Phosphorylation of Lysophosphatidylcholine Acyltransferase 2 at Ser³⁴ Enhances Platelet-activating Factor Production in Endotoxin-stimulated Macrophages

Abstract

Platelet-activating factor (PAF) is a potent proinflammatory phospholipid mediator that elicits various cellular functions under physiological and pathological conditions. We have recently identified two enzymes involved in PAF production: lysophosphatidylcholine acyltransferase-1 (LPCAT1) and LPCAT2. We found that LPCAT2 is highly expressed in inflammatory cells and is activated by lipopolysaccharide (LPS) treatment through Toll-like receptor 4. However, the molecular mechanism for the activation remains elusive. In this study, Phos-tag SDS-PAGE revealed the LPS-induced phosphorylation of LPCAT2. Furthermore, mass spectrometry and mutagenesis analyses identified Ser³⁴ of LPCAT2 as the phosphorylation site to enhance the catalytic activities. The experiments using inhibitors and siRNA against MAPK cascades demonstrated that LPCAT2 phosphorylation through LPS-TLR4 signaling may directly depend on MAPK-activated protein kinase 2 (MAPKAP kinase 2 or MK2). These findings develop a further understanding of both PAF production and phospholipid remodeling triggered by inflammatory stimuli. Specific inhibition of the PAF biosynthetic activity by phosphorylated LPCAT2 will provide a novel target for the regulation of inflammatory disorders.

Introduction

Platelet-activating factor (PAF³; 1-O-alkyl-2-acetyl-sn-glycero-3-phosphocholine) is a potent proinflammatory lipid mediator that triggers various cellular functions through its G protein-coupled receptor (PAF receptor) (1, 2). It is proposed that PAF is synthesized in various cells and tissues via two distinct pathways, the de novo and remodeling pathways. Through the remodeling pathway, PAF is rapidly synthesized in response to extracellular stimuli. Under such conditions, 1-O-alkyl-sn-glycero-3-phosphocholine (lyso-PAF), the precursor of PAF, is synthesized from 1-O-alkyl-2-arachidonoyl-sn-glycero-3-phosphocholine (1-alkyl phosphatidylcholine; PC) by the action of phospholipase A₂. Lyso-PAF is subsequently converted to PAF by acetyl-CoA:lyso-PAF acetyltransferase (lyso-PAFAT).

Endogenous lyso-PAFAT activity was initially demonstrated in 1980 (3) and partially characterized (4–7). Recently, we identified two molecular entities of lyso-PAFATs: a constitutively expressed lyso-PAFAT, LPCAT1 (lysophosphatidylcholine acyltransferase 1) (8), and an inducible lyso-PAFAT, LPCAT2 (9). In these previous reports, endogenous lyso-PAFAT in inflammatory cells was activated by prophlogistic stimuli (6) and LPCAT2 in mouse peritoneal macrophages was indeed activated by lipopolysaccharide (LPS) stimulation (9). However, the exact mechanisms for LPCAT2 activation remain unknown.

LPS activates TLR4 (Toll-like receptor 4), which plays a central role in the activation of the innate immune system. Through its association with different combinations of four adaptors, the TLR4 signaling pathway leads to the phosphorylation of MAPKs: p38, ERK, and JNK. Subsequently, activated p38 phosphorylates MK2 (MAPK-activated protein kinase 2) (10, 11), which can induce inflammatory cytokines (12, 13) and lipid mediators (14, 15).

LPCAT2 also possesses LPCAT activity to produce the major membrane phospholipid, PC, which mainly contains polyunsaturated fatty acids (PUFAs) at the sn-2 position. This biosynthetic pathway of phospholipids, known as Lands' cycle or remodeling pathway, is responsible for generating the membrane diversity (16). PUFAs in phospholipids may affect membrane curvature and fluidity and store lipid mediator precursors that are converted to eicosanoids, such as prostaglandins, leukotrienes, and lipoxins (1). PC plays an important role as a precursor of both eicosanoids and PAF.

