MS Method for Polyphosphoinositide Analyses

Anal. Chem. 2008, 80, 9226–9232

Reversed-Phase LC/MS Method for Polyphosphoinositide Analyses: Changes in Molecular Species Levels during Epidermal Growth Factor Activation in A431 Cells Hideo Ogiso and Ryo Taguchi* Department of Metabolome, Graduate School of Medicine, The University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan, and CREST, Japan Science and Technology Agency, Kawaguchi, Saitama 332-0012, Japan In studies on lipid metabolomics, liquid chromatography/ mass spectrometry (LC/MS) is a robust and popular technique. Although effective reversed-phase (RP) LC/ MS methods enabling the separation of phospholipid molecular species have been developed, RPLC methods to analyze phosphatidylinositol phosphates (PIPs) have not been reported. In this study, we developed conditions suitable for PIP analysis. Coupled with (diethylamino)ethyl (DEAE)-cellulose pretreatment, at least 1 pmol each of phosphatidylinositol monophosphates (PIP1), bisphosphates (PIP2), and triphosphates standards per ∼6 × 106 cultured cells could be measured. Using these methods, we detected elevated concentrations of more than 12 PIP1 species in epidermal growth factor (EGF)stimulated A431 cells, a human epidermoid carcinoma cell line. The PIP2 species detected were not elevated after stimulation. We also detected EGF-induced increases in the levels of several phosphatidic acid species using another set of methods. Our method sensitively determined PIPs within a biological sample and is thus suitable for analysis of phoisphoiniositide metabolism. Phospholipids (PLs) are the main components of cell membranes and perform important biological functions. PLs have a glycerophosphate backbone structure: the sn-3 carbon makes a phosphoester linkage to a polar headgroup, and the sn-1/2 carbon primarily forms a carboxylic acid ester (acyl) linkage to fatty acids with various side chains (called “molecular species”), such as arachidonate (abbreviated as 20:4), linoleate (18:2), and oleate (18: 1). The fatty acid structure in the sn-1/2 positions can differ in various tissues, cells, organelles, and membrane domains. Before the common use of mass spectrometry (MS), PLs were mainly measured by radioisotope (RI) labeling on thin-layer chromatography (TLC). At present, however, the biological importance of individual molecular species has led to increased focus on analytical methods for their characterization.1 Thus, effective MS methods, suitable for PL species analysis, have been developed and used in lipidomics.2-7 Although phosphatidylinositol mono* Corresponding author. Phone: +81 3 5841 3650. Fax: +81 3 5841 3430. E-mail: [email protected]. (1) Pulfer, M.; Murphy, R. C. Mass Spectrom. Rev. 2003, 22, 332–364. (2) Kim, H. Y.; Wang, T. C.; Ma, Y. C. Anal. Chem. 1994, 66, 3977–3982.

9226

Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

phosphates (PIP1), bisphosphates (PIP2), and triphosphates (PIP3) are minor components of membrane phospholipids, they act as precursor second messengers, such as inositol-1,4,5triphosphates (IP3) and diacylglycerol (DAG), and interact directly with various proteins involved in signal transduction.8 More specifically, phosphatidylinositol phosphates (PIPs) and their metabolites are versatile signaling components, regulating many cellular events, such as membrane fusion, vesicle transport, cell migration and polarity, apoptosis, oncogenesis, and calcium mobilization.9-12 A recent report demonstrated that PIP2 in the nucleus regulates gene expression.13 As PIPs are involved in many important cellular functions, their synthesis and degradation must be strictly regulated by a variety of metabolic enzymes, such as phosphoinositide kinases, phosphoinositide phosphatases, and phospholipases C (PLC). In many previous studies, PIPs were measured by RI labeling on TLC, or normal-phase (NP) liquid chromatography (LC) and ion-exchange LC after deacylation.14-18 In cytochemistry, antibodies or affinity proteins for the specific PIP classes are often used.19-23 Although these methods are very effective for trace quantification and spatiotemporal analysis of PIPs, respectively, (3) Ekroos, K.; Chernushevich, I. V.; Simons, K.; Shevchenko, A. Anal. Chem. 2002, 74, 941–949. (4) Han, X.; Yang, J.; Cheng, H.; Ye, H.; Gross, R. W. Anal. Biochem. 2004, 330, 317–331. (5) Houjou, T.; Yamatani, K.; Imagawa, M.; Shimizu, T.; Taguchi, R. Rapid Commun. Mass Spectrom. 2005, 19, 654–666. (6) Murph, M.; Tanaka, T.; Pang, J.; Felix, E.; Liu, S.; Trost, R.; Godwin, A. K.; Newman, R.; Mills, G. Methods Enzymol. 2007, 433, 1–25. (7) Brugger, B.; Erben, G.; Sandhoff, R.; Wieland, F. T.; Lehmann, W. D. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 2339–2344. (8) Lemmon, M. A. Nat. Rev. Mol. Cell Biol. 2008, 9, 99–111. (9) Di Paolo, G.; De Camilli, P. Nature 2006, 443, 651–657. (10) Halstead, J. R.; van Rheenen, J.; Snel, M. H.; Meeuws, S.; Mohammed, S.; D’Santos, C. S.; Heck, A. J.; Jalink, K.; Divecha, N. Curr. Biol. 2006, 16, 1850–1856. (11) Takenawa, T.; Itoh, T. IUBMB Life 2006, 58, 296–303. (12) De Matteis, M. A.; Di Campli, A.; Godi, A. Biochim. Biophys. Acta 2005, 1744, 396–405. (13) Mellman, D. L.; Gonzales, M. L.; Song, C.; Barlow, C. A.; Wang, P.; Kendziorski, C.; Anderson, R. A. Nature 2008, 451, 1013–1017. (14) Munnik, T.; de Vrije, T.; Irvine, R. F.; Musgrave, A. J. Biol. Chem. 1996, 271, 15708–15715. (15) Traynor-Kaplan, A. E.; Thompson, B. L.; Harris, A. L.; Taylor, P.; Omann, G. M.; Sklar, L. A. J. Biol. Chem. 1989, 264, 15668–15673. (16) Vadnal, R. E.; Parthasarathy, R. Biochem. Biophys. Res. Commun. 1989, 163, 995–1001. (17) Halstead, J. R.; Roefs, M.; Ellson, C. D.; D’Andrea, S.; Chen, C.; D’Santos, C. S.; Divecha, N. Curr. Biol. 2001, 11, 386–395. 10.1021/ac801451p CCC: $40.75  2008 American Chemical Society Published on Web 10/29/2008

