Regulation of phospholipase A2 activity in undifferentiated and ...

Vol. 269, No. 4, Issue of January 28, pp. 3117-3124,

T m JOURNAL OF BIOUXICAL CHE~~STRK

1994

Printed in USA.

0 1994 by The Amefican Society for Biochemistry and Molecular Biology, h e .

Regulation of Phospholipase A2 Activity in Undifferentiated and Neutrophil-like HL60 Cells LINKAGE BETWEEN IMPAIRED RESPONSES TO AGONISTS AND ABSENCE OF PROTEIN KINASE C-DEPENDENT PHOSPHORYLATION OF CYTOSOLIC PHOSPHOLIPASE Az* (Received for publication, April 26,

1993, and in revised form, October 18, 1993)

Mingzhao Xing$, Pamela L. Wilkins, Bradley K. McConnell, and Rafael Matteral From the Department of Physiology and Biophysics, Case Western Reserve University School of Medicine, Cleveland, Ohio 44106

We compared the regulation of cytosolic phospholi- sors of eicosanoids and platelet activatingfactor, respectively pase A2 (cPI&) activityin undifferentiated and neutro- (for review, see Refs. 1and 2). A cytosolic type of PLA2 (cPLA2, phil-like HL60 cells. AlthoughCas+-mobilizing P2-purin- as opposed to the secretory types, sPLA2s, see Ref. 3) has been ergic receptors are expressed in bothcelltypes, recently identified and purified from different cell types (4-9) arachidonic acid(AA)release stimulatedby P2-puriner- and cloned from human myelomonocytic U937 cells (10, 11). gic agonists was S-7-fold higher in the differentiated Compared with thesPLA2s, c P U 2 is characterized by its high cells. Similarly,the stimulation of AA release by AlF'; in molecular mass (85 versus 14 kDa for SPLAZS),low Ca2+ reintact cells or byATP and guanosine Sr-3-0-(thio)tri- quirement (micromolar versus millimolar for sPLA~s),high sephosphate (GTP+) in electropermeabilized cells was lectivity for arachidonic acid at sn-2 position on the phosphosignificantly higher in the differentiated cells. Treat- lipid substrate (versus no fatty acid selectivity at sn-2 position ment with phorbol 12-myristate 13-acetate (PMA) en- for sPLA2s), and insensitivity to disulfide-reducing agents (verhanced A23187-stimulated AA release in intact HL60 sus high sensitivity of sPLA2s to these agents). Recent evidence granulocytes with minimal effects in the undifferentisuggests thatcPLA2 is at least one of the enzymes responsible ated cells. Immunoblotting experiments showed similar levels ofcPLA2 and of agonist-mediated activation of for agonist-induced arachidonic acid release (12). The molecular mechanism underlying the regulation of remitogen-activated protein kinase in both cell types.Experiments measuring stimulation ofAA release by either ceptor-coupled PLA2 has received considerable attention. Permelittin,usingendogenouslylabeled intact cells, or tussis toxin-sensitive G proteins have been proposed to particiof PLA2, although receptor-mediated activation Cas+, using homogenatesand exogenous substrate, indi- patein cated that undifferentiated cells do not lack an activat- whether a G protein is directly coupled to the enzyme has not able P&. The stimulatoryeffects of GTPyS and Ca2+on been conclusively established (for review, see Refs. 13-15). ConAA release in homogenates from endogenously labeled siderable evidence is also available indicatingthe involvement G pro- of PKC in the regulation of PIA2 (e.g. Refs. 16-20). In this cells suggestedthat undifferentiated cells display tein-cPLA2 coupling. Basal and PMA-stimulated phos- context, we have demonstrated that cPLA2 can be phosphophorylation of CPLAa was detectedin differentiated, but rylated in a PKC-dependent manner in HL60 granulocytes not in undifferentiatedcells. However,the two cell types (21). This resultis consistent with theobservation that activadisplayed onlysubtle differences in the time coursesof tion of PKC enhanced receptor-stimulated phosphorylation of phosphorylation of mitogen-activated protein kinase cPLA2 overexpressed in Chinese hamster ovary cells (12). Retriggered by agonists and PMA. The observed defectin cently, two groups observed that in vitro phosphorylation of c P U 2 phosphorylation may represent the alteration cPLA2 by MAP kinase is accompanied by increased enzyme preventing agonist-mediated stimulationof AA release activity(22, 23). Since MAP kinase activation can occur in undifferentiated HL60 cells. through both PKC-dependent and PKC-independent mechanisms (24, 251, it has been proposed that at least one of the important cellular pathways in the stimulation ofCPLA,by PLA2' is the crucial enzymeinvolved in the production of AA receptors is the sequential activation of PLC (increasing the and alkylether-containing lysophosphatidylcholine,the precur- levels of diacylglycerol and inositol triphosphate/cytosolic phosphorylaCa2+),PKC and MAP kinase, leading to the final * This work was supported in part by American Heart Association- tion and activation of cPLA2 (21-23). Northeast Ohio Affiliate Grant 4802, CRC/American Cancer Society The human promyelocytic leukemic HL60 cell line can be Grant IRG-186, and National Institutes of Health Grant GM 46552OlAl (to R. M.). The costsof publication of this article were defrayedin induced to differentiate into granulocytes or monocytes by varipart by the payment of page charges. This article must therefore be ous agents (26-28). Several groups observed that expression of hereby marked "advertisement"in accordance with 18 U.S.C. Section PIA2 activity in HL60 cells is differentiation-dependent (291734 solely to indicate this fact. caused by a defective j In partial fulfillment of the research requirements corresponding to 32). Billah et al. (30) proposed that this is regulatorymechanism in the undifferentiated HL60 cells, the departmental Ph.D. program. fiTo whom correspondence should be addressed. "el.: 216-368-3180; which becomes developed upon cell differentiation. However, Fax: 216-368-5586. the natureof this alteration remainsundefined. In the present The abbreviations used are: PIA,, phospholipase A*; CPLA,, cytosolic PLA,; PLC, phospholipase C; PLD, phospholipase D; AA, arachi- report we have comparatively studied the regulation of cPLAz donic acid; HBSS, Hank's balanced salt solution; EPB, electroperme3'5'abilization buffer; Bt,cAMP, dibutyryl cyclic adenosine monophosphate; PMA, phorbol 12-myristate 13-acetate; PKC, protein teins, heterotrimeric guanine nucleotide-binding regulatory proteins; kinase C; MAP kinase, mitogen-activated protein kinase (or microtu- GTPyS, guanosine5'-0-(3-thiotriphosphate);PAGE, polyacrylamidegel bule-associated protein 2 kinase); BSA, bovine serum albumin; G pro- electrophoresis.