By mass spectrometry and mutagenesis studies, we demonstrated that LPCAT2 is activated by Ser³⁴ phosphorylation in mouse peritoneal macrophages and RAW264.7 cells with LPS treatment. Consensus sequence and experiments with an MK2 inhibitor and siRNA suggested that MK2 might directly phosphorylate and activate LPCAT2. These findings contribute to a better understanding of the regulatory mechanisms of PAF biosynthesis in inflammatory cells.

EXPERIMENTAL PROCEDURES

Materials

PC from frozen egg yolk, LPS from Salmonella minnesota, and anti-FLAG M2 antibody were from Sigma. Lyso-PAF was from Cayman Chemical Co. (Ann Arbor, MI). Arachidonoyl-CoA was from Avanti Polar Lipids (Alabaster, AL). [³H]Acetyl-CoA (129.5 GBq/mmol), horseradish peroxidase-linked anti-rabbit IgG, and horseradish peroxidase-linked anti-mouse IgG were from GE Healthcare. [1-¹⁴C]Arachidonoyl-CoA (2.22 GBq/mmol) was from Moravec Biochemicals (Brea, CA). Thin layer chromatography (TLC) silica gel plates (type 5721) were from Merck. Cell line Nucleofector kit V was from LONZA (Basel, Switzerland). Acetyl-CoA, DMSO, and acrylamide/bis (29:1) were from WAKO (Osaka, Japan). Phos-tag acrylamide was from NARD Institute, Ltd. (Hyogo, Japan). (5Z)-7-Oxozeanol was from TOCRIS Bioscience (Ellisville, MO). SB202474, SB203580, and MK2 inhibitor III were from Calbiochem. The siRNAs (ON-TARGETplus Non-targeting Pool D-001810-10-20 and ON-TARGETplus SMARTpool L-040135-00-0005) were from Thermo Scientific (Dharmacon) (Waltham, MA). Anti-MK2, anti-phospho-MK2, anti-p38 MAPK, and anti-phospho-p38 MAPK antibodies were from Cell Signaling Technology (Beverly, MA). The proteinase inhibitor mixture, EDTA-free Complete, was from Roche Applied Science.

Mice

C57BL/6J mice were obtained from Clea Japan, Inc. (Tokyo, Japan). Mice were maintained in a light-dark cycle with lights on from 0800–2000 h at 22 °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.

Isolation of Mouse Peritoneal Macrophages

Mouse peritoneal macrophages were isolated as previously described (6). Cells were cultured for 16 h before stimulation.

Preparation of Cell Lysates

Cells were pretreated with or without 20 μm MK2 inhibitor III, 20 μm SB203580 (p38 MAPK inhibitor), or 1 μm (5Z)-7-oxozeaenol (TAK1 (tumor growth factor-β-activated protein kinase 1) inhibitor) for 1 h and then stimulated with 100 ng/ml LPS for 30 min. After stimulation, cells (peritoneal macrophages or RAW264.7 cells) were washed with ice-cold buffer containing 20 mm Tris-HCl (pH 7.4), 0.3 m sucrose, and 1 mm sodium orthovanadate and collected in buffer containing 20 mm Tris-HCl (pH 7.4), 1 mm sodium orthovanadate, 5 mm 2-mercaptoethanol, and 1× EDTA-free Complete. Subsequently, cells were sonicated twice on ice for 30 s each time and centrifuged at 9,000 × g for 10 min at 4 °C to remove cellular debris, intact cells, and mitochondria.

For primary cultured mouse peritoneal macrophages, the resultant supernatant at 9,000 × g was centrifuged at 100,000 × g for 1 h at 4 °C. The resultant pellet was resuspended with ice-cold buffer containing 20 mm Tris-HCl (pH 7.4), 1 mm sodium orthovanadate, 5 mm 2-mercaptoethanol, 1× EDTA-free Complete. The concentration of each protein was measured by the Bradford method (17), using protein assay solution (Bio-Rad). Bovine serum albumin (fraction V, fatty acid-free; Sigma) served as a standard.

