5-Lipoxygenase products modulate the activity of the 85-kDa phospholipase A2 in human neutrophils.

Addition of submicromolar concentrations of arachidonic acid (AA) to human neutrophils induced a 2-fold increase in the activity of a cytosolic phospholipase A2 (PLA2) when measured using sonicated vesicles of 1-stearoyl-2-[14C]arachidonoylphosphatidylcholine as substrate. A similar increase in cytosolic PLA2 activity was induced by stimulation of neutrophils with leukotriene B4 (LTB4), 5-oxoeicosatetraenoic acid, or 5-hydroxyeicosatetraenoic acid (5-HETE). LTB4 was the most potent of the agonists, showing maximal effect at 1 nM. Inhibition of 5-lipoxygenase with either eicosatetraynoic acid or zileuton prevented the AA-induced increase in PLA2 activity but had no effect on the response induced by LTB4. Furthermore, pretreatment of neutrophils with a LTB4-receptor antagonist, LY 255283, blocked the AA- and LTB4-induced activation of PLA2 but did not influence the action of 5-HETE. Treatment of neutrophils with pancreatic PLA2 also induced an increase in the activity of the cytosolic PLA2; this response was inhibited by both eicosatetraynoic acid or LY 255283. The increases in PLA2 activity in response to stimulation correlated with a shift in electrophoretic mobility of the 85-kDa PLA2, as determined by Western blot analysis, suggesting that phosphorylation of the 85-kDa PLA2 likely underlies its increase in catalytic activity. Although stimulation of neutrophils with individual lipoxygenase metabolites did not induce significant mobilization of endogenous AA, they greatly enhanced the N-formylmethionyl-leucyl-phenylalanine-induced mobilization of AA as determined by mass spectrometry analysis. Our findings support a positive-feedback model in which stimulus-induced release of AA or exocytosis of secretory PLA2 modulate the activity of the cytosolic 85-kDa PLA2 by initiating the formation of LTB4. The nascent LTB4 is then released to act on the LTB4 receptor and thereby promote further activation of the 85-kDa PLA2. Since 5-HETE and LTB4 are known to prime the synthesis of platelet-activating factor, the findings suggest that 85-kDa PLA2 plays a role in platelet-activating factor synthesis.

Addition of submicromolar concentrations of arachidonic acid (AA) to human neutrophils induced a 2-fold increase in the activity of a cytosolic phospholipase A 2 (PLA 2 ) when measured using sonicated vesicles of 1-stearoyl-2-[ 14 C]arachidonoylphosphatidylcholine as substrate. A similar increase in cytosolic PLA 2 activity was induced by stimulation of neutrophils with leukotriene B 4 (LTB 4 ), 5-oxoeicosatetraenoic acid, or 5-hydroxyeicosatetraenoic acid (5-HETE). LTB 4 was the most potent of the agonists, showing maximal effect at 1 nM. Inhibition of 5-lipoxygenase with either eicosatetraynoic acid or zileuton prevented the AA-induced increase in PLA 2 activity but had no effect on the response induced by LTB 4 . Furthermore, pretreatment of neutrophils with a LTB 4 -receptor antagonist, LY 255283, blocked the AAand LTB 4 -induced activation of PLA 2 but did not influence the action of 5-HETE. Treatment of neutrophils with pancreatic PLA 2 also induced an increase in the activity of the cytosolic PLA 2 ; this response was inhibited by both eicosatetraynoic acid or LY 255283.
The increases in PLA 2 activity in response to stimulation correlated with a shift in electrophoretic mobility of the 85-kDa PLA 2 , as determined by Western blot analysis, suggesting that phosphorylation of the 85-kDa PLA 2 likely underlies its increase in catalytic activity. Although stimulation of neutrophils with individual lipoxygenase metabolites did not induce significant mobilization of endogenous AA, they greatly enhanced the N-formylmethionyl-leucyl-phenylalanine-induced mobilization of AA as determined by mass spectrometry analysis. Our findings support a positive-feedback model in which stimulus-induced release of AA or exocytosis of secretory PLA 2 modulate the activity of the cytosolic 85-kDa PLA 2 by initiating the formation of LTB 4 . The nascent LTB 4 is then released to act on the LTB 4 receptor and thereby promote further activation of the 85-kDa PLA 2 . Since 5-HETE and LTB 4 are known to prime the synthesis of platelet-activating factor, the findings suggest that 85-kDa PLA 2 plays a role in platelet-activating factor synthesis.