they cannot provide structural information related to the fatty acids. As some previous studies have suggested that PIPs function may depend on acyl structure, the distinctive molecular species of PIPs need to be analyzed quantitatively in order to fully understand the role of individual molecular species.24,25 To our knowledge, this is very difficult.26-28 Wenk et al.29 and Milne et al.25 have separately reported the MS analyses of the PIP species in biological samples. Wenk et al. showed differential changes in PIP1 species in yeast mutants deficient in enzymes involved in PIP1 metabolism. Furthermore, they detected elevated concentrations of specific PIP2 species in human fibroblasts from patients with Lowe syndrome. They did not report any PIP3 species in that study. On the other hand, Milne et al. showed that various PIP3 species were generated after differential agonist stimulation of macrophage (or macrophage-like) cells. As these two methods using MS rely on direct infusion analyses, ion suppression effects may affect the reliability of quantification. More recently, Pettitt et al. developed an NPLC/MS method to resolve PIP regioisomers (e.g., PI-4,5-P2, PI-3,4-P2, and PI-3,5-P2).30 They showed that PI3,5-P2 quintupled after salt-induced hyperosmotic stress in yeast and that PI-3,4,5-P3 (18:0/20:4) was present at significantly low levels after the thrombin stimulation of human platelets. The NPLC/MS method is able to resolve regioisomers, resembling the TLC method. But several problems in the development of MS assays for low-abundant PIPs have been demonstrated,31 and elution together with major and minor species of the same regioisomer may result in ion suppression effects, thus leading to underestimation of minor components. Reversed-phase (RP) LC is excellent for the separation of molecular species with different acyl structures and for detection of minor components. However, among PL classes, acidic phospholipids, such as phosphatidic acid (PA) and phosphatidylserine (PS), tend to elute as broad peaks under the RPLC conditions generally used in lipidomics studies. Furthermore, PIPs are not detected at all by RPLC. Thus, we recently developed an RPLC/ (18) Sasaki, T.; Irie-Sasaki, J.; Jones, R. G.; Oliveira-dos-Santos, A. J.; Stanford, W. L.; Bolon, B.; Wakeham, A.; Itie, A.; Bouchard, D.; Kozieradzki, I.; Joza, N.; Mak, T. W.; Ohashi, P. S.; Suzuki, A.; Penninger, J. M. Science 2000, 287, 1040–1046. (19) Rusten, T. E.; Stenmark, H. Nat. Methods 2006, 3, 251–258. (20) van Rheenen, J.; Song, X.; van Roosmalen, W.; Cammer, M.; Chen, X.; Desmarais, V.; Yip, S. C.; Backer, J. M.; Eddy, R. J.; Condeelis, J. S. J. Cell Biol. 2007, 179, 1247–1259. (21) Araki, N.; Egami, Y.; Watanabe, Y.; Hatae, T. Exp. Cell Res. 2007, 313, 1496–1507. (22) van Horck, F. P.; Lavazais, E.; Eickholt, B. J.; Moolenaar, W. H.; Divecha, N. Curr. Biol. 2002, 12, 241–245. (23) Sato, M.; Ueda, Y.; Takagi, T.; Umezawa, Y. Nat. Cell Biol. 2003, 5, 1016– 1022. (24) Schmid, A. C.; Wise, H. M.; Mitchell, C. A.; Nussbaum, R.; Woscholski, R. FEBS Lett. 2004, 576, 9–13. (25) Milne, S. B.; Ivanova, P. T.; DeCamp, D.; Hsueh, R. C.; Brown, H. A. J. Lipid Res. 2005, 46, 1796–1802. (26) Gunnarsson, T.; Ekblad, L.; Karlsson, A.; Michelsen, P.; Odham, G.; Jergil, B. Anal. Biochem. 1997, 254, 293–296. (27) Hsu, F. F.; Turk, J. J. Am. Soc. Mass Spectrom. 2000, 11, 986–999. (28) Nakamura, T.; Hatori, Y.; Yamada, K.; Ikeda, M.; Yuzuriha, T. Anal. Biochem. 1989, 179, 127–130. (29) Wenk, M. R.; Lucast, L.; Di Paolo, G.; Romanelli, A. J.; Suchy, S. F.; Nussbaum, R. L.; Cline, G. W.; Shulman, G. I.; McMurray, W.; De Camilli, P. Nat. Biotechnol. 2003, 21, 813–817. (30) Pettitt, T. R.; Dove, S. K.; Lubben, A.; Calaminus, S. D.; Wakelam, M. J. J. Lipid Res. 2006, 47, 1588–1596. (31) Postle, A. D.; Wilton, D. C.; Hunt, A. N.; Attard, G. S. Prog. Lipid Res. 2007, 46, 200–224.

