A Single Enzyme Catalyzes Both Platelet-activating Factor Production and Membrane Biogenesis of Inflammatory Cells

Platelet-activating factor (PAF) is a potent proinflammatory lipid mediator eliciting a variety of cellular functions. Lipid mediators, including PAF are produced from membrane phospholipids by enzymatic cascades. Although a G protein-coupled PAF receptor and degradation enzymes have been cloned and characterized, the PAF biosynthetic enzyme, aceyl-CoA:lyso-PAF acetyltransferase, has not been identified. Here, we cloned lyso-PAF acetyltransferase, which is critical in stimulus-dependent formation of PAF. The enzyme is a 60-kDa microsomal protein with three putative membrane-spanning domains. The enzyme was induced by bacterial endotoxin (lipopolysaccharide), which was suppressed by dexamethasone treatment. Surprisingly, the enzyme catalyzed not only biosynthesis of PAF from lyso-PAF but also incorporation of arachidonoyl-CoA to produce PAF precursor membrane glycerophospholipids (lysophosphatidylcholine acyltransferase activity). Under resting conditions, the enzyme prefers arachidonoyl-CoA and contributes to membrane biogenesis. Upon acute inflammatory stimulation with lipopolysaccharide, the activated enzyme utilizes acetyl-CoA more efficiently and produces PAF. Thus, our findings provide a novel concept that a single enzyme catalyzes membrane biogenesis of inflammatory cells while producing a prophlogistic mediator in response to external stimuli.

A G protein-coupled PAF receptor was cloned in our laboratory (17), and PAF acetylhydrolases have been cloned and characterized by others (18,19). Lyso-PAF acetyltransferase was initially demonstrated and partially characterized by Wykle et al. (14) in 1980. Since this first report, the enzyme activity has been detected in microsomes of rat spleen and lung as well as porcine leukocytes (14,20). Although several groups have attempted to identify and characterize the enzyme from various sources (2,4,14,21,22), it has not yet been purified nor cDNA-cloned.
PC is a major component of cellular membranes and also plays an important role as a PAF precursor. PC is formed from diacylglycerol by a de novo pathway, originally described by Kennedy (23) in 1961 but is also generated from LPC by a remodeling pathway. Rapid turnover of the sn-2 acyl moiety of glycerophospholipids was described by Lands (Lands' cycle) (24 -26) and is attributed to activation of phospholipases A 2 and lysophospholipid acyltransferases. Recently, we (27) and Chen et al. (28) independently cloned one of LPC acyltrans-* This work was supported in part by Grants-in-Aid from the Ministry of Education, Science, Culture, Sports, and Technology of Japan (to T. S. and S. I.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ferases designated LPCAT1, which is highly expressed in alveolar type II cells. Since PC synthesis occurs in a variety of different tissues, additional LPC acyltransferases may be present for membrane biogenesis. Using the previously reported LPCAT1 (27) and an extensive genomic data base search as well as 5Ј-and 3Ј-RACE, we have identified a lyso-PAF acetyltransferase gene. The enzyme is primarily expressed in inflammatory cells and is induced by LPS. Surprisingly, the enzyme also catalyzes incorporation of arachidonoyl-CoA to produce PAF precursor membrane glycerophospholipids (LPC acyltransferase activity). Thus, we designated this enzyme LysoPAFAT/LPCAT2. Although this enzyme possesses both acetyltransferase and acyltransferase activity, only the acetyltransferase activity was enhanced by acute inflammatory signals. To our knowledge, this is the first documentation of a cDNA for LysoPAFAT/LPCAT2, a critically important enzyme in the biogenesis of PAF and in membrane homeostasis of inflammatory cells. Cloning of LysoPAFAT/LPCAT2-The mLysoPAFAT/ LPCAT2 gene was identified based upon sequence similarity to the LPCAT1 gene (27) and LPA acyltransferase gene (29) through a comprehensive basic local alignment search tool (BLAST) search. A 1.6-kb cDNA clone encoding the full-length mLysoPAFAT/LPCAT2 (DDBJ accession number AB244716) was cloned by PCR amplification using the forward primer 5Ј-CTAGCTAGCCACCATGGATTACAAGGATGACGAT-GACAAGAACCGATGCGCCGAGGCGGCCGC-3Ј, the reverse primer 5Ј-CCGCTCGAGTCAGTCCACCTTTTTGTC-TGAGGTGCCCTC-3Ј, and a mouse spleen cDNA library as a template. The FLAG epitope (DYKDDDDK) was attached to the N terminus of mLysoPAFAT/LPCAT2 by PCR using a forward primer. Amplified PCR products were cloned into the pCXN2.1 vector, a slightly modified version of pCXN2 (30) with multiple cloning sites, and sequenced. Similarly, hLysoPAFAT/LPCAT2 cDNA (DDBJ accession number AB244718) was amplified by PCR and inserted into the pCXN2.1 vector.