Arachidonic acid (AA) 1 is the precursor for prostaglandins and leukotrienes (collectively named eicosanoids), which together with platelet-activating factor (PAF) are important lipid mediators involved in allergic and inflammatory reactions. The initial step in the production of eicosanoids is hydrolysis of sn-2-arachidonate from membrane phospholipids generating free AA; phospholipase A 2 s (PLA 2 ) with different characteristics have been implicated in this hydrolysis (1)(2)(3). The 14-kDa PLA 2 s (type I and II) are secretory enzymes, and although they show no apparent preference for hydrolysis of AA-containing phospholipids (4,5), they have been suggested to participate in the generation of eicosanoids after first being secreted by mobilizing AA from phospholipids on the outer leaflet of the plasma membrane (6 -12). The 85-kDa PLA 2 (13,14) is an intracellular enzyme with clear preference for AA-containing phospholipids (5,15,16). Furthermore, the 85-kDa PLA 2 translocates to membranes in response to submicromolar concentrations of Ca 2ϩ (13,17,18) and is also regulated by phosphorylation, which results in an increase in its catalytic activity (19 -21). These characteristics make the 85-kDa PLA 2 a likely candidate responsible for the stimuli-induced mobilization of AA and the subsequent generation of eicosanoids.
While many cell types release both cyclooxygenase and lipoxygenase metabolites, human neutrophils (PMN) release predominantly the 5-lipoxygenase products, leukotriene B 4 (LTB 4 ) and 5-hydroxy-6,8,11,14-eicosatetraenoic acid (5-HETE), upon stimulation. Besides release of these eicosanoids, exogenously added 5-HETE has been shown to "prime" PMN for release of AA and 5-lipoxygenase metabolites (22) as well as PAF (22)(23)(24) when challenged with another stimuli. Furthermore, ionophore-induced mobilization of cellular AA (25) and production of LTB 4 (26) have been shown to be enhanced by exogenously added AA and LTB 4 , respectively. Although an increase in activity of one or more PLA 2 types was assumed in the above studies, no measurements of the activity of PLA 2 were performed. In the present study, we have examined the effect of exogenously added free fatty acids and 5-lipoxygenase metabolites on the activity and phosphorylation, as determined by mobility shifts on Western blots, of the intracellular 85-kDa PLA 2 in PMN.  (27,28). Prostaglandin E 2 , prostaglandin D 2 , and prostaglandin I 2 were from Cayman Chemical Co. (Ann Arbor, MI). Antibody against the 85-kDa PLA 2 was generously provided by Dr. James Clark, Genetics Institute (Cambridge, MA). Horseradish peroxidase-conjugated goat antirabbit antibody and reagents for the enhanced chemiluminescence detection were from Amersham Corp. Nitrocellulose membranes were from Schleicher and Schuell (Keene, NH).

Materials
Preparation and Stimulation of Neutrophils-PMN were prepared from heparinized venous blood collected from healthy medication-free donors using dextran sedimentation followed by isolymph centrifugation and removal of remaining red blood cells by hypotonic lysis as described previously (29). The resultant cell population consisted of Ͼ95% PMN.
PMN (0.75-1.1 ϫ 10 7 cells/ml; 4 ml total volume) were incubated at 37°C in Dulbecco's phosphate-buffered saline (PBS) containing 1 mM CaCl 2 . Free fatty acids were dissolved in ethanol and diluted in PBS prior to the addition to cell suspensions, resulting in a final concentration of ethanol of Ͻ0.1%. LTB 4 , 5-oxoETE, 15-oxoETE, 5-HETE, and compound I, dissolved in methanol, were dried under nitrogen and resuspended in 25 l of BSA (2.5 mg/ml) prior to addition to the cells. Zileuton and ETYA were dissolved in Me 2 SO and added to cells resulting in a final Me 2 SO concentration of 0.1%. LY 255283 was dissolved in water. Pancreatic PLA 2 was diluted in PBS prior to the addition to cell suspensions. Control cells received appropriate vehicle (Me 2 SO and/or BSA), which had no effect on the PLA 2 activity compared with no addition. Ten volumes of ice-cold PBS without CaCl 2 were added to terminate incubations before suspensions were centrifuged at 300 ϫ g for 10 min. When cells were treated with exogenous PLA 2 , an additional wash of the cells with 40 ml of PBS was included. All subsequent procedures were performed at 4°C. Cells were resuspended in 1 ml of 80 mM KCl, 10 mM Hepes (pH 7.4), 1 mM EDTA, 1 mM EGTA, 40 g/ml leupeptin, 25 g/ml pepstatin A, 1 mM phenylmethylsulfonyl fluoride, 10 mM NaF, 0.2 mM NH 4 VO 3 , and 4 mM dithiothreitol (Buffer A) and broken by sonication for 2 ϫ 5 s at a power setting of 2 and 10% output with a probe sonicator (Heat System Inc.). Broken cells were centrifuged at 10 4 ϫ g for 10 min, and the residual supernatant was further centrifuged at 10 5 ϫ g for 60 min to obtain cytosol and membrane fractions. Glycerol was added to the cytosol fraction to a final concentration of 10% (v/v). Membrane fractions were resuspended by sonication for 5 s in 0.35 ml of Buffer A containing 10% glycerol. The cytosol fraction could be stored at either 4°C or Ϫ20°C for 3 weeks without any major loss of PLA 2 activity.