Site-directed Mutagenesis of LPCAT2

Mouse LPCAT2 mutants (S34A and S34D) were constructed by overlap extension PCR. The amplified PCR products were cloned into the pCXN2.1 vector, and the sequence was confirmed. The primer sets utilized were S34A (forward, CGC CAG GCG GCC TTC TTC CCG CCG C; reverse, GCG GCG GGA AGA AGG CCG CCT GGC G); and S34D (forward, CGC CAG GCG GAC TTC TTC CCG CCG C; reverse, GCG GCG GGA AGA AGT CCG CCT GGC G).

Transfection into RAW264.7 Cells

RAW264.7 cells (5 × 10⁶ cells), 100 μl of Nucleofector solution V, and 5 μg of each DNA of vector, FLAG-mLPCAT2, S34A, or S34D, were mixed. The mixture in the cuvette was set onto the Amaxa Nucleofector and electroporated with the program D-032. Then cells were seeded onto 6-cm dishes. Twenty-four hours after transfection, cells were stimulated with 100 ng/ml LPS for 30 min. The siRNA transfection was performed similarly. The mixture in the cuvette contained 120 pmol of siRNA.

Production of Anti-LPCAT2 and Anti-phospho-LPCAT2 Antibodies

Anti-LPCAT2 antiserum was generated at Immuno-Biological Laboratories (Gunma, Japan). The C-terminal peptide, SNKVSPESQEEGTSDKKVD, was used to immunize rabbits. Anti-LPCAT2 antibody was purified from the anti-LPCAT2 antiserum using activated thiol-Sepharose 4B binding to the LPCAT2 epitope. Anti-phospho-LPCAT2 antibody was generated by SCRUM (Tokyo, Japan) using a phosphopeptide, RQApSFFPPP (where pS represents phosphoserine) at the N terminus of LPCAT2.

Western Blot Analysis

Western blot analyses were performed as described previously (18). To detect the band shift, which represents phosphorylated protein, an SDS-polyacrylamide gel containing 50 μm Phos-tag acrylamide with 100 μm Mn²⁺ was used.

Assay of Lyso-PAF Acetyltransferase and LPCAT

Lyso-PAF acetyltransferase and LPCAT assays were performed as described previously (8, 9).

Quantitative Real-time PCR

Total RNAs were prepared using the RNeasy Mini Kit (Qiagen), and first strand cDNA was subsequently synthesized using Superscript III (Invitrogen). The PCRs were performed using Fast Start DNA Master SYBR Green I (Roche Applied Science). The primers for MK2 designed to amplify a 185-bp fragment were as follows: forward, GGA TCT TCG ACA AGA GAA CCC AG; reverse, GAG ACA CTC CAT GAC AAT CAG CA).

Software

All statistical calculations were performed using Prism 4 (GraphPad Software). Alignment of mammal LPCAT2 was performed using GENETYX-MAC version 13.0.6 (GENETYX Corp.). Sequences of mouse (BAF47695), human (BAF47696), bovine (XP_592529), dog (XP_854080), and rat (XP_001064713) LPCAT2 are available in the DDBJ/EMBL/GenBank^TM databases.

RESULTS

Phosphorylation of LPCAT2 by LPS Stimulation

To examine the different characteristics of the two lyso-PAFATs (LPCAT1 and LPCAT2), FLAG-tagged LPCAT1 and LPCAT2 were transiently transfected into the mouse macrophage cell line RAW264.7 using the Amaxa Nucleofector transfection kit V. Because RAW264.7 cells express TLR4 signaling molecules, cells were stimulated with LPS for 30 min, and the lyso-PAFAT activity was examined using the supernatant at 9,000 × g for 10 min. The lyso-PAFAT activities of LPCAT1 and LPCAT2 were measured by radioisotope assays. Although the LPCAT1 activity was unchanged after LPS stimulation, the LPCAT2 activity was enhanced 4-fold compared with non-stimulated LPCAT2 (Fig. 1A). Lyso-PAFAT activity in the vector-transfected cells was slightly increased by LPS stimulation, possibly due to the presence of endogenously expressed LPCAT2 in RAW264.7 cells.

