fMLP-Induced Arachidonic Acid Release in db-cAMP-Differentiated HL-60 Cells Is Independent of Phosphatidylinositol-4,5-bisphosphate-Specific Phospholipase C Activation and Cytosolic Phospholipase A2 Activation

Archives of Biochemistry and Biophysics Vol. 378, No. 2, June 15, pp. 246 –258, 2000 doi:10.1006/abbi.2000.1844, available online at http://www.idealibrary.com on

fMLP-Induced Arachidonic Acid Release in db-cAMPDifferentiated HL-60 Cells Is Independent of Phosphatidylinositol-4,5-bisphosphate-Specific Phospholipase C Activation and Cytosolic Phospholipase A 2 Activation Lutz Sternfeld, Frank The´venod, 1 and Irene Schulz Physiologisches Institut, Universita¨t des Saarlandes, D-66421 Homburg/Saar, Germany

Received January 4, 2000; and in revised form March 30, 2000

In inflammatory cells, agonist-stimulated arachidonic acid (AA) release is thought to be induced by activation of group IV Ca 2ⴙ-dependent cytosolic phospholipase A 2 (cPLA 2) through mitogen-activated protein kinase (MAP kinase)- and/or protein kinase C (PKC)-mediated phosphorylation and Ca 2ⴙ-dependent translocation of the enzyme to the membrane. Here we investigated the role of phospholipases in N-formylmethionyl-L-leucyl-L-phenylalanine (fMLP; 1 nM–10 ␮M)-induced AA release from neutrophil-like db-cAMP-differentiated HL-60 cells. U 73122 (1 ␮M), an inhibitor of phosphatidyl-inositol-4,5biphosphate-specific phospholipase C, or the membrane-permeant Ca 2ⴙ-chelator 1,2-bis[2-aminophenoxy]ethane-N,N,Nⴕ,Nⴕ-tetraacetic acid (10 ␮M) abolished fMLP-mediated Ca 2ⴙ signaling, but had no effect on fMLP-induced AA release. The protein kinase C-inhibitor Ro 318220 (5 ␮M) or the inhibitor of cPLA 2 arachidonyl trifluoromethyl ketone (AACOCF 3; 10 –30 ␮M) did not inhibit fMLP-induced AA release. In contrast, AA release was stimulated by the Ca 2ⴙ ionophore A23187 (10 ␮M) plus the PKC activator phorbol myristate acetate (PMA) (0.2 ␮M). This effect was inhibited by either Ro 318220 or AACOCF 3. Accordingly, a translocation of cPLA 2 from the cytosol to the membrane fraction was observed with A23187 ⴙ PMA, but not with fMLP. fMLP-mediated AA release therefore appeared to be independent of Ca 2ⴙ signaling and PKC and MAP kinase activation. However, fMLP-mediated AA release was reduced by ⬇45% by Clostridium difficile toxin B (10 ng/ml) or by 1-butanol; both block phospholipase D (PLD) activity. The inhibitor of phosphatidylcholine-specific phospholipase C (PC-PLC), D609 (100 ␮M), decreased fMLP-mediated AA 1 Present address: University of Manchester, School of Biological Sciences, G.38 Stopford Building, Oxford Road, Manchester M13 9PT, UK.

246

release by ⬇35%. The effect of D609 ⴙ 1-butanol on fMLPinduced AA release was additive and of a magnitude similar to that of propranolol (0.2 mM), an inhibitor of phosphatidic acid phosphohydrolase. This suggests that the bulk of AA generated by fMLP stimulation of dbcAMP-differentiated HL-60 cells is independent of the cPLA 2 pathway, but may originate from activation of PC-PLC and PLD. © 2000 Academic Press Key Words: arachidonic acid; HL-60; mitogen-activated protein kinase; phospholipase A 2; phospholipase C; phospholipase D; phophatidic acid phosphohydrolase.

Arachidonic acid (AA) 2 and its eicosanoid metabolites (e.g., prostaglandins and leukotrienes) are key molecules in inflammatory processes. Phospholipase A 2s (PLA 2s) are the crucial enzymes for the liberation 2 Abbreviations used: AA, arachidonic acid; AACOCF 3, arachidonyl trifluoromethyl ketone; BAPTA-AM, 1,2-bis[2-aminophenoxy]ethane-N,N,N⬘,N⬘-tetraacetic acid; BEL, (E)-6-(bromomethylene)-3-(1-naphthalenyl)-2H-tetrahydropyran-2-one; cPLA 2, group IV Ca 2⫹-dependent cytosolic phospholipase A 2; DAG, diacylglycerol; dbcAMP, dibutyryl cyclic AMP; DDT, dithiothreitol; fMLP, N-formylmethionyl-L-leucyl-L-phenylalanine; HBSS, Hanks’ balanced salt solution; IMDM, Iscove’s modified Dulbecco’s media; IP 3, inositol-1,4,5trisphosphate; PLA 2, phospholipase A 2; cPLA 2, Ca 2⫹-dependent cytosolic phospholipase A 2; iPLA 2, Ca 2⫹-independent phospholipase A 2; sPLA 2, Ca 2⫹-dependent secretory phospholipase A 2; MAFP, methyl arachidonyl fluorophosphonate; MAG, monoacylglycerol; MAP kinase, mitogen-activated protein kinase; PA, phosphatidic acid; PAF, platelet-activating factor; PA-P, phosphatidic acid phosphohydrolase; PC-PLC, phosphatidylcholine-specific phospholipase C; PIP 2-PLC, phosphatidylinositol-4,5-bisphosphate-specific phospholipase C; PKC, protein kinase C; PLD, phospholipase D; PMA, phorbol myristate acetate; TG, thapsigargin; ToxB, Clostridium difficile toxin B.

0003-9861/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

fMLP-INDUCED AA RELEASE IN HL-60 CELLS

of AA from the sn-2 position of membrane phospholipids that is believed to be the rate-limiting event in the generation of proinflammatory mediators. In recent years, it has become clear that PLA 2s represent a growing superfamily of enzymes. At least four groups of PLA 2s have been described in mammalian cells: (1) the 14-kDa Ca 2⫹-dependent secretory PLA 2s (sPLA 2; groups I, II, III, V, and IX); (2) two Ca 2⫹-independent PLA 2s (iPLA 2) of 80 – 85 and 40 kDa (VI and myocardial XII, respectively); (3) two Ca 2⫹-independent PLA 2s specific for platelet-activating factor (PAF); and (4) the 85-kDa Ca 2⫹-dependent cytosolic PLA 2 (cPLA 2; group IV) (1, 2). cPLA 2 has been cloned from human myelomonocytic U937 cells (3, 4). In addition, a Ca 2⫹independent cPLA 2-gamma with homology to cPLA 2 has been recently characterized (5). It has been suggested that cPLA 2 is responsible for G-protein-coupled receptor-mediated AA release (3, 6). However, little evidence supports a direct coupling of cPLA 2 to hormone receptors (7, 8); cPLA 2 may, as well, be regulated indirectly by receptor stimulation. A current model of the mechanisms of receptor-mediated activation of cPLA 2 (9) suggests that agonist-mediated receptor stimulation primarily activates a G-proteincoupled PIP 2-PLC, leading to the production of DAG and IP 3. The rise in these intracellular messengers causes the activation of PKC and mobilization of intracellular Ca 2⫹ as well as Ca 2⫹ influx from the extracellular space. The increase in cytosolic Ca 2⫹ promotes the translocation of cPLA 2 from the cytosol to the membrane where its phospholipid substrates are localized (4, 10). Activation of the enzyme occurs through Ser505 phosphorylation by MAP kinases and/or PKC (11, 12). One major flaw of this proposed signal cascade is that it is mainly based on in vitro studies with recombinant enzymes. Its relevance for agonist-stimulated AA release in inflammatory cells has not yet been proven. More recent studies have suggested that the 14-kDa secretory PLA 2s II and V (13), the Ca 2⫹-independent PLA 2 (14), or a 29-kDa cytosolic PLA 2 (15) could also be responsible for agonist-induced AA release. Furthermore AA could be generated by receptormediated activation of PIP 2-PLC (reviewed in (16)) or phosphatidylcholine-specific phospholipase C (PCPLC) (6) and subsequent release of AA from DAG by diacyl/monoacylglycerol lipases (DAG/MAG lipases) (see Fig. 5). Finally, AA could be produced through receptor-mediated activation of phospholipase D (PLD) (for review, see (17)) and the sequential actions of phosphatidic acid phosphohydrolase (PA-P) and DAG/ MAG lipases. In the present study, we used the HL-60 human promyelocytic leukemia cell line, which can be terminally differentiated into neutrophil-like cells by a variety of inducing agents, including dibutyryl cyclic AMP (reviewed in (18)). Such cells are widely used as