Protein content of the subcellular fractions from PMN was determined according to the method of Bradford (30) using BSA as standard.
PLA 2 Activity Assay-The assay for PLA 2 activity was performed with sonicated vesicles of 1-stearoyl-2-[1-14 C]arachidonoylphosphatidylcholine (400 pmol; 25,000 dpm; 400 pmol of substrate was found to be saturating with cytosol from 1 ϫ 10 6 PMN). The assay mixture (final volume of 525 l) contained 80 mM KCl, 10 mM Hepes (pH 7.4), 1 mM EDTA, 4.7 mM Ca 2ϩ (free concentration), 5 mM dithiothreitol, 200 g of BSA, and cytosol from 1 ϫ 10 6 cells (30 -40 g of protein). Following a 10-min incubation at 37°C in a shaking water bath, the reaction was stopped by extraction according to Bligh and Dyer (31) except 6 l of 10 M HCl were added to each extraction mixture. Phosphatidylcholine and free arachidonic acid were added as carrier lipids. Lipids were separated by thin-layer chromatography using hexane/ethyl ether/acetic acid (30:20:1 (v/v/v)) as a solvent system, and radioactivity in areas corresponding to free fatty acid and phospholipid was determined by liquid scintillation counting. Each sample was assayed in duplicate; with less than 10% difference in activity between duplicate samples. In experiments where PMN were treated with exogenous PLA 2 , the cytosol fraction was preincubated in the presence of 10 mM dithiothreitol for 10 min at 22°C followed by 4 min at 37°C prior to initiation of the reaction by addition of substrate. This preincubation was done in order to inhibit any residual activity of the added low molecular weight PLA 2 in the cytosol fraction and had no effect on the activity of PLA 2 in the cytosol fraction from control cells or cells stimulated with lipoxygenase metabolites.
Immunoblotting of the 85-kDa PLA 2 -Subcellular fractions from con-trol and stimulated PMN were prepared as described above. The 10 4 ϫ g pellet was resuspended by sonication for 5 s in 0.35 ml of Buffer A containing 10% glycerol. The 10 5 ϫ g pellet was washed once with Buffer A prior to the addition of 0.35 ml of Buffer A containing 10% glycerol and sonication for 5 s. The resuspended pellets and cytosol fractions were mixed with an equal volume of a 2 times concentration of Laemmli sample buffer (32) and boiled for 5 min. Samples were subjected to SDS-polyacrylamide gel electrophoresis (7.5% polyacrylamide) at 22 mA for 5 h, followed by electrotransfer of proteins onto a nitrocellulose membrane at 60 mA for 12 h. The membrane was blocked for 1 h with 5% nonfat milk in PBS containing 0.2% Tween 20. After washing with 0.2% Tween in PBS (3 times), the membrane was incubated for 2 h with a 1:1000 dilution of a polyclonal antibody against the 85-kDa PLA 2 (13), washed 3 times with 0.2% Tween 20 in PBS, and incubated for 1 h with a 1:1000 dilution of a horseradish peroxidaseconjugated antibody in 0.2% Tween 20 in PBS containing 1% nonfat milk. After five washes with 0.2% Tween 20 in PBS, protein was detected with the enhanced chemiluminescence system. Quantitation of Nonesterified Fatty Acids after Stimulation of Neutrophils-Incubations of control and stimulated PMN were terminated by the addition of 3 ml of chloroform/methanol 1:2 (v/v); 100 ng each of [ 2 H 8 ]AA and [ 2 H 3 ]stearic acid were added as internal standards. Lipids were extracted according to Bligh and Dyer (31), the chloroform phase was dried down under nitrogen, and the lipids were resuspended in hexane and passed through a silicic acid column (Prep-Sep, Fisher Scientific, Norcross, GA). Fatty acids were eluted with 4 ml of hexane/ ethyl ether (1:1 (v/v)), dried down and converted to pentafluorobenzyl esters by the addition of 30 l each of 20% (v/v) pentafluorobenzyl bromide in acetonitrile and 20% (v/v) diisopropylethylamine in acetonitrile, and heated in a sealed tube for 40 min at 40°C. After removal of solvent, the samples were resuspended in hexane, and fatty acids were quantitated by combined negative ion-chemical ionization gas chromatography/mass spectrometry using a Hewlett Packard model 5989 instrument as described previously (33). Quantities of free fatty acids were calculated from the signal intensity for each fatty acid by comparison with that of internal standards ([ 2 H 8 ]AA for AA and [ 2 H 3 ]stearic acid for oleic and linoleic acids).