MS method for improved detection of PA and PS.32 PIPs are still not measured by this improved method, probably because of the inability to remove undesired interactions with the RP column. Development of methods for separation of PIP species using RPLC is challenging but is necessary for detailed study. In the present study, we developed a sensitive method to determine PIP1, PIP2, and PIP3 with different fatty acid compositions, although regioisomers were not separated. We designed an acidified extraction and a (diethylamino)ethyl (DEAE)-cellulose treatment to obtain a PIP-rich fraction from the lipid extracts, and used RPLC/MS to separate them. Using these methods, we detected elevated concentrations of more than 12 PIP1 species in epidermal growth factor (EGF)-stimulated A431 cells, a human epidermoid carcinoma cell line. PIP2 species were also detected and were not elevated by EGF stimulation. Using another set of LC conditions, we also showed that EGF induced an increase in the levels of several PA species. Thus, the present method will be advantageous for the study of phosphoinositide metabolism in cellular responses and pathology. EXPERIMENTAL SECTION Materials. Phosphatidylinositol-4-phosphate (dipalmitoyl) ammonium salt [abbreviated as 16:0/16:0-PI-4-P1 ammonium salt or PIP1(16:0/16:0)], 16:0/16:0-PI-4,5-P2 ammonium salt [PIP2(16:0/ 16:0)], and 16:0/16:0-PI-3,4,5-P3 sodium salt [PIP3(16:0/16:0)] were obtained from Cayman Chemical (Ann Arbor, MI). 8:0/8: 0-PI-4,5-P2 sodium salt [PIP2(8:0/8:0)] was obtained from Wako Pure Chemicals (Osaka, Japan). 17:0-Lysophosphatidic acid sodium salt [LPA(17:0)] and 14:0/14:0-phosphatidic acid sodium salt [PA(14:0/14:0)] were purchased from Avanti Polar Lipids (Alabaster, AL) and Sigma-Aldrich (St. Louis, MO), respectively. Human epidermoid carcinoma A431 cells were purchased from American Tissue Culture Collection (Rockville, MD). Recombinant human EGF was purchased from Pepro Tech Inc. (Rocky Hill, NJ). All solvents were of HPLC or LC/MS grade, and other chemical reagents were of analytical grade. They were obtained from Wako Pure Chemicals. Ultrapure water was obtained from a Milli-Q water system (Millipore, Milford, MA). All glassware, such as microsyringes, stoppered test tubes, sample vials, glass wools, and Pasteur pipettes, was silanized by immersion in a 5% dimethyldichlorosilane solution (Wako Pure Chemicals), and metalware, such as microsyringes, was rinsed with 10 mM phosphoric acid, followed by methanol, before use. Phospholipid Solutions. PIP1(16:0/16:0), PIP2(16:0/16:0), and PIP3(16:0/16:0) were dissolved in methanol to prepare 1 pmol/µL solutions. LPA(17:0) and PA(14:0/14:0) were also dissolved in methanol to prepare 1 pmol/µL solutions. These standard solutions were stored at -70 °C. Before use, each solution was diluted to the desired concentration with methanol. PIP2(8:0/8:0) was dissolved in methanol/chloroform (9:1) to give a 1 nmol/µL solution, and this was stored at -70 °C. Cell Culture and Treatment. A431 cells were cultured in Dulbecco’s modified Eagle’s medium (Sigma-Aldrich) supplemented with 10% fetal bovine serum (Gibco/Invitrogen, Carlsbad, CA), 100 units/mL penicillin, and 100 µg/mL streptomycin (Nachalai Tesque, Kyoto, Japan) at 37 °C in a humidified atmosphere with 5% CO2. Cells were grown on a 100 mm plate to (32) Ogiso, H.; Suzuki, T.; Taguchi, R. Anal. Biochem. 2008, 375, 124–131.