247

model systems of neutrophils, since they exhibit neutrophil-specific functions, such as chemotaxis, lysosomal enzyme secretion, phagocytosis, and agonist-stimulated superoxide release. Differentiation is accompanied by the expression of cell surface receptors for the formylated chemotactic peptide fMLP of bacterial origin, which can activate various neutrophil-specific functions. Since several reports have proposed that G-protein-coupled receptors may functionally couple to cPLA 2, PIP 2-PLC, and PLD in db-cAMP-differentiated HL-60 cells (8, 19, 20), we have investigated the contribution of these phospholipases to fMLP-mediated arachidonic acid release in these cells. Interestingly, we found that fMLP-stimulated AA-release is independent of PIP 2-PLC-, MAP kinase-, and cPLA 2-dependent pathways. The bulk of fMLP-mediated AA release from db-cAMP-differentiated HL-60 cells appears to be the consequence of activation of PC-PLC- and PLD-dependent pathways. EXPERIMENTAL PROCEDURES

Materials Tritium-labeled arachidonic acid ([ 3H]AA, 100 Ci/mmol), enhanced chemiluminescence kit (ECL), and sheep anti-rabbit IgG were purchased from Amersham (UK); fura-2-AM and BAPTA-AM were purchased from Molecular Probes (Eugene, OR). Iscove’s modified Dulbecco’s media (IMDM) was obtained from Gibco (Eggenstein, Germany) and fetal calf serum from PAA Laboratories (Co¨lbe, Germany). U 73122, arachidonyl trifluoromethyl ketone (AACOCF 3), propranolol, and Pefabloc were from Biomol (Hamburg, Germany), methyl arachidonyl fluorophosphonate (MAFP) and bromoenol lactone (BEL) from SPI-Bio (Massy, France), and N-formylmethionylL-leucyl-L-phenylalanine (fMLP) and D609 from Sigma (Deisenhofen, Germany). Ro 318220 and thapsigargin (TG) were purchased from Calbiochem (Bad Soden, Germany) and PVDF membranes from DuPont (Germany). Clostridium difficile toxin B was a gift from Professor Dr. Aktories and Dr. Ingo Just (Institute for Pharmacology and Toxicology, Albert-Ludwigs-University Freiburg, Germany). The inhibitory peptide to sPLA 2, FLSYK, was synthesized by Dr. W. Nastainczyk (Department of Medical Biochemistry, University of Saarland, Homburg/Saar, Germany). Horseradish peroxidase-conjugated sheep anti-mouse and donkey anti-rabbit IgG were from Amersham-Buchler (Braunschweig, Germany). The polyclonal antibody against human cPLA 2 and the 85-kDa cPLA 2 protein standard were obtained from Genetics Institute (Cambridge, MA). The mouse monoclonal antibodies against p38-MAP kinase, phospho-p38-MAP kinase, and phospho-p42/44-MAP kinase as well as the rabbit polyclonal antibody to p42/44 MAP kinase were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). All other materials were obtained in reagent grade from commercial sources.

Methods Culture and differentiation of HL-60 cells. HL-60 cells (originally obtained from American Type Culture Collection) were used between passages 20 and 100. Cells were routinely cultured in Iscove’s modified Dulbecco’s media (IMDM) supplemented with 10% heat-inactivated fetal calf serum, 100 units/ml penicillin, and 100 ␮g/ml streptomycin at 37°C under 5% CO 2, at a density of 5–10 ⫻ 10 5 cells/ml. Before differentiation, cells were cultured for 24 h at a density of 5 ⫻ 10 5 cells/ml serum-free IMDM, supplemented with 5 ␮g/ml insulin, 5 ␮g/ml transferrin, 5 ng/ml sodium selenite, 100 units/ml penicillin, and 100 ␮g/ml streptomycin (IMDM-ITS) at 37°C with 5% CO 2.

248

´ VENOD, AND SCHULZ STERNFELD, THE

Differentiation into a neutrophil-like phenotype was induced by incubation in the presence of 0.5 mM db-cAMP for 72 h according to Chaplinski and Niedel (18). Preparation of cytosol and membrane fractions from HL-60 cells. Differentiated HL-60 cells (1–2 ⫻ 10 7) were washed with IMDM-ITS media, resuspended at a density of 5 ⫻ 10 5 cells/ml in IMDM-ITS media, and cultured for 60 min at 37°C under 5% CO 2 prior to stimulation. Following incubation with agonists, cells were placed on ice and centrifuged for 5 min at 250g (4°C). Cells were washed once with ice-cold buffer (DTT, 2 mM; trypsin inhibitor, 50 ␮g/ml; CaCl 2, 0.5 mM; MgCl 2, 0.5 mM; KCl, 2.5 mM; NaCl, 136 mM; Na 3VO 4, 0.1 mM; NaHPO 4, 1.3 mM; KH 2PO 4, 0.2 mM; pH 7.4) and resuspended in 2 ml Hepes buffer A (DTT, 2 mM; leupeptin, 0.1 mM; trypsin inhibitor, 50 ␮g/ml; Na 3VO 4, 0.1 mM; EGTA, 1 mM; EDTA, 1 mM; Hepes, 10 mM; mannose, 280 mM; KCl, 10 mM; pH 7). Cells were broken by N 2 cavitation in a nitrogen bomb (Parr Instruments, USA) with 5175 kPa. for 2 min. Under these conditions, free Ca 2⫹ concentrations between 50 and 100 nM were measured in the cytosolic fraction. Unbroken cells and nuclei were removed by centrifugation at 1000g for 5 min and the supernatant was centrifuged at 100,000g for 60 min. The 100,000g supernatant was defined as cytosolic fraction and the 100,000g pellet as membrane fraction. The membranes were resuspended by sonication in Hepes buffer. Measurement of [ 3H]Arachidonic acid release from db-cAMP-differentiated HL-60 cells. [ 3H]AA release was measured according to the method of Xing et al. (21). Briefly, differentiated HL-60 cells were harvested by centrifugation at room temperature (5 min, 250g) and diluted to approximately 5 ⫻ 10 6 cells/ml in culture medium (IMDMITS). They were incubated for 90 min with 1 ␮Ci/ml [ 3H]AA (100 Ci/mmol) at 37°C and 5% CO 2. Within 90 min more than 80% of the added [ 3H]AA was incorporated into the cellular phospholipid fraction, as detected by thin-layer chromatography. [ 3H]AA-labeled lipids following 90 min of incubation (data not shown) were found in phosphatidylcholine (15.5 ⫾ 1.1%), phosphatidylinositol (11.9 ⫾ 3.1%), and phosphatidylethanolamine (71.5 ⫾ 1.8%), but not in neutral lipids or phosphatidylserine. Incubation for 20 h left labeling of phosphatidylinositol nearly unchanged (7.7 ⫾ 3.3%) and decreased the amount of labeled phosphatidylcholine (6.1 ⫾ 2.1%), while the amount of labeled phosphatidylethanolamine increased (85.5 ⫾ 1.5%). These data are compatible with the results of a previous study (22). These changes affected neither the contribution of different phospholipases to fMLP-induced [ 3H]AA release nor the inhibitory effects of various drugs used in this study (e.g., U 73122, AACOCF 3, propranolol) (data not shown). At the end of the labeling phase, cells were washed twice with Hanks’ balanced salt solution (HBSS; CaCl 2, 1.3 mM; MgCl 2, 0.8 mM; and 0.3% fatty acid free BSA) and resuspended at a density of 10 7 cells/ml in HBSS. To monitor the incorporation of radioactivity, aliquots from each washing step were counted. Measurement of [ 3H]AA release was started by adding 150 ␮l of this suspension to 150 ␮l HBSS at 37°C containing unlabeled arachidonic acid to prevent a rapid reincorporation of released labeled [ 3H]AA (final concentration, 0.4 mM) and inhibitors or their carriers. With the exception of the DAG lipase inhibitor (RHC 80267, 30 min, 100 ␮M) incorporation of [ 3H]AA was not affected by the inhibitors tested. RHC 80267 decreased the percentage of incorporated [ 3H]AA into the cells by nearly 50%. Therefore the data showing a reduction of fMLP-mediated AA release in the presence of the DAG lipase inhibitor have not been included in this study, since they are difficult to interpret. After a preincubation time of 5–10 min, agonists were added. Previous studies have shown that agonistinduced AA release is maximal at 10 –15 min (21); therefore, if not indicated otherwise, [ 3H]AA measurements were carried out 10 min after stimulation. The reaction was stopped with 3 ml ice-cold Tris/ HCl solution (Tris, 50 mM; KCl, 100 mM; EGTA, 5 mM; and EDTA, 5 mM; pH 7.5). For the determination of total content of radioactivity, one aliquot of cell suspension (500 ␮l) was added to 2 ml scintillation cocktail (Ultima Flo, Packard). For determination of the released [ 3H]AA, the cell suspension was centrifuged at 1400g for 10