RESULTS
The cytosol fraction from PMN, disrupted in the presence of Ca 2ϩ -chelators, contains the predominant part of the PLA 2 activity when assayed with sonicated vesicles of arachidonoylphosphatidylcholine as substrate. Under these assay conditions (see "Experimental Procedures"), the cytosol fraction from control cells showed a specific PLA 2 activity of 0.26 Ϯ 0.04 pmol/g/10 min (mean Ϯ S.D. from 20 different PMN preparations); however, some variations between different preparations were seen (discussed below). While the 10 5 ϫ g pellet contained some activity, it amounted to only 2-4% of the activity found in the cytosol fraction. We were unable to detect any PLA 2 activity in the 10 4 ϫ g pellet. The activity of PLA 2 , both in the cytosol and the 10 5 ϫ g pellet, was totally Ca 2ϩ -dependent and resistant to dithiothreitol. These characteristics indicated that the PLA 2 we assay is similar or identical to the 85-kDa PLA 2 , which has previously been identified in several different cell types including PMN (34).
Addition of submicromolar concentrations of free AA to PMN resulted in a 1.8 -2-fold increase in PLA 2 activity in the cytosol fraction compared with that from control cells (Fig. 1). The AA response was maximal at 0.5 M with no further increase seen using 10 M AA. In contrast to the response seen with AA, no increase in PLA 2 activity was observed with linolenic acid (Fig.  1), linoleic acid, or oleic acid (not shown) in the range from 0.5 to 50 M. While Fig. 1 data were obtained with fatty acids added to the cells in ethanol (Ͻ0.1%), almost identical data were obtained when fatty acids were added to cells dissolved in Me 2 SO or complexed to BSA (not shown).
We next investigated the actions of selected AA metabolites. Stimulation of PMN with either LTB 4 , 5-HETE, or 5-oxoETE induced an increase in PLA 2 (Fig. 2). The potency among these agents varied with LTB 4 , the most potent of the metabolites, showing maximal effect at 1 nM. 5-oxoETE and 5-HETE were at least 50 times less active than LTB 4 , with 5-oxoETE being about 5-fold more potent than 5-HETE. In studies not shown, we found that 5 M 15-oxoETE and 10 nM compound I (an isomer of LTB 4 ) did not stimulate PMN to alter cytosolic PLA 2 activity. Likewise, 0.5 M prostaglandins D 2 , E 2 , and I 2 were inactive. 15-oxoETE, compound I and the prostaglandins do not stimulate PMN functional responses (27,35). Finally, pretreatment of PMN with 0.5 M prostaglandin E 2 for 2 min did not affect the ability of LTB 4 to activate PLA 2 (data not shown).
The time course of AA and LTB 4 -induced activation of PLA 2 is shown in Fig. 3. While the LTB 4 response was almost fully developed after 1 min of stimulation, the AA response was only marginal at this time point but reached the maximal response by 2 min. Thus, the pathway mediating PMN responses to AA is distinctly slower acting than that for LTB 4 .
Activation of protein kinase C by stimulation with phorbol myristate acetate has been shown to lead to an increase in the activity of the 85-kDa PLA 2 in PMN (36) as well as several other cell types (19,(37)(38)(39). Since unsaturated fatty acids can activate protein kinase C (40, 41), we considered the possibility that the AA-induced increase in PLA 2 activity might reflect the direct action of AA on protein kinase C. Alternatively, AA might act indirectly by first being converted to a lipoxygenase product and then triggering activation of PLA 2 via LTB 4 or putative 5-HETE receptors. We therefore investigated the effect of two competitive inhibitors of lipoxygenase on the AAinduced activation of PLA 2 . Pretreatment of PMN with either ETYA, an inhibitor of both cyclooxygenase and lipoxygenase, or zileuton, a selective inhibitor of 5-lipoxygenase, abolished the AA-induced activation of PLA 2 (Fig. 4). Neither of these inhibitors, at a concentration of 5 M, altered the basal activity of PLA 2 or the LTB 4 -induced activation of PLA 2 .