3117

3118

Phosphorylation of PLA2 during Granulocyte Differentiation

activity by P2 purinergic receptors in undifferentiated and differentiated HL60 cells. Our results indicate that thecoupling of P2 purinergic receptors to activation of cPLA2 in undifferentiated HL60 cells is defective and although thereceptor, cPLA2 and their cognate G protein seem normally expressed in both cell types, the PKC-dependent phosphorylation of cPLA2 is observed only in differentiated HL60 cells.

by volume) containing 200 pg/ml unlabeled AA. Samples were immediately vortexed, left on icefor 10 min, and centrifuged for 4 min a t 14,000 x g. Parallel experiments showed that the supernatants obtained following this procedure contained approximately 80%of the AA present in the samples and that a similar fraction of the phospholipid substrate remained in thepellets. The supernatants (50-pl aliquots) were spotted onto TLC plates and subjected to ascending chromatography using as solvent the upper phase of ethyl acetate:isooctane:water:acetic acid (55:75:100:10, by volume). Plates were developed with iodine and the areas corresponding to the AA were scraped and counted. EXPERIMENTALPROCEDURES Immunoprecipitation and Western Blotting of cPZ&"mmunoMaterials-Protein A-Sepharose, ATP, mepacrine, PMA, arachidonic precipitation of cPLA, from HL60 cell extracts and Western blotting acid, Tween20, Nonidet P-40, sodium orthovanadate, antimycin A, were performed as previously described (21). 2-deoxy-~-glucose,phenylmethylsulfonyl fluoride, diisopropyl fluoroWestern Blotting of MAP Kinase:Mobility Shift Assays-Undifphosphate, BSA, dried non-fat bovine milk, and theinsulin-transferrinferentiated or neutrophil-like HL60 cells were suspended at 2.5 x lo6 sodium selenite supplement for culture media were obtained from celldml in HBSS supplemented with 1 m g / d glucose and 0.5 mg/ml Sigma. GTPyS, UTP, Bt2cAMP, pepstatin A, and leupeptin were purBSA. Cell suspensions (5.5 ml) were preincubated at 37 "C for 1 min chased from Boehringer Mannheim. Synthetic melittin was from Pe- followed by addition of the indicated concentrations of nucleotides or ninsula Laboratories, Inc. [5,6,8,9,11,12,14,15-3H~achidonic acid and PMAand further incubation at 37 "C for varying times. One-ml aliquots lZ6I-goatanti-rabbit IgG antiserum were obtained from DuPont-New were taken a t the indicated times (1, 3, 6, and 10 min) transferred to England Nuclear. ty-32PlATPwas purchased from ICN Radiochemicals. microcentrifuge tubes and precipitated by addition of 110 pl of ice-cold Iscove's modified Dulbecco's medium was from Life Technologies Inc. 100%trichloroacetic acid. Tubes were allowed to stand on ice for10 min Bovine calf serum was from HyClone Laboratories. Staphylococcus au- followed by centrifugation for 1 min at 10,000 x g and 4 "C. Supernareus protein A-coated cells (Pansorbin Cells, standardized) and normal tants were aspirated and the pellets washed by addition of l ml of rabbit serum were purchased from Calbiochem. Thin layersilica plates ice-cold ethylic ether and centrifugation for 1 min at 10,000 x g and (LK5D) were from Whatman. The rabbit antiserum against recombi- 4 "C. The supernatants were removed by aspiration and the pellets nant human U937 cell cPLA2was kindly provided by Drs. L L . Lin and dried in a chemical hood. The dried pellets were resuspended in 200 p1 J. L. Knopf (Genetics Institute). of Laemmli sample buffer (36), boiled for 5 min, and centrifuged to Culture and Differentiation ofHL60 Cells-HL6O cells were cultured remove insoluble material. Aliquots of the samples (20 pl, representing and, when required, differentiated by BgcAMP into neutrophil-like approximately 2.5 x los cells) were subjected to SDS-PAGE (12%acrylcells, as described previously (33). amide, 0.16% bis-acrylamide, 30&gel). The electrophoresis was Labeling of Cellular Phospholipids with PHlAA and Assay for AA ended 1h after the tracking dye (bromphenol blue) had left the gels (to increase the separation between native and phosphorylated MAP kiRelease-HL6O cells were labeled with PHIAA for 90 min, as reported previously (33). The release of AA in intact cells was measured a s nase). Samples were transferred to nitrocellulose (pore size = 0.45 pm; described in Ref. 33. Cell suspensions were incubated for 8 min at 37 "C transference buffer was 95 m glycine, 12.5 n" Tris, and 10% methain the presence of 0.2 m exogenous AA (to prevent reacylation of nol) during 12-15 h at 30 V. Immunoblotting was carried out at 20 "C [3H]AA, Ref. 33). The radioactivity present in the extracellular media using a buffer consisting of 10 m TridHCl, pH 7.4, 0.05% Tween 20, following incubation was a valid measure of the release of PHIAA. This 0.15 M NaCl, and 1m EDTA (blotting buffer). Blocking was carried out was established by lipid extraction of the supernatants according to for 60 min using 3% dried milk in blotting buffer. Incubations with primary (Y2,OOO dilution of rabbit anti-rat ERK2; aIICp42, Ref. 37) and Bligh and Dyer (34) and ascending chromatography on silica thin layer secondary (Y2,OOO dilution of horseradish peroxidase-coupled donkey plates using as solvent the upper phase of the mixture ethyl acetate: isooctane:water:acetic acid, 45:25:50:10 (which resolves AA from differ- anti-rabbit IgG) antisera solutions (prepared using blocking buffer as ent oxidation metabolites, Ref. 35). Under these assay conditions (0.2 diluent) were also carried out for 60 min. The blots were washed with m unlabeled exogenous AA and 0.3% BSA), there was no significant blotting buffer previous to the blocking step and between the blocking conversion of [3H]AA into oxidation products; analysis of supernatants step and theincubations with primary and secondary antisera (1wash from stimulated cells showed that 90% of the radioactivity moving from of 15 min, followed by two washes of 5 min each). Detection was carried the origin co-migrated with the AA standard (visualized with iodine). out using an enhanced chemiluminescence kit (Amersham), following the instructions indicated by the supplier. Similar differences in the release oft3H1AAby undifferentiated and Labeling of Electropermeabilized HL60 Cells with [y-3zPlATPneutrophil-like HL60 cells (see "Results") were found by either directly counting the supernatants or following lipid extraction and thin layer Bt,cAMP-differentiated or undifferentiated HL60 cells were washed chromatography as described above (not shown). Assays using elec- twice with glucose-free Hank's balanced salt solution (HBSS), suspended in HBSS containing 5 p~ antimycin A and 6 m 2-deoxy-~tropermeabilized cells and homogenates derived from cells prelabeled with t3H1AAwere carried out a s described in Ref. 21 (incubations were glucose (1-1.5 x lo7 celldml) and incubated for 20 min at 37 "C, to deplete the cell of endogenous ATP (21). Cells were then washed twice carried out at 37 "C for 20 min in both assays). with ice-cold EPB (see previous paragraph) supplemented with 3 m g / d Release of AA Using Homogenates and Exogenous SubstratesUndifferentiated or neutrophil-like HL60 cells were washed and resus- BSA, 2 p~ pepstatin, 200 ~IMphenylmethylsulfonyl fluoride, 2 p g / d leupeptin, 2 m diisopropyl fluorophosphate, 10 m NaF, and 4 m pended (at lo7 celldml) in EPB buffer (20 m HEPES, pH 7.4, 120 potassium glutamate, 20 m potassium acetate, 3 m NaCl, 1 m Na3V04 (EPB plus protease and phosphatase inhibitors: "EPB+PPI"). The cells were exposed to a higher concentration of protease inhibitor by EGTA, 11pg/ml phenol red) supplemented with 3 m MgCl,, 1 m g / d glucose, 3 mg/ml BSA, 2 p~ pepstatin, 200 p~ phenylmethylsulfonyl resuspension (1-1.5 x lo7 celldml) in EPB+PPI containing 6 m diisofluoride, and 2 pg/ml leupeptin (EPB + protease inhibitors: EPB+PI). propyl fluorophosphate and incubation for 30 min at 4 "C. The cells The suspensions were further supplemented with 4 m diisopropyl were subsequently pelleted, resuspended in EPB+PPI, and electroperfluorophosphate, and incubated for 30 min at 4 "C. The cells were pel- meabilized as described before (21). Following electropermeabilization, leted, resuspended in EPB+PI (3-5 x lo7 celldml), and subjected to aliquots representing approximately 2-2.5 x lo7 cells were diluted into 1ml of EPB+PPI supplemented with 0.92 m CaC12 (resulting in apnitrogen cavitation at 500 p.s.i. for 20 min on ice. The cavitates were free Ca2+),3 m MgCl,, and 60-120 pCi of [y-32PlATP centrifuged for 8 min at 700 x g and 4 "C; the low speed supernatants proximately 1~IM thus obtained were kept on ice until the start of the assay. Enzyme (specific activity = 4,500 Ci/mmol, final concentration = 13.2-26.4 m), and left on ice for 20min. The cell mixtures were then preincubated at assays were carried outfollowing the guidelines described by Bonventre et al. (19). Reaction mixtures (prepared in microcentrifuge tubes) con- 37 "C for 2 min, followed by addition of100-150 m PMA or vehicle and furtherincubation for 8 min at 37 "C.Under these tained 0.7 nmol (approximately 50 nCi) of l-stear0yl,2-[l-~~C]arachi-(0.002% Me2SO), conditions (presence of phosphatase inhibitors) approximately 50% of donoyl phosphatidylcholine (previously evaporated and resuspended in remained as such following the 10-min incubation 3 pl of Me2SO), 42 pl of the low speed supernatants (representing the added [Y-~~PIATP 106200 pg of protein; supplemented with 0.3 m unlabeled arachi- (results not shown). Reactions were stopped by 25-fold dilution with donic acidand adjusted to pH 7.4), and 25 ploftest substances dissolved ice-cold 50 m Tris-HC1, pH 7.5, 100 m KCl, 5 m EGTA, and 5 m~ EDTA, supplemented with protease and phosphatase inhibitors, and in EPB+PI adjusted to pH 10.0 (final volume was 70 p1; the resulting pH of the mixtures was 9.0, the value at which the highest enzyme activi- pelleted by centrifugation (15 min a t 4 "C and 1,400 x g). The pellets ties were measured). Incubations were carried out at 37 "C for 30 min were extracted by resuspension in 400 pl of ice-cold extraction buffer and stopped by addition of an equal volume of ethano1:acetic acid (49:l consisting of 50 m Tris-HC1, pH 7.5, 1% Nonidet P-40, 150 l l l ~NaCl,