min and 500 ␮l of the supernatant was added to 2 ml scintillation cocktail. To make sure that [ 3H]AA released into the supernatant represented the totality of [ 3H]AA released from phospholipids, the composition of [ 3H]AA-labeled lipids in control and fMLP stimulated cells was measured according to Bligh and Deyer (23). Thin-layer chromatography showed that the bulk of the radioactivity in the supernatants from stimulated cells comigrated with the AA standard (data not shown). AA was also extracted by the method of Dole and Meinertz, which has a recovery for free AA of 95% (24). With this method, less than 3% of free AA was detected in the cells after stimulation with 10 ␮M fMLP (data not shown), which indicates that AA released by phospholipases is not stored in the cells but is rapidly and completely released into the supernatant. All samples were counted for 12 min in a liquid scintillation analyzer (tricarb 2100tr, Packard). In each experiment, measurements were carried out in triplicates. Measurement of cytosolic Ca 2⫹ concentration. Dibutyryl cAMPdifferentiated HL-60 cells (3–5 ⫻ 10 6 cells/ml) were loaded with 2 ␮M of the acetoxymethylester form of Fura-2 for 40 min at 37°C in modified tyrode solution (NaCl, 139 mM; KCl, 5 mM; MgCl 2, 1 mM; Hepes, 25 mM; glucose, 5 mM; CaCl 2, 1.5 mM; BSA, 1 mg/ml; pH 7.4). Thereafter cells were washed twice, incubated for an additional 10 min at 37°C, resuspended at a density of 5 ⫻ 10 6 cells/ml in the modified tyrode solution, and stored on ice. Three hundred microliters of the cell suspension were added to 2.7 ml of modified tyrode solution and preincubated for 5 min. Fluorescence was determined at 37°C under constant stirring using a Fluoromax-2 spectrofluorometer (Instruments S.A., Inc.). The excitation wavelengths were 340 and 380 nm; the emission wavelength was 505 nm. At the end of each experiment fluorescence signals were calibrated in the presence of 10 ␮g/ml digitonin to lyse the cells. For the determination of F max, Ca 2⫹ was added to a final concentration of the incubation buffer above 1.5 mM. F min was determined by the addition of 10 mM NaEGTA. Cytosolic Ca 2⫹ concentration was calculated as described by Grynkiewicz et al. (25). Immunoprecipitation of cPLA 2. cPLA 2 was immunoprecipitated using the method described by Xing and Mattera (8). Briefly, cell lysate (600 ␮g protein) was dissolved in a total volume of 500 ␮l of Hepes buffer A (see above for preparation of cellular fractions) with 1% Nonidet P-40 (NP-40) and 1 ␮l stock solution of polyclonal antibody against human cPLA 2 followed by agitation overnight at 4°C. Approximately 20 mg protein A–Sepharose suspended in 50 ␮l Hepes buffer A was added. After 1–3 h agitation the Sepharose beads were washed three times with Hepes buffer A containing 1% NP-40 and two more times in NP-40 free Hepes buffer A. The pellets were resuspended in 30 ␮l SDS-loading buffer (10 mM Tris, 1% mercaptoethanol, 1% SDS, 10% glycerol, and 0.015% bromphenol blue) and heated for 5 min at 95°C for SDS–PAGE. Western blotting. For Western blot analysis of cPLA 2, protein obtained by immunoprecipitation was separated on SDS–PAGE (6.25% acrylamide) Laemmli mini-gels and transferred onto PVDF membranes overnight. Blots were blocked for 8 h with 3% nonfat dry milk and incubated with polyclonal antibody against human cPLA 2 (1:3000 dilution). Following incubation with horseradish peroxidaseconjugated donkey anti-rabbit IgG (1:6000) for 1 h, the cPLA 2 was visualized using the enhanced chemiluminescence technique. For Western blots of p38 and p42/44 MAP kinases and phospho-MAP kinases, membrane proteins from db-cAMP-differentiated HL-60 cells, which had been stimulated with 10 ␮M fMLP for various time periods, rapidly frozen in liquid nitrogen and homogenized in the presence of the phosphatase inhibitor sodium pervanadate (2 mM), were separated using 12% acrylamide SDS–PAGE. Following transfer onto PVDF membranes as described above, primary antibodies were used at a dilution of 1:1000 and horseradish peroxidase-conjugated secondary antibodies at a dilution of 1:10,000. Data presentation. If not indicated otherwise, the figures show values for agonist-induced [ 3H]AA release in percentage of incorporated [ 3H]AA. It is defined as [ 3H]AA released in the presence of the

fMLP-INDUCED AA RELEASE IN HL-60 CELLS agonist minus basal release. In 80 experiments, basal release averaged 1.3 ⫾ 0.8% of incorporated [ 3H]AA and was not affected by the inhibitors tested, unless specifically indicated in the text. The data were analyzed by unpaired Student’s t test, and P ⬍ 0.05 was considered significant.

RESULTS

fMLP Increases Cytosolic [Ca 2⫹] by Activation of PIP 2-PLC In db-cAMP-differentiated HL-60 cells, ATP and fMLP are equally efficacious in stimulating PIP 2-PLC activity and Ca 2⫹ mobilization (26). The effects of a maximally effective concentration of ATP (0.1 mM) or fMLP (10 ␮M) on the Ca 2⫹ transient are shown in Fig. 1A (peak Ca 2⫹ signal for ATP, 351 ⫾ 98 nM cytosolic Ca 2⫹, n ⫽ 7; for 10 ␮M fMLP, 336 ⫾ 90 nM Ca 2⫹, n ⫽ 12). Figure 1B demonstrates the contribution of Ca 2⫹ released from intracellular stores and of Ca 2⫹ influx from the extracellular space to the Ca 2⫹ signal induced by 10 ␮M fMLP in HL-60 cells. In the presence of 1.5 mM extracellular Ca 2⫹, fMLP induced a very rapid and large increase in [Ca 2⫹] from the basal level of approximately 50 nM to a peak level of about 400 nM, due to Ca 2⫹ release from intracellular stores, followed by a Ca 2⫹ plateau attributed to Ca 2⫹ influx into the cell which slowly declined to about 250 nM within the time of observation (4 min). In the absence of extracellular [Ca 2⫹], the magnitude of the rise in cytosolic [Ca 2⫹] induced by fMLP was moderately reduced and followed by a rapid decay to the prestimulated level within 60 s (Fig. 1B). In the presence of 1.5 mM extracellular Ca 2⫹, preincubation of db-cAMP-differentiated HL-60 cells for 5 min with the PIP 2-PLC inhibitor U73122 (1 ␮M) abolished fMLP-induced increase in cytosolic [Ca 2⫹] (Fig. 1C). U 73122 was tested at concentrations ranging between 0.1 and 10 ␮M with an IC 50 of 0.54 ⫾ 0.17 ␮M (n ⫽ 3–9). The ineffective structural analog U 73343, when tested at a 10-fold higher concentration, had no effect on fMLP-induced increase in cytosolic [Ca 2⫹] (Fig. 1C). Since it has been reported for pancreatic acinar cells that U 73122 inhibits a TG-sensitive Ca 2⫹ ATPase of IP 3-sensitive Ca 2⫹ pools resulting in Ca 2⫹ depletion of these pools (27, 28), we tested whether U 73122 had any affect on the filling state of intracellular Ca 2⫹ pools in db-cAMP-differentiated HL-60 cells (inset of Fig. 1C). In the presence of 1.5 mM extracellular Ca 2⫹, the maximal cytosolic Ca 2⫹ concentration obtained by addition of 1 ␮M TG reached about 80% of the fMLP (10 ␮M)-induced Ca 2⫹ peak (Fig. 1C), which is comparable to the Ca 2⫹ signal obtained in the absence of extracellular Ca 2⫹ (see Fig. 1B). When 1 ␮M TG was added during the Ca 2⫹ plateau phase induced by fMLP, it failed to further release Ca 2⫹. This indicates that intracellular Ca 2⫹ stores affected by TG are identical to the fMLP-activatable intracellular Ca 2⫹ stores. Preincubation for 5 min with concentrations of

