The enzyme catalysing the conversion of PE (phosphatidylethanolamine) into PC (phosphatidylcholine), PEMT (PE N-methyltransferase), exists as two isoforms, PEMT-L (longer isoform of PEMT) and PEMT-S (shorter isoform of PEMT). In the present study, to compare the functions of the two isoforms of PEMT, we established HEK (human embryonic kidney)-293 cell lines stably expressing PEMT-L and PEMT-S. Both PEMT-L and PEMT-S were localized in the ER (endoplasmic reticulum). PEMT-L, but not PEMT-S, was N-glycosylated with high-mannose oligosaccharides. The enzymatic activity of PEMT-S was much higher than that of PEMT-L. By using novel enzymatic assays for measuring PC and PE, we showed that PEMT-L and PEMT-S expression remarkably increased the cellular PC content, whereas the PE content was decreased by PEMT-S expression, but was hardly affected by PEMT-L expression. The cellular content of phosphatidylserine was also reduced by the expression of PEMT-L or PEMT-S. MS analyses demonstrated that the expression of PEMT-S led to more increases in the molecular species of PC and PC-O (ether-linked PC) with longer polyunsaturated chains than that of PEMT-L, whereas the PC-O species with shorter chains were increased more by PEMT-L expression than by PEMT-S expression, suggesting a difference in the substrate specificity of PEMT-L and PEMT-S. On the other hand, various PE and PE-O species were decreased by PEMT-S expression. In addition, PEMT-L and PEMT-S expression promoted the proliferation of HEK-293 cells. Based upon these findings, we propose a model in which the enzymatic activity and substrate specificity are regulated by the glycosylated N-terminal region of PEMT-L localized in the ER lumen.

INTRODUCTION

PC (phosphatidylcholine) is the most abundant phospholipid in mammalian cell membranes, constituting 40–50% of total phospholipids [1]. All eukaryotic cells synthesize PC via the CDP (cytidine diphosphate)-choline pathway [2]. PC is also synthesized through the methylation of PE (phosphatidylethanolamine) by PEMT (PE N-methyltransferase) in a few bacteria, yeast and in the livers of mammals [24]. LPC (lysophosphatidylcholine) is acylated to PC by LPC acyltransferase activity [5]. PC is catabolized by phospholipases A2, C and D and by PS (phosphatidylserine) synthase 1. The second most abundant mammalian membrane phospholipid is PE, which constitutes ~45% of total phospholipids in the brain, but only ~20% of total phospholipids in the liver [1]. PE is generated in mammalian cells by two major pathways: the CDP-ethanolamine pathway and the PS decarboxylation pathway [1]. PE is also generated by the acylation of LPE (lysophosphatidylethanolamine) [5]. PE is involved in numerous cellular functions beyond serving a structural role in membranes.

PC-O (ether-linked PC), including plasmanylcholine and plasmenylcholine, and PE-O (ether-linked PE), including plasmanylethanolamine and plasmenylethanolamine, are characterized by the presence of an ether bond, but not an ester bond, at the sn-1 position of the glycerol backbone. Plasmenylcholine and plasmenylethanolamine are also called plasmalogens, and display a cis double bond on the alkyl chain, adjacent to the ether bond, forming a vinyl-ether linkage. Plasmalogens comprise approx. 18% of phospholipids in humans, but the plasmalogen content of individual tissues or cell types varies widely [6]. Plasmenylethanolamine is the more abundant form of PE-O. Although PC-O usually consists of plasmanylcholine, certain cells contain significant levels of plasmenylcholine. Plasmanylcholine is used for the formation of platelet-activating factor.

The human PEMT gene on chromosome 17p11.2 has nine exons and eight introns, and differential promoter usage generates three unique transcripts [7,8]. Transcript variant 1 (GenBank® accession number NM_148172) encodes PEMT-L (longer isoform of PEMT), whereas transcript variants 2 and 3 (GenBank® accession numbers NM_007169 and NM_148173 respectively) encode PEMT-S (shorter isoform of PEMT). PEMT expression is highest in the human liver, but transcripts are also detectable in the heart, testis, brain and skeletal muscle [8]. It has been reported that transcript variant 2 is the most abundant in human liver [8]. On the other hand, Resseguie et al. [9] have suggested that transcript variant 1 is the most abundant transcript (~80%). PEMT is localized to the ER (endoplasmic reticulum) and catalyses the sequential transfer of three methyl groups from AdoMet (S-adenosylmethionine) to PE to generate PC [2,4,10]. In the liver, the PEMT-controlled pathway accounts for ~30% of PC biosynthesis [11]. PEMT-knockout mice fed a choline-deficient diet exhibit severe liver failure within 3 days and die in 4–5 days [12]. PEMT compensates for the lack of dietary choline, which may occur during starvation, pregnancy or lactation. The polymorphism of the human PEMT gene is associated with susceptibility to steatosis [13,14], and patients with non-alcoholic steatohepatitis have a diminished PC/PE ratio within the liver [15]. Zhao et al. [16] have recently shown that the deletion of PEMT reduces hepatic secretion of very-low-density lipoproteins, alters the clearance rate of very-low-density lipoproteins from plasma and attenuates atherosclerosis in mice.

The function and structure of PEMT-L have not been characterized, in contrast with PEMT-S. In the present study, we compared the function of PEMT-L and PEMT-S using HEK (human embryonic kidney)-293 cell lines stably expressing PEMT-L and PEMT-S (HEK/PEMT-L and HEK/PEMT-S), because HEK-293 cells are commonly used for the study of phospholipid metabolism [1719] and do not endogenously express detectable levels of PEMT-L and PEMT-S proteins. We investigated the subcellular localization, glycosylation and enzymatic activity of PEMT-L and PEMT-S. We also studied the effects of PEMT-L and PEMT-S expression on the cellular contents of PC and PE using novel enzymatic methods, on the PC and PE molecular species by ESI–MS/MS (electrospray ionization–tandem MS) analyses, and on cell growth.

EXPERIMENTAL

Materials

Choline oxidase from Alcaligenes sp. was obtained from Wako Pure Chemical Industries. Amine oxidase (tyramine oxidase) from Arthrobacter sp. was provided by Asahi Kasei Pharma. PLD (phospholipase D) from Streptomyces chromofuscus and GPL-PLD (glycerophospholipid-specific PLD) from Streptomyces sp. were purchased from BIOMOL International. Horseradish peroxidase was obtained from Oriental Yeast. Amplex Red reagent was purchased from Molecular Probes. POPC (L-α-palmitoyl-oleoyl PC), PC from chicken egg, PC from soya bean, 14:0/14:0 PC, 16:0/18:1 plasmanylcholine, POPE (L-α-palmitoyl-oleoyl PE), PE from bovine liver, PE from soya bean, 18:1/18:1 plasmenylethanolamine, L-α-palmitoyl-oleoyl PS sodium salt, L-α-palmitoyl-oleoyl PA (phosphatidic acid) sodium salt, palmitoyl-oleoyl PG (phosphatidylglycerol) sodium salt, palmitoyl SM (sphingomyelin), L-α-mono-oleoyl PC and L-α-mono-oleoyl PE were purchased from Avanti Polar Lipids. 14:0/14:0 PE was purchased from Sigma–Aldrich. Dipalmitoyl PI (phosphatidylinositol) ammonium salt was purchased from Cayman Chemical. Endo H (endoglycosidase H) and PNGase F (peptide N-glycosidase F) were obtained from New England Biolabs. All other chemicals used were of the highest reagent grade.

Enzymatic measurement of PC

The three steps for the enzymatic measurement of PC are illustrated in Figure 1(A). (1) PC is hydrolysed to choline and PA by GPL-PLD, which does not react with SM or LPC. (2) Choline is oxidized by choline oxidase to betaine and two H2O2 molecules. (3) In the presence of peroxidase, H2O2 reacts with Amplex Red to produce highly fluorescent resorufin, which can be measured.

Enzymatic measurement of PC

Figure 1
Enzymatic measurement of PC

(A) Strategy for PC measurement. GPL-PLD catalyses the hydrolysis of PC to PA and choline. Oxidation of choline is catalysed by choline oxidase, which produces two H2O2 molecules. In the presence of peroxidase, Amplex Red reacts with H2O2 to produce highly fluorescent resorufin, which can be measured. (B and C) Standard curves for PC measurement. The POPC standard solution was added to reagent C1 and incubated at 37 °C for 30 min. Then, reagent C2 was added. After 30 min of incubation at room temperature, stop reagent was added. The fluorescence intensity was measured using a fluorescence microplate reader. The background fluorescence was 178±4, which was subtracted from each value. The lines were obtained by (B) linear regression analysis and (C) quadratic regression analysis. The correlation coefficients were (B) r=0.9995 and (C) r=0.9995. (D) Fluorescence changes in response to POPC, egg PC, soya bean PC and PC-O (16:0/18:1 plasmanylcholine) in PC measurement. The background fluorescence was 137±1, which was subtracted from each value. There were no statistically significant differences between POPC, egg PC, soya bean PC and PC-O. (E) Fluorescence changes in response to PC, SM and LPC in PC measurement. The background fluorescence was 261±1, which was subtracted from each value. The lines were obtained by linear regression analysis (PC, r=0.9960; SM, r=0.9973; LPC, r=0.9971). Results shown are means±S.D. of triplicate measurements.

Figure 1
Enzymatic measurement of PC

(A) Strategy for PC measurement. GPL-PLD catalyses the hydrolysis of PC to PA and choline. Oxidation of choline is catalysed by choline oxidase, which produces two H2O2 molecules. In the presence of peroxidase, Amplex Red reacts with H2O2 to produce highly fluorescent resorufin, which can be measured. (B and C) Standard curves for PC measurement. The POPC standard solution was added to reagent C1 and incubated at 37 °C for 30 min. Then, reagent C2 was added. After 30 min of incubation at room temperature, stop reagent was added. The fluorescence intensity was measured using a fluorescence microplate reader. The background fluorescence was 178±4, which was subtracted from each value. The lines were obtained by (B) linear regression analysis and (C) quadratic regression analysis. The correlation coefficients were (B) r=0.9995 and (C) r=0.9995. (D) Fluorescence changes in response to POPC, egg PC, soya bean PC and PC-O (16:0/18:1 plasmanylcholine) in PC measurement. The background fluorescence was 137±1, which was subtracted from each value. There were no statistically significant differences between POPC, egg PC, soya bean PC and PC-O. (E) Fluorescence changes in response to PC, SM and LPC in PC measurement. The background fluorescence was 261±1, which was subtracted from each value. The lines were obtained by linear regression analysis (PC, r=0.9960; SM, r=0.9973; LPC, r=0.9971). Results shown are means±S.D. of triplicate measurements.

Measurement was performed using a three-reagent system. Reagent C1 contained 150 units/ml GPL-PLD, 1.5 mM CaCl2, 50 mM NaCl and 50 mM Tris/HCl (pH 7.4). Reagent C2 contained 4 units/ml choline oxidase, 5 units/ml peroxidase, 300 μM Amplex Red, 0.2% Triton X-100, 50 mM NaCl and 50 mM Tris/HCl (pH 7.4). Amplex Red stop reagent was obtained from Molecular Probes. PC standard solutions were dissolved in an aqueous 1% Triton X-100 solution. At least 2.5 mM POPC, egg PC, soya bean PC and 16:0/18:1 plasmanylcholine could be completely solubilized with 1% Triton X-100.

The sample (10 μl) was added to reagent C1 (40 μl) and incubated at 37 °C for 30 min. After the incubation, reagent C2 (50 μl) was added. After 30 min of incubation at room temperature (20–25 °C), Amplex Red stop reagent (20 μl) was added. The fluorescence intensity was measured using a Fluoromark fluorescence microplate reader (Bio-Rad Laboratories), and the λex and λem filters were set at 544 and 590 nm respectively.

Enzymatic measurement of PE

There are three steps for the enzymatic measurement of PE (Figure 2A). (1) PLD hydrolyses PE to PA and ethanolamine. (2) Ethanolamine is oxidized by amine oxidase, which generates H2O2, ammonia and glycolaldehyde. (3) Finally, H2O2, in the presence of peroxidase, reacts with Amplex Red to generate resorufin.