FIGURE 1.

LPCAT2 activation and phosphorylation. RAW264.7 cells transfected with vector, FLAG-LPCAT1, or FLAG-LPCAT2 were stimulated with 100 ng/ml LPS for 30 min. A, LPS enhanced the activity of LPCAT2 but not that of LPCAT1. Open bars and closed bars indicate vehicle or LPS stimulation, respectively. Results are expressed as the mean ± S.D. (error bars) of an experiment performed in triplicate. Three independent experiments showed similar results. B, Western blot analysis was performed after Phos-tag SDS-PAGE. Only LPCAT2 showed the shifted band with LPS stimulation indicating its phosphorylation. C, alignment of LPCAT2 in various species. The newly identified phosphorylation site, Ser³⁴, was well conserved among mammals. For the accession numbers of LPCAT2, see “Experimental Procedures.” Statistical analyses were performed by analysis of variance and Tukey's multiple comparison test.

The mechanism of LPCAT2 activation was investigated using Phos-tag acrylamide gel electrophoresis. Phos-tag makes a complex with two Mn²⁺ ions and acts as a phosphate-binding molecule (19). The complex is used for phosphate affinity SDS-PAGE, which results in the mobility shift of the phosphorylated proteins. A shifted band of FLAG-LPCAT2, but not FLAG-LPCAT1, was observed after LPS stimulation (Fig. 1B). The upper band may represent the phosphorylated form of LPCAT2. This result suggests that LPCAT2 is phosphorylated and activated by extracellular stimuli.

To identify the phosphorylated amino acid residue(s) of LPCAT2, RAW264.7 cells stably overexpressing FLAG-LPCAT2 were established using Fugene HD in the presence of Geneticin. The cells were stimulated with LPS for 30 min, and the pellet at 100,000 × g for 1 h was analyzed by Phos-tag SDS-PAGE. The position corresponding to the shifted band in the Phos-tag Western blot was cut and subjected to in-gel trypsin digestion (20). After immobilized metal affinity chromatography enrichment of phosphopeptides (21), only one phospho-LPCAT2 peptide candidate (³²QApSFFPPPVPNPFVQQTTISASR⁵⁴) was detected by liquid chromatography-mass spectrometry (LTQ, Thermo Electron, San Jose, CA) (data not shown). Peptides containing unphosphorylated Ser³⁴ were not detected in the phosphopeptide-enriched fraction. The flow-through fraction of immobilized metal affinity chromatography contained several other unphosphorylated peptides derived from LPCAT2. Although the Mascot score was 38, which is not significant, these results suggest that Ser³⁴ of LPCAT2 is a candidate residue of the phosphorylation induced by LPS stimulation. Ser³⁴ of mouse LPCAT2 is well conserved among mammals, such as human, bovine, dog, and rat (Fig. 1C).

Site-directed Mutagenesis of LPCAT2

To confirm Ser³⁴ as the target of phosphorylation, site-directed mutagenesis of LPCAT2 was performed. Ser³⁴ was substituted for alanine (S34A) and aspartate (S34D). These constructs were transiently transfected into RAW264.7 cells using Amaxa, and the cells were stimulated with LPS for 30 min. In the Phos-tag Western blot analysis using the M2 anti-FLAG antibody, a mobility shift was detected in wild-type (WT) LPCAT2 but not in the S34A or S34D mutant (Fig. 2A).

FIGURE 2.