Expression of FLAG-mLysoPAFAT/LPCAT2 in CHO-K1 Cells-After 48 h of transfection with FLAG-tagged enzyme using Lipofectamine 2000 (Invitrogen), cells in 10-cm dishes were scraped into 1 ml of ice-cold buffer containing 20 mM Tris-HCl (pH 7.4), 300 mM sucrose, and a proteinase inhibitor mixture, Complete (Roche Applied Science), and then sonicated three times on ice for 30 s. After centrifugation for 10 min at 800 ϫ g, the supernatant was collected and centrifuged at 100,000 ϫ g for 1 h. The resulting pellet was resuspended in buffer containing 20 mM Tris-HCl (pH 7.4), 300 mM sucrose, and 1 mM EDTA. Protein concentration was measured by the method of Bradford (31), using a commercially prepared protein assay solution (Bio-Rad) and bovine serum albumin (fraction V, fatty acid-free; Sigma) as a standard.
Assay of Lyso-PAF Acetyltransferase-The lyso-PAF acetyltransferase activity was determined according to the method of Kume et al. (20), except for washing resin one time before and eight times after application of the reaction mixture. Briefly, 100 M [ 3 H]acetyl-CoA (1.11 GBq/mmol) and protein were incubated at 37°C for 10 min in the presence or absence of 20 M lyso-PAF (Cayman). Subsequently, the product was bound to C8 resin (Millipore), washed, and eluted. The difference between the radioactivities obtained in the presence and absence of lyso-PAF corresponded to lyso-PAF acetyltransferase activity.
Assay of Lysophospholipid Acetyltransferase and Acyltransferase by TLC-The acyltransferase activity was measured in two ways: (i) conversion of l. After incubation at 37°C for 10 min, the reaction was stopped by the addition of 0.3 ml of chloroform/methanol (1:2, v/v). Total lipids were extracted using the Bligh-Dyer method (32) and subsequently analyzed by TLC in chloroform/methanol/acetic acid/water (50:25:8:4, v/v/v/v). Bands at positions corresponding to the expected product were visualized by I 2 vapor, cut from the plate, placed in Microscinti-O (Packard Bioscience), and analyzed in a liquid scintillation counter LS6500 (Beckman).
Radioligand Binding Assay-The method of PAF-PAF receptor (PAFR) binding assay was described previously (7,33). Briefly, the membrane fraction containing 158 fmol of PAFR from hearts and skeletal muscles of PAFR transgenic mice (7,34) were mixed with 25 nM [ 3 H]WEB 2086 and the lipid extract in a 96-well plate. After incubation at 25°C for 90 min, receptor-bound [ 3 H]WEB 2086 was collected by filtration through a UniFilter-GF/C (PerkinElmer Life Sciences) using a MicroMate 196 simultaneous 96-well harvester (PerkinElmer Life Sciences), and the filter was washed and dried. Subsequently, the radioactivities were counted with a TopCount microplate scintillation counter (PerkinElmer Life Sciences).
Electrospray Ionization Mass Spectrometry Analysis of PAF-Extracted lipid from the acetyltransferase assay was identified by electrospray ionization mass spectrometry analysis. The analysis was performed using a 4000 Q-TRAP quadrupole-linear ion trap hybrid mass spectrometer (Applied Biosystems/ MDS Sciex, Concord, Canada) with an Ultimate 3000 high pressure liquid chromatography system (DIONEX Co.) combined with an HTC PAL autosampler (CTC Analytics, Zwingen, Switzerland). The extracted lipids were subjected to electrospray ionization mass spectrometry analysis by flow injection without liquid chromatography separation. The solvent was acetonitrile, methanol, 50 mM ammonium formate, pH 7.4 (v/v/v, 45/50/5), and the flow rate was 10 l/min. The scan range and speed were set at m/z 500 -600 and 1000 Da/s, respectively. The trap fill time was set at 5 ms, and the ion spray voltage was set at Ϫ4500 V in the negative ion mode. Nitrogen was used as curtain and collision gas. The declustering potential was set at 20 V to minimize in-source fragmentation. Both Q1 and Q3 resolution were set to unit mass. The collision energy used was varied according to the desired experiment. The method to identify phosphatidylcholine species was described previously (35).