Enzymatic measurement of PE

Figure 2
Enzymatic measurement of PE

(A) Strategy for PE measurement. PLD hydrolyses PE to PA and ethanolamine. Ethanolamine is oxidized by amine oxidase to generate H2O2. In the presence of peroxidase, Amplex Red reacts with H2O2 to produce highly fluorescent resorufin, which can be measured. (B and C) Standard curves for PE measurement. The POPE standard solution was added to reagent E1 and incubated at 37 °C for 30 min. Then, reagent E2 was added. After 30 min of incubation at room temperature, stop reagent was added. The fluorescence intensity was measured using a fluorescence microplate reader. The background fluorescence was 241±3, which was subtracted from each value. The lines were obtained by (B) linear regression analysis and (C) quadratic regression analysis. The correlation coefficients were (B) r=0.9978 and (C) r=0.9998. (D) Fluorescence changes in response to POPE, liver PE, soya bean PE and PE-O (18:1/18:1 plasmenylethanolamine) in PE measurement. The background fluorescence was 84±1, which was subtracted from each value. There were no statistically significant differences between POPE, liver PE, soya bean PE and PE-O. (E) Fluorescence changes in response to PE, PC, PS and LPE in PE measurement. The background fluorescence was 289±4, which was subtracted from each value. The lines were obtained by linear regression analysis (PE, r=0.9988; PC, r=0.5610; PS, r=0.9998; LPE, r=0.9945). Results shown are means±S.D. of triplicate measurements.

Figure 2
Enzymatic measurement of PE

(A) Strategy for PE measurement. PLD hydrolyses PE to PA and ethanolamine. Ethanolamine is oxidized by amine oxidase to generate H2O2. In the presence of peroxidase, Amplex Red reacts with H2O2 to produce highly fluorescent resorufin, which can be measured. (B and C) Standard curves for PE measurement. The POPE standard solution was added to reagent E1 and incubated at 37 °C for 30 min. Then, reagent E2 was added. After 30 min of incubation at room temperature, stop reagent was added. The fluorescence intensity was measured using a fluorescence microplate reader. The background fluorescence was 241±3, which was subtracted from each value. The lines were obtained by (B) linear regression analysis and (C) quadratic regression analysis. The correlation coefficients were (B) r=0.9978 and (C) r=0.9998. (D) Fluorescence changes in response to POPE, liver PE, soya bean PE and PE-O (18:1/18:1 plasmenylethanolamine) in PE measurement. The background fluorescence was 84±1, which was subtracted from each value. There were no statistically significant differences between POPE, liver PE, soya bean PE and PE-O. (E) Fluorescence changes in response to PE, PC, PS and LPE in PE measurement. The background fluorescence was 289±4, which was subtracted from each value. The lines were obtained by linear regression analysis (PE, r=0.9988; PC, r=0.5610; PS, r=0.9998; LPE, r=0.9945). Results shown are means±S.D. of triplicate measurements.

Measurement was performed using a three-reagent system. Reagent E1 contained 150 units/ml PLD, 1.5 mM CaCl2, 50 mM NaCl and 50 mM Tris/HCl (pH 7.4). Reagent E2 contained 4 units/ml amine oxidase, 5 units/ml peroxidase, 300 μM Amplex Red, 0.2% Triton X-100, 50 mM NaCl and 50 mM Tris/HCl (pH 7.4). PE standard solutions were dissolved in 1% Triton X-100 aqueous solution. At least 2.5 mM POPE, liver PE, soya bean PE and 18:1/18:1 plasmenylethanolamine could be completely solubilized with 1% Triton X-100.

Sample (10 μl) was added to reagent E1 (40 μl) and incubated at 37 °C for 30 min. After the incubation, reagent E2 (50 μl) was added. After 30 min of incubation at room temperature, Amplex Red stop reagent (20 μl) was added. The fluorescence intensity was measured using a fluorescence microplate reader and the λex and λem filters were set at 544 and 590 nm respectively.

Recombinant plasmid construction

The human PEMT gene (The I.M.A.G.E. Consortium clone number 3162576; GenBank® accession number BC_000557) was obtained from the A.T.C.C.. This PEMT gene contains the translation start sites of PEMT-L and PEMT-S and two SNPs (single nucleotide polymorphisms), rs7946 (V212M in PEMT-L or V175M in PEMT-S) and rs897453 (V95I in PEMT-L or V58I in PEMT-S). WT (wild-type) PEMT-L and WT PEMT-S were generated using a KOD-Plus mutagenesis kit (Toyobo). Using PCR, an oligonucleotide encoding an HA (haemagglutinin) (YPYDVPDYA)-tagged epitope was added to the 5′ end of WT PEMT-L or WT PEMT-S. These PCR products were ligated into the NheI and EcoRI sites of the pIRESneo3 mammalian expression vector (Clontech) to generate the plasmids pIRESneo3/HA-PEMT-L-WT and pIRESneo3/HA-PEMT-S-WT. pIRESneo3 contains the internal ribosome-entry site, which permits the translation of two open reading frames from one mRNA. This expression system facilitates the establishment of pools of stably transfected cell lines, whereby nearly all cells surviving in selective medium express the gene of interest, since the neomycin phosphotransferase gene is expressed under the control of the same promoter [20].

Cell cultures

HEK-293 cells were grown in MEM (minimum essential medium) supplemented with 10% (v/v) heat-inactivated FBS (fetal bovine serum) in a humidified incubator (5% CO2) at 37 °C.

Establishment of stable cell lines expressing PEMT-L and PEMT-S

HEK-293 cells were transfected with pIRESneo3/HA-PEMT-L-WT or pIRESneo3/HA-PEMT-S-WT using Lipofectamine™ reagent and PLUS™ reagent (Invitrogen) according to the manufacturer's instructions. Transfected cells were selected with 1.2 mg/ml G418 disulfate (Nacalai Tesque), and a large number of G418-resistant clones were pooled in one dish.

Expression of PEMT-L and PEMT-S

The expression of PEMT-L and PEMT-S was examined by Western blotting. Cells were lysed with PBS containing 1% Triton X-100 and protease inhibitors {100 μg/ml p-APMSF [(p-amidinophenyl)methanesulfonyl fluoride], 10 μg/ml leupeptin and 2 μg/ml aprotinin}. Cell lysate proteins were separated by SDS/PAGE on a 15% gel calibrated with Precision Plus WesternC protein standards (Bio-Rad Laboratories). These proteins were transferred on to PVDF membranes and immunoblotted with the monoclonal anti-HA antibody 12CA5 (Roche Applied Science)(1:3000 dilution) or polyclonal anti-PEMT antibody(Millipore) (1:1000 dilution). Protein–antibody complexes were detected by enhanced chemiluminescence using horseradish peroxidase-conjugated goat anti-mouse IgG (Invitrogen) and ECL Plus Western blotting detection reagents (GE Healthcare), and exposed to X-ray films. The film was scanned and the integrated optical densities of the bands were measured using ImageJ software (NIH).

Total RNA isolation was performed using an SV Total RNA Isolation System (Promega). RT (reverse transcription)–PCR for PEMT mRNA was performed using an Access Quick RT-PCR System (Promega) with a set of primers (5′-AAGACCCGCAAGCTGAGCA-3′ and 5′-AGTACATGGGGTTGTCCAGGA-3′) that bind to all three PEMT transcript variants [9]. Another primer pair (5′-TGAACGGGAAGCTCACTGG-3′ and 5′- TCCACCACCCTGTTGCTGTA-3′) was used for the detection of GAPDH (glyceraldehyde-3-phosphate dehydrogenase) mRNA. The amplified product was resolved by 3% (w/v) agarose gel electrophoresis.

Glycosylation of PEMT-L and PEMT-S

Cells were lysed with PBS containing 1% Triton X-100 and protease inhibitors (100 μg/ml p-APMSF, 10 μg/ml leupeptin and 2 μg/ml aprotinin). Digestion with Endo H and PNGase F was performed as described in the manufacturer's instructions. In brief, cell lysate (5 μg of protein) was treated with 2500 units of Endo H or 1000 units of PNGase F for 1 h at 37 °C. The deglycosylated proteins were separated by SDS/PAGE (15% gel) and immunodetected using a monoclonal anti-HA antibody (1:3000 dilution).

PEMT-activity assays

Cell homogenates (50 μg of protein) were assayed for PEMT activity using phosphatidylmonomethylethanolamine (Avanti Polar Lipids) as the methyl acceptor and S-adenosyl-L-[methyl-3H]methionine (MP Biomedicals) as the methyl donor using the method of Ridgway and Vance [21]. The PEMT specific activities were calculated by subtracting the non-specific activity in HEK-293 cells. The PEMT relative activities were determined by normalizing to protein expression levels.

Measurement of PC and PE content in cells

Cells were subcultured in poly-L-lysine-coated six-well plates at various cell densities in MEM supplemented with 10% (v/v) FBS. After incubation for 48 h, the cells were washed with fresh medium and incubated with MEM containing 0.02% BSA for 18 h at 37 °C. After incubation, the cells were chilled on ice, washed and scraped with ice-cold PBS. The cells were sonicated (model UR-20P, Tomy Seiko, power 10 for 40 s), and the cell protein concentration was measured using a BCA (bicinchoninic acid) protein assay kit (Thermo Scientific). Cellular lipids were extracted by the method of Bligh and Dyer [22] and dissolved in 1% Triton X-100 prepared just before use. The contents of PC and PE in the lipid extract from cells were measured by enzymatic assays.

Measurement of PS content in cells

Cells were subcultured in 10-cm dishes at a density of 9×106 cells per dish in MEM supplemented with 10% (v/v) FBS. After incubation for 48 h, the cells were washed with fresh medium and incubated with MEM containing 0.02% BSA for 18 h at 37 °C. After incubation, the cells were chilled on ice, washed and scraped with ice-cold PBS. The cells were sonicated (power 10 for 40 s) and the cell protein concentration was measured. Cellular lipids were extracted by the method of Bligh and Dyer [22]. The content of PS in the lipid extract from cells was determined using TLC and a phosphorus assay as described previously [23,24].

Mass-spectrometric analysis

Cells were subcultured in 10-cm dishes at a density of 9×106 cells in MEM supplemented with 10% (v/v) FBS. After incubation for 48 h, the cells were washed with fresh medium and incubated with MEM containing 0.02% BSA for 18 h at 37 °C. After incubation, the cells were chilled on ice, washed and scraped with ice-cold PBS. The cells were sonicated (power 10 for 40 s) and the cell protein concentration was measured. Cell homogenates corresponding to 2 mg of protein were subjected to lipid extraction using the method of Bligh and Dyer [22]. The non-naturally occurring lipid species 14:0/14:0 PC and 14:0/14:0 PE were added as internal standards [25]. The chloroform phase was dried under a gentle stream of nitrogen and dissolved in 10 mM ammonium acetate in methanol/chloroform (3:1). The extracted lipids were subjected to ESI–MS/MS analysis using an API3000 triple-quadrupole mass spectrometer (Applied Biosystems) by direct flow injection at a flow rate of 5 μl/min. The mass range of the instrument was set at m/z 600–1000 with a scan time of 1 s. The ion-spray voltage was set at 4500 V and the declustering potential was set at 30 V. Nitrogen was used as the curtain gas and as the collision gas. In the positive-ion mode, precursor ion scanning at m/z 184 and neutral loss scanning of 141 Da were used for PC and PE respectively [25,26]. The signal intensity of the internal standard (14:0/14:0 PC or 14:0/14:0 PE) was represented as 100%.

Statistical analysis

The statistical significance of differences between mean values was analysed using the unpaired Student's t test. Multiple comparisons were performed using the Bonferroni test following ANOVA. Differences were considered significant at P< 0.05. Unless indicated otherwise, results are given as means±S.E.M.

RESULTS AND DISCUSSION

PC measurement

Enzymatic assays have been reported for the quantification of various phospholipids, including PC [27,28], PA [29], PG [30], SM [31,32], LPC [33] and lysophosphatidic acid [29,34]. In addition, Ota et al. [35] recently described a preliminary study of an enzymatic method for PE measurement, while our present studies were in progress. Amplex Red is widely used to monitor H2O2 levels and a variety of oxidase-mediated reactions [29,32,36].

Using Amplex Red, we modified the method for the enzymatic measurement of PC reported by Hojjati and Jiang [28]. As shown in Figure 1(B), the standard curve for PC measurement was linear between 10 and 150 μM (r=0.9995). At lower concentrations of PC, there was a quadratic relationship between PC concentration and fluorescence intensity, and the detection limit was 1 μM (10 pmol in the reaction mixture) (Figure 1C), which was a 250-fold improvement compared with the colorimetric assay [29]. The fluorescence change in response to POPC was the same as that to egg PC or soya bean PC containing mixed acyl chains and to plasmanylcholine (a PC-O) (Figure 1D), indicating that this method can measure PC and PC-O regardless of the chain length and the number of double bonds. In this assay, other choline-containing phospholipids, SM and LPC, showed only negligible increases in fluorescence (~1% and ~2% of the increase induced by PC respectively) (Figure 1E), demonstrating that the activity of GPL-PLD towards SM or LPC is very low. In addition, the PC measurement was not affected by 100 μM PE, 100 μM PS, 100 μM PG, 100 μM PA, 100 μM PI or 100 μM LPE (results not shown).