In another set of experiments, we examined the effect of LY 255283, a competitive LTB 4  the AA-and LTB 4 -induced activation of PLA 2 , yet it had no effect on the basal activity of PLA 2 or the increase in PLA 2 activity induced by 5-HETE (Fig. 5) or 5-oxoETE (not shown). As little as 0.5 M LY 255283 was sufficient to block the AA-induced activation of PLA 2 .
The increase in catalytic activity of the 85-kDa PLA 2 has been linked to phosphorylation of the enzyme (19 -21, 36), which results in reduced mobility upon electrophoresis (19). When the cytosol fraction from control PMN was subjected to electrophoresis and immunoblotted with an antibody against the 85-kDa PLA 2 , two bands were detected with the lower band (higher mobility) being the more prominent (Figs. 6 -8). It should be pointed out that the distribution of immunodetected 85-kDa PLA 2 between the two bands in the cytosol fraction from different control preparations varied. In one control PMN preparation, an equal amount of protein was seen in the two bands, and this was accompanied by a somewhat higher specific activity of PLA 2 (0.32 pmol/g/10 min versus 0.23 pmol/ g/10 min in Fig. 6) as well as a lower -fold increase in response to stimulation (1.55-fold versus 1.9-fold in Fig. 6). These differences between PMN preparations could be due to partially activated cells in some preparations, resulting in a partial shift in mobility (phosphorylation) of the 85-kDa PLA 2 . Cytosol fractions from PMN stimulated with optimally active amounts of either LTB 4 , 5-HETE, or AA (as defined in PLA 2 assays) resulted in all of the immunodetected 85-kDa PLA 2 migrating with reduced mobility (Fig. 6). This suggests that the three stimuli induce phosphorylation of the 85-kDa PLA 2 . Fig. 6 also shows that the AA-induced shift in mobility (phosphorylation) could be blocked by pretreatment of the cells with the LTB 4receptor antagonist LY 255283; the blocking of the shift in mobility correlates with the loss of stimulation of the PLA 2 activity by AA (Fig. 5). The LTB 4 -induced shift in mobility of the 85-kDa PLA 2 was also inhibited by LY 255283, while that induced by 5-HETE was not affected (Fig. 7). Immunoblotting of the 10 5 ϫ g pellet from control and stimulated PMN revealed the same pattern as seen in the cytosol, i.e. two bands from control cells and one band with reduced mobility from stimu- Cytosol fraction (upper set) and 10 5 ϫ g pellet (lower set) were prepared as described under "Experimental Procedures." Twenty g of protein from cytosol fractions (6 ϫ 10 5 cell equivalents) and 6 g of protein from the 10 5 ϫ g pellets (3 ϫ 10 6 cell equivalents) were subjected to SDSpolyacrylamide gel electrophoresis. After electrotransfer to nitrocellulose membrane and immunoblotting with a polyclonal antibody against the 85-kDa PLA 2 , protein was detected with Amersham's enhanced chemiluminescence detection system. All 12 samples were run on the same gel and split into two parts after detection. Cont., control; LY, LY 255283; PLA 2 and PLA 2 -P depict the unphosphorylated and phosphorylated enzyme, respectively, as suggested by the shift in mobility. lated cells (Fig. 6). We emphasize that the amount of 10 5 ϫ g pellet subjected to immunoblotting was from 3 ϫ 10 6 cell equivalents, whereas the amount from the cytosol fraction was from 6 ϫ 10 5 cell equivalents. This indicates that the 10 5 ϫ g pellet constitutes only a minor part of the immunodetectable 85-kDa PLA 2 when PMN are broken in the presence of Ca 2ϩ -chelators. This is in agreement with our PLA 2 activity data. No immunodetectable 85-kDa PLA 2 could be found in the 10 4 ϫ g pellet (results not shown).
Having established that exogenously added AA induced activation of cytosolic PLA 2 , we hypothesized that an extracellular PLA 2 added to PMN could mimic the effect of AA. As seen in Fig. 8A, addition of pancreatic PLA 2 to PMN results in an increase in the activity of cytosolic PLA 2 ; this increase was to a large extent inhibited by pretreatment of the cells with LY 255283 or ETYA. In accordance with these activity results, pancreatic PLA 2 also induced a shift in mobility of the 85-kDa PLA 2 upon electrophoresis, and this shift was partially inhibited by pretreatment with LY 255283 and completely inhibited by ETYA (Fig. 8B).
