Lysophosphatidylcholine stimulates the release of arachidonic acid in human endothelial cells.

Lysophosphatidylcholine (lyso-PC) is a product of phosphatidylcholine hydrolysis by phospholipase A2 (PLA2) and is present in cell membranes, oxidized lipoproteins, and atherosclerotic tissues. It has the ability to alter endothelial functions and is regarded as a causal agent in atherogenesis. In this study, the modulation of arachidonate release by lyso-PC in human umbilical vein endothelial cells was examined. Incubation of endothelial cells with lyso-PC resulted in an enhanced release of arachidonate in a time- and concentration-dependent manner. Maximum arachidonate release was observed at 10 min of incubation with 50 microM lyso-PC. Lyso-PC species containing palmitoyl (C16:0) or stearoyl (C18:0) groups elicited the enhancement of arachidonate release, while other lysolipids such as lysophosphatidylethanolamine, lysophosphatidylserine, lysophosphatidylinositol, or lysophosphatidate were relatively ineffective. Lyso-PC-induced arachidonate release was decreased by treatment of cells with PLA2 inhibitors such as para-bromophenacyl bromide and arachidonoyl trifluoromethyl ketone. Furthermore, arachidonate release was attenuated in cells grown in the presence of antisense oligodeoxynucleotides that specifically bind cytosolic PLA2 mRNA. Treatment of cells with lyso-PC resulted in a translocation of PLA2 activity from the cytosolic to the membrane fractions of cells. Lyso-PC induced a rapid influx of Ca2+ from the medium into the cells, with a simultaneous enhancement of protein kinase C (PKC) activity in the membrane fractions. The lyso-PC-induced arachidonate release was attenuated when cells were preincubated with specific inhibitors of PKC (staurosporine and Ro31-8220) or a specific inhibitor of mitogen-activated protein kinase/extracellular regulated kinase kinase (PD098059). Taken together, the results of this study show that lyso-PC caused the elevation of cellular Ca2+ and the activation of PKC, which stimulated cytosolic PLA2 in an indirect manner and resulted in an enhanced release of arachidonate.

The release of arachidonate from phospholipids is the ratelimiting step in the synthesis of eicosanoids via the arachidonate cascade (1). Arachidonate and its metabolites possess diverse biological properties, many of which are related to vascular homeostasis (1). In endothelial cells, arachidonate is converted to prostacyclin, a potent vasodilator and platelet antiaggregator (2). Although different mechanisms have been proposed for the release of arachidonate in mammalian cells, the hydrolysis of the acyl chain at the sn-2 position of glycerophospholipids by phospholipase A 2 (PLA 2 ) 1 is regarded as the primary pathway for this reaction (1,3). In mammalian cells, several forms of PLA 2 have been identified. Those that have been purified and well characterized include the "type II" 14-kDa secretory PLA 2 (sPLA 2 ) and the "type IV" 85-kDa cytosolic PLA 2 (cPLA 2 ) (for reviews, see Refs. [3][4][5]. These two isoforms are products of distinct genes (5) and have different properties. The cPLA 2 preferentially hydrolyzes phospholipid substrates containing arachidonate at the sn-2 position (6), while sPLA 2 does not exhibit any preference with respect to substrate acyl composition. The sPLA 2 requires millimolar concentrations of Ca 2ϩ for maximum activity, while cPLA 2 contains a calciumdependent lipid binding domain and requires submicromolar levels for translocation to cellular membranes (6,7). In stimulated cells, cPLA 2 activity is enhanced by phosphorylation at serine 505 by mitogen-activated protein kinase (MAPK) (3,8). Protein kinase C (PKC) also appears to play a role in the regulation of PLA 2 activity, although PKC is not thought to directly phosphorylate cPLA 2 in vivo (3,9). Both isoforms are found in human endothelial cells and have been implicated in arachidonate release and prostacyclin production (10 -13).