PE measurement

An enzymatic method was also developed for measuring PE. A calibration reaction was performed using PE standard solutions to determine the detection sensitivity of the novel method. As shown in Figure 2(B), there was a linear relationship from 50 to 250 μM PE (r=0.9978), and a much better linear fit was obtained without forcing the curve through zero. At lower concentrations of PE, the curve was slightly non-linear and fitted a quadratic regression equation (r=0.9998) (Figure 2C). The detection limit was as low as 1 μM (10 pmol in the reaction mixture). Because the activity of GPL-PLD towards PE was not high enough to completely hydrolyse PE, we used PLD, but not GPL-PLD, for PE measurement. There were no significant differences in the fluorescence changes in response to POPE, liver PE, soya bean PE and plasmenylethanolamine (a PE-O) (Figure 2D), indicating that this PE measurement is not affected by the chain length, the number of double bonds or the linkage type (ester or ether). Next, we investigated whether this assay also detects LPE, because PLD hydrolyses LPE to release ethanolamine. As expected, PE and LPE increased the fluorescence to the same extent when normalized to moles (Figure 2E). This measurement cannot distinguish between PE and LPE. On the other hand, using other amine-containing phospholipids, PC and PS, led to no increase in fluorescence (Figure 2E), indicating that choline and serine are not substrates of amine oxidase from Arthrobacter. In addition, 100 μM PC, 100 μM PS, 100 μM PG, 100 μM PA, 100 μM PI, 100 μM SM or 100 μM LPC had no influence on the PE measurement (results not shown).

PC and PE content in cultured cells

Choline oxidase for PC measurement and amine oxidase for PE measurement react with choline and various amines respectively to produce H2O2 [35,37]. Cultured cells contain considerable amounts of choline and amines, confounding PC and PE measurements. The concentration of choline and amines can be determined by enzymatic assays using reagents C2 and E2 without reagents C1 and E1 respectively. The method of Bligh and Dyer [22] has been widely used for lipid extraction, followed by quantification of PC and PE using TLC or HPLC [15,24]. The lipid extract from HEK-293 cells obtained by the method of Bligh and Dyer [22] did not contain a detectable amount of choline or amines that reacted with choline oxidase or amine oxidase. Therefore lipid extraction from cells is recommended for enzymatic measurements of PC and PE, although the PC and PE concentrations in cell lysates can be determined without lipid extraction by subtracting the concentrations of contaminating choline and amines respectively.

We sought to determine the accuracy of PC and PE measurements by recovery studies in which known quantities of POPC or POPE were added to the cellular lipid extract (Table 1). The mean recoveries of PC and PE in concentrations of 12.5–75.0 μM were 101.3±0.6% and 99.6±0.7% respectively. These results indicated that in PC and PE assays, there was no interference of hydrophobic compounds extracted from the cells.

Table 1
Recovery of PC and PE added to the cellular lipid extract

POPC or POPE standard solution was added to the lipid extract from HEK-293 cells. The concentrations of PC and PE were measured by enzymatic assays. The endogenous concentrations of PC and PE in the cellular lipid extract were 49.5 and 18.4 μM respectively.

 Added (μM) Measured (μM) Expected (μM) Recovery (%) 
PC 49.5 – – 
 12.5 63.8 62.0 102.9 
 25.0 76.2 74.5 102.3 
 37.5 87.9 87.0 101.0 
 50.0 99.6 99.5 100.1 
 75.0 124.5 124.5 100.0 
PE 18.4 – – 
 12.5 31.3 30.9 101.3 
 25.0 43.9 43.4 101.0 
 37.5 55.3 55.9 98.8 
 50.0 66.8 68.4 97.7 
 75.0 92.6 93.4 99.1 
 Added (μM) Measured (μM) Expected (μM) Recovery (%) 
PC 49.5 – – 
 12.5 63.8 62.0 102.9 
 25.0 76.2 74.5 102.3 
 37.5 87.9 87.0 101.0 
 50.0 99.6 99.5 100.1 
 75.0 124.5 124.5 100.0 
PE 18.4 – – 
 12.5 31.3 30.9 101.3 
 25.0 43.9 43.4 101.0 
 37.5 55.3 55.9 98.8 
 50.0 66.8 68.4 97.7 
 75.0 92.6 93.4 99.1 

This enzymatic assay for PE quantifies the total concentration of PE and LPE. The content of LPE in cellular membranes is generally much lower than that of PE [38]. The extraction of lysophospholipids using the method of Bligh and Dyer [22] requires acidic conditions [39]. Thus, in the cellular lipid extract prepared using the normal method of Bligh and Dyer [22], the LPE concentration may be very low and negligible for total PE measurement.

These enzymatic methods for PC and PE measurement are simple, rapid, specific, sensitive and high-throughput. Both PC and PE measurement requires only a 10 μl sample volume, and the detection limits are as low as 10 pmol. All enzymes and substrates are available commercially. These assays are applicable to the quantification of PC and PE in sparse cell cultures and may be extended to determine PC and PE levels in subcellular organelles or in animal tissues and fluids.

Expression of PEMT-L and PEMT-S in HEK-293 cells

To study the function of human PEMT-L and PEMT-S, HEK-293 cell lines stably expressing PEMT-L (HEK/PEMT-L) and PEMT-S (HEK/PEMT-S) were established. A HA-tag was fused to the N-termini of PEMT-L and PEMT-S for immunodetection using an anti-HA antibody. It has been demonstrated that HA-tagged PEMT is enzymatically active [40]. PEMT-L and PEMT-S were expressed as proteins of ~26 kDa and ~19 kDa respectively, as assessed by SDS/PAGE (Figure 3A). The expression level of PEMT-L was 4.4-fold higher than that of PEMT-S. The expression of PEMT-L or PEMT-S in the stable cell lines was also detected using polyclonal anti-PEMT antibody, but endogenous PEMT-L or PEMT-S were not detected in the host HEK-293 cells (Figure 3B). However, endogenous PEMT mRNA was barely detected in the host HEK-293 cells by RT–PCR (Figure 3C).

Expression of PEMT-L and PEMT-S in HEK-293 cells

Figure 3
Expression of PEMT-L and PEMT-S in HEK-293 cells

(A) Immunoblot analysis of PEMT-L and PEMT-S. Cell lysates (15.9 μg of protein) from HEK-293, HEK/PEMT-L and HEK/PEMT-S cells were separated by SDS/15% PAGE. The N-termini of PEMT-L and PEMT-S were fused to the HA-tag, and PEMT-L and PEMT-S were detected with an anti-HA antibody. (B) PEMT-L and PEMT-S were detected with a polyclonal anti-PEMT antibody. (C) Expression of PEMT mRNA in HEK-293, HEK/PEMT-L and HEK/PEMT-S cells was assessed by RT–PCR. (D) Glycosylation of PEMT-L and PEMT-S. Cell lysates (1.33 μg of protein) from HEK/PEMT-L and HEK/PEMT-S cells were treated without (−) or with Endo H (H) or PNGase F (F). The ~24 kDa bands were derived from the deglycosylated PEMT-L. The ~28 kDa bands were derived from Endo H. In (A, B and D) the molecular mass in kDa is indicated. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Figure 3
Expression of PEMT-L and PEMT-S in HEK-293 cells

(A) Immunoblot analysis of PEMT-L and PEMT-S. Cell lysates (15.9 μg of protein) from HEK-293, HEK/PEMT-L and HEK/PEMT-S cells were separated by SDS/15% PAGE. The N-termini of PEMT-L and PEMT-S were fused to the HA-tag, and PEMT-L and PEMT-S were detected with an anti-HA antibody. (B) PEMT-L and PEMT-S were detected with a polyclonal anti-PEMT antibody. (C) Expression of PEMT mRNA in HEK-293, HEK/PEMT-L and HEK/PEMT-S cells was assessed by RT–PCR. (D) Glycosylation of PEMT-L and PEMT-S. Cell lysates (1.33 μg of protein) from HEK/PEMT-L and HEK/PEMT-S cells were treated without (−) or with Endo H (H) or PNGase F (F). The ~24 kDa bands were derived from the deglycosylated PEMT-L. The ~28 kDa bands were derived from Endo H. In (A, B and D) the molecular mass in kDa is indicated. GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

It has been demonstrated previously that the human PEMT-S protein transiently expressed in normal rat kidney cells co-localizes with protein disulfide isomerase, an ER marker, and that a C-terminal lysine residue is essential for targeting PEMT to the ER [10]. We examined the subcellular localization of PEMT-L and PEMT-S in HEK-293 cells by immunofluorescence confocal microscopy using a rat anti-HA monoclonal antibody and a rabbit anti-protein disulfide isomerase polyclonal antibody. As expected, both isoforms of PEMT co-localized with protein disulfide isomerase, the ER protein (results not shown).

PEMT-L has a 37-amino-acid sequence upstream of the N-terminus of PEMT-S. An asparagine residue at position 13 in the N-terminal region of PEMT-L is predicted to be N-glycosylated, whereas there is no putative N-glycosylation site in PEMT-S. Endo H and PNGase F cleave high-mannose structures and nearly all types of N-glycan chains respectively. As shown in Figure 3(D), PEMT-L, but not PEMT-S, expressed in HEK-293 cells migrated faster (~24 kDa) after treatment with Endo H or PNGase F. The anti-HA antibody cross-reacted with Endo H, and the ~28 kDa bands were derived from Endo H. These results indicated that the N-terminal region of PEMT-L with a high-mannose oligosaccharide resides in the lumen of the ER.

To assess the enzymatic activity of PEMT-L or PEMT-S, protein homogenates from HEK/PEMT-L and HEK/PEMT-S cells were assayed for PEMT activity. The PEMT-specific activities were calculated by subtracting the non-specific activity in the host HEK-293 cells, which did not express detectable PEMT proteins (Figure 3B). HEK/PEMT-S cells displayed greater PEMT activity than HEK/PEMT-L cells (Figure 4A), despite the lower expression level of PEMT-S compared with PEMT-L (Figure 3A). The relative activity of PEMT-L normalized to the expression level was 12.2% of that of PEMT-S (Figure 4B). These results suggested that the N-terminal region of PEMT-L in the ER lumen negatively regulates enzymatic activity, although the binding site on PEMT for AdoMet is localized to the cytosolic surface of the ER [41]. It is also likely that the activity of PEMT-L is attenuated by another protein that interacts with its N-terminal region.

Enzymatic activities of PEMT-L and PEMT-S

Figure 4
Enzymatic activities of PEMT-L and PEMT-S

(A) Homogenates (50 μg of protein) from HEK-293, HEK/PEMT-L and HEK/PEMT-S cells were assayed for PEMT activity. (B) PEMT relative activity normalized to protein expression levels. The activity of PEMT-S was taken as 100%. Each bar represents the mean ± S.E.M. of three measurements. The absence of an error bar signifies an S.E.M. value smaller than the graphic symbol. §P< 0.05 indicated a significant difference between HEK/PEMT-L and HEK/PEMT-S cells.

Figure 4
Enzymatic activities of PEMT-L and PEMT-S

(A) Homogenates (50 μg of protein) from HEK-293, HEK/PEMT-L and HEK/PEMT-S cells were assayed for PEMT activity. (B) PEMT relative activity normalized to protein expression levels. The activity of PEMT-S was taken as 100%. Each bar represents the mean ± S.E.M. of three measurements. The absence of an error bar signifies an S.E.M. value smaller than the graphic symbol. §P< 0.05 indicated a significant difference between HEK/PEMT-L and HEK/PEMT-S cells.