Short Term LPS Stimulation-After transfection of RAW264.7 cells with LysoPAFAT/LPCAT2 using Lipofectamine 2000, cells were pretreated with or without 20 M SB 203580 for 1 h and subsequently stimulated with 100 ng/ml LPS for 30 min. For preparation of cell extracts, the cells were scraped into 600 l of an ice-cold buffer containing 20 mM Tris-HCl, (pH 7.4), 50 mM ␤-glycerophosphate, 1 mM sodium orthovanadate, 5 mM 2-mercaptoethanol, 20 M 4-amidinophenylmethanesulfonyl fluoride, and Complete, and the col-lected cells were sonicated twice on ice for 30 s. Intact cells, cellular debris, and mitochondria were removed by centrifugation at 9000 ϫ g for 10 min at 4°C. Enzyme activities were measured as described above.
Isolation and Stimulation of Mouse Peritoneal Cells-Mouse peritoneal macrophages induced by thioglycollate (Difco) was prepared as described in detail previously (11). The cells were treated with 100 ng/ml LPS, 0.8 M ODN1826, or 1 g/ml poly(I:C), in the presence or absence of 100 nM DEX or 100 nM estradiol-17␤ for 16 h. After treatment, the cells were washed with an ice-cold buffer containing 20 mM Tris-HCl (pH 7.4) and 300 mM sucrose. Cell extracts were prepared by the same method as described for experiments using RAW264.7 cells, and the enzyme activity was measured.
After treatment with microbial components for 16 h, total RNA was collected using the Absolutely RNA RT-PCR miniprep kit. Likewise, at 4 h after intraperitoneal injection with 2 ml of 2% casein, peritoneal exudate neutrophils were harvested from the peritoneal cavity, and their total RNA was prepared.
Statistics-Data are presented as mean Ϯ S.E. or S.D. p values less than 0.05 were considered statistically significant. All statistical calculations were performed using Prism 4 (GraphPad Software) and StatView-J, version 5.0 (Abacus Concepts, Berkeley, CA).
Mice-C57BL/6J mice were obtained from Clea Japan, Inc. (Tokyo, Japan). Mice were maintained in a light-dark cycle with light from 8:00 to 20:00 at 21°C. Mice were fed with a standard laboratory diet and water ad libitum. All animal studies were conducted in accordance with the guidelines for Animal Research at The University of Tokyo and were approved by the University of Tokyo Ethics Committee for Animal Experiments.
Tissue Distribution and Subcellular Localization of mLyso-PAFAT/LPCAT2 mRNA-The tissue distribution of mLyso-PAFAT/LPCAT2 was analyzed by quantitative PCR. We found the highest level of mLysoPAFAT/LPCAT2 expression in resident macrophages, casein-induced neutrophils, followed by skin, colon, spleen, and thioglycollate-induced macrophages ( Fig. 2A). To facilitate immunocytochemical analysis of mLysoPAFAT/LPCAT2, we constructed an expression vector encoding FLAG-tagged mLysoPAFAT/LPCAT2. It was transfected into CHO-K1, and the enzyme distribution was examined by confocal microscopy after 48 h. FLAG-mLyso-PAFAT/LPCAT2 exhibited significant enzyme activities (Fig. 3, A and B). Cells were stained for ER-Golgi using DiOC 6 (3). The subcellular distribution pattern of FLAG-mLysoPAFAT/LPCAT2 was similar to that of DiOC 6 (3), suggesting that the enzyme is present mainly in the ER and Golgi (Fig. 2B).