Lysophosphatidylcholine (lyso-PC) is a product of phosphatidylcholine hydrolysis by PLA 2 . This lysophospholipid possesses detergent properties at high concentrations (14) but is quickly metabolized or reacylated within cells (15,16). Lyso-PC is a normal constituent of blood plasma (17), vascular tissue (18), and lipoproteins (19,20), but its levels are greatly elevated in hyperlipidemia (21), atherosclerotic tissue (18), oxidized lipoproteins (19,20), and ischemic hearts (22). A growing body of evidence has implicated lyso-PC in the pathogenesis of cardiovascular diseases. For example, lyso-PC in oxidized low density lipoproteins impairs vascular relaxation (20,23,24) and induces mitogenesis of macrophages (25). Lyso-PC is chemotactic for monocytes (26) and T lymphocytes (27). In endothelial cells, lyso-PC can induce the expression of genes for various growth factors (28,29) and cellular adhesion molecules (30,31). The perturbation of vascular endothelial function and recruitment of various cell types to sites of lesion have been implicated as early events in atherogenesis (32,33). Thus, given its many biological properties, lyso-PC has been postulated to be an important causal agent in inflammation and atherosclerosis (34,35).
The interactions among phosphatidylcholine, fatty acids, lyso-PC, and sPLA 2 have been examined in in vitro kinetic studies (36). An abrupt increase in PLA 2 activity after an initial lag period was observed in these studies. This pattern of activity was attributed to the accumulation of fatty acid and lysophospholipids, which together altered the organization of substrate vesicles (36). In light of the many biological effects of lyso-PC, we hypothesize that it can modulate PLA 2 in intact cells. In the present study, the effects of lyso-PC on the release of arachidonate in endothelial cells was examined. The involvement of Ca 2ϩ , PKC, and MAPK in the modulation of PLA 2 activity was examined.

EXPERIMENTAL PROCEDURES
Materials-Medium 199 with Hanks' salt and L-glutamine, heatinactivated fetal calf serum, and other standard culture reagents were obtained from Life Technologies, Inc. Type I collagenase was obtained from Worthington. Endothelial cell growth supplement was obtained from Collaborative Biomedical Products (Bedford, MA). Phorbol 12myristate 13-acetate, staurosporine, para-bromophenacyl bromide, and all other chemicals were purchased from Sigma. PD098059 was a product of Calbiochem. [5,6,8,11,12,14, H]arachidonate (230.5 Ci/mmol) was obtained from NEN Life Science Products, and 1-stearoyl-2-[1-14 C]arachidonoyl-L-3-phosphatidylcholine (55 mCi/mmol) was obtained from Amersham Corp. Arachidonoyl trifluoromethyl ketone (AACOCF 3 ) and H89 were obtained from Biomol Inc. (Plymouth Meeting, PA). Ro31-8220 was a gift from Roche Research Center (Welwyn Garden City, Hertfordshire, United Kingdom). Lysophospholipids and all lipid standards were obtained from Serdary Research Laboratory (London, Ontario, Canada). Thin layer chromatography plates (silica gel G) were products of Fisher. Anti-cPLA 2 polyclonal antibody was a generous gift from Drs. J. L. Knopf and L-L. Lin of the Genetics Institute (Boston, MA). Anti-human sPLA 2 monoclonal antibody was a product of Upstate Biotechnology Inc. (Lake Placid, NY).
Cell Culture-Endothelial cells were harvested from human umbilical veins using Type I collagenase as described previously (37,38). The cells were grown in flasks or culture dishes pretreated with 0.2% gelatin, in medium 199 (pH 7.4) supplemented with 25 mM HEPES, 30 g/ml endothelial cell growth supplement, 90 g/ml heparin, 10% fetal calf serum, 100 units/ml penicillin, 100 g/ml streptomycin, and 1.25 g/ml Fungizone. The cells were subcultured at a 1:3 ratio using 0.05% trypsin to free the cells from the culture ware. Near-confluent cell monolayers from the third passage were used for all experiments.
Radiolabeling and Challenge of Cells-Cells were radiolabeled as described previously (39). Cell monolayers grown to near-confluence in 35-mm culture dishes were incubated for 20 h with 1 Ci/ml [ 3 H]arachidonate in medium 199 containing 10% fetal calf serum. The cells were washed three times with HEPES-buffered saline (140 mM NaCl, 4 mM KCl, 5.5 mM glucose, 10 mM HEPES, 1.5 mM CaCl 2 , and 1.0 mM MgCl 2 , pH 7.4) containing 0.025% (w/v) essentially fatty acid-free bovine serum albumin. Aliquots of lysophospholipids were dissolved in chloroform/ methanol (2:1, v/v). The solvent was evaporated under N 2 , and the lysophospholipid samples were then resuspended in HEPES-buffered saline containing bovine serum albumin.