Effect of PEMT expression on PC and PE contents in HEK-293 cells

To examine the effect of PEMT expression on PC and PE cellular content, we quantified PC and PE content in the lipid extracts from HEK-293 cells HEK/PEMT-L cells and HEK/PEMT-S cells at various cell densities using our new enzymatic assays. As shown in Figure 5(A), the PC content in HEK-293 cells increased in a cell-density-dependent manner, which was consistent with our previous results obtained from the colorimetric PC assay [29]. The increase in the density of HEK-293 cells was also accompanied by an increase in the cellular PE content (Figure 5B). The PE content in HEK-293 cells at a high cell density (77.1 μg of protein/cm2) was 1.3-fold higher than that at a low cell density (6.17 μg of protein/cm2) (P<0.005). On the other hand, the PE/PC ratio in HEK-293 cells decreased with increasing cell density (Figure 5C). We have also reported previously that the cellular content of PA is negatively correlated with cell density [29]. However, the relationships between cell density and metabolism of phospholipids, including PC and PE, have been scarcely investigated. These findings raise the possibility that signalling from cell–cell adhesion regulates the enzyme activities involved in the synthesis and/or degradation of PC and PE. The large amounts of PC and PE at high cell density may be required for various complicated structures of the cell membranes. Further studies are necessary to clarify the mechanisms for PC and PE biosynthesis dependent on cell density.

Effect of cell density and PEMT expression on PC, PE and PS contents in HEK-293 cells

Figure 5
Effect of cell density and PEMT expression on PC, PE and PS contents in HEK-293 cells

HEK-293 cells (open circles), HEK/PEMT-L cells (filled circles) and HEK/PEMT-S cells (filled triangles) were incubated in six-well plates in MEM containing 0.02% BSA for 18 h at 37 °C. (A) PC content, (B) PE content and (C) PE/PC ratio of the cells were determined by enzymatic measurements of PC and PE. (D) HEK-293 cells, HEK/PEMT-L cells and HEK/PEMT-S cells were incubated in 10-cm-diameter dishes in MEM containing 0.02% BSA for 18 h at 37 °C. The PS content in the cells was determined using TLC and a phosphorus assay. There was no difference in the densities of HEK-293 cells, HEK/PEMT-L cells and HEK/PEMT-S cells (55.2±2.6, 58.3±0.6 and 55.7±2.0 μg of protein/cm2 respectively). Results shown are means±S.E.M. of three measurements. *P< 0.05 indicated a significant difference between HEK-293 and HEK/PEMT-L cells at similar cell densities. #P< 0.05 indicated a significant difference between HEK-293 and HEK/PEMT-S cells at similar cell densities. §P<0.05 indicated a significant difference between HEK/PEMT-L and HEK/PEMT-S cells at similar cell densities.

Figure 5
Effect of cell density and PEMT expression on PC, PE and PS contents in HEK-293 cells

HEK-293 cells (open circles), HEK/PEMT-L cells (filled circles) and HEK/PEMT-S cells (filled triangles) were incubated in six-well plates in MEM containing 0.02% BSA for 18 h at 37 °C. (A) PC content, (B) PE content and (C) PE/PC ratio of the cells were determined by enzymatic measurements of PC and PE. (D) HEK-293 cells, HEK/PEMT-L cells and HEK/PEMT-S cells were incubated in 10-cm-diameter dishes in MEM containing 0.02% BSA for 18 h at 37 °C. The PS content in the cells was determined using TLC and a phosphorus assay. There was no difference in the densities of HEK-293 cells, HEK/PEMT-L cells and HEK/PEMT-S cells (55.2±2.6, 58.3±0.6 and 55.7±2.0 μg of protein/cm2 respectively). Results shown are means±S.E.M. of three measurements. *P< 0.05 indicated a significant difference between HEK-293 and HEK/PEMT-L cells at similar cell densities. #P< 0.05 indicated a significant difference between HEK-293 and HEK/PEMT-S cells at similar cell densities. §P<0.05 indicated a significant difference between HEK/PEMT-L and HEK/PEMT-S cells at similar cell densities.

The PC content in both HEK/PEMT-L and HEK/PEMT-S cells were similarly higher than the PC content in HEK-293 cells between the cell densities examined (Figure 5A). HEK/PEMT-S cells exhibited slightly, but significantly, lower PE content than HEK-293 cells at similar cell densities (P< 0.05), whereas the cellular PE content was not affected by the expression of PEMT-L (Figure 5B). Consequently, the PE/PC ratio decreased in the order HEK-293 cells>HEK/PEMT-L cells>HEK/PEMT-S cells (Figure 5C).

The medium used for the culture of animal cells usually lacks ethanolamine, and cultured cells produce PE largely through the PS decarboxylation pathway [1,42]. We also determined whether PEMT-L and PEMT-S expression influenced the PS content in HEK-293 cells. HEK/PEMT-L and HEK/PEMT-S cells showed 17.9% and 23.1% decreases in PS content compared with HEK-293 cells respectively (Figure 5D); the densities of HEK-293 cells, HEK/PEMT-L cells and HEK/PEMT-S cells were the same (55.2±2.6, 58.3±0.6 and 55.7±2.0 μg of protein/cm2 respectively). There was no significant difference in the PS content in HEK/PEMT-L and HEK/PEMT-S cells.

The expression of PEMT-L and PEMT-S resulted in the generation of a substantial amount of PC, but had no or little effect on the cellular PE content. The observation that PEMT-L and PEMT-S expression led to decreased PS content in the cells suggested that PS decarboxylase compensated for the loss of PE due to PEMT-L and PEMT-S. The reduction in PE and PS content induced by PEMT expression is probably suppressed in vivo, where the ethanolamine supply is sufficient, because both PE derived from ethanolamine and PE derived from PS are converted into PC as a result of PEMT expression [43]. Although the activity of PEMT-S in the cell homogenate was remarkably higher than that of PEMT-L (Figure 4A), the increases in PC content were similar in HEK/PEMT-L and HEK/PEMT-S cells (Figure 5A). It is conceivable that PEMT-S expression enhanced the degradation of PC or down-regulated the activity of the CDP-choline pathway compared with PEMT-L expression. The possibility that the activity of PEMT-S in living cells was more limited than that of PEMT-L cannot be eliminated.

Effect of PEMT expression on molecular species of PC and PE

It has been reported previously that purified rat PEMT prefers the long-chain polyunsaturated PE as a substrate for PC synthesis [44]. In McA-RH7777 rat hepatoma cells expressing rat PEMT, the major species of PC derived from the CDP-choline pathway were 34:1 and 36:2 PC, whereas PC derived from the methylation pathway contained significantly more long-chain polyunsaturated PC species (36:2, 38:6, 38:5, 38:4 and 40:6 PC) [11]. To reveal the PC species increased by PEMT-L and PEMT-S expression, we performed ESI–MS/MS, analysing molecular species of PC in HEK-293, HEK/PEMT-L and HEK/PEMT-S cells (Supplementary Figure S1 at http://www.BiochemJ.org/bj/432/bj4320387add.htm). The ESI–MS/MS analysis performed only determined the total number of C atoms and double bonds in the acyl, alkyl or alkenyl chains. In this analysis, the signal intensities did not allow direct comparison between different molecular species that were not detected with equal efficiency. Relative abundances of individual PC molecular species, in comparison with the signal intensity of the internal standard 14:0/14:0 PC [45], are summarized in Table 2. There was no significant difference in the densities of HEK-293, HEK/PEMT-L and HEK/PEMT-S cells (58.6±0.8, 55.0±1.6, and 57.1±0.9 μg of protein/cm2 respectively). Both HEK/PEMT-L and HEK/PEMT-S cells exhibited higher intensities of all PC-O species and three PC species (36:2, 38:4 and 40:2) compared with HEK-293 cells. HEK/PEMT-L cells contained more PC-O species with shorter chains (32:2, 32:1, 32:0, 34:1 and 34:0) than HEK/PEMT-S cells, whereas PC-O species with a longer polyunsaturated acyl chain (36:5, 36:4, 38:4, 40:7, 40:6, 40:5 and 40:4) were more abundant in HEK/PEMT-S cells than in HEK/PEMT-L cells. PC species containing longer chains with six or five double bonds (38:6, 38:5, 40:6 and 40:5) were significantly increased in HEK/PEMT-S cells, but not HEK/PEMT-L cells. The PEMT-S-induced increase in PC species with longer polyunsaturated chains, but not shorter chains, was in agreement with the substrate specificity of rat PEMT [11,44], which exists only as a short isoform. Interestingly, the levels of species containing longer chains with three double bonds (38:3 PC, 40:3 PC and 40:3 PC-O) were higher in HEK/PEMT-L cells than in HEK/PEMT-S cells. These results suggested that the specificities for PC and PC-O species are different between PEMT-L and PEMT-S and depend on the chain length and the number of double bonds. Levels of PC species with C36 acyl chains except for 36:2 PC were similar between the three cell lines. On the other hand, PC species with C30 or C32 acyl chains (30:0, 32:2 and 32:1) were lower in both HEK/PEMT-L and HEK/PEMT cells than in HEK-293 cells. PC species with shorter chains are probably not substrates for PEMT-L or PEMT-S, and the loss of these PC species by PEMT-L and PEMT-S expression may be attributed to down-regulation of PC synthesis through the CDP-choline pathway or degradation by phospholipases.

Table 2
Analysis of PC molecular species in HEK-293, HEK/PEMT-L and HEK/PEMT-S cells by ESI–MS/MS

HEK-293, HEK/PEMT-L and HEK/PEMT-S cells on 10-cm dishes were incubated in MEM containing 0.02% BSA for 18 h at 37 °C. There was no significant difference in the densities of HEK293, HEK/PEMT-L and HEK/PEMT-S cells (58.6±0.8, 55.0±1.6, and 57.1±0.9 μg of protein/cm2 respectively). The total lipids were extracted from cells normalized to the same protein content (2 mg) and analysed by ESI–MS/MS. PC in the lipid extract from cells was detected by precursor-ion scanning of m/z 184 in the positive-ion mode. The signal intensity of the internal standard (14:0/14:0 PC) is represented as 100%. Results shown are means±S.E.M. of three measurements.