Site-directed mutagenesis of LPCAT2. RAW264.7 cells transfected with vector, wild-type (WT), S34A, or S34D were stimulated with 100 ng/ml LPS for 30 min. A, in Phos-tag Western blot analysis, only WT with LPS stimulation showed the shifted band. The two mutants showed no shift. Lyso-PAFAT (B) and LPCAT (C) activities were measured. S34A and S34D did not show the activation, and the activity of S34D was already high without LPS. Open bars and closed bars indicate vehicle or LPS stimulation, respectively. Results are expressed as the mean ± S.D. (error bars) of an experiment performed in triplicate. Four independent experiments were performed with similar results. Statistical analyses were performed by analysis of variance and Tukey's multiple comparison test.

Next, we examined the effect of phosphorylation on the dual activities of LPCAT2 (lyso-PAFAT and LPCAT). Both activities of mutants were measured by radioisotope assays. Lyso-PAFAT and LPCAT activities were enhanced in WT LPCAT2 with LPS stimulation (Fig. 2, B and C). The enzyme activity of S34A was similar to WT but was not increased by LPS stimulation. In contrast, S34D exhibited a higher enzyme activity than WT, but no further stimulation was observed (Fig. 2B). The expression level of each mutant was similar to that of WT (Fig. 2A). These results indicate that both lyso-PAFAT and LPCAT activities were enhanced by the Ser³⁴ phosphorylation of LPCAT2.

Signaling Pathway for LPCAT2 Phosphorylation

To investigate the time course of LPCAT2 phosphorylation, thioglycolate-induced murine peritoneal macrophages were stimulated with LPS for varying times (0–120 min). Each microsomal protein (pellet at 100,000 × g for 1 h) was analyzed by Western blot using anti-LPCAT2 and anti-phospho-LPCAT2 antibodies. The amount of total LPCAT2 was nearly equal among the samples. The most intense phospho-LPCAT2 signal was detected at 15–30 min and decreased as the incubation continued until 120 min (Fig. 3). This is consistent with lyso-PAFAT activation in our previous study (6). Similarly, MK2 phosphorylation reached a peak at 15–30 min. The consensus phosphorylation sequence (HydXRXXS; where Hyd represents a hydrophobic residue) of MK2 substrates (22) is conserved around Ser³⁴ (VPRQAS) in LPCAT2 (Fig. 1C). These results suggest that LPCAT2 is one of the protein substrates of MK2. Murine MK2 has two splice variant proteins (23), and thus MK2 appeared at the positions of 45 and 55 kDa by the Western blot.

FIGURE 3.

Time course for LPCAT2 phosphorylation. Mouse thioglycolate-induced peritoneal macrophages were stimulated with 100 ng/ml LPS for the indicated periods. Microsomal proteins (100,000 × g pellets for 60 min) were analyzed by Western blot. Phosphorylated LPCAT2 and MK2 appeared within 15–30 min. The arrowheads indicate MK2 splice variants. Three independent experiments were performed with similar results.

The signal transduction pathway for LPCAT2 phosphorylation was studied using several inhibitors of TLR4 signaling molecules: TAK1 (tumor growth factor-β-activated kinase-1), p38 MAPK, and MK2 (see Fig. 6). RAW264.7 cells overexpressing FLAG-LPCAT2 were pretreated with each inhibitor for 1 h and stimulated with LPS for 30 min. Treatment with (5Z)-7-oxozeaenol (a TAK1 inhibitor) abolished the phosphorylation of p38 MAPK, MK2, and LPCAT2 (Fig. 4A). SB203580 (a p38 MAPK inhibitor) also inhibited the phosphorylation of MK2 and LPCAT2, whereas the inactive analogue, SB202474, did not affect their phosphorylation (Fig. 4B). Pyrrolopiridine (MK2 inhibitor III) (24) treatment diminished the phosphorylation of LPCAT2 (Fig. 4C). Combined with the consensus sequence of the MK2 substrates, these data strongly suggested that LPS-induced phosphorylation of LPCAT2 is dependent on MK2, a downstream kinase of TAK1 and p38 (see Fig. 6).

FIGURE 6.