3119

Phosphorylation of PLA2 during Granulocyte Differentiation 1 m~ EGTA, and 1m~ EDTA, supplemented with protease and phosphatase inhibitors as for EPB+PPI (see above), followed by incubation on ice for 40 min. The samples were then centrifuged for 10 min at 16,000 x g and 4 "C and the supernatants saved as T-labeled cell extracts for the immunoprecipitation protocols. Immunoprecipitation,Electrophoresis, and Autoradiography of Phosphorylated cPZ&"mmunoprecipitation of 32P-labeledcPLA2 was carried out using a modification of a previously reported protocol (21). The first step consisted in removing (preclearing) unrelated phosphoproteins that could interact with either irrelevant IgGs present in the rabbit anti-cPLA2antiserum or the protein A-coated beads.To this end, aliquots of 350 pl of 32P-labeledcell extracts were incubated with 7 pl of normal rabbit serum for 1.5 h at 4 "C, followed by addition of 8 mg of protein A-Sepharose (precoated with 3% milk for 2 h and washed by centrifugation) and further incubation for 1.5 h at 4 "C. The samples were then centrifuged at 16,000 x g and 4 "C for 10 min. The supernatants were transferred to new microcentrifuge tubes and subjected to two additional rounds of centrifugation to completely remove the protein A-Sepharose beads. Thesupernatants ("precleared extracts") were subdivided in two equal aliquots and incubated with either normal rabbit serum or rabbit anti-cPLA, antiserum (1pl of s e d 1 0 0 111 of extracts) for 1.5 h at 4 "C. The antigen-antibody complexes thus formed were precipitated by incubation at 4 "C for 1.5 h with 4 mgof milkcoated protein A-Sepharose and subsequent centrifugation at 16,000 x g and 4 "C for 10 min. The pellets, containing the immune complexes bound to protein A, were washed four times with 50 m~ Tris-HC1, pH 7.5,350 m~ NaC1,5 m~ EDTA, and 0.5% Nonidet P-40, supplemented with protease and phosphatase inhibitors, and twice using this solution without NonidetP-40. The washed pellets were resuspended in Laemmli sample buffer (36), boiledfor10 min, and centrifuged at 16,000 x g for 10 min a t room temperature. The supernatants were subsequently subjected to SDS-PAGE (10% acrylamide).At the end of electrophoresis the gels weredried and exposed at -70 "C using Kodak X-AR films and intensifying screens. This protocol significantly reduced, sometimes completelyeliminating, the amount of the nonspecific 32P-labeledproteins present in the autoradiograms. In the autoradiograms showing carry-over of nonspecific phosphoproteins (see Fig. @I, an additional band of cPLA2was detected in the lanes corresponding to the pellets obtained after immunoprecipitation with the anti-cPLA2 antiserum (compared with those obtained with normal rabbit serum). Protein A-Sepharose was preferred in these studies over Pansorbin; experiments with the latter resulted sometimes in the precipitation of nonspecific phosphoproteinsof apparent molecular mass similar to that of cPLA2. Presentation of Datu-The release ofL3H1AA was expressed as the percentage of the total radioactivity incorporated into intact or electropermeabilized cells or cell homogenates, after subtraction of basal (defined as the release of 13HIAA measured a t 4 "C). Data are representative of results obtained in at least three experiments.

RESULTS Defective Coupling of P,-Purinergic Receptors to AA Release in Undifferentiated HL60 Cells-Both undifferentiated and neutrophil-like HL60 cells express a Pz-purinergicreceptor coupled to PLC activation and Ca2+ mobilization (21, 33, 3841). The Pa-purinergicreceptor is also coupled to theactivation of PLAz (21,33,38) andPLD (42) in differentiated HL60 cells. The similar EC50 values measured in the concentration-response curves corresponding to the above mentioned effects of ATP and UTP suggest that the same subtype of Pz-purinergic receptor is expressed in both differentiated and undifferentiated HL60 cells. Since previous studies using agents thatbypass receptor occupancy (mainly calcium ionophore A23187) revealed no activation of AA release in undifferentiated HL60 cells (29-32), we investigated whether coupling between Pzpurinergic receptors and cPLAz was also altered in thesecells. As shown in Fig. 1, the release of AA in differentiated HL60 cells was increased by a 5-7 factor upon stimulation by 10-20 ATP or UTP. This was followed bya partial decline at higher nucleotide concentrations due to the proposed interaction with a n inhibitory subtype of Pz-purinergic receptor (33). On the contrary, under the same conditions, the stimulation of AA release by ATP or UTP representedonly a duplication (or less) of the restingactivity measured inundifferentiated cells. Simi-