9227

a confluence of 30% to ∼80% and were then starved for 24 h in serum-free medium supplemented with 1 mg/mL bovine serum albumin (Cohn Fraction V, Wako Pure Chemicals), followed by stimulation with 10 nM human EGF for the indicated period (0, 0.5, 2, 5, or 30 min). Incubation was terminated by replacing medium with ice-cold phosphate-buffered saline, followed by addition of ice-cold organic solvent (methanol or butanol). Extraction of Phospholipids. For PIP analyses, cultured cells (3 to ∼6 × 106 per 100 mm plate) were scraped with 2 mL of ice-cold methanol, and 50 pmol of PIP1(16:0/16:0), PIP2(16:0/ 16:0), and PIP3(16:0/16:0) were added as internal standards. Then, 2 µL of 1 nmol/µL PIP2(8:0/8:0) as an adsorption protectant, 2 mL of 1 N HCl, 0.15 mL of 2 M NaCl, and 2 mL of chloroform were added to each suspension, followed by vigorous shaking for 3 min while cooling with ice. Subsequent procedures were carried out at room temperature. Samples were centrifuged at 1000g for 5 min, and the lower layer was collected. After addition of 1.5 mL of methanol, each sample was applied to a DEAE-cellulose (Wako Pure Chemicals) column. For acidic phospholipid analyses, cultured cells (3 to ∼6 × 106 per 100 mm plate) were scraped with 2 mL of ice-cold butanol, and 50 pmol of LPA(17:0) and PA(14: 0/14:0) were added as an internal standard. Then, 2 mL of water was added to each suspension, followed by vigorous shaking for 3 min while cooling with ice. Subsequent procedures were preformed at room temperature. Samples were centrifuged at 1000g for 5 min, and the upper layer was collected. After the addition of 1 mL of methanol and 1 mL of chloroform, each sample was applied to a DEAE-cellulose column. For phosphatidylcholine (PC) analysis, the butanol extraction was performed as described above. The obtained butanol solution was dried under diminished pressure at 25 °C, followed by redissolving with 0.5 mL of 2-propanol/methanol/water (5:1:4). Fractionation Using a DEAE-Cellulose Column. A 500 µL bed of DEAE-cellulose was prepared as described previously,32 except the NaOH form was used instead of the acetic acid form. For PIP analyses, a stepwise elution was performed as follows. Column-bound lipids were washed with chloroform (1 mL), chloroform/methanol (1:1) (3 × 1 mL), and chloroform/methanol/ 28% aqueous ammonia/acetic acid (200:100:3:0.9) (3 × 1 mL) in series. Highly acidic lipids, including PIPs, LPA, and LPS, etc., were eluted with chloroform/methanol/HCl (6:6:1) (3 × 1 mL). After addition of 1.5 mL of water and 0.113 mL of 2 M NaCl, the solution was vigorously shaken and centrifuged to collect the lower layer. After addition of 2 µL of 1 nmol/µL PIP2(8:0/8:0), each sample was dried under nitrogen gas and was redissolved in 0.05 mL of methanol/70% ethylamine (100:0.065), followed by addition of 0.03 mL of 1 M ammonium bicarbonate containing 0.2 mM phosphoric acid. The obtained PIP-rich fractions were analyzed within a day of preparation. For analyses of other acidic phospholipids, stepwise elution was performed as described previously,32 except with modification to effectively recover LPA. Column-bound lipids were washed with chloroform (1 mL) and chloroform/methanol (1:1) (3 × 1 mL) in series. Acidic lipids, containing PA, PS, and phosphatidylinositol (PI), and their lyso forms, etc., were eluted with chloroform/ methanol/28% aqueous ammonia/acetic acid (200:100:10:6.7) (3 × 1 mL). After removing organic solvent from each eluate under nitrogen gas, 0.05 mL of water and 0.1 mL of 2-propanol were 9228