249

U 73122 varying between 0.1–1 ␮M gradually decreased the fMLP-induced Ca 2⫹ signal. However, subsequent addition of TG (1 ␮M) during the plateau phase of the respective fMLP-induced Ca 2⫹ signal caused an additional Ca 2⫹ release from internal Ca 2⫹ stores (Fig. 1C, inset). The sum of the heights of the Ca 2⫹ peaks obtained with fMLP and TG at varying U 73122 concentrations was comparable to the maximal Ca 2⫹ signal obtained by the addition of TG alone. These data indicate that the effect of U 73122 (1 ␮M) on the Ca 2⫹ signal is not due to depletion of intracellular Ca 2⫹ pools. Another way to prevent increase in cytosolic free [Ca2⫹] induced by fMLP is the use the membrane-permeant Ca 2⫹ chelator BAPTA-AM. When BAPTA-AM (10 ␮M) was added to the bath, basal cytosolic [Ca 2⫹] decreased to 27 ⫾ 7 nM [Ca 2⫹] (n ⫽ 5; P ⬍ 0.001). When cells were incubated in a nominally Ca2⫹-free solution, no increase in cytosolic [Ca 2⫹] could be induced by the addition of 10 ␮M fMLP (Fig. 1D). As described by Hofer et al. (29) BAPTA-AM is taken up into cytosol and Ca2⫹ stores and decreases the Ca 2⫹ concentration in the stores. The absence of an fMLP-mediated Ca2⫹ signal after preincubation with 10 ␮M BAPTA-AM in nominally Ca2⫹-free solutions must be explained by emptying of the stores and buffering of Ca 2⫹ in stores and cytosol. When 1.5 mM Ca 2⫹ was present in the extracellular space, basal cytosolic [Ca 2⫹] was not different from that at low extracellular [Ca 2⫹]. The addition of fMLP (10 ␮M) led to a slow increase in cytosolic [Ca2⫹] in the presence of BAPTAAM, which was caused by Ca2⫹ influx from the extracellular space (Fig. 1D). A Functional cPLA 2 Is Expressed in db-cAMPDifferentiated HL-60 Cells In the presence of millimolar concentrations of extracellular [Ca 2⫹], the Ca 2⫹ ionophore A23187 causes an increase in cytosolic [Ca 2⫹], which leads to a translocation of cPLA 2 from the cytosol to cellular membranes where its phospholipid substrates are present (4, 10). As can be seen in Fig. 2A, the addition of 10 ␮M A23187 to db-cAMP-differentiated HL-60 cells incubated in 1.3 mM extracellular Ca 2⫹ induced a small release of AA (2.6 ⫾ 1.2%; n ⫽ 13; P ⬍ 0.001). These values are similar to those determined by others (8, 30). Preincubation for 5 min with the protein kinase C activator PMA (0.2 ␮M) had no stimulatory effect on AA release (0.2 ⫾ 0.2%; n ⫽ 12), indicating that phosphorylation of cPLA 2 by PKC-dependent pathways, though essential, is not sufficient to activate cPLA 2 (7). However, a combination of both stimuli (10 ␮M A23187 ⫹ 0.2 ␮M PMA) strongly stimulated AA release (10.5 ⫾ 4.3%; n ⫽ 12; P ⬍ 0.001) (Fig. 2A). This is thought to be the consequence of phosphorylationdependent stimulation of the enzymatic activity and

250

´ VENOD, AND SCHULZ STERNFELD, THE

fMLP-INDUCED AA RELEASE IN HL-60 CELLS

Ca 2⫹-dependent translocation of activated cPLA 2 to the membranes (9). As shown in Fig. 2B, a cPLA 2 (its apparent molecular weight on SDS–PAGE gels is about 110 kDa) could be immunoprecipitated from the lysate of db-cAMP-differentiated HL-60 cells with a polyclonal antiserum directed against the 85-kDa cPLA 2 protein (4) (see lane with cPLA 2 standard). Under control conditions, cPLA 2 was almost exclusively located in the cytosol of dbcAMP-differentiated HL-60 cells (control cytosol as compared to control membranes). Following incubation of cells with 10 ␮M A23187 and 0.2 ␮M PMA for 1, 5, or 10 min as described above, a substantial amount of cPLA 2 protein was translocated to the membrane fraction of HL-60 cells (Fig. 2B, “membranes” A23187 ⫹ PMA) that could account for the increase of [ 3H]AA release shown in Fig. 2A. fMLP-Induced [ 3H]AA Release Is Independent of PIP 2-PLC Stimulation, of Increase in Cytosolic [Ca 2⫹], and of PKC Activation When 10 ␮M fMLP was added to db-cAMP-differentiated HL-60 cells, which had been incubated in a medium with 1.3 mM Ca 2⫹, they rapidly released incorporated [ 3H]AA (Fig. 3). Release was significantly increased (3.3 ⫾ 0.5% of incorporated [ 3H]AA) at the earliest time point determined after addition of the agonist, i.e., 1 min, and reached a plateau within 5 min (Fig. 3A). Release was 6.6 ⫾ 1.3% of incorporated [ 3H]AA 10 min after addition of 10 ␮M fMLP (n ⫽ 12; Fig. 3C). When a dose–response analysis of fMLP-induced [ 3H]AA release was performed, curve fitting of the data yielded an apparent K d for fMLP of ⬇20 nM and maximal stimulation of [ 3H]AA release was observed at ⱖ1 ␮M fMLP (Fig. 3B). This is similar to the results obtained by Cockcroft and Stutchfield (20). The amount of [ 3H]AA release induced by a maximally effective concentration of 10 ␮M fMLP was comparable in experiments in which extracellular Ca 2⫹ had been omitted and in which extracellular [Ca 2⫹] had been adjusted to submicromolar concentrations by addition of 10 ␮M EGTA (6.2 ⫾ 0.6%; n ⫽ 3; Fig. 3C). When cells were incubated in a medium containing 1.3 mM

251

extracellular Ca 2⫹ in the presence of the phospholipase C inhibitor U 73122 (1 ␮M; see Fig. 1), [ 3H]AA release induced by 10 ␮M fMLP was not different from that of controls without inhibitor (controls, 7.6 ⫾ 1.1%, versus U 73122, 7.7 ⫾ 1.0%; n ⫽ 6; n.s.). This effect was observed as early as 1 min after the addition of fMLP (Fig. 3A) and was independent of the concentration of the agonist used (Fig. 3B). When cells were incubated in nominally Ca 2⫹-free medium in the presence of 10 ␮M BAPTA-AM to clamp cytosolic [Ca 2⫹] to about 30 nM (see Fig. 1C), fMLP-stimulated [ 3H]AA release was not significantly different from that of the respective controls, in which contaminating extracellular [Ca 2⫹] had been chelated with EGTA (BAPTA-AM, 5.5 ⫾ 0.7%, versus EGTA, 6.2 ⫾ 0.6%; n ⫽ 3; n.s.) (Fig. 3C). Since PKC activation of p38 and p42/44 MAP kinases leads to phosphorylation of cPLA 2 (11), we tested the effect of the PKC inhibitor Ro 318220 (5 ␮M) on fMLPinduced [ 3H]AA release. Ro 318220 had no effect on fMLP-induced [ 3H]AA release (controls, 6.5 ⫾ 1.7%, versus Ro 318220, 6.4 ⫾ 1.6%; n ⫽ 4; n.s.) (Fig. 2A). In contrast, 5 ␮M Ro 318220 almost completely inhibited the activation of cPLA 2 induced by 10 ␮M A23187 ⫹ 0.2 ␮M PMA (1.6 ⫾ 1.2%; n ⫽ 3; P ⫽ 0.005; Fig. 2A). Immunoblotting of fMLP-stimulated db-cAMP-differentiated HL-60 cells using antibodies to the phosphorylated forms of p38 and p42/44 MAP kinases revealed an increased phosphorylation of MAP kinases that was detectable 15 s after addition of the agonist (Fig. 4A). Maximal phosphorylation was observed 15–30 s and 1–2 min after addition of fMLP (10 ␮M) for p42/44 and p38 MAP kinases, respectively, and returned to control levels 5 min after addition of fMLP (Fig. 4A). The PKC inhibitor Ro 318220 (5 ␮M) and the PIP 2-PLC inhibitor U 73122 (1 ␮M) reduced fMLP-induced phosphorylation of MAP kinases by about 40 and 60%, respectively (Fig. 4B), suggesting that fMLP activates p38 and p42/44 MAP kinases through Ca 2⫹- and PKC-dependent and -independent pathways and that Ca 2⫹- and PKC-dependent pathways are not involved in fMLPinduced [ 3H]AA release. In addition, propranolol (0.2 and 0.6 mM) had no effect on fMLP-mediated phosphorylation of MAP kinases (data not shown). More