Measurement of Arachidonate Release-The arachidonate released from cells was determined as described previously (39). Briefly, the lysophospholipid was added to the cell culture and incubated for the prescribed period. The buffer was then removed and acidified with 50 l of glacial acetic acid. A 0.8-ml aliquot was used for lipid extraction in a solvent mixture consisting of chloroform/methanol/water (4:3:2, by volume). Oleic acid was added as an internal fatty acid standard. The free fatty acid fraction in the organic phase was resolved by thin layer chromatography in a solvent system consisting of hexane/diethyl ether/ acetic acid (70:30:1, v/v). The fatty acid fraction was visualized by iodine vapor, and its radioactivity was determined by liquid scintillation counting.
Binding of Lyso-PC to Endothelial Cells-Endothelial cells were cultured on 60-mm plates and incubated with medium 199 containing 100 nM [ 14 C]lyso-PC (57 nCi/nmol) for 15 min. The medium was removed, and the cells were incubated for 15 min with medium 199 (control) or medium 199 containing 10 M lyso-PC (a 100-fold excess of nonradioactive lyso-PC). The media were subsequently removed, and the cells were dislodged from the culture dish in HEPES-buffered saline. Samples were taken for protein determination or scintillation counting.
Immunoblotting Analysis of Phospholipase A 2 -Immunoblotting analysis of cPLA 2 or sPLA 2 was performed as described previously (39). Cell lysates containing approximately 50 g of protein were subjected to sodium dodecylsulfate, 7.5% polyacrylamide gel electrophoresis. The protein fractions from the gels were transferred to nitrocellulose membranes and then allowed to react with a polyclonal anti-cPLA 2 antibody or with an anti-sPLA 2 antibody. The nitrocellulose membranes were then exposed to a goat anti-rabbit antibody that was coupled to horseradish peroxidase. The cPLA 2 or sPLA 2 bands were detected on film using a Western blotting detection reagent kit (from Amersham), which yields a fluorescent compound via a reaction catalyzed by the peroxidase.
Oligonucleotide Treatment-The antisense oligonucleotides for group II PLA 2 (ASsA 2 , 5Ј-GAT CCT CTG CCA CCC ACA CC-3Ј) (40) and for cPLA 2 (AScA 2 , 5Ј-GTA AGG ATC TAT AAA TGA CAT-3Ј) (11) with phosphorothioate linkages were synthesized by the University Core DNA Services, University of Calgary (Alberta, Canada). Complementary sense oligomers were used as controls. Seventy-two hours prior to challenge with lyso-PC, the cells were incubated with medium containing 10 M oligonucleotides. The cells were supplied with fresh medium containing 10 M oligonucleotides at 24-h intervals thereafter. The presence of oligonucleotides did not affect cell viability or arachidonate labeling.
Determination of Phospholipase A 2 Activity-Cells were lysed by sonication in a buffer containing 50 mM Tris-HCl (pH 8.0), 1 mM EDTA, 10 M leupeptin, 10 M aprotinin, 20 mM NaF, and 10 mM Na 2 HPO 4 . Cell lysates were centrifuged at 100,000 ϫ g for 60 min. The supernatant was designated as the cytosolic fraction, while the pellet was designated as the membrane fraction and resuspended in the buffer described above. PLA 2 activity in the subcellular fractions was determined by the hydrolysis of 1-stearoyl-2-[1-14 C]arachidonoyl-sn-glycero-3-phosphocholine to yield free radiolabeled arachidonate. The assay mixture contained 50 mM Tris-HCl (pH 8.0), 1.5 mM CaCl 2 , 0.9 nmol of 1-stearoyl-2-[1-14 C]arachidonoyl-sn-glycero-3-phosphocholine (100,000 dpm/assay), and approximately 10 g of protein in a final volume of 100 l. The reaction mixtures were incubated at 37°C for 30 min, and the reactions were terminated by the addition of 1.5 ml of chloroform/ methanol (2:1, by volume). Total lipid was extracted, and the radioactivity of the arachidonate released was determined as described above. The amounts of protein in the samples were determined by the bicinchoninic acid method (41).