   Signal intensity (%) 
m/z PC PC-O† HEK-293 HEK/PEMT-L HEK/PEMT-S 
704.9¶ 30:1  10.05±0.20 8.54±0.36* 8.57±0.04# 
706.9 30:0  10.70±0.04 9.25±0.14* 8.86±0.24# 
716.9  32:2 3.98±0.04 5.41±0.16* 4.57±0.06#§ 
718.9  32:1 25.3±0.2 39.5±0.8* 31.8±0.2#§ 
720.9  32:0 17.7±0.2 26.3±0.0* 22.2±0.1#§ 
730.9 32:2  5.47±0.17 4.52±0.25* 3.65±0.02#§ 
732.9 32:1  46.3±0.5 43.6±0.5* 41.0±0.5#§ 
734.9 32:0  14.8±0.2 14.3±0.5 11.7±0.3#§ 
743.0  34:3 0.890±0.004 1.726±0.053* 1.542±0.069# 
745.0  34:2 11.4±0.2 20.5±0.4* 20.0±0.2# 
747.0  34:1 32.6±0.5 57.1±0.2* 54.2±0.6#§ 
749.0  34:0 10.5±0.3 18.8±0.3* 17.4±0.2#§ 
757.0 34:3  1.277±0.030 1.143±0.065 0.985±0.025# 
759.0 34:2  22.6±0.4 23.2±0.4 22.6±0.3 
761.1 34:1  77.3±0.8 80.5±0.2 83.5±1.1# 
763.9‡ 34:0  3.39±0.10 3.48±0.09 3.76±0.08 
767.1  36:5 0.759±0.013 1.204±0.042* 1.435±0.018#§ 
769.1  36:4 1.60±0.05 2.74±0.11* 3.25±0.07#§ 
771.1  36:3 2.86±0.06 6.49±0.14* 6.56±0.07# 
773.1  36:2 10.5±0.0 25.5±0.6* 26.7±0.2# 
775.1  36:1 12.3±0.1 26.1±0.3* 25.7±0.5# 
777.1  36:0 2.51±0.04 4.73±0.05* 4.66±0.06# 
781.1 36:5  0.394±0.013 0.438±0.015 0.407±0.011 
783.1 36:4  1.25±0.08 1.18±0.13 1.11±0.05 
785.1 36:3  3.52±0.06 3.95±0.18 3.73±0.11 
787.1 36:2  32.6±0.5 39.9±0.5* 44.2±0.7#§ 
789.1 36:1  22.7±0.4 24.0±1.1 26.0±0.5 
791.1 36:0  3.30±0.10 3.75±0.17 3.81±0.07 
793.1  38:6 1.34±0.03 2.34±0.06* 2.61±0.10# 
795.1  38:5 3.23±0.06 6.49±0.30* 7.02±0.14# 
797.1  38:4 1.42±0.02 3.09±0.06* 3.65±0.11#§ 
799.1  38:3 1.41±0.04 2.15±0.14* 1.99±0.05# 
801.1  38:2 2.28±0.02 4.79±0.34* 4.87±0.05# 
803.1  38:1 2.19±0.03 3.95±0.14* 3.88±0.10# 
805.1 38.7 38:0 0.698±0.025 0.947±0.062* 0.993±0.018# 
807.1 38:6  0.498±0.014 0.541±0.019 0.615±0.029# 
809.1 38:5  0.642±0.016 0.638±0.015 0.801±0.012#§ 
811.1 38:4  0.874±0.012 1.001±0.029* 0.981±0.021# 
813.1 38:3  1.50±0.06 1.85±0.11* 1.46±0.01§ 
815.1¶ 38:2  4.64±0.05 6.22±0.18* 6.47±0.05# 
817.1 38:1  2.02±0.06 2.00±0.08 2.21±0.03 
819.1 38:0 40:7 0.696±0.016 1.125±0.067* 1.517±0.017#§ 
821.1  40:6 0.910±0.019 1.761±0.076* 2.100±0.050#§ 
823.1  40:5 0.0571±0.015 1.178±0.025* 1.603±0.018#§ 
825.1  40:4 0.297±0.005 0.640±0.020* 0.725±0.018#§ 
827.1  40:3 0.400±0.002 0.736±0.017* 0.662±0.004#§ 
829.1  40:2 0.803±0.011 1.468±0.076* 1.548±0.019# 
831.1 40:8 40:1 0.752±0.012 1.078±0.037* 1.167±0.021# 
833.1 40:7 40:0 0.568±0.024 0.713±0.036* 0.813±0.001# 
835.1 40:6  0.304±0.008 0.342±0.024 0.395±0.015# 
837.1 40:5  0.218±0.006 0.244±0.015 0.287±0.012# 
839.1 40:4  0.182±0.011 0.218±0.013 0.186±0.006 
841.1 40:3  0.250±0.007 0.417±0.051* 0.259±0.005§ 
843.1 40:2  0.506±0.016 0.698±0.020* 0.728±0.024# 
845.1 40:1  0.311±0.008 0.311±0.007 0.381±0.008#§ 
   Signal intensity (%) 
m/z PC PC-O† HEK-293 HEK/PEMT-L HEK/PEMT-S 
704.9¶ 30:1  10.05±0.20 8.54±0.36* 8.57±0.04# 
706.9 30:0  10.70±0.04 9.25±0.14* 8.86±0.24# 
716.9  32:2 3.98±0.04 5.41±0.16* 4.57±0.06#§ 
718.9  32:1 25.3±0.2 39.5±0.8* 31.8±0.2#§ 
720.9  32:0 17.7±0.2 26.3±0.0* 22.2±0.1#§ 
730.9 32:2  5.47±0.17 4.52±0.25* 3.65±0.02#§ 
732.9 32:1  46.3±0.5 43.6±0.5* 41.0±0.5#§ 
734.9 32:0  14.8±0.2 14.3±0.5 11.7±0.3#§ 
743.0  34:3 0.890±0.004 1.726±0.053* 1.542±0.069# 
745.0  34:2 11.4±0.2 20.5±0.4* 20.0±0.2# 
747.0  34:1 32.6±0.5 57.1±0.2* 54.2±0.6#§ 
749.0  34:0 10.5±0.3 18.8±0.3* 17.4±0.2#§ 
757.0 34:3  1.277±0.030 1.143±0.065 0.985±0.025# 
759.0 34:2  22.6±0.4 23.2±0.4 22.6±0.3 
761.1 34:1  77.3±0.8 80.5±0.2 83.5±1.1# 
763.9‡ 34:0  3.39±0.10 3.48±0.09 3.76±0.08 
767.1  36:5 0.759±0.013 1.204±0.042* 1.435±0.018#§ 
769.1  36:4 1.60±0.05 2.74±0.11* 3.25±0.07#§ 
771.1  36:3 2.86±0.06 6.49±0.14* 6.56±0.07# 
773.1  36:2 10.5±0.0 25.5±0.6* 26.7±0.2# 
775.1  36:1 12.3±0.1 26.1±0.3* 25.7±0.5# 
777.1  36:0 2.51±0.04 4.73±0.05* 4.66±0.06# 
781.1 36:5  0.394±0.013 0.438±0.015 0.407±0.011 
783.1 36:4  1.25±0.08 1.18±0.13 1.11±0.05 
785.1 36:3  3.52±0.06 3.95±0.18 3.73±0.11 
787.1 36:2  32.6±0.5 39.9±0.5* 44.2±0.7#§ 
789.1 36:1  22.7±0.4 24.0±1.1 26.0±0.5 
791.1 36:0  3.30±0.10 3.75±0.17 3.81±0.07 
793.1  38:6 1.34±0.03 2.34±0.06* 2.61±0.10# 
795.1  38:5 3.23±0.06 6.49±0.30* 7.02±0.14# 
797.1  38:4 1.42±0.02 3.09±0.06* 3.65±0.11#§ 
799.1  38:3 1.41±0.04 2.15±0.14* 1.99±0.05# 
801.1  38:2 2.28±0.02 4.79±0.34* 4.87±0.05# 
803.1  38:1 2.19±0.03 3.95±0.14* 3.88±0.10# 
805.1 38.7 38:0 0.698±0.025 0.947±0.062* 0.993±0.018# 
807.1 38:6  0.498±0.014 0.541±0.019 0.615±0.029# 
809.1 38:5  0.642±0.016 0.638±0.015 0.801±0.012#§ 
811.1 38:4  0.874±0.012 1.001±0.029* 0.981±0.021# 
813.1 38:3  1.50±0.06 1.85±0.11* 1.46±0.01§ 
815.1¶ 38:2  4.64±0.05 6.22±0.18* 6.47±0.05# 
817.1 38:1  2.02±0.06 2.00±0.08 2.21±0.03 
819.1 38:0 40:7 0.696±0.016 1.125±0.067* 1.517±0.017#§ 
821.1  40:6 0.910±0.019 1.761±0.076* 2.100±0.050#§ 
823.1  40:5 0.0571±0.015 1.178±0.025* 1.603±0.018#§ 
825.1  40:4 0.297±0.005 0.640±0.020* 0.725±0.018#§ 
827.1  40:3 0.400±0.002 0.736±0.017* 0.662±0.004#§ 
829.1  40:2 0.803±0.011 1.468±0.076* 1.548±0.019# 
831.1 40:8 40:1 0.752±0.012 1.078±0.037* 1.167±0.021# 
833.1 40:7 40:0 0.568±0.024 0.713±0.036* 0.813±0.001# 
835.1 40:6  0.304±0.008 0.342±0.024 0.395±0.015# 
837.1 40:5  0.218±0.006 0.244±0.015 0.287±0.012# 
839.1 40:4  0.182±0.011 0.218±0.013 0.186±0.006 
841.1 40:3  0.250±0.007 0.417±0.051* 0.259±0.005§ 
843.1 40:2  0.506±0.016 0.698±0.020* 0.728±0.024# 
845.1 40:1  0.311±0.008 0.311±0.007 0.381±0.008#§ 
*

P< 0.05, significant difference between HEK-293 and HEK/PEMT-L cells.

#P< 0.05, significant difference between HEK-293 and HEK/PEMT-S cells.

§

P< 0.05, significant difference between HEK/PEMT-L and HEK/PEMT-S cells.

PC-O includes plasmanylcholine and plasmenylcholine.

These peaks overlap with the isotopic peaks of SM species.

The isotopic peak (m/z 763.9) was used for the evaluation of 34:0 PC because the mono-isotopic peak of 34:0 PC (m/z 763.1) is obscured by the adjacent intense peak (Supplementary Figure S1 at http://www.BiochemJ.org/bj/432/bj4320387add.htm).

We also profiled PE molecular species in HEK-293, HEK/PEMT-L and HEK/PEMT-S cells by ESI–MS/MS (Table 3 and Supplementary Figure S2 at http://www.BiochemJ.org/bj/432/bj4320387add.htm). The expression of PEMT-S resulted in decreased intensities of PE-O species with shorter chains (32:2, 32:0, 34:2, 34:1, 34:0, 36:4, 36:3, 36:2 and 36:0) and various PE species, but had no effect on the intensities of PE species with saturated chains (30:0, 34:0 and 36:0). Both PE and PE-O species having C40 chains with four, five or six double bonds (40:6, 40:5 and 40:4) were not altered by PEMT-S expression. The reduction of PE and PE-O species were less in HEK/PEMT-L cells than in HEK/PEMT-S cells, and several PE-O species (36:1, 36:0, 38:3 and 40:5) were increased in HEK/PEMT-L cells. There were many PE-O species that were similar between HEK-293 and HEK/PEMT-S cells, despite the increase in all PC-O species by PEMT-S expression. On the other hand, levels of PE species (34:2, 34:1, 36:4, 36:2, 36:1, 38:3 and 38:1) were lower in HEK/PEMT-S cells than in HEK-293 cells, although PC species with the same acyl chains were not changed by PEMT-S expression. Both PC and PE species with C32 acyl chains (32:2, 32:1 and 32:0) were decreased in HEK/PEMT-S cells. Both the CDP-ethanolamine pathway and PS decarboxylation are able to produce all PE species [46]. However, the CDP-ethanolamine pathway preferentially synthesized PE with mono- or di-unsaturated acyl chains at the sn-2 position, whereas PS decarboxylase generated species with mainly polyunsaturated acyl chains at the sn-2 position [46]. In HEK/PEMT-S cells cultured in the medium without ethanolamine, the decrease in PE species containing longer polyunsaturated chains might be preferentially compensated by PS decarboxylase. The disagreement between increased PC species and decreased PE species in HEK/PEMT-S cells may be largely ascribed to the remodelling pathway of deacylation and reacylation.

Table 3
Analysis of PE molecular species in HEK-293, HEK/PEMT-L and HEK/PEMT-S cells by ESI–MS/MS

HEK-293, HEK/PEMT-L and HEK/PEMT-S cells on 10-cm dishes were incubated in MEM containing 0.02% BSA for 18 h at 37 °C. There was no significant difference in the densities of HEK-293, HEK/PEMT-L and HEK/PEMT-S cells (58.6±0.8, 55.0±1.6 and 57.1±0.9 μg of protein/cm2 respectively). Total lipids were extracted from cells normalized to the same protein content (2 mg) and analysed by ESI–MS/MS. PE in the lipid extract from cells was detected by neutral-loss scanning of 141 Da in the positive-ion mode. The signal intensity of the internal standard (14:0/14:0 PE) was taken as 100%. Results shown are means±S.E.M. of three measurements.