FIG. 1. Agonist-induced increase in [Ca 2⫹] i in db-cAMP-differentiated HL-60 cells. [Ca 2⫹] i in Fura-2 loaded cells was measured as described under Experimental Procedures. Changes in [Ca 2⫹] i were induced by adding aliquots of the agonist stock solutions (1000-fold concentrated) to the cell suspensions. The arrows indicate additions of agonists (fMLP, ATP) or BAPTA-AM. Experiments are representative of 4 –32 similar experiments. (A) Effect of fMLP (10 ␮M) or ATP (100 ␮M) on [Ca 2⫹] i of HL-60 cells. Cells were incubated in a modified tyrode solution containing 1.5 mM extracellular [Ca 2⫹]. (B) Effect of extracellular Ca 2⫹ concentration (1.5 mM Ca 2⫹ versus nominally [Ca 2⫹]-free) on fMLP (10 ␮M)-induced changes in [Ca 2⫹] i. (C) Effects of the phosphatidylinositol-4,5-bisphosphate-specific phospholipase C (PIP 2-PLC) inhibitor U 73122 (1 ␮M) and of the ineffective structural analog U 73343 (10 ␮M) on fMLP-induced changes of [Ca 2⫹] i in differentiated HL-60 cells. The inhibitors were added 5 min prior to addition of fMLP. In the experiment with U 73122 TG (1 ␮M) was given 60 s after fMLP. The inset summarizes the effect of preincubation with different concentrations of U 73122 (0 –1 ␮M) on fMLP- and thapsigargin (TG; 1 ␮M)-induced changes of [Ca 2⫹] i. Depletion of intracellular Ca 2⫹ pools by TG was induced after Ca 2⫹ release by fMLP (for further explanations, see Results). (D) fMLP-induced changes of [Ca 2⫹] i in differentiated HL-60 cells incubated with the membrane-permeant Ca 2⫹-chelator BAPTA-AM (10 ␮M). Measurements were either performed in an extracellular solution containing 1.5 mM Ca 2⫹ or in a nominally [Ca 2⫹]-free solution.

252

´ VENOD, AND SCHULZ STERNFELD, THE

FIG. 2. Effects of the inhibitors of protein kinase C (Ro 318220) and cPLA 2/iPLA 2 (AACOCF 3) on fMLP-induced or A23187 ⫹ PMA-induced [ 3H]AA release (A) and Ca 2⫹-dependent translocation of cPLA 2 to the membrane fraction of fMLP-stimulated or A23187 ⫹ PMA-stimulated db-cAMP-differentiated HL-60 cells (B). (A) Cells were incubated for 5 min with the indicated compounds or their solvents (0.1% v/v ethanol or DMSO) for the respective controls. The protein kinase C inhibitor Ro 318220 (5 ␮M) and the inhibitor of cPLA 2/iPLA 2, arachidonyl trifluoromethyl ketone (AACOCF 3, 10 ␮M), were added 5 min prior to addition of fMLP (10 ␮M) or A23187 (10 ␮M) ⫹ PMA (0.2 ␮M). Means ⫾ SD of 4 –13 different cell preparations. (B) Western blotting of immunoprecipitated cPLA 2 extracted from HL-60 cell membranes and cytosol using the anti-cPLA 2 antibody (Genetics Institute, Cambridge, MA). Dibutyrl cAMP-differentiated HL-60 cells were incubated with fMLP (10 ␮M), A23187 (10 ␮M) ⫹ PMA (0.2 ␮M) or with DMSO (0.1%) for 10 min prior to separation of cytosol and membranes. Purified recombinant 85-kDa cPLA 2 (cPLA 2 standard) was also loaded in order to define the position of this protein on the gel. Negative controls include omission of primary antibody or omission of HL-60 cell fractions. One representative experiment out of four similar ones is shown. Similar results were obtained when cells had been stimulated for 1 or 5 min with fMLP prior to separation of cytosol and membranes (not shown).

fMLP-INDUCED AA RELEASE IN HL-60 CELLS

253

important was to know whether activation of MAP kinases is involved at all in fMLP-induced AA release—regardless of their activation through Ca 2⫹- and PKC-dependent and -independent pathways. In summary, these data suggest that fMLP-induced [ 3H]AA release is independent of the activation of PIP 2-PLC, of increased cytosolic [Ca 2⫹], and of PKC-mediated and/or MAP kinase-mediated phosphorylation. It seems therefore unlikely that Ca 2⫹-, PKC-, and MAP kinasedependent activation of cPLA 2 mediate fMLP-induced [ 3H]AA release in db-cAMP-differentiated HL-60 cells. fMLP-Induced [ 3H]AA Release Does Not Involve Stimulation of cPLA 2

FIG. 3. Effects of inhibitors of cPLA 2 (AACOCF 3 ) and iPLA 2 , (MAFP), of phosphatidic acid phosphohydrolase (propranolol), of PIP 2 -PLC (U 73122), and of Ca 2⫹ chelators on fMLP-induced [ 3 H]AA release in db-cAMP-differentiated HL-60 cells. [ 3 H]AAprelabeled db-cAMP-differentiated HL-60 cells were incubated for 5 min with the inhibitors or DMSO (0.1%) before addition of the agonist fMLP (A, B). (A) Time course analysis of fMLP-stimulated (10 ␮M) [ 3 H]AA release in the absence (E) and presence of inhibitors (, 25 ␮M MAFP; ƒ, 10 ␮M AACOCF 3 ; ■, 0.2 mM propranolol; 䊐, 1 ␮M U 73122). The data shown represent release of [ 3 H]AA after subtraction of basal release. Basal release amounted to 1.2 ⫾ 0.7% at 1 min, 1.8 ⫾ 0.5% at 3 min, 2.2 ⫾ 0.8% at 5 min, 2.4 ⫾ 0.9 at 7 min, and 2.5 ⫾ 0.9% at 10 min. (B) Dose–response curves of fMLP-stimulated [ 3 H]AA release during a period of 10 min from HL-60 cells without (E) and with inhibitors (, 25 ␮M MAFP; ƒ, 10 ␮M AACOCF 3 ; ■, 0.2 mM propranolol; 䊐, 1 ␮M U 73122) are plotted. Basal release amounted to 2.2 ⫾ 0.8%. (C) HL-60 cells were incubated for 10 min in nominally Ca 2⫹ -free solution with or without 10 ␮M of membrane-permeant Ca 2⫹ che-

To investigate the role of cPLA 2 in fMLP-induced [ 3H]AA release, we examined the effect of AACOCF 3, a trifluoromethyl ketone analog of AA that can inhibit the 85-kDa cPLA 2 and, with lower affinity, the Ca 2⫹independent PLA 2 (iPLA 2), but not the 14-kDa lowmolecular-weight forms of PLA 2 (31–34). As shown in Fig. 3A and Table I, 10 ␮M AACOCF 3, which inhibits more than 50% of [ 3H]AA released from A23187-challenged U937 cells (31), had no effect on fMLP-induced [ 3H]AA release (controls, 7.6 ⫾ 1.1%, versus AACOCF 3, 7.7 ⫾ 0.9%; n ⫽ 7; n.s.). This indicates that cPLA 2 is not responsible for fMLP-mediated AA release in dbcAMP-differentiated HL-60 cells. A higher concentration of AACOCF 3 (30 ␮M) did not have an inhibitory effect on the kinetics of [ 3H]AA release induced by 10 ␮M fMLP or on the dose–response curve for fMLPinduced [ 3H]AA release (data not shown). In our hands, 10 ␮M AACOCF 3 partially inhibited [ 3H]AA release induced by 10 ␮M A23187 ⫹ 0.2 ␮M PMA (from 10.5 ⫾ 4.3% to 6.5 ⫾ 1.5%; n ⫽ 4; P ⬍ 0.03; Fig. 2A), which confirms that AACOCF 3 specifically inhibits Ca 2⫹- and PKC-activated cPLA 2 expressed in db-cAMP-differentiated HL-60 cells. Additional evidence that this cPLA 2 is not involved in fMLP-induced [ 3H]AA release from db-cAMP-differentiated HL-60 cells was obtained from studies with the polyclonal antiserum to cPLA 2: In three different experiments we did not observe any translocation of cPLA 2 to the membrane fraction of db-cAMP-differentiated HL-60 cells in the presence of fMLP (10 ␮M for 1, 5, or 10 min) (Fig. 2B). The 14-kDa secretory (group II) PLA 2 is expressed in db-cAMP-differentiated HL-60 cells (data not shown). It becomes activated once released into the extracellular space, where Ca 2⫹ is present at millimolar concentrations (1). Its contribution to fMLP-induced [ 3H]AA release in db-cAMP-differentiated HL-60 cells can be excluded, since no inhibition of fMLP-induced [ 3H]AA

lator (BAPTA-AM) prior to addition of fMLP (10 ␮M). Bars indicate fMLP-stimulated [ 3H]AA release during a time span of within 10 min. Values are means ⫾ SD of 3–12 different experiments.