Monitoring of Intracellular Ca 2ϩ -Changes in cytosolic free Ca 2ϩ were monitored using the fluorescent Ca 2ϩ indicator fura-2 as described previously (42). Briefly, monolayers grown on microscope coverslips were incubated in medium with 5 M fura-2/AM for 30 min. Fura-2/AM is permeable to cells, and once inside the cells the compound is hydrolyzed by endogenous esterases to yield the cell-impermeable fura-2. The cells on the coverslip were transferred into a cuvette, rinsed with HEPES-buffered saline containing 0.025% bovine serum albumin, and immersed in the same buffer. Fluorescent signals were monitored on a SPEX fluorescence spectrophotometer at the excitation and emission wavelengths of 340 and 380 nm, respectively. Cells were then challenged with lyso-PC or A23187 for 10 min, and the ratio of the fluorescence at the two wavelengths was monitored as an indicator of changes in cytosolic Ca 2ϩ levels. The isosbestic (cross-over) point of fura-2 remained constant during lyso-PC treatment.
Determination of PKC Activity-Cells were sonicated in buffer B (50 mM Tris-HCl, pH 7.5, 5 mM EDTA, 10 mM EGTA, 0.25 M sucrose, 0.3% ␤-mercaptoethanol, 10 M benzamidine, 1 mM PMSF, 10 g/ml leupeptin, and 10 g/ml aprotinin) and were centrifuged at 1500 ϫ g for 10 min. The supernatants were subjected to ultracentrifugation at 100,000 ϫ g for 60 min to obtain the soluble and membrane fractions. Approximately 15-30 g of protein from these fractions were used to determine PKC activity using a PKC assay kit (Amersham), which is based on the incorporation of 32 P from [␥-32 P]ATP into a PKC-specific substrate peptide.
Statistical Analysis-The data were analyzed using a two-tailed independent Student's t test. The level of statistical significance was defined as p Ͻ 0.05.

Lyso-PC Stimulates Arachidonate Release in Endothelial
Cells-To determine the effect of lyso-PC on arachidonate re-lease, human umbilical vein endothelial cells were labeled with [ 3 H]arachidonate in medium 199 containing 10% fetal calf serum for 20 h. The cells were rinsed, and then incubated with HEPES-buffered saline containing 0.025% bovine serum albumin and 0 or 50 M lyso-PC for various time periods (Fig. 1A). Lyso-PC elicited a time-dependent arachidonate release, which reached a maximum at 10 min of incubation, after which arachidonate release was slightly diminished. A nominal amount of bovine serum albumin was required to bind the arachidonate that is released into the buffer. The optimal concentration of lyso-PC for the induction of arachidonate release was determined at bovine serum albumin concentrations ranging from 0.025 to 0.1% (w/v) (4 -16 M albumin). The effect of lyso-PC on arachidonate release was affected by the albumin concentration (Fig. 1B). Higher concentrations of lyso-PC was required to elicit a stimulation of arachidonate release at higher albumin concentrations. For example, at 0.025% albumin, the maximal stimulation of arachidonate release was observed at 50 M lyso-PC. Lyso-PC at this concentration has been found to be nonlethal to endothelial cells (43), and we confirmed cell viability under the incubation conditions by the exclusion of trypan blue dye. Hence, these conditions were routinely used in subsequent experiments.
Release of Arachidonate by Long Chain Lyso-PC and Other Lysolipids-Initial experiments on the effect of lyso-PC on arachidonate release were performed using lyso-PC derived from egg lecithin. Since egg lysolecithin contains mainly saturated acyl species, we tested the ability of palmitoyl (C 16:0 )-and stearoyl (C 18:0 )-lyso-PC to stimulate arachidonate release. Fig.  2 shows that lyso-PC containing palmitoyl and stearoyl chains induced a high release of arachidonate. To determine if the stimulation of arachidonate release is specific to lyso-PC or if it is a property common to all lysolipids, we tested the effect of other lysophospholipids such as lysophosphatidylethanolamine, lysophosphatidylserine, lysophosphatidylinositol, and lysophosphatidate on arachidonate release. As shown in Fig. 2, lysolipids with head groups other than choline were minimally effective in the stimulation of arachidonate release. Based on these results, lyso-PC containing a palmitoyl (C 16:0 ) chain was used in subsequent experiments.