   Signal intensity (%) 
m/z PE PE-O† HEK-293 HEK/PEMT-L HEK/PEMT-S 
662.7 30:1  0.339±0.024 0.211±0.021* 0.220±0.027# 
664.7 30:0  0.440±0.042 0.607±0.081 0.503±0.065 
674.7  32:2 0.983±0.023 0.646±0.010* 0.420±0.038#§ 
676.7  32:1 1.144±0.041 1.145±0.051 0.989±0.108 
678.7  32:0 0.526±0.040 0.504±0.059 0.248±0.022#§ 
688.7 32:2  1.365±0.020 0.697±0.086* 0.774±0.109# 
690.8 32:1  5.90±0.03 5.12±0.10* 4.20±0.16#§ 
692.8 32:0  1.178±0.015 1.180±0.014 0.946±0.014#§ 
700.8  34:3 0.741±0.040 0.647±0.093 0.473±0.058 
702.8  34:2 5.53±0.15 4.61±0.24* 3.17±0.07#§ 
704.8  34:1 5.45±0.18 5.87±0.08 4.06±0.10#§ 
706.8  34:0 1.061±0.052 1.086±0.081 0.597±0.059#§ 
714.8 34:3  0.604±0.001 0.648±0.052 0.296±0.046#§ 
716.8 34:2  13.30±0.46 9.47±0.68* 6.57±0.08#§ 
718.8 34:1  27.5±0.6 24.6±1.1 20.9±0.0#§ 
720.8 34:0  3.23±0.09 3.05±0.27 2.48±0.17 
724.8  36:5 1.63±0.12 1.65±0.04 2.00±0.19 
726.8  36:4 1.64±0.04 1.74±0.02 1.03±0.06#§ 
728.8  36:3 3.10±0.11 3.44±0.24 2.02±0.11#§ 
730.8  36:2 4.80±0.20 4.65±0.17 3.33±0.06#§ 
732.8  36:1 4.43±0.16 5.65±0.23* 3.72±0.09§ 
734.8  36:0 0.700±0.026 0.832±0.031* 0.436±0.016#§ 
738.9 36:5  0.618±0.041 0.662±0.057 0.807±0.156 
740.9 36:4  1.79±0.01 1.41±0.03* 1.03±0.04#§ 
742.9 36:3  3.29±0.08 2.60±0.10* 1.62±0.08#§ 
744.9 36:2  34.3±1.0 28.5±0.8* 21.0±0.2#§ 
746.9 36:1  18.9±0.5 16.2±0.0* 14.8±0.3# 
748.9 36:0  2.84±0.05 2.90±0.19 2.60±0.14 
750.9  38:6 2.47±0.06 2.71±0.19 2.91±0.03 
752.9  38:5 2.37±0.04 2.57±0.18 2.31±0.17 
754.9  38:4 1.49±0.08 1.63±0.04 1.56±0.01 
756.9  38:3 1.417±0.044 1.873±0.036* 0.974±0.045#§ 
758.9  38:2 1.164±0.020 1.519±0.161 0.953±0.058§ 
760.9  38:1 0.970±0.069 1.231±0.220 0.907±0.069 
762.9 38:7 38:0 0.470±0.013 0.448±0.078 0.352±0.030 
764.9 38:6  1.09±0.10 1.19±0.09 1.17±0.01 
766.9 38:5  3.17±0.11 2.41±0.13* 2.58±0.13# 
768.9 38:4  5.87±0.14 4.99±0.16* 4.32±0.20# 
770.9 38:3  2.89±0.10 3.01±0.05 1.45±0.04#§ 
772.9 38:2  3.73±0.15 3.91±0.01 3.03±0.10#§ 
774.9 38:1  1.77±0.03 1.35±0.01* 1.45±0.04# 
776.9 38:0 40:7 1.19±0.06 1.35±0.01 1.45±0.04 
778.9  40:6 1.35±0.05 1.27±0.01 1.37±0.05 
780.9  40:5 0.602±0.015 0.844±0.061* 0.741±0.063 
782.9  40:4 0.602±0.034 0.655±0.081 0.584±0.063 
784.9  40:3 0.536±0.032 0.564±0.042 0.327±0.018#§ 
786.9  40:2 0.399±0.019 0.454±0.025 0.292±0.033§ 
788.9 40:8 40:1 0.360±0.017 0.407±0.021 0.278±0.029§ 
790.9 40:7 40:0 1.32±0.14 1.16±0.15 1.35±0.07 
792.9 40:6  1.86±0.09 1.54±0.03 1.79±0.11 
794.9 40:5  1.26±0.07 1.12±0.03 1.18±0.03 
796.9 40:4  0.797±0.059 0.811±0.140 0.604±0.072 
798.9 40:3  0.596±0.050 0.558±0.119 0.411±0.046 
800.9 40:2  1.221±0.072 1.063±0.048 0.884±0.057# 
802.9 40:1  0.751±0.055 0.641±0.039 0.556±0.023# 
   Signal intensity (%) 
m/z PE PE-O† HEK-293 HEK/PEMT-L HEK/PEMT-S 
662.7 30:1  0.339±0.024 0.211±0.021* 0.220±0.027# 
664.7 30:0  0.440±0.042 0.607±0.081 0.503±0.065 
674.7  32:2 0.983±0.023 0.646±0.010* 0.420±0.038#§ 
676.7  32:1 1.144±0.041 1.145±0.051 0.989±0.108 
678.7  32:0 0.526±0.040 0.504±0.059 0.248±0.022#§ 
688.7 32:2  1.365±0.020 0.697±0.086* 0.774±0.109# 
690.8 32:1  5.90±0.03 5.12±0.10* 4.20±0.16#§ 
692.8 32:0  1.178±0.015 1.180±0.014 0.946±0.014#§ 
700.8  34:3 0.741±0.040 0.647±0.093 0.473±0.058 
702.8  34:2 5.53±0.15 4.61±0.24* 3.17±0.07#§ 
704.8  34:1 5.45±0.18 5.87±0.08 4.06±0.10#§ 
706.8  34:0 1.061±0.052 1.086±0.081 0.597±0.059#§ 
714.8 34:3  0.604±0.001 0.648±0.052 0.296±0.046#§ 
716.8 34:2  13.30±0.46 9.47±0.68* 6.57±0.08#§ 
718.8 34:1  27.5±0.6 24.6±1.1 20.9±0.0#§ 
720.8 34:0  3.23±0.09 3.05±0.27 2.48±0.17 
724.8  36:5 1.63±0.12 1.65±0.04 2.00±0.19 
726.8  36:4 1.64±0.04 1.74±0.02 1.03±0.06#§ 
728.8  36:3 3.10±0.11 3.44±0.24 2.02±0.11#§ 
730.8  36:2 4.80±0.20 4.65±0.17 3.33±0.06#§ 
732.8  36:1 4.43±0.16 5.65±0.23* 3.72±0.09§ 
734.8  36:0 0.700±0.026 0.832±0.031* 0.436±0.016#§ 
738.9 36:5  0.618±0.041 0.662±0.057 0.807±0.156 
740.9 36:4  1.79±0.01 1.41±0.03* 1.03±0.04#§ 
742.9 36:3  3.29±0.08 2.60±0.10* 1.62±0.08#§ 
744.9 36:2  34.3±1.0 28.5±0.8* 21.0±0.2#§ 
746.9 36:1  18.9±0.5 16.2±0.0* 14.8±0.3# 
748.9 36:0  2.84±0.05 2.90±0.19 2.60±0.14 
750.9  38:6 2.47±0.06 2.71±0.19 2.91±0.03 
752.9  38:5 2.37±0.04 2.57±0.18 2.31±0.17 
754.9  38:4 1.49±0.08 1.63±0.04 1.56±0.01 
756.9  38:3 1.417±0.044 1.873±0.036* 0.974±0.045#§ 
758.9  38:2 1.164±0.020 1.519±0.161 0.953±0.058§ 
760.9  38:1 0.970±0.069 1.231±0.220 0.907±0.069 
762.9 38:7 38:0 0.470±0.013 0.448±0.078 0.352±0.030 
764.9 38:6  1.09±0.10 1.19±0.09 1.17±0.01 
766.9 38:5  3.17±0.11 2.41±0.13* 2.58±0.13# 
768.9 38:4  5.87±0.14 4.99±0.16* 4.32±0.20# 
770.9 38:3  2.89±0.10 3.01±0.05 1.45±0.04#§ 
772.9 38:2  3.73±0.15 3.91±0.01 3.03±0.10#§ 
774.9 38:1  1.77±0.03 1.35±0.01* 1.45±0.04# 
776.9 38:0 40:7 1.19±0.06 1.35±0.01 1.45±0.04 
778.9  40:6 1.35±0.05 1.27±0.01 1.37±0.05 
780.9  40:5 0.602±0.015 0.844±0.061* 0.741±0.063 
782.9  40:4 0.602±0.034 0.655±0.081 0.584±0.063 
784.9  40:3 0.536±0.032 0.564±0.042 0.327±0.018#§ 
786.9  40:2 0.399±0.019 0.454±0.025 0.292±0.033§ 
788.9 40:8 40:1 0.360±0.017 0.407±0.021 0.278±0.029§ 
790.9 40:7 40:0 1.32±0.14 1.16±0.15 1.35±0.07 
792.9 40:6  1.86±0.09 1.54±0.03 1.79±0.11 
794.9 40:5  1.26±0.07 1.12±0.03 1.18±0.03 
796.9 40:4  0.797±0.059 0.811±0.140 0.604±0.072 
798.9 40:3  0.596±0.050 0.558±0.119 0.411±0.046 
800.9 40:2  1.221±0.072 1.063±0.048 0.884±0.057# 
802.9 40:1  0.751±0.055 0.641±0.039 0.556±0.023# 
*

P< 0.05, significant difference between HEK-293 and HEK/PEMT-L cells.

#P< 0.05, significant difference between HEK-293 and HEK/PEMT-S cells.

§

P< 0.05, significant difference between HEK/PEMT-L and HEK/PEMT-S cells.

PE-O includes plasmanylethanolamine and plasmenylethanolamine.

The CDP-ethanolamine pathway was used for the synthesis of both PE and PE-O in the heart, liver and kidney, whereas the PS decarboxylase pathway was used solely for PE, but not PE-O, synthesis [47]. In McA-RH7777 cells and CHO (Chinese-hamster ovary) cells, plasmenylethanolamine species are exclusively synthesized through the CDP-ethanolamine pathway, and PS decarboxylation only marginally contributes to the formation of plasmenylethanolamine [46]. In contrast, a serine residue serves as a precursor of the head group of PS, PE and PE-O in cultured glioma cells [48]. Minor amounts of serine-linked ether-phospholipids can be found in animal cells (<0.2% of total phospholipid mass) [6]. In the present study, HEK-293 cells were cultured in MEM containing no ethanolamine. It is unclear whether PE-O in HEK-293 cells without exogenous ethanolamine was synthesized through the PS decarboxylation and/or the CDP-ethanolamine pathway using ethanolamine or phosphoethanolamine released from PE hydrolysed by PLD or phospholipase C respectively.

Plasmanylcholine is synthesized through the CDP-choline pathway. Plasmenylethanolamine, but not plasmanylcholine, is the precursor for plasmenylcholine. Alkenylglycerol derived from plasmenylethanolamine is converted into plasmenylcholine. Lee et al. [49] suggested that AdoMet-dependent methylation of plasmenylethanolamine in the synthesis of plasmenylcholine is a minor contributing pathway in neonatal rat myocytes. The synthesis of plasmenylcholine by the methylation pathway has also been demonstrated in rat brain and rabbit myocardial membranes [50,51]. Although the brain and heart contain high levels of plasmalogens [6], the expression of PEMT in these tissues is relatively low compared with that in the liver [8]. Vance [52] reported that rat hepatocytes secrete nascent lipoproteins containing a relatively large amount of plasmenylethanolamine, although there was little plasmenylethanolamine in the hepatocytes. Plasmenylethanolamine and plasmenylcholine in low-density lipoproteins may function as antioxidants [53]. The molar ratio of plasmenylcholine to plasmenylethanolamine in the serum correlates with the particle size of low-density lipoproteins, suggesting that the declining conversion from plasmenylethanolamine into plasmenylcholine is related to the appearance of atherogenic small dense low-density lipoproteins [54]. Lack of PEMT reduces the molar ratio of PC/PE in nascent very-low-density lipoproteins and the atherosclerotic lesion area [16]. Therefore the formation of plasmenylcholine from plasmenylethanolamine by PEMT may be related to the atherogenicity of lipoproteins.

Effect of PEMT expression on cell growth

To explore the relationship between cell growth and PEMT expression in HEK-293 cells, we compared the proliferation rates of HEK-293, HEK/PEMT-L and HEK/PEMT-S cells. Figure 6 shows the exponential increase in cell density. HEK/PEMT-L and HEK/PEMT-S cells proliferated faster than HEK-293 cells. The doubling times of HEK/PEMT-L and HEK/PEMT-S cells between days 1 and 4 (43.3±0.6 and 46.5±0.8 h respectively) were markedly shorter than that of HEK-293 cells (64.8±2.5 h) (P<0.001). Thus the expression of PEMT-L and PEMT-S promoted the growth of HEK-293 cells. It has been reported that PEMT expression in McA-RH7777 rat hepatoma cells induces cell growth inhibition and apoptosis, and that the expression of PEMT decreases the PE content, but does not affect PC content in the hepatoma cells [55,56]. In contrast, the expression of PEMT in CHO cells has been shown to have no significant effect on cell growth or on the cellular levels of PC and PE [43,55,57]. In addition, Emoto et al. [58,59] have shown that cell growth during cytokinesis ceases for mutant cells that have a specific reduction in cellular PE level, indicating that PE is an essential molecule for the completion of cytokinesis. These findings suggest that a severe loss of cellular PE leads to the suppression of cell growth, whereas an increase in cellular PC production promotes cell growth, probably due to the enhanced formation of cell membranes. PC-O and its metabolites may also be associated with increased cell growth.

Effect of PEMT expression on cell proliferation

Figure 6
Effect of PEMT expression on cell proliferation

HEK-293 cells (open circles), HEK/PEMT-L cells (filled circles) and HEK/PEMT-S cells (filled triangles) were cultured in six-well plates in MEM containing 10% (v/v) FBS at 37 °C. Results shown are means±S.E.M. of three measurements.