A proposed scheme of the molecular mechanism underlying LPCAT2 activation. The results of the present study indicate that LPCAT2 phosphorylation under LPS stimulation depends on the MyD88 (myeloid differentiation primary response gene 88), TAK1, p38α, and MK2 signaling pathway. Ser³⁴ is the only phosphorylated site of LPCAT2 that enhances its catalytic activities. ER, endoplasmic reticulum.

FIGURE 4.

Signaling pathway for LPCAT2 phosphorylation. RAW264.7 cells transfected with LPCAT2 were preincubated with or without each inhibitor for 1 h and subsequently stimulated with 100 ng/ml LPS for 30 min. See also Fig. 5. Supernatants (9,000 × g for 10 min) were subjected to Western blot analysis. 1 μm (5Z)-7-oxozeaenol (A), 20 μm SB203580 (B), and MK2 inhibitor III (C) abolished LPCAT2 phosphorylation. SB202474 is an inactive analogue of SB203580. The arrowheads indicate MK2 splice variants. Three independent experiments were performed with similar results.

Suppression of LPCAT2 Phosphorylation by MK2 siRNA

The involvement of MK2 in LPCAT2 phosphorylation was further examined using the MK2 knockdown (MK2-KD) of RAW264.7 cells stably expressing LPCAT2. MK2 siRNA was transiently transfected into RAW264.7 cells by the Amaxa Nucleofector transfection kit V. After 48 h, the cells were treated with LPS for 30 min. The level of MK2 mRNA expression was decreased by 70–80% in MK2 siRNA-transfected cells (MK2-KD) compared with cells transfected with negative control (NC) siRNA (Fig. 5A). The supernatant at 9,000 × g for 10 min was analyzed by Western blot using anti-MK2, anti-phospho-MK2, anti-LPCAT2, and anti-phospho-LPCAT2 antibodies. The amounts of total MK2 and phosphorylated MK2 were decreased in MK2-KD cells (Fig. 5B), consistent with their MK2 mRNA levels. Although the amount of total LPCAT2 was nearly equal in both MK2-KD and NC cells, that of phospho-LPCAT2 was significantly diminished in MK2-KD cells.

FIGURE 5.

MK2-dependent LPCAT2 phosphorylation. MK2 siRNA was transfected into RAW264.7 cells stably expressing WT LPCAT2. The levels of MK2 mRNA (A) and MK2 protein (B) were decreased in the MK2-KD cells. Levels of phospho-MK2 and phospho-LPCAT2 were lower in MK2-KD than in negative control cells (NC). C and D, LPS stimulation induced neither lyso-PAFAT nor LPCAT activation in MK2-KD. The arrowheads indicate MK2 splice variants. The open bars and closed bars indicate vehicle and LPS stimulation, respectively. Results are expressed as the mean ± S.D. (error bars) of an experiment performed in triplicate. Five independent experiments were performed with similar results. Statistical analyses were performed by analysis of variance and Tukey's multiple comparison test.

We also performed enzymatic assays and examined the effect of MK2 siRNA on LPCAT2 activation. Both lyso-PAFAT and LPCAT activities were enhanced by LPS stimulation in the NC cells; however, both activations were abolished in MK2-KD cells (Fig. 5, C and D). These results are consistent with the effect of the MK2 inhibitor on LPCAT2 phosphorylation (Fig. 4C) and thus indicate the MK2-dependent phosphorylation of LPCAT2.

DISCUSSION

Here, we present the activation mechanism of PAF biosynthetic enzyme by endotoxin stimulation. In response to inflammatory stimuli, LPCAT2 was phosphorylated and activated in mouse peritoneal macrophages and RAW264.7 cells. Mass spectrometry and mutagenesis analyses identified Ser³⁴ of LPCAT2 as the phosphorylation site to enhance the enzymatic activities. MK2 inhibitor and siRNA suppressed LPCAT2 phosphorylation, suggesting that LPCAT2 might be directly phosphorylated by MK2 to promote PAF and PC biosynthesis (Fig. 6).