added to about 0.05 mL of residue. Fractionated samples were stored at -70 °C until use. Reversed-Phase Liquid Chromatography. The LC system consisted of two Shimadzu LC10Avp-micro pumps, a DGU-14AM degassing unit, an SCL-10Avp controller (Shimadzu, Kyoto, Japan), and an HTC PAL autosampler (CTC Analysis, Zwingen, Switzerland) equipped with a nonmetallic injector unit and a 20 µL PEEK tubing sample loop. To wash a portion of the injector, two rinse solvents were used: the first rinse solvent was 2-propanol/water (1:1) supplemented with 0.1% formic acid, 0.0 14% ammonia and 0.05% phosphoric acid, and the second was methanol. After the injector was washed with these rinse solvents several times, the sample (typically 5 µL) was injected. The LC system was controlled with Chromeleon software (Dionex, Sunnyvale, CA). For PIP analyses, lipids were separated on a Waters X-Bridge C8 (3.5 µm, 150 mm × 1.0 mm i.d.) column (Waters, Milford, MA). Mobile phase A was methanol/water/70% ethylamine (50:50:0.13). Mobile phase B was 2-propanol/70% ethylamine (100:0.13). The gradient was 95%/5% (A/B) to 10%/90% (A/B) over 15 min, after which it was held for 1 min, before returning to 95%/5% (A/B) over 1 min, followed by holding for 10 min for reequilibration. Flow rate was 30 µL/min, and the chromatography was performed at 25 °C. Typically, 5 µL of sample was injected. Other acidic phospholipid analyses were preformed as described previously.32 Briefly, lipids were separated on a Waters X-Bridge C18 (3.5 µm, 150 mm × 1.0 mm i.d.) column (Waters). Mobile phase A was 2-propanol/methanol/water (5:1:4) supplemented with 0.2% formic acid, 0.028% ammonia, and 5 µM phosphoric acid. Mobile phase B was 2-propanol supplemented with 0.2% formic acid and 0.028% ammonia. Gradient elution was performed as described previously, except for that 10%/90% (A/B) was held for 10 min instead of 5 min. Mass Spectrometry. The LC system described above was coupled online to an LTQ Orbitrap mass spectrometer (ThermoFisher Scientific, Waltham, MA) equipped with an electrospray ionization (ESI) source. Ion spray voltage was set to -4.0 kV in negative ion mode. The heated capillary temperature was set to 350 °C for the PIP analyses or 300 °C for other PL analyses. Other parameters were set according to the manufacturer’s recommendations. The MS system was controlled using Xcalibur software. Phospholipids were measured by a full scan of parent ion (MS1) within m/z 380-1500 in negative ion mode using an Orbitrap FT-MS (Fourier transform mass spectrometry) analyzer with a resolution of 60 000. Mass accuracy was within 3 ppm. MS/ MS (MS2) measurements at a collision energy setting of 35% were performed with an LTQ-IT (ion trap) analyzer when needed. Each molecular species was identified by the exact mass value and LC retention time. Furthermore, the PIP1-, PIP2-, and PIP3-specific fragments (m/z 321, 401, and 481, respectively) observed in MS2 were used for identification. When the specific fragment was not observed, we did not identify the PIP species. RESULTS AND DISCUSSION Development of Reversed-Phase LC/MS Conditions for PIP Analysis. With the use of the traditional method, in which RI-labeled PIPs were separated by TLC, total amounts of each class of PIP1, PIP2, PIP3, and some of their regioisomers have often been estimated. As MS is a potent technique in lipid metabolomics, MS coupled with NPLC having the same separation

Figure 2. Recovery rates and calibration curves verifying the developed method. PIPs were measured by a full scan of MS1 on an Orbitrap FT-MS analyzer. Standard PIP1, PIP2, and PIP3 were added to a cell suspension. Recovery rates were examined at 50 pmol/plate. The three sets of experiments for examining the recovery rate were performed in different days.

Figure 1. Reconstructed ion current (RIC) chromatograms of standard and naturally occurring PIPs species on RPLC/MS. Cell lysate added inner standard PIPs was analyzed. Each [M - H]- ion was detected by an Orbitrap FT-MS analyzer. TIC: total ion current.

mode as TLC, has been increasing used for PIP analyses.30 To date, PIPs have not been measured by the RPLC method because of undesired adsorption onto the RP columns, such as C18. In this study, we developed RPLC conditions suitable for MS analyses of individual PIP species. We first examined standard PIP1(16: 0/16:0), PIP2(16:0/16:0), and PIP3(16:0/16:0) solutions using a C18 column. When an alkaline mobile phase containing 10 mM ethylamine (pH > 11) was used with an alkaline-resistant C18 column, PIPs were eluted as narrow peaks. However, significant adsorption of PIPs on the C18 column made quantification difficult. We found that these PIPs were eluted without adsorption onto an alkaline-resistant C8 column. LC/MS conditions were further optimized to measure the PIP standards and were then developed as described in the Experimental Section. The temperature of the heated capillary of the ESI source was raised to 350 °C in order to detect monovalent anions of PIPs. Although the extracted lipids were eluted within a short time on the RPLC, due to the mobile phase over pH 11, lipid species were eluted without peak tailings (Figure 1). After optimization the elution peaks became very narrow, but PIP regioisomers were still not separated. On the other hand, we speculate that an alkaline mobile phase of more than pH 11 causes unstable ESI, judging from the fact that intensities of the other phospholipids detected under the same conditions vary. To overcome this issue, the mean value of two measurements was needed. Lipid Extraction from Cultured Cells and Sample Pretreatment. As Pettitt et al. described previously,30 use of silanized glassware is essential for handling of PIPs, particularly when working with dilute solutions. PIP extraction was performed by the commonly utilized HCl-acidified extraction method.14,15,29,33,34 (33) Nasuhoglu, C.; Feng, S.; Mao, J.; Yamamoto, M.; Yin, H. L.; Earnest, S.; Barylko, B.; Albanesi, J. P.; Hilgemann, D. W. Anal. Biochem. 2002, 301, 243–254.