O t ATP

-8

+e

-7

-6

ACUTP+

A

-5

-3

-4

-2

Nucleotides (log M) FIG.1. Effects of ATP and UTP on AA release in intactundifferentiatedanddifferentiated HL60 cells. L3H1AA-labeled cells were assayed for AA release in HBSS containing 2 m~ CaC12, 1 m~ MgCI,, and increasing concentrations ofATP (circles) or UTP (triangles). Incubations were camed out for 8 min at 37"C.Open and closed symbols represent undifferentiated and Bt&MP-differentiated HL60 cells, respectively. The calculated free Ca2+concentration ranged from 2 m~ (in the absence of nucleotide) to 1.7 m~ (in the presence of 0.67 m~ nucleotide, the highest concentration shown in the figure). Under this condition the chelation of Ca2+by nucleotides cannot account for the inhibitory phase of the concentration-responsecurve (Ref. 33). The data in the figure are presented as thepercentage of the total 3H radioactivity incorporated into the cell, after subtraction of the basal value (defined as the release of [3HIAAmeasuredat 4 "C). Thevalues of total incorporated SH activity added per tube and the basal release of PHIAA were 84,610 and 385 cpm for undifferentiated, and 71,759 and 950 cpm for differentiated cells, respectively. For details, see "Experimental Procedures."

lar differences in agonist-mediated release of r3H1AA bythe two cell types were found when the lipidic extracts of the supernatants were subjected to chromatography in a system that allows separation of this compound from its oxidation metabolites. Differences inthe radioactivity of thesupernatants obtained under our assay conditions (200 p~ exogenous unlabeled AA and 0.3% BSA) cannot be attributed to modifications in the rate of metabolism of f3H]AA. Likewise, differences in the rateof reacylation of the two cell types cannot account for the significant differences in the releaseof AA. The addedAA (40 nmol, reaction volume was 200 pl) represents approximately a 2-3:l molar excess over the total massof phospholipids present in the cells (15 and 10 nmoV106 undifferentiated and neutrophil-like HL60 cells, respectively, Refs. 30 and 31). This excess becomes 40:l when the mass of mobilized AA (and available reacylation sites) is estimated from the fraction of incorporated L3H]AA that is released following agonist stimulation (10% of 10 nmol/106 cells). The addedAA is probably in a n even larger excess over that released from the cells (approximately 400:l) considering that "newly" incorporated AA (e.g. during pulse labeling) is a n order of magnitude more "releasable" than that present in the endogenous cellularpool (10 and 1%, respectively, Ref. 43). We can therefore estimate a release of approximately 0.1 nmoY1O' cells, in agreement with that measured in stimulated human neutrophils (44). Consistent with these calculations, we found no differences in thelevels of unlabeled AA le& in thecell supernatants following cell incuat 37 "C or at 4 "C (not shown). bation with or without agonists The release of AA in differentiated HL60 cells appears to represent CPLAZactivity (21). Thereforethe results shown in Fig. 1 could indicate either the lack of expression of cPI& or a defect in the coupling of the Pz-purinergic receptor to this enzyme in undifferentiated HL60 cells. Comparative studies us-

3120

Phosphorylation of PLA, during Granulocyte Differentiation

TABLE I Release of P4CIAAfiom l-stearqyl,2-[1-14Clarachidonoyl phosphatidylcholine in homogenates fiom undifferentiated and neutrophil-like HL60 cells

in undifferentiated HL60 cells reflects an alteration in receptor-effector coupling, as opposed to the lack of expression of the relevant effectors. G Protein-dependent Regulation of cPLA2 Activity in UndifActivities measured in the low speed supernatants were expressed as dpm of P4C1AA released per mg of protein. Data shown represent the ferentiated a n d Differentiated HL60 Cells-As a first step in average f S.D. of triplicate determinations.For details see “Experimen- elucidating the mechanism underlying the defective coupling tal Procedures.” between membrane receptors and cPLA2 in undifferentiated HL60 cells, we examined the effects of G protein activators on Neutrophil-like Undifferentiated AA release. In experiments using intact cells we observed that Basal 1,136f 62 1,698 f 90 addition of AlF, significantly stimulated AA release in HL60 Control 2,035f 627 1,620 f 127 2 ~ free l l Ca2+ ~ 23,195f 533 16,664 f 2,713 granulocytes, but not in undifferentiated HL60 cells (Fig. 3A). Similarly, addition of GTPyS resulted inATP-dependent stimulation of AA release in electropermeabilized HL60 granuloing both differentiated and undifferentiated HL60 cells were cytes, as we reported before (21),with minor effects in undifthen designed to further investigate the molecular mechanisms ferentiated HL60 cells (Fig. 3B).Also, ATPper se did not show responsible for the above mentioned alteration inAA release. significant effects on AA release inpermeabilized undifferentiA Functional cPLAz Is Expressed in Undifferentiated HL60 ated HL60 cells, in contrast with the observations using differCells-We measured the effects of melittin, a membrane-per- entiated cells (Fig. 3B ). In 12 experiments performed with electurbing peptide that appearsto activate some forms of PLAz by tropermeabilized HL60 granulocytes, in the presence of 1 facilitating enzyme-phospholipid interactions (45-47),on the free Ca2+and 3 IMI total M$+, addition of 1 mM ATP resulted release of AA in intact undifferentiated and neutrophil-like in an 104 * 38% increase in AA release over control, while the cells. Synthetic melittin induced comparable increases on AA combination of ATP and GTPyS induced a maximal increase of release in undifferentiated andneutrophil-like HMO cells (not 244 2 48% over control. In contrast, in five experiments conshown); the effects of melittin were blocked by mepacrine, a ducted with undifferentiated HL60 cells under the same concompound that has been used as PLA2 inhibitor in numerous ditions, 1 IMI ATP only elicited a 21 2 23% increase in AA release and the maximal stimulation by the combination ofATP studies (48, 49). Although Ca2+ionophore A23187 failed to induce AA release and GTPyS was 77 35% over control. We have previously in intact undifferentiated HL60 cells (29-32),total sonicates proposed that cPLAz in differentiated HL60 cells is normally from these cells showed Ca2+-dependent activities comparable under an inhibitory constraint, and that the release of this to those measured in similar preparationsfrom HL60 granu- constraint by PKC-dependent phosphorylation of cPLA2 is a locytes derived by MezSO treatment (30).This was interpreted prerequisite for the maximal activation of cPLAz byG proteins as evidence for the presence of PIAz, but thelack of its regu- and Ca2+ (21).This proposal was based on experimental evilatory mechanism, in undifferentiated HL60 cells (30).During dence including the ATP requirement for the effect of GTP$ in electropermeabilized HL60 granulocytes (21).Under this hythe course of this study we also measured the ability of low speed (700 x g) supernatants preparedfollowing nitrogen cavi- pothesis, the altered ability of AlFa or GTP@ to stimulate AA tation of undifferentiated or Bt2-cAMP-differentiated (neutro- release in intactor electroporated undifferentiated HL60 cells, phil-like) cells to release AA from a n exogenous substrate (1- respectively (Fig. 3), could be interpreted as resulting from stearoyl,2-[l-14C]arachidonoyl phosphatidylcholine). The either theinsufficient expressionof the cognate G proteinor a n results obtained showed that thespecific activity present in the alteration in themechanisms responsible for the phosphorylahomogenates derived from the undifferentiatedcells was infact tion of cPLA2 in undifferentiatedcells. To distinguish between higher than that in the corresponding fraction from the HL60 these two possibilities, we studied the effect of GTPyS on AA granulocytes (Table I). This result confirmed the claim by Bil- release incell homogenates. This activity is ATP-independent, lah et al. (30)that undifferentiated cells do not lack a Ca2+- as opposed to that in electropermeabilized cells, suggesting activatable PLAz activity (similar specific activities weremeas- that the inhibitory constraint that prevents the activation of ured by these authors in totalsonicates from undifferentiated cPLAz by G proteins and Ca2+ in HL60 granulocytes can be and Me2SO-induced cells). Regardless of the mechanism trig- released by either mechanical disruption of the cell or phosgering the recovery of PLAz activity upon homogenization of phorylation (21).As shown in Fig. 4, both Ca2+ and GTPyS undifferentiated HL60 cells, these data do not completely rule stimulate AA release in a n ATP-independent manner in caviout the absence of receptor-coupled cPLAz in these cells. We tates prepared from differentiated or undifferentiated HL60 therefore compared the levels of cPLAz in undifferentiated and cells. The observation of the effects of GTPyS in cell cavitates neutrophil-like HL60 cells by immunoprecipitation of cell ex- extends the observation of Ca2+-dependent stimulation of AA tracts with a rabbit anti-cPLA2 antiserum, followed by SDS- release in cell homogenates from undifferentiated HMO cells PAGE, transference to nitrocellulose and immunoblotting (the (30).These results indicate a comparable G protein-effector above mentioned anti-cPLA, and 1251-labeledgoat anti-rabbit coupling in differentiated and undifferentiated HL60 cells and IgG were the primary and secondary antisera, respectively). suggest that thephosphorylation of cPLAz, a proposed prerequisite for the activation of cPLAz by G proteins in intactHL60 The results included in Fig. 2 showed that: ( a ) the system provided a n increasing signal when increasing concentrations granulocytes (21),may be deficient in undifferentiated HL60 of anti-cPLA2 antiserum and a constant concentration of re- cells. The Phosphorylation of cPLA2 a n d the Effects of PMA on AA combinant cPLA2 standard were added to the extraction mixtures, and that the immunoprecipitation only occurred in the Release Are Altered in Undifferentiated HL60 Cells-We have ( b )cPLA2 present observed the phosphorylation of cPLAz in electropermeabilized presence of the specific antiserum (panel C); in cell extracts from HL60 granulocytes can be specifically de- HL60 granulocytes subjected to PMA treatment and proposed tected when theanti-cPLAz antiserum is present in the immu- that this phosphorylation plays an importantrole in receptornoprecipitation mixtures (panel B ) ; (c) thelevels of cPLAz in mediated regulation of cPLAz (21).We therefore compared the extracts from differentiated and undifferentiated HL60 cells regulation of cPLA, by PKC in intactundifferentiated and neutrophil-like HL60 cells. In agreement with previous observaare similar (panel A). Taken together, these results suggest that theinability of Pz-purinergic agonist to induce AA release tions (29-32),Ca2+ionophore A23187 stimulated AA release in