Substrate Selectivity of mLysoPAFAT/LPCAT2-We next examined the acetyltransferase activity of mLysoPAFAT/ LPCAT2 using a variety of 20 M lysophospholipid acceptors and 100 M [ 3 H]acetyl-CoA as a donor. mLysoPAFAT/ LPCAT2 had detectable acetyltransferase activity toward LPC (Fig. 3A) and alkyl-LPC (Fig. 3B). The enzyme had a higher activity toward C16 lysophospholipids than C18 at the sn-1 moiety, which served as an acetyl acceptor (Fig. 3B). The acetyltransferase activity of mLysoPAFAT/LPCAT2 was linear for the first 40 min at 37°C, and the enzyme exhibited calcium-dependent activity with a pH optimum around 7.4 (data not shown). To confirm whether the product of the acetyltransferase reaction is PAF, we first performed a competitive receptor binding assay using PAFR prepared from PAFR-transgenic mice (7, 34). Lipid extracts from the acetyl-transferase reaction of mLy-soPAFAT/LPCAT2 with lyso-PAF competed for binding of [ 3 H]WEB2086 to PAFR, indicating that the enzyme indeed produced PAF (supplemental Fig. 1). More definitively, the reaction product was identified by mass spectrometry (MS) using a 4000 Q-TRAP mass spectrometer (supplemental Fig. 2).
Enzyme Activation by an Inflammatory Stimulus-Next, we investigated the activation of the enzyme by an inflammatory stimulus, such as LPS. To examine the response to LPS stimulation, mLysoPAFAT/LPCAT2 was transfected into the macrophage cell line, RAW264.7, which expresses the LPS receptor, Toll-like receptor 4 (TLR4), and cells were stimulated by LPS for 30 min in the presence or absence of the p38 mitogen-activated protein kinase inhibitor SB 203580 (Tocris Cookson). The acetyltransferase activity of mLysoPAFAT/LPCAT2 was increased by LPS stimulation, but the effect was decreased in the presence of SB 203580; the acyltransferase activity of the enzyme was unchanged (Fig. 5). The endogenous lyso-PAF acyltransferase in RAW264.7 cells was activated by LPS, and this activation was blocked in the presence of SB 203580 (data not shown).
Induction of mLysoPAFAT/LPCAT2 mRNA-Next, we examined induction of mLysoPAFAT/LPCAT2 mRNA in response to long term treatment with Toll-like receptor agonists. Mouse thioglycollate-induced macrophages were treated with LPS (a TLR4 ligand), ODN1826 (a TLR9 ligand), or poly(I:C) (a TLR3 ligand) for 16 h in the presence or absence of DEX or estradiol-17␤. As shown in Fig. 6A, the endogenous lyso-PAF acetyltransferase activity was enhanced 2.4-and 2.2-fold by LPS and ODN1826 treatment, respectively. Moreover, augmentation of the enzyme activity by LPS treatment was suppressed in the presence of DEX but not estradiol-17␤. Similar results were obtained for the LPC acyltransferase activity (data not shown). The enzyme activation by ODN1826 also   The mLysoPAFAT/LPCAT2 mRNA levels were increased 7.3-and 4.8-fold by LPS and ODN1826 treatment, respectively. Furthermore, induction of mLysoPAFAT/LPCAT2 by LPS was repressed by DEX treatment (Fig. 6B). ODN1826 also enhanced mLysoPAFAT/LPCAT2 levels, which tended to be reduced by DEX treatment. The expression level of mLyso-PAFAT/LPCAT2 was not changed by poly(I:C) treatment, similar to the lack of effect on the enzyme activity (Fig. 6A). Induction of an IFN␥-inducible gene (IP-10) used as a positive control was observed by PCR after poly(I:C) stimulation under these conditions (data not shown).

DISCUSSION
This is the first report of isolation of a cDNA for LysoPAFAT/LPCAT2, a critically important enzyme in the biosynthesis of PAF. After long term treatments with a TLR4 or TLR9 agonist, the expression level of mLysoPAFAT/LPCAT2 mRNA was up-regulated, but not with a TLR3 agonist. Surprisingly, this enzyme catalyzed not only PAF biosynthesis (lyso-PAF acetyltransferase) but also generation of membrane glycerophospholipids (LPC acyltransferase), which are major membrane constituents and precursors of PAF. In the resting conditions, the enzyme prefers arachidonoyl-CoA to produce membrane lipids. However, under acute inflammatory stimulation by activating the TLR4, only the acetyltransferase activity of the enzyme was enhanced, and PAF production was augmented (Fig. 7).