Figure 3. Two-dimensional map of retention times (RT) vs m/z values of PIPs in A431 cells stimulated with EGF for 2 min. The PIPrich fraction was measured by a full scan of MS1 on an Orbitrap FTMS analyzer. Additional lines were drawn to indicate relationships between acyl structure and RT.

When the lipid extract from cultured cells was directly applied to the RPLC/MS method developed here, PIPs were not effectively measured because of their affinity with other insoluble lipids, thereby resulting in insolubility in mobile phase A. Furthermore, at least one of the PIP species exhibited ion suppression by coelution with lysophosphatidylethanolamine. Thus, we developed a pretreatment procedure using solid-phase extraction with DEAEcellulose for PIP fractionation.35 After the total lipid extract was loaded onto the DEAE-cellulose column, stepwise elution was initiated with chloroform in order to wash out neutral lipids. Among the bound lipids, basic lipids such as PC and phosphatidylethanolamine were eluted with chloroform/methanol (1:1), and acidic lipids such as PA, PS, phosphatidylglycerol, phosphatidylinositol, and cardiolipin were then eluted with chloroform/ (34) Dawson, R. M.; Eichberg, J. Biochem. J. 1965, 96, 634–643. (35) Kiselev, G. V. Biochim. Biophys. Acta 1982, 712, 719–721.


9229

Figure 4. PIP1 and PIP2 molecular species profiles for EGF-stimulated A431 cells. After addition of EGF, A431 cells were incubated at 37 °C for 0 to 30 min. PIP species in the PIP-rich fraction were measured by a full scan of MS1 on an Orbitrap FT-MS analyzer. (A) Profile of the PIP1 species. (B) Profile of PIP2 species. Values represent means ( standard deviation of triplicate determinations.

Figure 5. PIP3 species detected after EGF stimulation for 2 min. (A) Selective detections by exact mass of PIP3(36:2) (m/z 1101.449), the PIP3-specific fragment ion (m/z 481), and neutral loss of H3PO4 (m/z 98). PLs were measured with an LTQ-IT analyzer and an Orbitrap FT-MS analyzer. (B) MS/MS spectrum for precursor ion of m/z 1101.4 [PIP3(36:2)]. (C) Quantification of PIP3 species. Concentrations were estimated based on ratio to PIP3(16:0/16:0) and by detection of neutral loss of m/z 98, which corresponds to the retention time when the m/z 481 fragment was detected. Values represent means ( standard deviation of triplicate determinations.

methanol/water/28% aqueous ammonia/acetic acid (200:100:3: 0.9). Finally, highly acidic lipids such as PIPs, lysophosphatidylserine (LPS), and LPA were eluted with chloroform/methanol/HCl (6: 6:1). To remove HCl, water was added to the eluate, followed by the liquid-liquid extraction. Each collected chloroform solution was dried under nitrogen gas and was redissolved in 0.05 mL of methanol/70% ethylamine (100:0.065) with sonication. As PIPs are 9230


labile in alkaline solution, the solution was neutralized with 0.03 mL of 1 M ammonium bicarbonate aqueous solution supplemented with 0.2 mM phosphoric acid. PIPs in this solution were stably measured after at least a day in a polypropylene (PP) vial on an autosampler maintained at 10 °C. When methanol was used as a solvent to redissolve the sample, PIP recovery was impaired and reproducibility decreased because of adsorption onto the glass

Figure 6. Molecular species profiles of PA, PI, and PC for A431 cells. After addition of EGF, A431 cells were incubated at 37 °C for 0 to 30 min. PA and PI species in the acidic PL-rich fraction were measured by a full scan of MS1 on an Orbitrap FT-MS analyzer. PC species were only measured in control cells. Values represent means ( standard deviation of triplicate determinations.

