.; Curtis, J. M. The direct determination of double bond positions in lipid mixtures by liquid chromatography/inline ozonolysis/mass spectrometry. Anal. Chim. Acta 2013, 762, 68− 75. (27) Liebisch, G.; Vizcaíno, J. A.; Köfeler, H.; Trötzmüller, M.; Griffiths, W. J.; Schmitz, G.; Spener, F.; Wakelam, M. J. O. Shorthand notation for lipid structures derived from mass spectrometry. J. Lipid Res. 2013, 54, 1523−1530. (28) Norris, C.; Fong, B.; MacGibbon, A.; McJarrow, P. Analysis of phospholipids in rat brain using liquid chromatography-mass spectrometry. Lipids 2009, 44, 1047−1054. (29) Hus, F. F.; Turk, J. Characterization of phosphatidylinositol,phosphatidylinositol-4-phosphate, and phosphatidylinositol-4,5-bisphosphate by electrospray ionization tandem mass spectrometry: A mechanistic study. J. Am. Soc. Mass Spectrom. 2000, 11, 986−999. (30) Thomas, M. C.; Mitchell, T. W.; Harman, D. G.; Deeley, J. M.; Nealon, J. R.; Blanksby, S. J. Ozone-induced dissociation: Elucidation of double bond position within mass-selected lipid ions. Anal. Chem. 2008, 80, 303−311. (31) Xiong, Y.; Zhao, Y.; Goruk, S.; Oilund, K.; Field, C. J.; Jacobs, R. L.; Curtis, J. M. Validation of an LC−MS/MS method for the quantification of choline-related compounds and phospholipids in foods and tissues. J. Chromatogr., B 2012, 911, 170−179. (32) Lísa, M.; Cífková, E.; Holcapek, M. Lipidomic profiling of biological tissues using off-line two dimensional high-performance liquid chromatography−mass spectrometry. J. Chromatogr., A 2011, 1218, 5146−5156. (33) Chen, S.; Li, K. W. Mass Spectrometric Identification of molecular species of phosphatidylcholine and lysophosphatidylcholine extracted from shark liver. J. Agric. Food Chem. 2007, 55, 9670−9677. (34) Hus, F. F.; Turk, J.; Thukkani, A. K.; Messner, M. C.; Wildsmith, K. R.; Ford, D. A. Characterization of alkylacyl, alk-1enylacyl and lyso subclasses of glycerophosphocholine by tandem quadrupole mass spectrometry with electrospray ionization. J. Mass Spectrom. 2003, 38, 752−763.

1451


Characterization of Phospholipids by Two-Dimensional Liquid

Recommend Documents