In order to verify that the observed increase of the 85-kDa PLA 2 activity was due to the exogenously added PLA 2 rather than possible lipopolysaccharide (LPS) contamination in the pancreatic PLA 2 preparation, we used a number of approaches. Polymyxin B is an antibiotic that inactivates endotoxin by binding the lipid A portion of LPS. Incubation of pancreatic PLA 2 (2 g/ml) with 1 mg/ml polymyxin B for 1 h did not decrease its ability to activate the 85-kDa PLA 2 compared with untreated pancreatic PLA 2 (data not shown). A second approach was to treat the pancreatic lipase with Pronase, which consists of a mixture of several proteolytic enzymes including endopeptidases and exopeptidases. Therefore, Pronase should hydrolyze the PLA 2 enzyme while leaving the endotoxin intact. Treatment of pancreatic PLA 2 (2 g/ml) with 100 units of Pronase for 1 h resulted in a total loss of exogenous PLA 2 ability to activate cytosolic PLA 2 , indicating that the added PLA 2 was responsible for the increase of 85-kDa PLA 2 activity (data not shown). Finally, we were able to directly measure the amount of endotoxin contamination in the pancreatic PLA 2 preparation using a chromogenic limulus amebocyte lysate assay (Bio-Whitaker) sensitive to 10 pg/ml LPS. The standard addition of 2 g/ml pancreatic PLA 2 was shown to contain less than 50 pg/ml LPS. This amount of LPS has been shown to be below the threshold needed for the activation of cytosolic PLA 2 in PMN (36), strengthening the argument that the observed activation of cytosolic PLA 2 was due to exogenously added PLA 2 , not LPS. In addition to these treatments, boiling the pancreatic lipase for 15 min did not destroy its activity against phospholipid vesicles or its ability to activate the 85-kDa PLA 2 in PMN.
Phosphorylation of the 85-kDa PLA 2 and the concomitant increase in activity, determined in an in vitro assay, may not necessarily reflect a mobilization of cellular AA. We therefore examined the effect of lipoxygenase metabolites on the mobilization of AA from PMN using gas chromatography/mass spectrometry. In order to prevent reacylation of mobilized AA, as well as metabolism to lipoxygenase products, BSA (at a final concentration of 2 mg/ml) was added to cells 15 s after stimuli. Although the lipoxygenase products were added in BSA-containing vehicle (2.5 mg/ml), the final concentration of BSA with the cells was only 0.03 mg/ml, which had no significant effect on AA metabolism. In a pilot study, we found that a powerful stimulus, 1 M fMLP, did not result in any detectable increase in free AA above that of control if BSA (final concentration 2 mg/ml) was added to the cells 3 min after stimulation (data not shown). However, using the same protocol altered by the addi-  5-HETE). Twenty g of protein from cytosol fractions (0.6 ϫ 10 6 cell equivalents) were subjected to SDS-polyacrylamide gel electrophoresis. After electrotransfer to nitrocellulose membrane and immunoblotting with a polyclonal antibody against the 85-kDa PLA 2 , protein was detected with Amersham's enhanced chemiluminescence detection system. Positive controls for the LTB 4and 5-HETE-induced shifts in mobility were shown in Fig. 6.   FIG. 8. Effect of pancreatic PLA 2 on PMN 85-kDa PLA 2 . Cells (1.1 ϫ 10 7 /ml; 4 ml total) were pretreated with 10 M LY 255283, 5 M ETYA, or no pretreatment for 3 min prior to addition of pancreatic PLA 2 (2 g/ml) and incubation for 12 min. Cells were washed twice with 30 ml of PBS followed by resuspension in buffer A, sonication, and preparation of cytosol fraction. A, cytosol fraction was assayed for PLA 2 activity. Results are mean values Ϯ S.D. from four different PMN preparations and are expressed as -fold increase above the activity in cytosol fraction from control cells. B, 20 g of protein (6 ϫ 10 5 cell equivalents) from cytosol fractions were subjected to SDS-polyacrylamide gel electrophoresis, electrotransferred to nitrocellulose membrane, and immunoblotted with a polyclonal antibody against the 85-kDa PLA 2 followed by detection of protein with Amersham's enhanced chemiluminescence detection system. Representative results of three independent experiments are shown. Cont., control; LY, LY 255283; pan PLA 2 , pancreatic PLA 2 . tion of BSA (final concentration 2 mg/ml) 15 s after fMLP stimulation resulted in accumulation of AA (Table I). PMN stimulated with either LTB 4 , 5-HETE, or 5-oxoETE did not induce any major mobilization of AA when added alone. This was true whether these stimuli were added 3.25 min (Table I) or 15 s (not shown) before addition of BSA (final concentration, 2 mg/ml). On the other hand, all three lipoxygenase metabolites substantially enhanced the fMLP-induced mobilization of AA. The priming effect of LTB 4 on the fMLP-induced mobilization of AA was completely blocked by LY 255283 (results not shown). Although the amount of mobilized AA mass varied, the same pattern was