wall, despite silanization. In addition, the high salt concentration was essential to avoid adsorption onto the injector portion. Development of quantitative analysis for PIPs was difficult due to their undesired adsorption onto the surface of resin packed in the separation column and other materials, such as metal and glass. In practice, standard PIP3 methanol solution (1pmol/3 mL) was dried under nitrogen gas and then was redissolved in methanol in a silanized glass tube. The recovery rate of this treatment is only 7%. When the standard PIP3 solution supplemented with 2 nmol of PIP2 (8:0/8:0) was dried and redissolved, the rate was increased up to 60%, which is still a loss of 40%. If methanol containing 0.1 M HCl is used as a redissolving solvent, higher recovery is expected, but this solvent would disrupt LC/ MS measurement. To reduce the undesired adsorption of PIPs, we have taken various countermeasures, such as rinsing metals with phosphoric acid, using an alkaline mobile phase coupled with a C8 stationary phase, adding ammonium bicarbonate and phosphoric acid to the sample redissolved in alkaline solvent, and using PP vials, in addition to using silanized glassware, as described in the Experimental Section. We believe that the adsorptive properties cause the relatively low recoveries, around 50% (described below), even after optimization. On the other hand, for other acidic phospholipid analyses, a butanol extraction method was used.6 Other fractions obtained from the HCl-acidified extraction described above were not used for the PL analyses, as we were concerned about acid-induced degradation, such as the undesired conversion of lysophosphatidylcholine to LPA.36 Total lipid extract was applied to the DEAEcellulose column. After washing out the basic lipids with chloroform/methanol (1:1), acidic lipids were eluted with chloroform/ methanol/water/28% aqueous ammonia/acetic acid (200:100:10: 6.7). In order to verify these conditions, we determined the recovery rate and constructed calibration curves with respect to each (36) Ishida, M.; Imagawa, M.; Shimizu, T.; Taguchi, R. J. Mass Spectrom. Soc. Jpn. 2005, 53, 217–226.

standard lipid, whose content is vanishingly small in mammalian cells. A total of 50 pmol of PIP1(16:0/16:0), PIP2(16:0/16:0), and PIP3(16:0/16:0) were added to untreated A431 cells (∼6 × 106 cells/plate) before and after the extraction process, which included organic solvent extraction and DEAE-cellulose fractionation. These two types of sample were measured by the LC/MS method, and peak intensities were compared in order to determine the rate of lipid recovery. The calculated recovery rates were 55% for PIP1, 55% for PIP2, and 42% for PIP3 (Figures 1 and 2). To construct calibration curves, 0, 1, 2, 5, 20, 50, and 100 pmol of PIPs were added to untreated cultured cells (∼6 × 106 cells/plate) before extraction without addition of an internal standard. Approximate linear correlations between peak intensities and additive amounts were observed (Figure 2), except for PIP3 at less than 5 pmol, as PIP2(8:0/8:0), which was added as adsorption-protective agent, included a few picomoles of PIP3(16:0/16:0) as an impurity. When a full scan measurement using the FT-MS analyzer of the LTQ Orbitrap was used for detection, the expected detection limit of each PIPs was lower than 1 pmol/plate. Although the recovery rates are relatively low and inconsistent in interday experiments, the linearity achieved without using an inner standard in intraday experiment indicates that our method is suitable for practical use with respect to the application of biological samples including an internal standard (Figure 2). In the same way, the conditions for the other acidic PL analyses were verified using standard LPA(17: 0) and PA(14:0/14:0) solutions. For both PLs, the recovery rates were sufficiently high, more than 90%, excellent linear correlations were achieved, and the expected detection limit was lower than 1 pmol/plate. Molecular Species Analysis of PIPs in A431 Cells Stimulated with EGF. For application to biological samples, we analyzed the changes in the levels of lipid mediators in A431 cells stimulated with EGF. EGF reportedly alters the levels of PIP1, PIP2, PIP3, and PA in A431 cells,20,21,37,38 but data regarding its effects on individual molecular species have not been reported. (37) Pike, L. J.; Eakes, A. T. J. Biol. Chem. 1987, 262, 1644–1651.