Phosphorylation of PLADduring Granulocyte Differentiation A

IMMUNOBLOTS B

3121

C

218-

-218

-lo5

-100 -71

4 3 -28

R"I

r

-18

--v-18

18-

FIG.2. Immunoprecipitation and Western blotting of cPLA, from HLBO cell extracts.A, extracts from undifferentiated and BtcAMPdifferentiated HMOcellswereimmunoprecipitatedwithanti-cPLA, antiserum, separated bySDS-PAGE, and transferred tonitrocellulose membranes. Detectionwas carried out by incubation witha 1:1,000dilution of rabbit anti-cPLA,antiserum, followed by treatment with 12sI-labeled goat anti-rabbit IgG antiserum. In addition tothe cPLA, band, the secondary antiserum recognizes the rabbitIgG light andheavy chains present in the immunoprecipitated samples. Duplicate lanes are shown for the cell extracts derived from both differentiated Coif.) and undifferentiated (Undif.)cells. The amounts of extracted protein subjectedto preclearing/immunoprecipitationcorresponding to each lane were: 2.55 and 2.30 mg (for the duplicate of undifferentiated cells) and 2.25 and 2.40 mg (for the duplicate correspondingto the differentiated cells). cPLA2 st., 30 ng of cPLA, standard (bacterial recombinantcPLA,,Ref. 10) either directly added to the gel (left lane, no immunoprecipitation) or subjected to immunoprecipitation by the anti-cPLA, antiserum (right lane). For details, see "Experimental Procedures."B, control experiment showing the specific immunoprecipitation of cPLA, From neutrophil-like HL60 cell extracts by the anti-cPLA, antiserum. This panel also shows the signals obtained when 30 ng of cPLA, were either directly added to the gel (no immunoprecipitation)or added to extraction buffer and subjectedto the immunoprecipitation procedure, inthe presence or absence of anti-cPLA,. Detection was carried outas indicated for panel A. C, signals obtained when a constant concentration of cPLA, standard (20 nghbe) and increasing concentrationsof anti-cPLA, antiserum were added to extraction buffer and subjectedto the immunoprecipitation protocol (result obtained with 2 pl of preimmune serum per tube is also shown). Numbers at the bottom of lanes indicate the microliters of rabbit anti-cPLA, or preimmune antiserum added to the mixtures. Lanes labeled + and ++ show the signals obtained when 20 and 40 ng of cPLAz standard were directly added to the gels. Detection was carried out as indicated for panel A.

differentiated HL60 cells, but not in undifferentiated cells (Fig. 5). PMA, although inactive per se, significantly enhanced the effect of A23187 in differentiated HMO cells, consistent with similar effects in other cell types (e.g. Refs. 16-20), suggesting the involvement of PKC in the regulation of cPLA2. However, under the same conditions, PMA displayed weaker effects in the undifferentiated HL60 cells. These results suggested an altered PKC-mediated regulation of PLA2 in undifferentiated HL60 cells. To test this hypothesis, we compared the phosphorylation of cPLA2 in electropermeabilized undifferentiated and neutrophil-like HL60 cells using a n anti-cPLA2antiserum. As shown in Fig. 6 A , PMA strongly stimulated the phosphorylation of cPLA2 in electropermeabilized HL60 granulocytes, but failed to have any effect in theundifferentiated cells. It is worth mentioning that under these experimental conditions there wasoften clearly detectable phosphorylation of cPLA2 in differentiated HL60 cells even in the absence of PMA (with PMA usually significantly enhancing this reaction) but we never observed phosphorylation of cPLA2 in undifferentiated HL60 cells, regardless of the presence or absence of the PKC activator. These results suggest an alteration in the PKC-dependent phosphorylation mechanism for cPLA2 in undifferentiated HL60 cells. Agonist-induced Changes in MAP Kinase Phosphorylation in Undifferentiated andNeutrophil-like HL60 Cells-The lack of phosphorylation of PLA2 in the undifferentiated cells, and the reported in vitro phosphorylation and activation of PLA2 by MAP kinase (ERK2, Refs. 22 and 231,prompted us to study the phosphorylation of this kinase following incubation of cells in the presence of nucleotides or PMA. Activation of MAP kinases, resulting from the phosphorylation of both threonine andtyrosine residues, causes a slight reduction in themobility of these proteins during SDS-PAGE (Refs. 50 and 51; shift was from

approximately 39.6 to 40.2 kDa in our system). Aliquots were taken a t different times from cell suspensions incubated in the presence or absence of nucleotides or PMA; analysis of the samples showed that these compounds were able to trigger phosphorylation of ERK2 in both undifferentiated and differentiated cells (Fig. 7,top and bottom panels). Some minor differences were, however, evident: the responses triggered by ATP and UTP in the HL60 granulocytes were higher and more sustained than those measured in the undifferentiated cells. Also, UTP, which was more efficacious than ATP in triggering release of [3H]AA from HL60 granulocytes (Fig. 11, induced in these cells a higher and more sustained phosphorylation of MAP kinase compared to that driven by ATP (Fig. 7, bottom panel ). DISCUSSION