seen in three separate experiments, showing that 5-oxoETE and 5-HETE enhanced the fMLP-induced mobilization of AA to a somewhat greater extent than LTB 4 . We also examined the amounts of free linoleic and oleic acid in PMN. In contrast to AA, these fatty acids were found in higher amounts in control cells (25 and 400 ng/10 7 cells for linoleic acid and oleic acid, respectively) and did not change significantly after stimulation (results not shown). DISCUSSION A wide variety of different stimuli such as growth factors, hormones, and cytokines have been shown to induce phosphorylation of the 85-kDa PLA 2 , which is accompanied by about a 2-fold increase in the activity of the enzyme when assayed in vitro (42)(43)(44)(45)(46)(47). The 85-kDa PLA 2 has been shown to act as a substrate for protein kinase C in vitro; however, such phosphorylation does not result in any increase in catalytic activity (45,48). Instead, mitogen-activated protein kinase has been proposed to be the kinase responsible for the phosphorylation resulting in an increase in activity of the 85-kDa PLA 2 (45,49). Activation of the 85-kDa PLA 2 has been linked to release of eicosanoids in several different cell types (19,21,36,42,43), which emphasizes the importance of this PLA 2 . The 85-kDa PLA 2 of PMN has been examined in only a few studies, even though the cells are prominent sources of eicosanoids in hostdefense and inflammatory reactions.
In the present study, we have demonstrated that stimulation of PMN with 5-lipoxygenase metabolites results in a 2-fold increase in the activity of a cytosolic PLA 2 when assayed using sonicated vesicles of 1-stearoyl-2-[ 14 C]arachidonoylphosphatidylcholine as substrate. Immunoblot analysis of cytosol fractions from LTB 4 -and 5-HETE-stimulated PMN using a polyclonal antibody against the 85 kDa PLA 2 revealed a shift in electrophoretic mobility of the 85-kDa PLA 2 . This shift has been shown to be due to phosphorylation of the enzyme (19). The decrease in electrophoretic mobility of the 85-kDa PLA 2 in response to stimulation of PMN correlated in all cases with an increase in the enzymatic activity of PLA 2 in the cytosol fraction. This correlation, together with the characteristics of the PLA 2 , i.e. its total Ca 2ϩ -dependence and resistance to dithiothreitol, indicate that the activity measured in the cytosol fraction is due solely to the 85-kDa PLA 2 . LTB 4 (50) and apparently 5-HETE (51) bind to G-protein coupled plasma membrane receptors. Several lines of evidence suggest that these two receptors are distinctly different (50,51). Indeed, we find that the LTB 4 -induced activation of PLA 2 was inhibited by pretreatment of PMN with a LTB 4 -receptor antagonist LY 255283, whereas the 5-HETE-induced activation of PLA 2 was unaffected by this pretreatment. Our results thus support the previous conclusion that the PMN recognition systems for 5-HETE and LTB 4 are distinctive. We presume, as indicated in earlier work (28,52) that 5-HETE and 5-oxoETE operate through an identical cell-activating mechanism. Maximal increase in PLA 2 activity was seen with 1 nM LTB 4 , while 500 nM 5-HETE and between 50 and 500 nM 5-oxoETE were required to achieve the same level of response. These differences in concentration required to activate the 85-kDa PLA 2 correlate well with their potencies in stimulating PMN function (28,51,52).
AA as well as other unsaturated fatty acids such as linolenic and linoleic acid have been shown to activate protein kinase C (40,41). The increase in PLA 2 activity seen in the cytosol fraction of PMN stimulated with AA could therefore be due to a direct activation of protein kinase C. However, the fact that linolenic acid, even at concentrations 100 times that of AA, did not result in any increase in PLA 2 activity argues that AA does not act here through protein kinase C. Instead, our results strongly suggest that the increase in PLA 2 activity induced by exogenous AA is mediated via metabolism of AA to a 5-lipoxygenase product prior to activation of PLA 2 . This conclusion is supported by results showing that either ETYA or zileuton, both competitive inhibitors of 5-lipoxygenase, completely suppress the AA response. Furthermore, the AA response was also inhibited by the LTB 4 receptor antagonist LY 255283. The time-course of PLA 2 activation in response to LTB 4 and AA, in which maximal responses were observed at 1 and 2 min, respectively, is also consistent with a model in which conversion of AA to a 5-lipoxygenase metabolite is required for the activation of the PLA 2 . AA can be rapidly incorporated into neutral lipids and phospholipids and could become unavailable for metabolism by lipoxygenase. However, since 1 nM LTB 4 was sufficient to induce a maximal response, only a small fraction (Ͻ0.1%) of the added AA needs to be converted to LTB 4 in order to account for the AA-induced activation of the 85-kDa PLA 2 .