9231

The question of whether specific molecular species are involved in the response of cells to stimulation has attracted much attention. First, we examined the PIP species in A431 cells (3 × 106 cells) after 10 nM EGF treatment for 0, 0.5, 2, 5, or 30 min. As shown in Figures 1 and 3, naturally occurring PIPs species were separately detected. On a two-dimensional map of retention times (RT) versus m/z values, approximately linear relationships were observed between acyl chain length/unsaturation degree and RT, as is the way with general RP separation of PLs. The profile for PIP1 was essentially same as that for PIP2 (Figure 4). When quantification of individual PIP1 and PIP2 species was performed based on the ratio to PIP1 and PIP2(16:0/16:0), added as internal standards, the cellular concentrations of the major 36:2 species of PIP1 and PIP2 were 40 and 65 pmol per 3 × 106 cells, respectively. On the other hand, the major PIP species was 38:4 in animal samples.25,30 This difference is probably due to the culture medium containing fewer polyunsaturated fatty acids, such as arachidonic acid. The EGF-stimulated increase was 2-fold of control levels in almost all PIP1 species that peaked at 2 min and failed to alter the levels of almost all PIP2 species. Considering that PIP2 hydrolysis producing IP3 is activated upon EGF stimulation,20 PIP2 species content may be kept constant by hydrolysis of newly synthesized and unbound PIP2. In the first study using MS1 detection by the FT-MS analyzer, any PIP3 species were not detected, even after EGF stimulation. We then specifically examined PIP3 species in A431 cells (6 × 106 cells) after 10 nM EGF treatment for 0 and 2 min using multiple reaction monitoring for the detection of specific PIP3 species. PIP3 species was detected based on the PIP3-specific fragment at m/z 481 and the neutral loss of phosphoric acid (Figure 5, parts A and B). The detected structure was a 36:2containing species at significantly low levels near the detection limit, whereas other species were not detected in the A431 cells (Figure 5). This 36:2-containing species was the major component of the PIP1 and PIP2 classes (Figure 5C). The cellular contents of PIP3 species generated after stimulation was estimated to be about 0.3 pmol per 6 × 106 cells, which is only 0.1% that of the corresponding PIP2 species. Stimulation resulted in a 2-fold increase in the quantity of this PIP3(36:2) species. These results apparently suggest that the PIP metabolism activated by EGF stimulation is not significantly dependent on acyl structures. As regioisomers cannot be separately analyzed in our study, the possibility of separate increases in regioisomers is undeniable. The observations that total PIP1 contents increased after EGF stimulation and that total PIP2 contents do not increase were in good agreement with earlier results obtained using other methods.38 (38) Dadabay, C. Y.; Patton, E.; Cooper, J. A.; Pike, L. J. J. Cell Biol. 1991, 112, 1151–1156. (39) Wang, X.; Devaiah, S. P.; Zhang, W.; Welti, R. Prog. Lipid Res. 2006, 45, 250–278. (40) Cook, S. J.; Wakelam, M. J. Biochem. J. 1992, 285 (1), 247–253. (41) Kaszkin, M.; Seidler, L.; Kast, R.; Kinzel, V. Biochem. J. 1992, 287 (1), 51–57. (42) Slaaby, R.; Jensen, T.; Hansen, H. S.; Frohman, M. A.; Seedorf, K. J. Biol. Chem. 1998, 273, 33722–33727. (43) Zhao, C.; Du, G.; Skowronek, K.; Frohman, M. A.; Bar-Sagi, D. Nat. Cell Biol. 2007, 9, 706–712.

9232


Molecular Species Analysis of Phosphatidic Acids in A431 Cells Stimulated with EGF. PLC hydrolysis of PIP2 forms both IP3 and DAG, and the generated DAG is rapidly phosphorylated by DAG kinase to produce PA in the polyphosphoinositide (PPI) cycle.39 Furthermore, PA is spatiotemporally produced as a result of EGF-induced phospholipase D2 (PLD2) activation.40-43 Thus, we further explored the effects of EGF stimulation on each PA species. For this purpose, we designed a butanol extraction method for total lipids containing acid-labile and hydrophilic phospholipids, a DEAE-cellulose treatment to remove neutral and alkaline lipids from the extracts, and RPLC/MS for acidic PL analysis. When quantification of individual PA species was performed based on the ratio to PA(14:0/14:0), added as an internal standard, the cellular concentrations of the major PA(34: 2) and PA(36:2) species were 42 and 43 pmol per 3 × 106 cells, respectively (Figure 6). The EGF-stimulated increase was about 1.7-fold of control levels in most PA species at 2 min, except for 36:1-containing species. Elevation of PA levels remained for 30 min (Figure 6). The observation that total PA contents increase after EGF stimulation was largely in agreement with earlier results using other methods.37 The 38:4 species were scarcely detected, although cellular concentrations of the same species of PIP1 and PIP2 were relatively high at 26 and 36 pmol per 3 × 106 cells, respectively (Figure 4). And the molecular species profiles of PA were similar to those of PC. These data suggest that PLD2 heavily contributes to PA production in A431 cells. The profiles of PIP1/2 species without stimulation (Figure 4, parts A and B) closely resemble that of PI species (Figure 6). However, these PI and PIP species were separately extracted and analyzed using different methods, and give similar profiles, thus suggesting that certain PIP species, for example, saturated fatty acid containing species, are not preferentially isolated by the methods developed in this study. On the other hand, the fact that the molecular species profile of PI marginally differed from those of PIP1/2 (Figures 4 and 6) may suggest that PIP1 production is partly dependent on the acyl structure. Further studies are required to investigate these issues. CONCLUSION To date, use of the RP separation mode has been considered inappropriate for PIP analysis; we have developed an RPLC/MS method for molecular species analyses of PIPs. Our LC/MS method, coupled with an effective extraction procedure and the DEAE-cellulose pretreatment, substantially reduced the ion suppression effect. As a result, when compared with the previous method using MS, the sensitive analysis of PIPs is possible. The present methods will be advantageous for studies into phosphoinositide metabolism in cellular responses and pathology. ACKNOWLEDGMENT This study was performed with the help of Special Coordination funds from the Ministry of Education, Culture, Sports, Science and Technology of the Japanese Government. Received for review July 13, 2008. Accepted September 29, 2008. AC801451P

MS Method for Polyphosphoinositide Analyses

Recommend Documents