Lyso-PC is an amphiphilic molecule and can incorporate into lipid membranes. Thus, we performed binding studies as described under "Experimental Procedures" to determine the nature of the association of lyso-PC with the endothelial cells. Cells were labeled with [ 14 C]lyso-PC (100 nM, 57 nCi/nmol), followed by incubation with control medium (without lyso-PC) or medium containing 10 M nonradioactive lyso-PC. The majority of radioactivity remained associated with cells in the presence of excess lyso-PC (results not shown), indicating that the association of lyso-PC with cells is mainly nonspecific.
Involvement of Phospholipase A 2 in Lyso-PC-induced Arachidonate Release-To determine whether the release of arachidonate is mediated by PLA 2 , we examined the effects of the PLA 2 inhibitors para-bromophenacyl bromide (pBPB) and arachidonoyl trifluoromethyl ketone (AACOCF 3 ), the latter of which specifically inhibits the cPLA 2 (44). As shown in Table I, arachidonate release was significantly inhibited in those cells that were preincubated with these inhibitors prior to challenge with lyso-PC. The inhibition of arachidonate release by up to 62% by AACOCF 3 indicates that the cPLA 2 may be involved in the arachidonate release induced by lyso-PC. However, sPLA 2 is also present in endothelial cells and may also participate in arachidonate release (10).
To further delineate the type of PLA 2 that was involved in the lyso-PC-induced arachidonate release, we used antisense oligonucleotides toward cPLA 2 and sPLA 2 . These oligonucleotides were designed to bind specifically to the respective mRNAs and prevent the translation and synthesis of the enzyme protein (11,40). Complementary sense oligonucleotides were used as negative controls. Cells were grown in the presence of sense or antisense oligonucleotides to either PLA 2 isoform for 3 days prior to challenge with lyso-PC. Treatment of the cells with either sense or antisense oligonucleotides did not alter the total incorporation of [ 3 H]arachidonate. However, lyso-PC-induced arachidonate release was significantly attenuated in cells grown in the presence of antisense oligonucleotides for cPLA 2 , compared with cells grown without oligonucleotides or with sense oligonucleotides (Fig. 3). The level of cPLA 2 protein after the treatment with antisense cPLA 2 oligonucleotides was determined by immunoblotting analysis with a polyclonal antibody for cPLA 2 . The level of cPLA 2 protein was decreased (40% reduction) by the antisense oligonucleotide treatment (Fig. 4A). In cells treated with antisense sPLA 2 oligonucleotides, the lyso-PC-induced arachidonate release was not significantly affected (Fig. 3), despite a 35% decrease in the sPLA 2 protein level in those cells (Fig. 4B).
Taken together, the inhibition of arachidonate release by antisense oligonucleotides for cPLA 2 and by AACOCF 3 indicate that the cPLA 2 is involved in the lyso-PC-induced arachidonate release. Therefore, we assayed in vitro the PLA 2 activity in subcellular fractions from control cells and from cells treated with lyso-PC. As shown in Table II, specific PLA 2 activity was decreased by 52% in the cytosolic fractions of lyso-PC-treated cells, while activity in the membrane fractions was increased by 33%. The increase in the membrane PLA 2 activity is consistent with the notion that the enzyme translocated to cell membranes. The membrane-bound enzyme is regarded as the active form, since the cell membrane contains the endogenous phospholipid substrates of cPLA 2 (3). This notion was supported by an in vitro study, in which PLA 2 activity was determined after the treatment of cells with the antisense cPLA 2 oligonucleotides. The total PLA 2 activity was reduced in those cells, with a corresponding (30%) decrease in the activity associated with the membrane fractions.