Figure 6
Effect of PEMT expression on cell proliferation

HEK-293 cells (open circles), HEK/PEMT-L cells (filled circles) and HEK/PEMT-S cells (filled triangles) were cultured in six-well plates in MEM containing 10% (v/v) FBS at 37 °C. Results shown are means±S.E.M. of three measurements.

Membrane topology of PEMT

Shields et al. [40] have demonstrated the cytosolic orientation of the C-terminus of PEMT and the luminal orientation of loops A and C by endoproteinase-protection analysis, and proposed a topological model of PEMT in which four transmembrane regions span the membrane such that both the N- and C-termini are localized external to the ER. However, our present results showed the luminal orientation of the PEMT-L N-terminal region modified with high-mannose-type N-linked oligosaccharides. Hydropathy analysis of PEMT-L using the method of Kyte and Doolittle [60] predicted the presence of four hydrophobic regions (Supplementary Figure S3 at http://www.BiochemJ.org/bj/432/bj4320387add.htm). The prediction of transmembrane domains using the computer program SOSUI [61] suggested that PEMT-L contains three transmembrane helices. The HA-tag appended to the N-terminus of PEMT-L did not change the number or length of the predicted transmembrane domains. Therefore we propose a new model of PEMT topology where three transmembrane regions (amino acids 83–105, 131–153 and 194–216) are integrated in the ER membrane, and the first hydrophobic region (amino acids 43–65) is localized in the ER lumen or associated with the ER membrane, but does not span the membrane (Figure 7). It is possible that the first hydrophobic region forms a helical hairpin in the membrane, because three proline residues are located in this region. In this model, the N- and C-termini of PEMT are localized in the ER lumen and cytosol respectively. This model does not contradict the observations from the endoproteinase-protection analysis reported by Shields et al. [40]. In addition, the length or glycosylation in the N-terminal region may modulate the enzymatic activity and substrate specificity of PEMT. The pool of cellular AdoMet exists mainly in the cytosol [62], and two AdoMet-binding motifs of PEMT are positioned towards the cytosolic face of the second and third transmembrane helices. It is unclear which residues are critical for the binding of PE to PEMT or how the PEMT N-terminal region localized in the ER lumen regulates the enzymatic reaction.

Topological model of PEMT

Figure 7
Topological model of PEMT

Transmembrane domains were predicted using the program SOSUI [61]. Three transmembrane domains (amino acids 83–105, 131–153 and 194–216) span the ER membrane, and the first hydrophobic region (amino acids 43–65) is a cytosolic or membrane-associated domain. The N- and C-termini are localized in the ER lumen and cytosol respectively. NL and NS indicate the N-termini of PEMT-L and PEMT-S respectively. The C-terminus is indicated by C. The residue Asn13 is N-glycosylated with a high-mannose oligosaccharide.

Figure 7
Topological model of PEMT

Transmembrane domains were predicted using the program SOSUI [61]. Three transmembrane domains (amino acids 83–105, 131–153 and 194–216) span the ER membrane, and the first hydrophobic region (amino acids 43–65) is a cytosolic or membrane-associated domain. The N- and C-termini are localized in the ER lumen and cytosol respectively. NL and NS indicate the N-termini of PEMT-L and PEMT-S respectively. The C-terminus is indicated by C. The residue Asn13 is N-glycosylated with a high-mannose oligosaccharide.

In conclusion, PEMT-S showed higher activity than PEMT-L. By using novel enzyme-based fluorimetric assays for measuring PC and PE, we demonstrated that the cellular PC content, PE content and PE/PC ratio were highly dependent on cell density, and that PEMT-L and PEMT-S expression markedly increased PC content in HEK-293 cells, whereas the PE content was significantly decreased by the expression of PEMT-S, but not PEMT-L. We also showed that the cellular PS was decreased by the expression of PEMT-L or PEMT-S. ESI–MS/MS analyses revealed that the levels of PC and PC-O species with longer polyunsaturated chains were higher in HEK/PEMT-S cells than HEK/PEMT-L cells, whereas PEMT-L expression increased the levels of PC-O species with shorter chains more than PEMT-S expression, suggesting that there is a difference in the substrate specificity of PEMT-L and PEMT-S. In addition, bothPEMT-L and PEMT-S expression enhanced the proliferation of HEK-293 cells. Our results also suggest a model in which the enzyme activity and specificity of PEMT are controlled by its N-terminal region localized in the ER lumen.

We thank Dr Kazumitsu Ueda (Graduate School of Agriculture and Institute for Integrated Cell-Material Sciences, Kyoto University, Kyoto, Japan) for invaluable discussions.

Abbreviations

     
  • AdoMet

    S-adenosylmethionine

  •  
  • p-APMSF

    (p-amidinophenyl)methanesulfonyl fluoride

  •  
  • CDP

    cytidine diphosphate

  •  
  • CHO

    Chinese-hamster ovary

  •  
  • Endo H

    endoglycosidase H

  •  
  • ER

    endoplasmic reticulum

  •  
  • ESI–MS/MS

    electrospray ionization–tandem MS

  •  
  • FBS

    fetal bovine serum

  •  
  • HA

    haemagglutinin

  •  
  • HEK

    human embryonic kidney

  •  
  • LPC

    lysophosphatidylcholine

  •  
  • LPE

    lysophosphatidylethanolamine

  •  
  • MEM

    minimum essential medium

  •  
  • PA

    phosphatidic acid

  •  
  • PC

    phosphatidylcholine

  •  
  • PC-O

    ether-linked PC

  •  
  • PE

    phosphatidylethanolamine

  •  
  • PE-O

    ether-linked PE

  •  
  • PEMT

    PE N-methyltransferase

  •  
  • PEMT-L

    longer isoform of PEMT

  •  
  • PEMT-S

    shorter isoform of PEMT

  •  
  • PG

    phosphatidylglycerol

  •  
  • PI

    phosphatidylinositol

  •  
  • PLD

    phospholipase D

  •  
  • GPL-PLD

    glycerophospholipid-specific PLD

  •  
  • PNGase F

    peptide N-glycosidase F

  •  
  • PS

    phosphatidylserine

  •  
  • POPC

    L-α-palmitoyl-oleoyl PC

  •  
  • POPE

    L-α-palmitoyl-oleoyl PE

  •  
  • RT

    reverse transcription

  •  
  • SM

    sphingomyelin

  •  
  • WT

    wild-type

AUTHOR CONTRIBUTION

Shin-ya Morita designed the research, performed the experiments, analysed the data and wrote the manuscript. Atsuko Takeuchi helped with MS analysis. Shuji Kitagawa provided scientific expertise.

FUNDING

This work was supported in part by a Grant-in-Aid for Young Scientists (B) from MEXT (Ministry of Education, Culture, Sports, Science and Technology, Japan) [grant number 22790053], a Japan Heart Foundation Young Investigator's Research Grant, and by a grant from Hyogo Science and Technology Association.