Characterization of Lyso-PAF Acetyltransferase-mLyso-PAFAT/LPCAT2 possessed lyso-PAF acetyltransferase activity but did not show 1-alkyl-LPA acetyltransferase activity, which catalyzes the first step of the de novo PAF biosynthesis pathway (Fig. 3, A and B) (2,4). Using a heterologous overexpression system, we found that the enzyme was predominantly localized to the ER-Golgi complex (Fig. 2B). The exact localization of the endogenous protein in native cells remains to be determined. Additionally, LysoPAFAT/LPCAT2 possessed putative EFhand-like motifs and showed calcium-dependent activity. These correlations remain to be further clarified. Using hLysoPAFAT/LPCAT2 siRNA transfection of HEK293 cells, both the mRNA and the enzyme activity of endogenous lyso-PAF acetyltransferase were significantly reduced (Fig. 3C), indicating that LysoPAFAT/LPCAT2 constitutes a major (50 -80%) lyso-PAF acetyltransferase, in HEK293 cells. It is possible that other enzyme(s) are present also to catalyze PAF production in other tissues and cells.
LPC Acyltransferase Activity of LysoPAFAT/LPCAT2-mLyso-PAFAT/LPCAT2 exhibited not only significant lyso-PAF acetyltransferase activity but also LPC acyltransferase activity (Fig. 4) in vitro, indicating that mLysoPAFAT/LPCAT2 can produce both PAF and PC. These data agree well with the previous studies showing that the lyso-PAF acetyltransferase activity in neutrophils was competed by long chain acyl-CoAs (40,41). Recently, we and another group identified an LPC remod-  eling enzyme, designated LPCAT1, which is highly expressed in lung (27,28). In contrast, LysoPAFAT/LPCAT2 is predominantly expressed in inflammatory cells with modest expression in skin, brain, and colon. Because PC is biosynthesized in all cell types, a different class of LPCATs may exist in addition to LPCAT1 (27,28) and LysoPAFAT/LPCAT2 (present study).
An Acute Inflammatory Response-Upon acute inflammatory stimulation by LPS, the acetyltransferase activity of mLysoPAFAT/LPCAT2 was enhanced (Fig. 5). Similar activation has been observed with endogenous lyso-PAF acetyltransferase in mouse peritoneal macrophages (11). Nixon et al. reported that lyso-PAF acetyltransferase in human neutrophils is directly activated by p38 mitogen-activated protein kinase (42). It is possible that LysoPAFAT/LPCAT2 is modified and activated by phosphorylation after LPS-stimulation. In contrast, the LPC acyltransferase activity of the enzyme was not changed by short term LPS stimulation (Fig. 5). The cellular signaling pathway mediating LysoPAFAT/LPCAT2 activation by LPS remains to be elucidated. In addition to putative conformational changes of the enzyme, both enzyme activities may be regulated by the ratio of two substrates (acetyl-CoA and arachidonoyl-CoA) in cells.
LPS-induced accumulation of LysoPAFAT/LPCAT2 increases acyltransferase activity in addition to the acetyltransferase activity. The biological significance of up-regulation of the acyltransferase activity may be related to the fact that under long term LPS stimulation, cytosolic and secretory phospholipase A 2 are activated, leading to increased release of free fatty acids and lysophospholipids from PC (52,53). Lysophospholipids are toxic to cells because of their detergent effects. Alternatively, active membrane remodeling is required during inflammatory responses, such as phagocytosis or chemotaxis. LysoPAFAT/LPCAT2 may play an important role in the regulation of lysophospholipid and PAF levels and in the storage of PC as PAF precursor membrane glycerophospholipids.  LysoPAFAT/LPCAT2 activity is increased by two distinct pathways in mouse macrophages. Although LysoPAFAT/LPCAT2 catalyzes both acetyltransferase and acyltransferase, only acetyltransferase activity was enhanced under acute inflammatory conditions. GR, glucocorticoid receptor; p38, p38 mitogen-activated protein kinase; 20:4-CoA, arachidonoyl-CoA. See the first paragraph under "Discussion" for details.
Conclusion-We have isolated a new enzyme that catalyzes PAF production and membrane biogenesis (LysoPAFAT/ LPCAT2). Further studies are needed to elucidate the roles of mLysoPAFAT/LPCAT2 in vivo and to determine its potential as a novel therapeutic target for various diseases involving PAF biosynthesis. It will be important to characterize both acetyltransferase and acyltransferase activities of the enzyme, including identification of binding sites for each substrate (acetyl-CoA and arachidonoyl-CoA) and differential regulation of individual enzyme activity. Molecular cloning and characterization of this first LysoPAFAT/LPCAT2 will enable us to better understand the biochemical mechanisms underlying PAF and phospholipid biosynthesis in inflammatory cells.