In separate experiments, the direct effect of lyso-PC on PLA 2 activity was examined by adding lyso-PC to the PLA 2 assays. Lyso-PC had no effect on PLA 2 activity when added directly to the enzyme assays (results not shown). Thus, the enhancement of PLA 2 activity observed in the membrane fraction of lyso-PCtreated cells was probably due to an indirect mechanism occurring in the intact cells. Therefore, we investigated the potential roles of Ca 2ϩ and protein kinases on the induction of arachidonate release by lyso-PC.
Extracellular Ca 2ϩ Is Required for Arachidonate Release by Lyso-PC-Since both sPLA 2 and cPLA 2 are regulated by Ca 2ϩ and since Ca 2ϩ stimulates the translocation of cPLA 2 to cell membranes (7), the role of Ca 2ϩ in arachidonate release was investigated. Cells were challenged with lyso-PC in the pres-  ence of 0 -1.5 mM Ca 2ϩ . As shown in Fig. 5, the induction of arachidonate release by lyso-PC was progressively suppressed at the lower Ca 2ϩ concentrations. Arachidonate release was completely abolished when Ca 2ϩ was absent from the buffer (the calcium-free buffer also contained 1 mM EDTA and 1 mM EGTA). Thus, the lyso-PC-induced arachidonate release was dependent on the Ca 2ϩ concentration in the buffer. Lyso-PC has been shown to cause increases in intracellular Ca 2ϩ concentrations (43). In the current study, treatment of the cells with 50 M lyso-PC in the presence of 1.5 mM Ca 2ϩ in the buffer caused an approximately 3-fold increase in the intracellular Ca 2ϩ level (Fig. 5, inset). In the absence of extracellular Ca 2ϩ , neither lyso-PC nor the calcium ionophore bromo-A23187 was able to cause any change in cell Ca 2ϩ . It appears that an influx of Ca 2ϩ from an extracellular source was a prerequisite for the induction of arachidonate release by lyso-PC and that the rise of cellular Ca 2ϩ was not derived from an intracellular pool. Involvement of PKC and MAPK in the Stimulation of Arachidonate Release by Lyso-PC-In addition to Ca 2ϩ , phosphoryla-tion events have also been shown to play a role in the regulation of cPLA 2 activity in a number of cell types (3,45,46). Thus, we investigated whether PKC or MAPK are involved in the lyso-PC-induced arachidonate release. Cells were pretreated with the PKC inhibitor staurosporine (47) or Ro31-8220 (48) prior to challenge with lyso-PC. For comparison, we also investigated the involvement of the cAMP-dependent protein kinase A by using the protein kinase A inhibitor H89 (49). Staurosporine and Ro31-8220 inhibited the arachidonate release induced by lyso-PC by up to almost 70% (Table III). In contrast, H89 did not cause any significant inhibition, up to a dose (1 M) far exceeding its K i (0.05 M) for protein kinase A (49). Lyso-PC has previously been shown to modulate PKC activity in both cell-free and cell-based assays (50 -52). Consistent with these findings, we observed that treatment of the cells with lyso-PC for 5 min caused a 58% increase in PKC activity in the membrane fraction of the cells (Table IV).
The phosphorylation of cPLA 2 by MAPK results in enhanced phospholipase activity, and MAPK is thought to be responsible for cPLA 2 phosphorylation in vivo (8,45,46). Thus, we investigated the involvement of the MAPK pathway by determining the release of arachidonate in the presence of PD098059, an inhibitor of the MAPK/extracellular regulated kinase kinase (53). Cells were preincubated with the indicated concentrations of PD098059 for 30 min prior to challenge with lyso-PC (Table  V). Pretreatment of cells with PD098059 resulted in as much as a 38% decrease in the arachidonate release induced by lyso-PC, relative to the arachidonate release caused by lyso-PC alone.