254

´ VENOD, AND SCHULZ STERNFELD, THE

FIG. 4. Effect of the inhibitors of PIP 2-PLC (U 73122, 1 ␮M) and protein kinase C (Ro 318220, 5 ␮M) on phosphorylation of p38 and p42/44 MAP kinases in fMLP-stimulated db-cAMP-differentiated HL-60 cells. (A) Representative time course of fMLP-induced (10 ␮M) phosphorylation of p38 and p42/44 MAP kinases by immunoblot analysis using antibodies to the unphosphorylated and phosphorylated forms of the MAP kinases. Dephosphorylation was prevented by rapid freezing and homogenization of cells in the presence of 2 mM sodium pervanadate. One out of four similar experiments is shown. (B) Dibutyryl cAMP-differentiated HL-60 cells were incubated for 5 min with the inhibitors or DMSO (0.1%) before addition of the agonist fMLP (10 ␮M). Incubation of cells was stopped 60 s after addition of fMLP by rapid freezing, and immunoblot analysis was performed. Signals from different experiments were scanned, and the intensity (optical density) of the chemiluminescence signals was quantified on a Bioprofil computer-assisted imaging and scanning system (Vilber-Lourmat; Marne La Valle´e, France). Data are plotted as percentages of the optical density of phosphorylated MAP kinases with 10 ␮M fMLP (means ⫾ SD of three to five different experiments). Statistical analysis was by unpaired Student’s t test.

release was observed when extracellular [Ca 2⫹] was adjusted to micro- or submicromolar concentrations (see Fig. 3C). Moreover, the inhibitory peptide to group II sPLA 2, FLSYK (35), up to 100 ␮g/ml had no effect on fMLP-stimulated [ 3H]AA release (controls, 6.4 ⫾ 0.6%, versus FLSYK, 6.7 ⫾ 0.4%; n ⫽ 3; n.s.; Table I).

A Ca 2⫹-independent PLA 2 (iPLA 2) has been shown to be involved in receptor-mediated AA release from smooth muscle cells (36), which could be completely inhibited by 10 ␮M (E)-6-(bromomethylene)-3-(1-naphthalenyl)-2H-tetrahydropyran-2-one (BEL). BEL possesses a 1000-fold higher selectivity for the inhibition

fMLP-INDUCED AA RELEASE IN HL-60 CELLS TABLE I fMLP-stimulated AA release (% of total content) Inhibitor (concentration)

Selectivity

U 73122 (1 ␮M) Ro 318220 (5 ␮M) AACOCF 3 (10 ␮M) MAFP (25 ␮M) BEL (50 ␮M) FLSYK (150 ␮M) ToxB (10 ng/ml) 1-Butanol (40 mM) Propranolol (200 ␮M) D609 (100 ␮M) D609 (100 ␮M) ⫹ 1-butanol (40 mM)

PIP 2-PLC PKC cPLA 2/iPLA 2 cPLA 2/iPLA 2 iPLA 2 sPLA 2 PLD PLD PA-P PC-PLC PC-PLC PLD

⫺ Inhibitor ⫹ Inhibitor

n

6.3 ⫾ 1.1 6.4 ⫾ 1.6 6.9 ⫾ 1.7 7.5 ⫾ 1.3 7.6 ⫾ 1.4 6.4 ⫾ 0.6 5.3 ⫾ 0.4 5.3 ⫾ 1.0 6.8 ⫾ 1.4 5.3 ⫾ 1.3

6.5 ⫾ 1.6 6.3 ⫾ 1.4 7.2 ⫾ 2.1 5.6 ⫾ 1.4* 6.0 ⫾ 1.5 6.7 ⫾ 0.4 3.5 ⫾ 0.2* 2.7 ⫾ 1.0* 1.7 ⫾ 1.2* 3.1 ⫾ 1.6*

6 4 7 12 7 3 3 7 13 6

4.7 ⫾ 1.3

1.6 ⫾ 0.7

3

Note. Preincubation of cells for 4 h was necessary with D609 (100 ␮M) and ToxB (10 ng/ml) or for 5 min with all other compounds listed in Table I. * P ⬍ 0.05 by unpaired Student’s t test.

of iPLA 2 versus cPLA 2 or sPLA 2 (37). As shown in Table I, BEL (50 ␮M) had no inhibitory effect on fMLPinduced [ 3H]AA release (controls, 7.6 ⫾ 1.4%, versus BEL, 6.0 ⫾ 1.5%; n ⫽ 7; n.s.). Methyl arachidonyl fluorophosphonate (MAFP), a less specific inhibitor of

255

iPLA 2 in P388D 1 macrophages (38), which also inhibits cPLA 2 (39), was tested with preincubation periods of 5–30 min at concentrations between 1 and 100 ␮M. We found 5 min and 25 ␮M to be maximally effective. The IC 50 10.7 ⫾ 1.6 ␮M is comparable to the data shown by others (38). When tested at the maximally effective concentration of 25 ␮M weakly reduced fMLP-induced [ 3H]AA release (see Figs. 3A and 3B and Table I) (controls, 7.5 ⫾ 1.3%, versus MAFP, 5.6 ⫾ 1.4%; n ⫽ 8; P ⬍ 0.01). In summary, these data indicate that none of the PLA 2s investigated in HL-60 cells (cPLA 2, sPLA 2, or iPLA 2) plays a major role in fMLP-stimulated [ 3H]AA release. fMLP-Induced [ 3H]AA Release Involves Activation of PC-PLC and PLD As can be seen from the scheme shown in Fig. 5, phosphatidylcholine (PC) can be hydrolyzed by PLD into phosphatidic acid (PA) and choline and by PC-PLC into DAG and phosphocholine. AA could be generated from both PA and DAG by the sequential actions of phosphatidic acid phosphohydrolase (PA-P) and DAG/ MAG lipases. We therefore tested whether the PLD pathway might have any relevance for fMLP-induced [ 3H]AA release. Membrane-bound PLD activity in HL-60 cells appears to be regulated by the small GTPbinding protein RhoA (40). RhoA can be inactivated by

FIG. 5. Phospholipid breakdown pathways likely involved in AA release in db-cAMP-differentiated HL-60 cells. fMLP-stimulated pathways responsible for AA release in HL-60 cells are indicated by continuous lines; pathways represented by broken lines are not involved in fMLP-stimulated AA release. Dotted lines indicate regulatory pathways. Crossbars show sites of inhibitor action. For further details, see text.

256

´ VENOD, AND SCHULZ STERNFELD, THE

ADP ribosylation through the ADP-ribose transferase activity of C. difficile toxin B (ToxB) (41). Since toxin B also affects other small GTP-binding proteins, such as Rac or Cdc42 (41), we chose a second approach to investigate the role of PLD in fMLP-mediated AA release. It is known that PLD preferentially utilizes primary alcohols as phosphate acceptors in lieu of water, which prevents the generation of PA from PC (42). The data shown in Table I demonstrate that incubation of db-cAMP-differentiated HL-60 cells with 10 ng/ml ToxB inhibit fMLP-stimulated [ 3H]AA release by 35% (from 5.3 ⫾ 0.4 to 3.5 ⫾ 0.2%; n ⫽ 3; P ⬍ 0.05). Similarly, in the presence of 0.3% 1-butanol (but not 2-butanol; data not shown), fMLP-induced [ 3H]AA release was reduced from 5.3 ⫾ 1.0 to 2.7 ⫾ 1.0% (n ⫽ 7; P ⬍ 0.01; Table I). Addition of an inhibitor of the more distal step in this pathway, i.e., PA-P (see Fig. 5), propranolol (0.2 mM) (43), reduced fMLP-stimulated [ 3H]AA release by about 75–90% (see Figs. 3A and 3B and Table I). Propranolol was tested with preincubation periods of 5–30 min and concentrations between 1 and 500 ␮M. We found 5 min and 200 ␮M to be maximally effective. The IC 50 ⫽ 84.4 ⫾ 11.1 ␮M for inhibition of fMLP-mediated AA release by propranolol is in accordance with its effect on PA-P (44). This suggests that activation of PLD is not the only pathway responsible for fMLP-induced production of AA. We therefore considered the possibility that fMLP might also activate PC-PLC to generate DAG, which could be salvaged as PA by the action of DAG kinase before being metabolized back to DAG and AA by PA-P and DAG/ MAG lipases, respectively (see Fig. 5). A significant contribution to AA production by PC-PLC activation is documented in Table I, showing more than 40% inhibition of fMLP-stimulated [ 3H]AA release with an inhibitor of PC-PLC, D609 (100 ␮M) (45) (from 5.3 ⫾ 1.3 to 3.1 ⫾ 1.1%; n ⫽ 6; P ⬍ 0.01). D609 was tested at concentrations of 100 and 250 ␮M and preincubation times between 15 min and 6 h. We found 100 ␮M D609 and 3 h preincubation to be maximally effective. When both, the PLD and the PC-PLC pathways were blocked by combining D609 (100 ␮M) and 1-butanol (0.3%), fMLP-stimulated [ 3H]AA release was reduced to a similar extent as by addition of 0.2 mM propranolol alone, namely, from 4.7 ⫾ 1.3 to 1.6 ⫾ 0.7% (n ⫽ 3; P ⬍ 0.05), suggesting that DAG produced in the PC-PLC pathway can be converted to PA by the action of a DAG kinase and metabolized back to DAG by a propranololinhibitable PA phosphohydrolase. DISCUSSION