In 1980, the lyso-PAFAT activity as the PAF biosynthetic enzyme was reported (3). Since then, several groups have attempted to characterize the enzyme. Lyso-PAFAT is rapidly activated in response to extracellular stimuli, such as calcium ionophore (4), acid stress (7), and LPS (16). However, neither the lyso-PAFAT cDNA sequence nor the mechanism of lyso-PAFAT activation had been elucidated. Recently, we identified two types of lyso-PAFATs: LPCAT2, which is an inducible lyso-PAFAT (9), and LPCAT1, which has constitutive lyso-PAFAT activity (8). LPCAT2 mRNA in macrophages is also up-regulated by LPS treatment for 16 h (9). The difference between LPCAT1 and LPCAT2 resembles that of cyclooxygenase-1 and -2 to produce prostaglandins (25, 26). In mouse peritoneal macrophages, LPCAT2 is activated within 30 min by LPS stimulation (Fig. 3), consistent with the characteristics of endogenous lyso-PAFAT (6, 9).

In this study, phosphorylated LPCAT2 was detected with the Phos-tag Western blot by mobility shift (Figs. 1B and 2A). Through mass spectrometric analysis of the phosphorylated enzyme, Ser³⁴ was identified as a phosphorylation site. Both the band shift and the activation were observed in WT LPCAT2, whereas the S34A mutant displayed neither characteristic (Fig. 2). Because mutagenesis at Ser³⁴ did not abolish the basal activities, it is proposed that Ser³⁴ is located in a regulatory region of LPCAT2. Moreover, the mutagenesis study indicated that Ser³⁴ was the only target of the phosphorylation that led to the enzymatic activation of LPCAT2. Furthermore, Ser³⁴ phosphorylation enhanced both the lyso-PAFAT and LPCAT activities of LPCAT2 (Fig. 2).

The activation of LPCAT2 in LPS-stimulated RAW264.7 cells was dependent on MK2 located downstream of p38 MAPK. Both p38α and p38δ are mainly expressed in macrophages (27), and p38α and p38β signals are inhibited by SB203580. Thus, Figs. 3 and 4 suggest an LPCAT2 phosphorylation mediated by p38α-MK2 axis. MK2 induces the phosphorylation of its substrates with the consensus sequence (HydXRXXS) (12, 22). Near the N terminus of LPCAT2, ²⁹VPRQAS³⁴ was detected as corresponding to the consensus sequence. Thus, MK2 may directly phosphorylate LPCAT2, although it is possible that other kinases are present to link the two proteins. Future determination of the three-dimensional structure of LPCAT2 should definitively clarify this activation mechanism.

LPCAT2 has lyso-PAFAT and LPCAT activities, both of which are enhanced by LPS stimulation. The endogenous LPCAT activity in RAW264.7 cells was much higher than its lyso-PAFAT activity (Fig. 2, B and C). Thus, activated LPCAT2 may function as a lyso-PAFAT to produce PAF. It is also possible that the LPCAT activity of LPCAT2 plays an important role in the storage of phospholipid precursors of PAF and eicosanoids (1). LPCAT2 catalyzes the membrane biogenesis (LPCAT activity) of inflammatory cells while producing PAF (lyso-PAFAT activity) in response to external stimuli. Further studies are needed to elucidate the physiological and pathological importance of these dual activities.

This is the first report on the posttranslational modification of lysophospholipid acyltransferases functioning in Lands' cycle. Our results showed that LPCAT2, a member of the lysophospholipid acyltransferases, can produce lipid mediator and may contribute to membrane dynamics in response to extracellular stimuli.

This study will aid in the development of new anti-inflammatory drugs that inhibit PAF production by exogenous insults while maintaining the constitutive levels of the mediator. Because of the physiologically important roles of PAF, PAF receptor antagonists have encountered several adverse effects during drug development. Inhibition of inducible PAF production by phospho-LPCAT2, but not unphospho-LPCAT2 or LPCAT1 (constitutive lyso-PAFAT), could serve as a potential target of medical interventions. These findings improve our understanding of both inflammatory responses and membrane biogenesis.