It has been shown that lyso-PC can modulate the activities of adenylyl cyclase and guanylyl cyclase, leading to alterations of cAMP and cGMP levels within cells (54 -56). To determine whether these cyclic nucleotides can affect the release of arachidonate induced by lyso-PC, cells were pretreated with forskolin, a direct activator of adenylyl cyclase, or with cellpermeable analogs of cAMP (8-bromo-cAMP) or cGMP (8-bromo-cGMP) prior to lyso-PC exposure. These compounds had little effect on lyso-PC-stimulated arachidonate release (data not shown). This result is consistent with the observation that H89 had no effect on arachidonate release (Table III) and supports the noninvolvement of cAMP-or cGMP-dependent protein kinases in the lyso-PC-induced arachidonate release. DISCUSSION The present study was conducted to study the effects of lyso-PC on the release of arachidonate in endothelial cells. We  Fig. 1. Values represent means Ϯ S.D. of three separate experiments. Inset, cells were challenged with 50 M lyso-PC in HEPES-buffered saline containing 0.025% bovine serum albumin, and the intracellular Ca 2ϩ was monitored using fura-2 as described under "Experimental Procedures." A typical trace is shown.

TABLE II
Effect of lyso-PC on PLA 2 activity in endothelial cells Cells were treated with 0 M (Ϫ) or 50 M (ϩ) lyso-PC in HEPESbuffered saline containing 0.025% bovine serum albumin. Cells were lysed, and PLA 2 activity was assayed in the cytosolic and membrane fractions as described under "Experimental Procedures." Results are expressed as mean Ϯ S.D. of three separate experiments.

Lyso-PC-induced Arachidonate Release in Endothelial Cells
found that exposure of the cells to lyso-PC containing long saturated acyl chains induced a dose-dependent increase in the release of arachidonate and that the effect was mediated through PLA 2 . Our findings support a model in which the induction of arachidonate release by lyso-PC is dependent on Ca 2ϩ influx and the activation of PKC. These processes result in the stimulation of cPLA 2 activity to give rise to an enhanced arachidonate release. A major pathway for arachidonic acid release from agoniststimulated cells is via hydrolysis of phospholipids by PLA 2 (4). Within the PLA 2 subtypes, the preference for arachidonatecontaining substrates and the low (intracellular concentrations) requirement for Ca 2ϩ of cPLA 2 have led many investigators to believe that this isoform is the main enzyme responsible for arachidonate release (3). The present study shows that the arachidonate release induced by lyso-PC was inhibited in a dose-dependent manner by the PLA 2 inhibitors pBPB and the cPLA 2 -specific AACOCF 3 . Furthermore, the arachidonate release by lyso-PC was also attenuated in cells grown in the presence of antisense oligonucleotides for the cPLA 2 . The observed enhancement of membrane-associated PLA 2 activity is consistent with an activation and translocation of cPLA 2 to membranes. The decrease in soluble PLA 2 activity was not quantitatively recovered in the membrane fraction of the cell lysates. This result is not entirely surprising, since it was previously documented that the membrane component may interfere with the cPLA 2 activity when assayed in vitro (57). This phenomenon was also observed with other enzymes upon their association with membranes and was attributed to a reduced accessibility of exogenous radioactive substrate to the enzyme and/or to a "dilution" effect on the exogenous radioactive substrate; i.e. the presence of endogenous membrane phospholipids lowered the effective specific activity of the radioactive substrate (57)(58)(59). The normal Ca 2ϩ concentration in endothelial cells is approximately 70 nM (60,61), and in the current study lyso-PC was found to cause a 3-fold increase in the intracellular Ca 2ϩ level. The increased Ca 2ϩ level induced by lyso-PC is similar to those Ca 2ϩ concentrations that were shown to cause the association of cPLA 2 with membranes (7,57). We conclude that the cPLA 2 is involved in the lyso-PCstimulated arachidonate release.
The contribution of sPLA 2 was also considered. Although antisense oligonucleotides for this isoform caused a reduction in sPLA 2 protein, a corresponding attenuation of the lyso-PCinduced arachidonate release was not detected in those cells. This result suggests that the sPLA 2 does not contribute significantly to the arachidonate release stimulated by lyso-PC. This finding is in accord with studies in which hormone-stimulated arachidonate release and eicosanoid production was attributed to cPLA 2 but not sPLA 2 (11)(12)(13). It is interesting to note that the involvement of both the cytosolic and secretory PLA 2 subtypes in the release of arachidonate for prostacyclin synthesis has also been reported (10).