In the present study we have investigated the pathways that are involved in the fMLP-stimulated arachidonic acid release. The data indicate that cPLA 2 and PIP 2-PLC activation are not required for generation of AA stimulated by fMLP. Preliminary studies using

inhibitors of various phospholipases suggest that fMLP-stimulated AA production could be mediated by ⬇45% by PLD activation and by ⬇35% through the PC-PLC pathway. Agonist-stimulated AA release from db-cAMP-differentiated HL-60 cells has been described by several investigators (8, 20, 21, 30). These studies showed that AA release could be induced by a rise in cytosolic free Ca 2⫹ concentration either by addition of a Ca 2⫹ ionophore or by permeabilization of cells (8, 30). This effect of Ca 2⫹ was drastically enhanced by activation of PKC with phorbol esters, such as PMA (8, 30), or by addition of fMLP (21). Phosphorylation of Ca 2⫹-dependent cytosolic PLA 2 with PMA could be directly demonstrated (8, 30). From studies with purified enzymes (11, 12) or MDCK cells (46), it has been deduced that phosphorylation at Ser-505 is mediated by MAP kinase and that the stimulatory effect of PKC must occur upstream of MAP kinase to promote phosphorylation of cPLA 2 (9). From these data and the observation that agonists, such as fMLP or ATP, activate PIP 2-PLC which leads to release of IP 3 and DAG and subsequent Ca 2⫹ release and PKC activation, it has been concluded that cPLA 2 is involved in agonist-induced AA release from HL-60 cells (9). To our knowledge, however, receptor-operated, agonist-induced translocation of cPLA 2 from the cytosol to the membrane fraction that would be responsible for AA release has not yet been demonstrated in db-cAMP-differentiated HL-60 cells. Our present results on Ca 2⫹- and PMA-stimulated AA release in db-cAMP-differentiated HL-60 cells (Fig. 2A) confirm the above-mentioned findings that PMAinduced activation of PKC in the presence of elevated cytosolic Ca 2⫹ concentrations can activate a pathway that leads to stimulation of cPLA 2 activity. However, this pathway does not seem to be activated by stimulation with fMLP. Although fMLP stimulated IP 3-induced Ca 2⫹ release (Figs. 1B and 1C), U 73122, an inhibitor of PIP 2-PLC, which suppressed fMLP-induced Ca 2⫹ release (Fig. 1C), had no effect on fMLPstimulated AA release (Figs. 3A and 3B and Table 1). Furthermore chelation of extra- and intracellular free [Ca 2⫹], which prevents fMLP-induced rise in cytosolic free [Ca 2⫹] (Fig. 1D), had no significant effect on fMLPinduced AA release (Fig. 3C). Inhibition of PKC activity with Ro 318220 abolished A23187 ⫹ PMA-induced AA release (Fig. 2A). U 73122 and Ro 318220 decreased fMLP-induced activation of p38 and p42/44 MAP kinases by about 50% (Fig. 4B). However, these compounds had no effect on the fMLP-induced AA release (Figs. 2A, 3A, and 3B). It is unlikely that fMLP-induced activation of cPLA 2 operates through the Ca 2⫹- and PKC-independent pathways of MAP kinase phosphorylation (e.g., tyrosine kinases) to release AA (47), since direct inhibition of cPLA 2 by AACOCF 3 did not block fMLP-induced AA release (Figs. 3A and 3B and Table 1). We therefore assume that fMLP-stimulated activa-

fMLP-INDUCED AA RELEASE IN HL-60 CELLS

tion of the PIP 2-PLC pathway that may lead to IP 3 and DAG production, increase in cytosolic free [Ca 2⫹], activation of MAP kinases, and increased cPLA 2 activity does not have any major role in fMLP-induced AA release from db-cAMP-differentiated HL-60 cells. Since our data suggest that Ca 2⫹-dependent cPLA 2 is not involved in fMLP-stimulated AA release from intact cells, we looked for other phospholipases A 2 that could lead to production of both AA and 2-lysophosphatidyl choline from phosphatidylcholine. Recently, a Ca 2⫹-independent PLA 2 (iPLA 2) has been shown to be involved in receptor-mediated AA release from smooth muscle cells (36). This enzyme could be completely inhibited by BEL or MAFP (36). As shown in Table I, fMLP-stimulated AA release was decreased by BEL or MAFP, maximally by about 25%. These results indicate that iPLA 2 plays a minor role in fMLP-stimulated AA release. The remaining AA, which is released in response to fMLP, should be produced by other pathways. It had been previously shown that fMLP stimulates PLD activity in endoplasmic and plasma membranes of db-cAMP-differentiated HL-60 cells (48). PLD activity in both membranes and cytosol of HL-60 cells depends on the presence of the small GTP-binding proteins ADP ribosylation factor (ARF) and RhoA (40). RhoA can be inactivated by ADP ribosylation through the ADP-ribose transferase of ToxB (41). Furthermore, primary alcohols prevent PLD-mediated generation of phosphatidic acid from phospholipids (42). If the PLD pathway plays a role in fMLP-induced AA release from HL-60 cells, this should occur by sequential hydrolysis of PC/PE to PA, to DAG, and finally to AA (see Fig. 5). Indeed, inhibition of the PLD-stimulated pathway by ToxB reduced fMLP-stimulated AA release by 35% (Table I). Similarly, with 1-butanol, fMLP-stimulated AA release was inhibited by 45% (Table I). PC-PLC activation also contributes somewhat to fMLP-induced AA production, as demonstrated by the effect of the PCPLC inhibitor D609, which reduced fMLP-mediated AA release by about 40%. Both D609 and 1-butanol decreased fMLP-induced AA release to an extent similar to that of the blocker of PA-P, propranolol (see Table I), which indicates that the bulk of AA produced by fMLP-stimulation of db-cAMP-differentiated HL-60 cells originates from two sources: DAG generated from PC-PLC activation and PA generated from PLD activation (see Fig. 5). We show that HL-60 cells contain a functionally expressed cPLA 2. Liu and Levy (49) report that this enzyme is necessary for proliferation of HL-60 cells. This raises the question of why cPLA 2 does not contribute to fMLP-mediated AA release. Hirabayashi et al. (50) demonstrated that the duration of intracellular Ca 2⫹ response is the critical factor for translocation and activation of cPLA 2. In our hands the fMLP-mediated rise of cytosolic Ca 2⫹ concentration reached a peak

257

value of ca. 400 nM during the initial phase (first 30 s) and a value of ca. 200 nM during the plateau phase lasting for minutes. With the Ca 2⫹ ionophore A23187, a higher cytosolic Ca 2⫹ concentration will be obtained that is also longer lasting. It is possible that long lasting Ca 2⫹ stimuli are necessary for proliferation via activation of cPLA 2 in HL-60 cells. However we show that cPLA 2 is not necessary for fMLP-mediated AA release. What could be the physiological significance of sustained AA production originating from PC-PLC and PLD activation? Similar to neutrophils, HL-60 cells are capable of metabolizing arachidonate to bioactive eicosanoids which may trigger acute inflammatory reactions in surrounding tissues and vasculature (1). But AA and/or its eicosanoid metabolites may also affect the function of the inflammatory cells in an autocrine manner and activate transcription factors (51, 52). This has been directly demonstrated in the U937 and J774 mononuclear and T-lymphoid Jurkat cell lines for the transcription factors AP-1 (51) and NF-␬B (52), respectively, which upregulate the expression of several genes critical for immunoinflammatory reactions, such as cytokines, cPLA 2, and mitogen-inducible cyclooxygenase II, and thereby enhance the inflammatory response of the cells. ACKNOWLEDGMENTS This work was supported by a grant from the Bundesministerium fu¨r Bildung, Wissenschaft, Forschung und Technologie (“Rheumaforschungszentrum Saar,” 01 VM 9310). We thank Professor Dr. Klaus Aktories and Dr. Ingo Just (Institute for Pharmacology and Toxicology, Albert-Ludwigs-University Freiburg, Germany) for kindly providing C. difficile toxin B; Dr. Wolfgang Nastainczyk (Department of Medical Biochemistry, University of Saarland, Homburg/Saar, Germany) for synthesizing the inhibitory peptide of sPLA 2, FLSYK; and Ms. B. Kohler for excellent technical assistance.