G proteins have been proposed to play a role in receptormediated regulation of PLA2 in many cells (reviewed in Refs. 13-15).In thiscontext, it hasbeen proposed that theactivation of cPLA2by receptor-coupled G proteins may be mediated through two distinct pathways: 1) direct interaction of G proteins with cPLA2; 2) G protein activation of PLC, leading to increased levels of diacylglycerol and inositol triphosphate/ cytosolic Ca2+, activationof PKC, and phosphorylation and activation of cPLA2(14).Two groups haverecently reported the in vitro phosphorylation and activation of cPLA2 by MAP kinase (22, 23). This suggests that thein vivo activation of cPLA2 by PKC may be mediated by MAP kinase, since the latter can be activated by PKC-dependent pathways in many cells (24, 25). However, several groups also observed the in vitro phosphorylation of cPLA2 by PKC (9, 22, 231, although the functional significance of this modification is controversial, as it was

Phosphorylation of PLA, during Granulocyte Differentiation

3122 A

0

CTRL

A

0

A

AIF,’

Own: Undlller. Calla Clomad: Mller. Calls

0

20

10

40

30

TIME ( min )

B 0 CTRL 0

A

ATP

A

0 0

.-c .c

0

s

v

AZP 8 GTPyS

Open: Undlfler. Cella Cloaed: Differ. Cella

I

I

I

lrM+Ca*’ 50pM GTPyS

FIG.4. AA release in homogenates prepared from undifferentiated or differentiated HL60 cells. [3HlAA-labeled cells were homogenized in EPB and assayed for AA release in EPB containing 3 m~ MgClz in the presence or absence of 50 p~ GTPyS and 1 p~ free Ca2+. The total incorporated 3H activity and the basal release of l3H1AA per assay tube were 129,348 and 310 cpm for undifferentiated, and 96,705 and 452 cpm for differentiated cells, respectively. For details, see “Experimental Procedures.” Open and closed bars represent undifferentiated and differentiated HL60 cells, respectively.

h

i

50pM GTPyS 1pM Ca”

COntrOl

0

A 0

W

0

v)

a W

CTRL ~

25

1

2

A V

0

3

%

~

PMA

A

A23187+PMA T

Open: Undiffw. Cells Closed: Oiller. Cells

c

0

J

W

a

a

s I Y

FREE Ca2+(log M)

AlF,

on FIG.3. A,Time course correspondingto the effects of AA release in intact undifferentiated or differentiated HLBO cells. AA release was measured in intact cells in the absence (CTRL, ircles) or presence (triangles) of 10 n” NaF/15 p~ Acls. Samples were collected a t indicated times and assayed for PHIAA release. Total cellincorporated 3H activity and the basal release of L3HIAAper assay tube were 46,463 and 724 cpm for undifferentiated, and 35,202 and 1088 cpm for differentiated cells, respectively.B , effects of Ca2+,ATP, and GTPyS on AA release in electropermeabilizedcells. PHIAA-labeledundifferentiated or differentiated HL60 cells were electropermeabilized and assayed for AA release in EPB containing 3 m~ MgClz and increasing concentrations of free Ca2+,in the absence (circles)or presence of 1 m~ ATP (triangles)or 1 m~ ATP plus 100 p~ GTPyS (squares). The total cell-incorporated 3H activity and the basal release of PHIAA per assay tube were 168,650 and 5,585 cpm forundifferentiated, and 128,543 and 8,757 cpmfor differentiated cells, respectively. For more details see ”Experimental Procedures.” Open and closed symbols represent undifferentiated and differentiated HL60 cells, respectively, in both panels.

found to increase cPLAz activity in some (23), but not all (9,22), protocols. Previous studies have examined, in particular, the regulation of PLAz in HL60 granulocytes (14, 21, 33, 38). The stimulation of AA release by chemotactic or P2-purinergicreceptor agonists was inhibited completely or largely, respectively, by pretreatment of these cells with pertussis toxin (21, 38). Our recent studies demonstrated the importance of PKC-dependent phosphorylation in theregulation of HL60 granulocyte cPLA, (21).

0

30 10

20

40

TIME ( min ) FIG.5. Effects of A23167 and PMA on AA release in intactundifferentiated and differentiated HL80 cells. Cells were labeled with L3H3AA and assayed for AA release in the absence (CTRL, circles) or presence of 5 p~ A23187 (squares),100 n~ PMA (triangles),or their combination (inuerted triangles). Open and closed symbols represent undifferentiated and differentiated HL60 cells, respectively. The total incorporated 3H activity and the basal release of PHUA per assay tube were 68,360 and 1,192 cpm for undifferentiated, and 97,647 and 1,372 cpm for differentiated cells, respectively.

We have also proposed that HL60 granulocyte cPLAz is under an inhibitory constraint that prevents its activation by G protein and Ca2+, and that this constraint can be released by PKC-dependent phosphorylation of cPLAz or by mechanical disruption of the cell (21). This may represent aunique mechanism for the regulation of cPLA, in that both, G protein and phosphorylation, are required for the expression of maximal enzyme activity. The present study demonstrates that the coupling of Pzpurinergic receptors to AA release in intact undifferentiated HL60 cells is altered when comparedto that inHL60 granulocytes (33,38) (Fig. 1). This deficiency is reminiscent of the findings by a number of groups (Refs. 29-32, also this study) showing that the Ca2+ ionophore A23187 cannot induce AA