References

References
1
Vance
J. E.
Phosphatidylserine and phosphatidylethanolamine in mammalian cells: two metabolically related aminophospholipids
J. Lipid Res.
2008
, vol. 
49
 (pg. 
1377
-
1387
)
2
Li
Z.
Vance
D. E.
Phosphatidylcholine and choline homeostasis
J. Lipid Res.
2008
, vol. 
49
 (pg. 
1187
-
1194
)
3
Arondel
V.
Benning
C.
Somerville
C. R.
Isolation and functional expression in Escherichia coli of a gene encoding phosphatidylethanolamine methyltransferase (EC 2.1.1.17) from Rhodobacter sphaeroides
J. Biol. Chem.
1993
, vol. 
268
 (pg. 
16002
-
16008
)
4
Vance
D. E.
Li
Z.
Jacobs
R. L.
Hepatic phosphatidylethanolamine N-methyltransferase, unexpected roles in animal biochemistry and physiology
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
33237
-
33241
)
5
Shindou
H.
Hishikawa
D.
Harayama
T.
Yuki
K.
Shimizu
T.
Recent progress on acyl CoA: lysophospholipid acyltransferase research
J. Lipid Res.
2009
, vol. 
50
 (pg. 
S46
-
S51
)
6
Nagan
N.
Zoeller
R. A.
Plasmalogens: biosynthesis and functions
Prog. Lipid Res.
2001
, vol. 
40
 (pg. 
199
-
229
)
7
Walkey
C. J.
Shields
D. J.
Vance
D. E.
Identification of three novel cDNAs for human phosphatidylethanolamine N-methyltransferase and localization of the human gene on chromosome 17p11.2
Biochim. Biophys. Acta
1999
, vol. 
1436
 (pg. 
405
-
412
)
8
Shields
D. J.
Agellon
L. B.
Vance
D. E.
Structure, expression profile and alternative processing of the human phosphatidylethanolamine N-methyltransferase (PEMT) gene
Biochim. Biophys. Acta
2001
, vol. 
1532
 (pg. 
105
-
114
)
9
Resseguie
M.
Song
J.
Niculescu
M. D.
da Costa
K. A.
Randall
T. A.
Zeisel
S. H.
Phosphatidylethanolamine N-methyltransferase (PEMT) gene expression is induced by estrogen in human and mouse primary hepatocytes
FASEB J.
2007
, vol. 
21
 (pg. 
2622
-
2632
)
10
Shields
D. J.
Lingrell
S.
Agellon
L. B.
Brosnan
J. T.
Vance
D. E.
Localization-independent regulation of homocysteine secretion by phosphatidylethanolamine N-methyltransferase
J. Biol. Chem.
2005
, vol. 
280
 (pg. 
27339
-
27344
)
11
DeLong
C. J.
Shen
Y. J.
Thomas
M. J.
Cui
Z.
Molecular distinction of phosphatidylcholine synthesis between the CDP-choline pathway and phosphatidylethanolamine methylation pathway
J. Biol. Chem.
1999
, vol. 
274
 (pg. 
29683
-
29688
)
12
Walkey
C. J.
Yu
L.
Agellon
L. B.
Vance
D. E.
Biochemical and evolutionary significance of phospholipid methylation
J. Biol. Chem.
1998
, vol. 
273
 (pg. 
27043
-
27046
)
13
Song
J.
da Costa
K. A.
Fischer
L. M.
Kohlmeier
M.
Kwock
L.
Wang
S.
Zeisel
S. H.
Polymorphism of the PEMT gene and susceptibility to nonalcoholic fatty liver disease (NAFLD)
FASEB J.
2005
, vol. 
19
 (pg. 
1266
-
1271
)
14
Dong
H.
Wang
J.
Li
C.
Hirose
A.
Nozaki
Y.
Takahashi
M.
Ono
M.
Akisawa
N.
Iwasaki
S.
Saibara
T.
Onishi
S.
The phosphatidylethanolamine N-methyltransferase gene V175M single nucleotide polymorphism confers the susceptibility to NASH in Japanese population
J. Hepatol.
2007
, vol. 
46
 (pg. 
915
-
920
)
15
Li
Z.
Agellon
L. B.
Allen
T. M.
Umeda
M.
Jewell
L.
Mason
A.
Vance
D. E.
The ratio of phosphatidylcholine to phosphatidylethanolamine influences membrane integrity and steatohepatitis
Cell Metab.
2006
, vol. 
3
 (pg. 
321
-
331
)
16
Zhao
Y.
Su
B.
Jacobs
R. L.
Kennedy
B.
Francis
G. A.
Waddington
E.
Brosnan
J. T.
Vance
J. E.
Vance
D. E.
Lack of phosphatidylethanolamine N-methyltransferase alters plasma VLDL phospholipids and attenuates atherosclerosis in mice
Arterioscler. Thromb. Vasc. Biol.
2009
, vol. 
29
 (pg. 
1349
-
1355
)
17
Morita
S. Y.
Kobayashi
A.
Takanezawa
Y.
Kioka
N.
Handa
T.
Arai
H.
Matsuo
M.
Ueda
K.
Bile salt-dependent efflux of cellular phospholipids mediated by ATP binding cassette protein B4
Hepatology
2007
, vol. 
46
 (pg. 
188
-
199
)
18
Chen
Y. Q.
Kuo
M. S.
Li
S.
Bui
H. H.
Peake
D. A.
Sanders
P. E.
Thibodeaux
S. J.
Chu
S.
Qian
Y. W.
Zhao
Y.
, et al. 
AGPAT6 is a novel microsomal glycerol-3-phosphate acyltransferase
J. Biol. Chem.
2008
, vol. 
283
 (pg. 
10048
-
10057
)
19
Jain
S.
Zhang
X.
Khandelwal
P. J.
Saunders
A. J.
Cummings
B. S.
Oelkers
P.
Characterization of human lysophospholipid acyltransferase 3
J. Lipid Res.
2009
, vol. 
50
 (pg. 
1563
-
1570
)
20
Newman
E. A.
Muh
S. J.
Hovhannisyan
R. H.
Warzecha
C. C.
Jones
R. B.
McKeehan
W. L.
Carstens
R. P.
Identification of RNA-binding proteins that regulate FGFR2 splicing through the use of sensitive and specific dual color fluorescence minigene assays
RNA
2006
, vol. 
12
 (pg. 
1129
-
1141
)
21
Ridgway
N. D.
Vance
D. E.
Phosphatidylethanolamine N-methyltransferase from rat liver
Methods Enzymol.
1992
, vol. 
209
 (pg. 
366
-
374
)
22
Bligh
E. G.
Dyer
W. J.
A rapid method of total lipid extraction and purification
Can. J. Biochem. Physiol.
1959
, vol. 
37
 (pg. 
911
-
917
)
23
Zhou
X.
Arthur
G.
Improved procedures for the determination of lipid phosphorus by malachite green
J. Lipid Res.
1992
, vol. 
33
 (pg. 
1233
-
1236
)
24
Stone
S. J.
Cui
Z.
Vance
J. E.
Cloning and expression of mouse liver phosphatidylserine synthase-1 cDNA: overexpression in rat hepatoma cells inhibits the CDP-ethanolamine pathway for phosphatidylethanolamine biosynthesis
J. Biol. Chem.
1998
, vol. 
273
 (pg. 
7293
-
7302
)
25
Leidl
K.
Liebisch
G.
Richter
D.
Schmitz
G.
Mass spectrometric analysis of lipid species of human circulating blood cells
Biochim. Biophys. Acta
2008
, vol. 
1781
 (pg. 
655
-
664
)
26
Taguchi
R.
Houjou
T.
Nakanishi
H.
Yamazaki
T.
Ishida
M.
Imagawa
M.
Shimizu
T.
Focused lipidomics by tandem mass spectrometry
J. Chromatogr. B Analyt. Technol. Biomed. Life Sci.
2005
, vol. 
823
 (pg. 
26
-
36
)
27
Takayama
M.
Itoh
S.
Nagasaki
T.
Tanimizu
I.
A new enzymatic method for determination of serum choline-containing phospholipids
Clin. Chim. Acta
1977
, vol. 
79
 (pg. 
93
-
98
)
28
Hojjati
M. R.
Jiang
X. C.
Rapid, specific, and sensitive measurements of plasma sphingomyelin and phosphatidylcholine
J. Lipid Res.
2006
, vol. 
47
 (pg. 
673
-
676
)
29
Morita
S. Y.
Ueda
K.
Kitagawa
S.
Enzymatic measurement of phosphatidic acid in cultured cells
J. Lipid Res.
2009
, vol. 
50
 (pg. 
1945
-
1952
)
30
Artiss
J. D.
McGowan
M. W.
Strandbergh
D. R.
Epstein
E.
Zak
B.
Enzymic colorimetric determination of phosphatidylglycerol in amniotic fluid
Clin. Chem.
1984
, vol. 
30
 (pg. 
534
-
537
)
31
Blaton
V.
De Buyzere
M.
Spincemaille
J.
Declercq
B.
Enzymic assay for phosphatidylcholine and sphingomyelin in serum
Clin. Chem.
1983
, vol. 
29
 (pg. 
806
-
809
)
32
He
X.
Chen
F.
McGovern
M. M.
Schuchman
E. H.
A fluorescence-based, high-throughput sphingomyelin assay for the analysis of Niemann–Pick disease and other disorders of sphingomyelin metabolism
Anal. Biochem.
2002
, vol. 
306
 (pg. 
115
-
123
)
33
Kishimoto
T.
Soda
Y.
Matsuyama
Y.
Mizuno
K.
An enzymatic assay for lysophosphatidylcholine concentration in human serum and plasma
Clin. Biochem.
2002
, vol. 
35
 (pg. 
411
-
416
)
34
Aoki
J.
Taira
A.
Takanezawa
Y.
Kishi
Y.
Hama
K.
Kishimoto
T.
Mizuno
K.
Saku
K.
Taguchi
R.
Arai
H.
Serum lysophosphatidic acid is produced through diverse phospholipase pathways
J. Biol. Chem.
2002
, vol. 
277
 (pg. 
48737
-
48744
)
35
Ota
H.
Tamezane
H.
Sasano
Y.
Hokazono
E.
Yasuda
Y.
Sakasegawa
S.
Imamura
S.
Tamura
T.
Osawa
S.
Enzymatic characterization of an amine oxidase from Arthrobacter sp. used to measure phosphatidylethanolamine
Biosci. Biotechnol. Biochem.
2008
, vol. 
72
 (pg. 
2732
-
2738
)
36
Zhou
M.
Diwu
Z.
Panchuk-Voloshina
N.
Haugland
R. P.
A stable nonfluorescent derivative of resorufin for the fluorometric determination of trace hydrogen peroxide: applications in detecting the activity of phagocyte NADPH oxidase and other oxidases
Anal. Biochem.
1997
, vol. 
253
 (pg. 
162
-
168
)
37
Ohta-Fukuyama
M.
Miyake
Y.
Emi
S.
Yamano
T.
Identification and properties of the prosthetic group of choline oxidase from Alcaligenes sp
J. Biochem.
1980
, vol. 
88
 (pg. 
197
-
203
)
38
Weerheim
A. M.
Kolb
A. M.
Sturk
A.
Nieuwland
R.
Phospholipid composition of cell-derived microparticles determined by one-dimensional high-performance thin-layer chromatography
Anal. Biochem.
2002
, vol. 
302
 (pg. 
191
-
198
)
39
Wakelam
M. J.
Pettitt
T. R.
Postle
A. D.
Lipidomic analysis of signaling pathways
Methods Enzymol.
2007
, vol. 
432
 (pg. 
233
-
246
)
40
Shields
D. J.
Lehner
R.
Agellon
L. B.
Vance
D. E.
Membrane topography of human phosphatidylethanolamine N-methyltransferase
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
2956
-
2962
)
41
Shields
D. J.
Altarejos
J. Y.
Wang
X.
Agellon
L. B.
Vance
D. E.
Molecular dissection of the S-adenosylmethionine-binding site of phosphatidylethanolamine N-methyltransferase
J. Biol. Chem.
2003
, vol. 
278
 (pg. 
35826
-
35836
)
42
Voelker
D. R.
Organelle biogenesis and intracellular lipid transport in eukaryotes
Microbiol. Rev.
1991
, vol. 
55
 (pg. 
543
-
560
)
43
Lee
M. W.
Bakovic
M.
Vance
D. E.
Overexpression of phosphatidylethanolamine N-methyltransferase 2 in CHO-K1 cells does not attenuate the activity of the CDP-choline pathway for phosphatidylcholine biosynthesis
Biochem. J.
1996
, vol. 
320
 (pg. 
905
-
910
)
44
Ridgway
N. D.
Vance
D. E.
Specificity of rat hepatic phosphatidylethanolamine N-methyltransferase for molecular species of diacyl phosphatidylethanolamine
J. Biol. Chem.
1988
, vol. 
263
 (pg. 
16856
-
16863
)
45
Sato
H.
Taketomi
Y.
Isogai
Y.
Miki
Y.
Yamamoto
K.
Masuda
S.
Hosono
T.
Arata
S.
Ishikawa
Y.
Ishii
T.
, et al. 
Group III secreted phospholipase A2 regulates epididymal sperm maturation and fertility in mice
J. Clin. Invest.
2010
, vol. 
120
 (pg. 
1400
-
1414
)
46
Bleijerveld
O. B.
Brouwers
J. F.
Vaandrager
A. B.
Helms
J. B.
Houweling
M.
The CDP-ethanolamine pathway and phosphatidylserine decarboxylation hatidylethanolamine molecular species
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
28362
-
28372
)
47
Arthur
G.
Page
L.
Synthesis of phosphatidylethanolamine and ethanolamine plasmalogen by the CDP-ethanolamine and decarboxylase pathways in rat heart, kidney and liver
Biochem. J.
1991
, vol. 
273
 (pg. 
121
-
125
)
48
Xu
Z. L.
Byers
D. M.
Palmer
F. B.
Spence
M. W.
Cook
H. W.
Serine utilization as a precursor of phosphatidylserine and alkenyl-(plasmenyl)-, alkyl-, and acylethanolamine phosphoglycerides in cultured glioma cells
J. Biol. Chem.
1991
, vol. 
266
 (pg. 
2143
-
2150
)
49
Lee
T. C.
Qian
C. G.
Snyder
F.
Biosynthesis of choline plasmalogens in neonatal rat myocytes
Arch. Biochem. Biophys.
1991
, vol. 
286
 (pg. 
498
-
503
)
50
Mozzi
R.
Siepi
D.
Andreoli
V.
Porcellati
G.
The synthesis of choline plasmalogen by the methylation pathway in rat brain
FEBS Lett.
1981
, vol. 
131
 (pg. 
115
-
118
)
51
Mogelson
S.
Sobel
B. E.
Ethanolamine plasmalogen methylation by rabbit myocardial membranes
Biochim. Biophys. Acta
1981
, vol. 
666
 (pg. 
205
-
211
)
52
Vance
J. E.
Lipoproteins secreted by cultured rat hepatocytes contain the antioxidant 1-alk-1-enyl-2-acylglycerophosphoethanolamine
Biochim. Biophys. Acta
1990
, vol. 
1045
 (pg. 
128
-
134
)
53
Jurgens
G.
Fell
A.
Ledinski
G.
Chen
Q.
Paltauf
F.
Delay of copper-catalyzed oxidation of low density lipoprotein by in vitro enrichment with choline or ethanolamine plasmalogens
Chem. Phys. Lipids
1995
, vol. 
77
 (pg. 
25
-
31
)
54
Maeba
R.
Maeda
T.
Kinoshita
M.
Takao
K.
Takenaka
H.
Kusano
J.
Yoshimura
N.
Takeoka
Y.
Yasuda
D.
Okazaki
T.
Teramoto
T.
Plasmalogens in human serum positively correlate with high-density lipoprotein and decrease with aging
J. Atheroscler. Thromb.
2007
, vol. 
14
 (pg. 
12
-
18
)
55
Cui
Z.
Houweling
M.
Vance
D. E.
Suppression of rat hepatoma cell growth by expression of phosphatidylethanolamine N-methyltransferase-2
J. Biol. Chem.
1994
, vol. 
269
 (pg. 
24531
-
24533
)
56
Tessitore
L.
Sesca
E.
Bosco
M.
Vance
D. E.
Expression of phosphatidylethanolamine N-methyltransferase in Yoshida ascites hepatoma cells and the livers of host rats
Carcinogenesis
1999
, vol. 
20
 (pg. 
561
-
567
)
57
Houweling
M.
Cui
Z.
Vance
D. E.
Expression of phosphatidylethanolamine N-methyltransferase-2 cannot compensate for an impaired CDP-choline pathway in mutant Chinese hamster ovary cells
J. Biol. Chem.
1995
, vol. 
270
 (pg. 
16277
-
16282
)
58
Emoto
K.
Kobayashi
T.
Yamaji
A.
Aizawa
H.
Yahara
I.
Inoue
K.
Umeda
M.
Redistribution of phosphatidylethanolamine at the cleavage furrow of dividing cells during cytokinesis
Proc. Natl. Acad. Sci. U.S.A.
1996
, vol. 
93
 (pg. 
12867
-
12872
)
59
Emoto
K.
Umeda
M.
An essential role for a membrane lipid in cytokinesis: regulation of contractile ring disassembly by redistribution of phosphatidylethanolamine
J. Cell Biol.
2000
, vol. 
149
 (pg. 
1215
-
1224
)
60
Kyte
J.
Doolittle
R. F.
A simple method for displaying the hydropathic character of a protein
J. Mol. Biol.
1982
, vol. 
157
 (pg. 
105
-
132
)
61
Hirokawa
T.
Boon-Chieng
S.
Mitaku
S.
SOSUI: classification and secondary structure prediction system for membrane proteins
Bioinformatics
1998
, vol. 
14
 (pg. 
378
-
379
)
62
Farooqui
J. Z.
Lee
H. W.
Kim
S.
Paik
W. K.
Studies on compartmentation of S-adenosyl-L-methionine in Saccharomyces cerevisiae and isolated rat hepatocytes
Biochim. Biophys. Acta
1983
, vol. 
757
 (pg. 
342
-
351
)

Supplementary data