REFERENCES 1. Dennis, E. A. (1997) Trends. Biochem. Sci. 22, 1–2. 2. Balsinde, J., and Dennis, E. A. (1997) J. Biol. Chem. 272, 16,069 –16,072. 3. Clark, J. D., Lin, L. L., Kriz, R. W., Ramesha, C. S., Sultzman, L. A., Lin, A. Y., Milona, N., and Knopf, J. L. (1991) Cell 65, 1043–1051. 4. Sharp, J. D., White, D. L., Chiou, X. G., Goodson, T., Gamboa, G. C., McClure, D., Burgett, S., Hoskins, J., Skatrud, P. L., Sportsman, J. R., Becker, G. W., Kank, L. H., Roberts, E. F., and Kramer, R. M. (1991) J. Biol. Chem. 266, 14,850 –14,853. 5. Underwood, K. W., Song, C., Kriz, R. W., Chang, X. J., Knopf, J. L., and Lin, L. L. (1998) J. Biol. Chem. 273, 21,926 –21,932. 6. Exton, J. H. (1994) Biochim. Biophys. Acta 1212, 26 – 42. 7. Lin, L. L., Lin, A. Y., and Knopf, J. L. (1992) Proc. Natl. Acad. Sci. USA 89, 6147– 6151. 8. Xing, M., and Mattera, R. (1992) J. Biol. Chem. 267, 25,966 – 25,975. 9. Clark, J. D., Schievella, A. R., Nalefski, E. A., and Lin, L. L. (1995) J. Lipid. Mediat. Cell Signal. 12, 83–117.

258

´ VENOD, AND SCHULZ STERNFELD, THE

10. Channon, J. Y., and Leslie, C. C. (1990) J. Biol. Chem. 265, 5409 –5413. 11. Lin, L. L., Wartmann, M., Lin, A. Y., Knopf, J. L., Seth, A., and Davis, R. J. (1993) Cell 72, 269 –278. 12. Nemenoff, R. A., Winitz, S., Qian, N. X., Van Putten, V., Johnson, G. L., and Heasley, L. E. (1993) J. Biol. Chem. 268, 1960 – 1964. 13. Murakami, M., Kudo, I., and Inoue, K. (1993) J. Biol. Chem. 268, 839 – 844. 14. Lehman, J. J., Brown, K. A., Ramanadham, S., Turk, J., and Gross, R. W. (1993) J. Biol. Chem. 268, 20,713–20,716. 15. Du, X., Harris, S. J., Tetaz, T. J., Ginsberg, M. H., and Berndt, M. C. (1994) J. Biol. Chem. 269, 18,287–18,290. 16. Rhee, S. G., and Bae, Y. S. (1997) J. Biol. Chem. 272, 15,045– 15,048. 17. Exton, J. H. (1997) J. Biol. Chem. 272, 15,579 –15,582. 18. Chaplinski, T. J., and Niedel, J. E. (1982) J. Clin. Invest. 70, 953–964. 19. Pai, J. K., Siegel, M. I., Egan, R. W., and Billah, M. M. (1988) J. Biol. Chem. 263, 12,472–12,477. 20. Cockcroft, S., and Stutchfield, J. (1989) Biochem. J. 263, 715– 723. 21. Xing, M., The´venod, F., and Mattera, R. (1992) J. Biol. Chem. 267, 6602– 6610. 22. Heung, Y. M., and Postle, A. D. (1995) FEBS Lett. 364, 250 –254. 23. Bligh, E. G., and Deyer, W. J. (1959) Can. J. Biochem. Physiol. 37, 911–917. 24. Dole, V. P., and Meinertz, H. (1960) J. Biol. Chem. 235, 2595– 2599. 25. Grynkiewicz, G., Poenie, M., and Trien, R. Y. (1985) J. Biol. Chem. 260, 3440 –3450. 26. Cockcroft, S., and Stutchfield, J. (1989) FEBS Lett. 245, 25–29. 27. Willems, P. H., Van de Put, F. H., Engbersen, R., Bosch, R. R., Van Hoof, H. J., and de Pont, J. J. (1994) Pflugers Arch. 427, 233–243. 28. Mogami, H., Lloyd Mills, C., and Gallacher, D. V. (1997) Biochem. J. 324, 645– 651. 29. Hofer, A. M., Landolfi, B., Debellis, L., Pozzan, T., and Curci, S. (1998) Embo J. 17, 1986 –1995. 30. Xing, M., Wilkins, P. L., McConnell, B. K., and Mattera, R. (1994) J. Biol. Chem. 269, 3117–3124. 31. Riendeau, D., Guay, J., Weech, P. K., Laliberte, F., Yergey, J., Li, C., Desmarais, S., Perrier, H., Liu, S., Nicoll Griffith, D., and Street, I. P. (1994) J. Biol. Chem. 269, 15,619 –15,624.

32. Bartoli, F., Lin, H. K., Ghomashchi, F., Gelb, M. H., Jain, M. K., and Apitz Castro, R. (1994) J. Biol. Chem. 269, 15,625–15,630. 33. Ackermann, E. J., Conde Frieboes, K., and Dennis, E. A. (1995) J. Biol. Chem. 270, 445– 450. 34. Ma, Z., Ramanadham, S., Hu, Z., and Turk, J. (1998) Biochim. Biophys. Acta 1391, 384 – 400. 35. Tseng, A., Inglis, A. S., and Scott, K. F. (1996) J. Biol. Chem. 271, 23,992–23,998. 36. Wolf, M. J., Wang, J., Turk, J., and Gross, R. W. (1997) J. Biol. Chem. 272, 1522–1526. 37. Hazen, S. L., Zupan, L. A., Weiss, R. H., Getman, D. P., and Gross, R. W. (1991) J. Biol. Chem. 266, 7227–7232. 38. Balsinde, J., and Dennis, E. A. (1996) J. Biol. Chem. 271, 6758 – 6765. 39. Balsinde, J., Balboa, M. A., Insel, P. A., and Dennis, E. A. (1999) Annu. Rev. Pharmacol. Toxicol. 39, 175–189. 40. Siddiqi, A. R., Smith, J. L., Ross, A. H., Qiu, R. G., Symons, M., and Exton, J. H. (1995) J. Biol. Chem. 270, 8466 – 8473. 41. Ohguchi, K., Banno, Y., Nakashima, S., Kato, N., Watanabe, K., Lyerly, D. M., and Nozawa, Y. (1996) Infect. Immun. 64, 4433– 4437. 42. Kobayashi, M., and Kanfer, J. N. (1987) J. Neurochem. 48, 1597– 1603. 43. Eichberg, J., Gates, J., and Hauser, G. (1979) Biochim. Biophys. Acta 573, 90 –106. 44. Jamal, Z., Martin, A., Gomez Munoz, A., and Brindley, D. N. (1991) J. Biol. Chem. 266, 2988 –2996. 45. Mu¨ller Decker, K. (1989) Biochem. Biophys. Res. Commun. 162, 198 –205. 46. Xing, M., and Insel, P. A. (1996) J. Clin. Invest. 97, 1302–1310. 47. Durstin, M., Durstin, S., Molski, T. F., Becker, E. L., and Sha’afi, R. I. (1994) Proc. Natl. Acad. Sci. USA 91, 3142–3146. 48. Whatmore, J., Morgan, C. P., Cunningham, E., Collison, K. S., Willison, K. R., and Cockcroft, S. (1996) Biochem. J. 320, 785– 794. 49. Liu, Y., and Levy, R. (1997) Biochim. Biophys. Acta 1355, 270 – 280. 50. Hirabayashi, T., Kume, K., Hirose, K., Yokomizo, T., Iino, M., Itoh, H., and Shimizu, T. (1999) J. Biol. Chem. 274, 5163–5169. 51. Mollinedo, F., Gajate, C., and Flores, I. (1994) J. Immunol. 153, 2457–2469. 52. Camandola, S., Leonarduzzi, G., Musso, T., Varesio, L., Carini, R., Scavazza, A., Chiarpotto, E., Baeuerle, P. A., and Poli, G. (1996) Biochem. Biophys. Res. Commun. 229, 643– 647.