Lyso-PC, with the participation of diacylglycerol, phosphatidylserine, and Ca 2ϩ (50 -52), has been shown to modulate PKC activity both in vitro and in vivo. The increase in intracellular Ca 2ϩ caused by lyso-PC may contribute to the enhancement of membrane-associated PKC activity. Although cPLA 2 is an in vitro substrate for PKC, the direct phosphorylation of cPLA 2 by PKC does not result in enhanced phospholipase activity (8), nor has PKC been shown to directly phosphorylate cPLA 2 in vivo. However, PKC is a known activator of the p42/p44 MAPK signaling cascade via phosphorylation of Raf-1 (62). Our results with the MAPK/extracellular regulated kinase kinase 1 inhibitor PD098059 implicate the involvement of the p42/p44 MAPK cascade in the arachidonate release by lyso-PC. The concentrations of PD098059 used in this study (up to 30 M) were similar to those used to almost completely inhibit the activation of MAPK/extracellular regulated kinase kinase 1 and to inhibit the activation of p42 MAPK by up to 80% (53,63). The partial inhibition of arachidonate release by PD098059 suggests that lyso-PC may also act through pathways other than the recruitment of p42/p44 MAPK. For example, the p38 MAPK is thought to participate in the activation of cPLA 2 by agonists (64,65).
Reported plasma concentrations of lyso-PC range from approximately 130 -150 M in healthy subjects (66, 67) to 1.7 mM in hyperlipidemic patients (21), while reported concentrations of human serum albumin ranged from approximately 185 to 850 M (68, 69). These figures would correspond to theoretical molar lyso-PC:albumin ratios in serum that range from 0.15 to 9.2. In our studies, arachidonate release was stimulated by lyso-PC at concentrations that correspond to lyso-PC:albumin ratios of 6.2-25. We selected a lyso-PC:albumin ratio of 12.5 (50 M lyso-PC and 0.025% or approximately 4 M albumin) for the subsequent experiments, because this ratio represented the lowest lyso-PC concentration that elicited the maximum effect on arachidonate release. Although the complexities of other serum components would probably complicate the in vivo situation, the lyso-PC:albumin ratios used in this study may mimic conditions found in physiological or pathophysiological situations.
Lyso-PC is a natural amphiphile and incorporates into lipid membranes and affects membrane fluidity and permeability (14,70,71). Indeed, lyso-PC (at concentrations higher than those used in this study) has been used as an agent for permeabilizing cells (72). However, the detergent properties of lyso-PC do not fully account for its myriad biological effects. For example, lyso-PC increases intracellular Ca 2ϩ levels (43,73,74) and yet inhibits receptor-mediated Ca 2ϩ mobilization (52,74). It exhibits both vasorelaxant (75) and vasoconstrictive (23) properties. It perturbs nitric oxide synthase mRNA and protein levels in endothelial cells, and up-or down-regulation is dependent on the concentration and incubation conditions used in each study (76 -78). Lyso-PC increases the expression of various growth factors and adhesive molecules in endothelial cells  (28 -31), and it was shown recently that lysophosphatidylcholine can modulate gene expression independently of PKC and MAPK (31,79,80). In our study, lyso-PC was the only lysolipid tested that stimulates arachidonate release, despite the fact that other lysolipids also possess detergent properties (81,82). Furthermore, the stimulation of arachidonate release by lyso-PC parallels earlier observations that the ability of lyso-PC to stimulate PKC is unique among lysolipids (50,51).
The findings of the current study demonstrate a novel role of lyso-PC in the modulation of endothelial cell functions. It is clear from this study that lyso-PC may modulate a pathway that is responsible for its generation in vivo. It is therefore tempting to speculate that the activation of phosphatidylcholine hydrolysis by PLA 2 could be regulated via a positive feedback mechanism that is mediated by its product lyso-PC. Lyso-PC may thus function as an intracellular messenger molecule. A prolonged activation and overactivation of the system may subsequently create adverse effects to the cells. Hence, the physiological consequences of the stimulation of arachidonate release in endothelial cells by lyso-PC will be an interesting area for further study. Since arachidonate and its metabolites have many biological properties related to vascular homeostasis, the perturbation of arachidonate release by lyso-PC may be a further mechanism whereby this lysolipid could contribute to vascular dysfunction.