Phosphorylation of PLAZ during Granulocyte Differentiation

3123

PLA2 (33, 38) and PLD (42) in differentiated cells. Therefore the inability ofATPto stimulateAA release in undifferentiated HL60 cells is apparently not due to the lack of the expression of 5.. ' functional P2-purinergic receptors. Undifferentiated HL60 cells I f express a functional PLA2, as indicated by the release of AA 208-215 from exogenous substrates measured inhomogenates (Table I and Ref. 30). The observation that melittin triggers similar 1004PLAP- . * -105 increases inAA release inundifferentiated and neutrophil-like 71cells seems consistent with the previous result; however, it -70 should be interpreted with caution in view of the different effects of this peptide on low and high molecular weight phos434 3 pholipases A2 (47). Immunoprecipitation and Western blotting experiments clearly demonstrated the similarlevels of expression of cPLA2 in undifferentiated and neutrophil-like HL60 28cells (Fig. 2). It can therefore be concluded that the coupling -28 between purinergic receptorand cPLA2 is deficient in undifferentiated HL60 cells. This situation differs from the increased 181 -18 PLA2 activity measured in extracts of cells undergoing adipoW f . + + + + - - - + + Dif.. P M A cyte differentiation (52). Dif. - - - + + + + + NRS To elucidate the mechanism responsible for the altered couPMA - - + + - - + + + anti.cPLA, pling between purinergic receptors and PLA2 in undifferentiNRS + - + - + - + ated HL60 cells, we inspected the pathways leading to PLA2 anti.cPLA, - + + + - + activation either via G proteins or PKC-mediated phosphoryFIG.6. Phosphorylation of cPLA, in electropermeabilizedUII- lation. We observed similar stimulatoryeffects of GTPyS on AA differentiated and differentiated HL60 cells. A, cells were subjected to electropermeabilization and incubated in the presence of release in both differentiated and undifferentiated HL60 cell [+j2PlATP, 150 TIM PMA, or solvent (0.002% dimethyl sulfoxide)for 8 homogenates (Fig. 4). This evidence indicates that the G promin at 37 "C. Extracts from the labeledcellswereprepared, =pre- tein related to PLA2 activation is expressed in both differenticleared" and subjected to immunoprecipitation inthe presence of antiated andundifferentiated HL60 cells. This observation appears cPLA, antiserum or normal rabbit serum (NRS). Apparent molecular consistent with the demonstration of similar levels of multiple weights of prestained standards are indicated. For experimental details see "Experimental Procedures." B , phosphorylation of cPLA, in elec- forms of Gi, in undifferentiated and dibutyryl cyclic AMP-diftropermeabilized cells in thepresence of 100 n~ PMA. Theextracts were ferentiated HL60 cells (53). On the contrary, we failed to detect precipitated with either normal rabbit serumor anti-cPLA,. The auto- PKC-dependent phosphorylation of cPLA2 in undifferentiated radiogram is representative of the results obtained in several experi- HMO cells, as opposed to the results obtained with the HL60 mentswhere the bandcorrespondingtophosphorylatedcPLA, apgranulocytes. This alterationcould be explained by insufficient peared as anadditionalbandwhen the immunoprecipitationwas carried out in the presence of anti-cPLA2 (right lane) but not with levels of PKC isoforms andor alterations in the expression or normal rabbit serum (left lane). activation of MAP kinase. Undifferentiated HL60 cells express relatively low levels of a- and PI1 isozymes of PKC ( y is exIMMUNOBLOTS anti-MAP2 kinase tremely low or absent) (54, 551, while differentiation into a neutrophil-like phenotype results in significant expression of the y and in5-7-fold increases in thea-and PI1 PKC isoforms (54). Interestingly, in Madin-Darby canine kidney cells cPLA2mediated AA release was shown to depend on the normal exI I I pression of the a, but not the P, isozyme of PKC (20). It is Dif. therefore possible that the deficient coupling of P2-purinergic receptors to activation of cPLA2in undifferentiated HL60 cells is related to a particular patternof expression of PKC isoforms. I I I The expression of particular isoforms of PKC, that do not. con0 3 6 1 0 0 1 3 6 1 0 0 1 3 6 1 0 0 1 3 6 1 0 vey the phosphorylation and activation of cPLA2, may also be CTRL. 25 uM ATP 25 uM UTP 150 nM PMA responsible for the inability of PMA to enhance the effects of FIG.7. Nucleotide- and PMA-induced changes in MAP2 kinase A23187 in undifferentiated HL60 cells (Fig. 5 ) , and for the phosphorylation in undifferentiated and neutrophil-likeHL60 minor effects of AIF; and GTPyS on the releaseof AA in intact cells. The phosphorylation state of MAP kinase was determined by and electropermeabilized undifferentiated HL60 cells (Fig. 3, A virtue of the decrease in mobility in SDS-PAGE gelsresulting from this covalent modification(50,51).Aliquots from each mixture were taken at and B ) . These conclusions must be integrated with our obserthe indicated times (0,1, 3, 6, and 10 min; the 1-min time point corre- vations on MAP kinase phosphorylation. Immunoblotting exsponding to the controls is not shown) pelleted,subjected to SDS-PAGE periments usinga n anti-ERK2 antiserum showed that agonists (12% acrylamide, 0.16% bis-acrylamide), followed by transference to or PMA triggered phosphorylation of MAP kinase in both unnitrocellulose membranes and imunoblotting. The primary antiselvm was a rabbit anti-rat ERK2 ((rIICp42, Ref. 37). The top panel (Undif.) differentiated and differentiated HL60 cells (Fig. 7). At this shows the signals corresponding to 2.5 x lo5 undifferentiated HL60 time, we cannot ascertain whether therelatively minor differcells; the bottom panelCoif.), illustrates the signals corresponding to the ences in the fraction of MAP2 kinase that underwent phossame number of HL60 granulocytes. For details, see "Experimental phorylation in response to ATP or UTP, and/or in the time Procedures." course of this modification (lower and less persistent for the undifferentiated cells), mayexplain (atleastinpart)the release in undifferentiated HL60 cells. The purinergic recep- marked differences in the releaseof [3HlAAby both cell types. tors are expressed in both differentiated and undifferentiated In this regard, there seems to be some direct correlation beHL60 cells, as suggested by the similar potencies of various tween a higher, more persistent, phosphorylation of MAP2 kireceptor agonists in the stimulationof PLC in both neutrophil- nase and the release of arachidonic acid when the effects of like and undifferentiated cells (3941) and in the stimulation of UTP andATP in HL60 granulocytes are compared (see Figs. 1 PHOSPHORYLATION A