Previous work has shown that 5-HETE potentiates ionophore-induced formation of PAF, 5-HETE, and LTB 4 in PMN (22). Furthermore, synthesis of PAF, which in many cases occurs in parallel with eicosanoids, can be induced by a nonmetabolized bioactive analog of PAF and to a lesser extent by LTB 4 (24). In this context, it is noteworthy that stimulation of PMN with PAF (10 nM, 5 min) induces a 2-fold increase in the activity of PLA 2 in the cytosol fraction as well as reduced mobility of 85 kDa PLA 2 upon electrophoresis. 2 Using mass spectrometry, we have found that LTB 4 , 5-oxoETE, and 5-HETE substantially enhance the fMLP-induced mobilization of AA in intact PMN, while these 5-lipoxygenase products alone resulted in little or no mobilization of AA. Since all three lipoxygenase products induced an increase in the activity of PLA 2 when assayed in vitro, our data suggest that phosphorylation alone is insufficient to mobilize any major amount of AA in PMN. It appears that in addition to phosphorylation, an increase in intracellular Ca 2ϩ may explain the striking increase in AA release observed when fMLP is added as a second stimulus. This is in agreement with previous work on PMN showing that tumor necrosis factor (53) and granulocyte-macrophage colony-stimulating factor (54, 55) enhanced the fMLPinduced mobilization of AA and formation of lipoxygenase products but had little effect alone. Both of these cytokines have been shown, although not in PMN, to phosphorylate the 85-kDa PLA 2 (43,56). The involvement of secretory low molecular weight PLA 2 in the mobilization of AA and the subsequent generation of eicosanoids has been suggested in several different cell types such as mast cells (6,9), endothelial cells (8,10), HL-60 cells (7), mesangial cells (10), human neutrophils (12), and in the macrophage cell line P388D 1 (11). Some of the studies revealed formation of cyclooxygenase products in response to treatment of intact cells with low molecular weight PLA 2 alone (8 -10), while others required addition of a second stimulus such as antigen or Ca 2ϩ -ionophore to yield measurable products (6,7). Our results demonstrate that addition of low molecular weight PLA 2 can activate the intracellular 85-kDa PLA 2 in PMN. This activation could be inhibited to a large extent by either LY 255283 or ETYA, suggesting that one or more lipoxygenase metabolites is responsible for the communication between the two types of PLA 2 . In PMN, Shimizu et al. (12) found no formation of lipoxygenase products in response to addition of either pancreatic or venom PLA 2 (0.14 g/ml; 1.5 ϫ 10 7 cells), although each enhanced the zymosan-induced formation of LTB 4 . Since we have shown (Fig. 2) that even small amounts of LTB 4 (4 pmol/4 ϫ 10 7 cells) will activate the 85-kDa PLA 2 , LTB 4 may have been formed also in the study by Shimizu et al. (12) but below the detection limit. In mast cells, Fonteh et al. (9) has shown that addition of Naja naja PLA 2 resulted in low but significant formation of LTB 4 (5-10 pmol/5 ϫ 10 6 cells). Recently, Reddy et al. (57) reported that neutrophils contain an LTB 4 -dependent PLA 2 that requires extracellular Ca 2ϩ for activation. These findings, which were based on studies of intact cells, add new support for our conclusions.
In summary, our results support a positive feedback model for regulating the 85-kDa PLA 2 . We propose that 5-lipoxygenase products derived from initially released AA activate the enzyme and prime for increased AA mass release. Secretory PLA 2 , following its release by exocytosis, can liberate AA from the plasma membrane and through its conversion to lipoxygenase products can provide cross-talk between the secreted and cytosolic PLA 2 s. In our experimental system, AA released by extracellular PLA 2 appears to act mainly through formation of LTB 4 and its receptor to activate the 85-kDa PLA 2 , but 5-ox-oETE and 5-HETE also activated the enzyme. Further work will be necessary to determine the relative roles of secretory and intracellular PLA 2 s in the mobilization of AA and the subsequent formation of eicosanoids.