6

-

-

-

-

-

3 124

Phosphorylation of PLA2 during Granulocyte Differentiation

and 7, bottom panel 1. However, the following facts deserve con- 13. Insel, P.A., Weiss, B,A., Slivka, S. R., Howard, M. J.,Waite, J. J., and Godson, C. A. (1991)Biochem. Soc. P a m . 19,329-333 sideration: ( a ) PMA treatment enhances M A P kinase phos- 14. Cockcroft, S., Nielson, C. P., and Stutchfield, J. (1991)Biochem.Soc. 'Pans. 19, phorylation in both cells types (Fig. 7);( b )PMA-induced phos333-336 phorylation ofPLA2 was detected only in the differentiated 15. Axelrod, J. (1990)Biochem. Soc. Pans. 18, 503-507 R. M., and Axelrod, J. (1987)Proc. Natl. Acad. Sci. U. S. A. 84, 6374cells (Fig. 6 ) ;( c ) PMA treatment, per se, does not trigger release 16. Burch, 6378 of arachidonic acid in the differentiated cells (Fig. 5 and Ref. 17. Murayama, T.,Kajiyama, Y., and Nomura, Y. (1990)J . Biol. Chem. 266,42904295 21); (dl nucleotides cannot trigger substantial release of AA in 18. Ha, A. K, and Klein, D.C. (1987)J. Biol. Chem. 262, 11764-11770 undifferentiated cells (Fig. 11, in spite of triggering MAP kinase 19. Bonventre, J. V., Gronich, J. H., and Nemenoff, R. A. (1990)J . B i d . Chem. 266, 4934-4938 activation (Fig. 71, PLC activation and Ca2+mobilization (Ref. 20. Godson, C., Weiss, B. A, and Insel, P. A. (1990)J. B i d . Chem. 266,83698372 41); (e) in vitro studies have observed that MAP2 kinase can 21. Xing, M.,and Mattera, R. (1992)J. B i d . Chem. 267, 2596625975 phosphorylate PLA2 inducing a concomitant increase (2-3-fold) 22. Lin, L.-L., Wartmann, M., Lin, A. Y., Knopf, J. L., Seth, A,, and Davis, R. J. (1993)Cell 72,269-278 in theactivity of this enzyme (Refs.22 and 23).Accordingly, it 23. Nemenoff, R. A., Winitz, S., Qian, N.-X., Van Putten, V,Johnson, G. L., and seems reasonable to conclude that MAP kinase activation and Heasley, L. E.(1993)J . B i d . Chem. 268,1960-1964 Ca2+mobilization must act in concert with the recruitment of 24. Cobb, M. H., Boulton, T.G., and Robbins, D. J. (1991)CeN Regul. 2,965-978 25. Posada, J., and Cooper, J. A. (1992)Mol. Bioi. Cell 3, 583-592 additional PKC-dependent stimulatory factors (andlor removal 26. Collins, J. S., Gallo, R. C., and Gallagher, R. E. (1977)Nature 270,347349 of inhibitory factors) for phosphorylation of PLA, (and activa- 27. Collins, J. S., Ruscetti, F. W., Gallagher, R. E., and Gallo, R. C. (1978)P r w . Natl. Acad. Sei. U. S. A. 76, 2458-2462 tion) ta occur in. uiuo. Finally, the similar activities measured in G.,Santoli, D. E., and Damsky, C. ( 1979)Proc. Natl. Acad. Sei. U. S. A. homogenates and subcellular fractions of u n ~ f f e ~ n t iand a t ~ 28. Rovera, 76,2779-2783 neutrophil-like HMO cells (Ref. 30, Table I and Fig. 4 of this 29. Bonser,R. W., Siegel. M.I., McConnell, R. T., and Cuatrecasas, P. (1981) Biocfiem. Bwphys. Res. Commun. 93,614-620 report) indicate that both cell types contain a similar "poten30. Billah, M.M., Eckel, S., Myers, R. F., and Siegel, M. I. (19861J. Bwt. Chem. tial" PIA2activity, supporting the proposal that theabove men261,5824-5831 '6 6 , tioned PKC-de~ndentr ~ ~ t m e ofn factors t (in addition ta 31. Suga,K., Kawasaki, T.,Blank, M. L., and Snyder, F. (1990)J. Biol. Chem.2 12363-12371 MAP kinase activation) represents the locus of the alteration 32. Garcia, M.C., Garcia, C., Gijon, M. A., Fernandez-Gallardo, S., Mollinedo, F., that minimizes agonist-mediated AA release in differ en tiand Sanchez Crespo, M. (1991)Biochem. J. 273,573578 33. Xing, M., Thevenod, F., and Mattera,R. (1992)J. Biol. Chem. 267,6602-6610 ated HL60 cells. Aeknow~e~men~s-We thank L.-L. Lin and J. L. Knopf forthe generous gift of anti-cPLA, antiserum and M. s. Simonson and M. J. D u m (Division of Nephrology, University Hospitals of Cleveland) for providing us theanti-ERK2 antiserum. We also thank S. Schomisch for advice in the MAP kinase immunoblotting experiments and R. Walenga for helpful discussions and lipid standards.

REFERENCES 1. Irvine, R.F. (1982)Biochem. J. 204,3-16 2. Hanahan, D. J. (1986)Annu. Rev. Biochem. Mi, 483-509 3. Dennis, E. A,, Rhee, S. G., Billah, M. M., and Hannun,Y. A. (1991)FASEB J . s,2068-2077 4. Clark, J. D., Milona, N., and Knopf, J. L. (1990)Proc. Acad. Sei. U. S. A. 87, 770%7712 5. Gronich, J. H., Boventre, J. V., and Nemenoff, R. A. (1990)Biochem. J. 271, 37-43 6. Kramer, R. M., Roberts, E. F., Manetta, J., and Putnam, J. E. (1991)J. B i d . Chem. 266,5266-5272 7. Leslie, C. C. (1991)J. B i d . Chem. 266, 11366-11371 8. Takayama, K., Kudo, I., Kim, D. K., Nagata, K., Nozawa, Y., and Inoue, K. (1991)FEBS Lett. 282,326-330 9. Wijkander, J., and Sundler, R. (1991)Eur. J. Biochem. 202,873480 10. 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 66, 1043-1051 11. Sharp, J. D.,White, D. L., Chiou, X.G., Godson, T., Gamboa, G . C., McClure, D., Burgett, S., Hoskins, J.,Skatrud, P. L., Sportsman, J. R., Becker, G . W., Kang. L. H., Roberts. E. F., and Kramer, R. M. (1991)J. Biol. Chem. 266, 1485%14853 12. Lin, L.-L., Lin, A. Y., and Knopf, J. L.(1992)P m . Natl. Acad. Sci. U. S. A. 89, 6147-6151

34. Bligh, E. G., and Dyer, W. J. (1959)Can. J. Bioehem. Physid. 37,911-917 35. Hang, S-C. L.,and Levine, L. (1976)Proe. Natl. h a d . Sei. U. S.A. 73, 173% 1734 36. Laemmli, U. K. (1970)Nature 227,680-685 37. Wang, Y., Simonson, M. S., Pouyssegur, J., and Dunn, M. J. (1992)Biochem. J. 287,589-594 38. Cockcroft, S., and StutcMeld, J. (1989)Biochem. J. 263, 715-723 39. Cowen, D. S., Lazarus, H. M., Shurin, S. B., Stoll, S. E., and Dubyak, G . R. (1988)J . Clin. Inuest. 83,1651-1660 40. Cowen, D. S., Sanders, M., and Dubyak, G . (1990)Biochim. Biophys. Acta 1063,195-203 41. Dubyak, G . R., Cowen, D. S., and Meuller, L.M. (1988)J. Biol. Chem. 285, 18106-18117 42. Xie, M., Jacobs, L. S., and Dubyak, G. R. (1991)J. Clin. Inuest. 88, 45-54 43. F'urth, E. E.,Hurtubise, V., Schott, M. A., and Laposata, M. (1989)J. B i d . Chem. 264, 18494-18501 44. Winkler, J. D., Sung, C-M., Hubbard, W. C., and Chilton, F. H. (1993)Biochem. J. 291,825431 45. Mollay, C., and Kreil, G . (1974)FEBS Lett. 46, 141-144 46. Shier, W. T.(1979)P m . Natl. Acad. Sci. U. S. A. 76, 195-199 47. Steiner, M. R., Bomalski, J. S., and Clark, M. A. (1993)Biochim. Biophys.Acta 1166,124-130 48. Flower, R. J., and Blackwell, G . L. (1976)Biochem. Pharmacol. 26,285-291 49. Yorio, T., and Bentley, P. J. (1978)Nature 271, 79-81 50. Leevers, S. J., and Marshall, C. J. (1992)EMBO J. 11, 569-574 51. de Vries-Smits, A. M. M., Burgering, B. M. Th., Leevers, S. J., Marshall, C. J., and Bos, J. L. (1992)Nature 367, 602-604 52. Gao, G., and Serrero, G . (1990)J. B i d . Chem. 266,2431-2434 53. Murphy, P. M., Eide, B., Goldsmith, P.,Brann, M., Gierschik, I?, Spiegel, A., and Malech, H. L. (1987)FEBS Lett. 221,8146 54. Makowske, M., Ballester, R., Cayre, Y., and Rosen, 0. M. (1988)J . B i d . Chem. 2% 34023410 55. Hoceuar, B. A., and Fields, A. I? (1991)J. Bid. Chem.Zse, 2833