Phosphatidylcholine Hydrolysis by Phospholipase D Determines ...

3 downloads 0 Views 1MB Size Report
Phosphatidylcholine Hydrolysis by Phospholipase D Determines. Phosphatidatle and Diglyceride Levels in Chemotactic. Peptide-stimulated Human Neutrophils.
Vol. 264. No . 29, Issue of' October 15, PP. 17069-17077,1989 Printed in U.S.A.

THEJOURNAL OF BIOLOGICAL CH~EMISTRY 0 1989 by The American Society for Biochemistry and Molecular Biology, Inc

Phosphatidylcholine Hydrolysisby Phospholipase D Determines Phosphatidatle and Diglyceride Levelsin Chemotactic Peptide-stimulated Human Neutrophils INVOLVEMENT OF PHOSPHATIDATE PHOSPHOHYDROLASE IN SIGNAL TRANSDUCTION* (Received for publication, May 1, 1989)

M. Motasilm Billah, Stephen Eckel, TheodoreJ. Mullmann, Robert W.Egan, and MarvinI. Siege1 From the Department of Allergy and Inflammation, Schering-Plough Corporation, Bloomfield, New Jersey 07003 Many cells produce DG' and PA upon specific stimulation Human neutrophils have been labeled in l-O-alkylphosphatidylcholine (alkyl-PC) with ""P by incubation (1, 2). DG together with Ca2+ activates proteinkinase C and with alkyl-[32P]lysoF'C.Upon stimulation with the che- modulates various cell responses including degranulation (3). motacticpeptide,formylMet-Leu-Phe(fMLP),these PA is a potent Ca2+ionophore in several systems and pos32P-labeled cells produce l-O-alkyl-[32P]phosphatidic sesses fusogenic properties (2). Formation of PA and DG by acid (alkyl-[32P]PA),and, in the presence of ethanol, 1- stimulated cells could occur via several distinct pathways. In O-alkyl-[32P]phosph~didylethanol (alkyl-[32PlPEt). many cells, activation of specific receptors leads to rapid Because the cellular ATP contains no "'P, a l k ~ l - [ ' ~ P ] stimulation of phospholipase C that degrades phosphoinosiPA and alkyl-[32P]PEt must be formed from alkyl-['"P] tides (4). The resultant DG is then phosphorylated by DG PC byphospholipase D (PLD)-catalyzed hydrolysis and kinase to PA (1,2). PA is also formed by direct action of PLD transphosphatidylati.on,respectively. Analyses of the on phospholipids (5-10) or de m u 0 involving glycerol kinase sn-1 bonds by selective hydrolysis and mass measurements reveal that the PA and formed PEt during stim- and acyltransferases (11).PA, thus formed (5-11), can then be dephosphorylated by PA phosphohydrolase (12) to produce ulation contain both ester and ether bonds with distributions similar tothat in the endogenous PC. Further- DG. Human neutrophils stimulated with various agents includmore, in neutrophills labeled in alkyl-[32P]PC, the specific activities of the diradyl-PA and diradyl-PEt ing fh4LP produce PA (5-9) and DG (6, 8, 13-17).DG formed during stimulation are similar to thatof dira- formation occurs in a biphasic manner (16, 17). The early, dyl-PC. These results demonstrate that the fMLP-in- small rise is detectable only by radiolabeling procedures (16) duced PLD utilizes diradyl-PC as the major substrate. and is followed by a large, sustained increase quantifiable by It is further conclud'ed that, at earlytimes (30 s), P A mass measurements (8, 13-17). While the early rise might be and PEt are both folrmed almost exclusively by PLD. due to the action of phospholipase C on phosphoinositides, Following stimulation with fMLP, neutrophils double- the late increase appears to derive from sources other than labeled in alkyl-PC! by incubationwith ["Hlalkyl- phosphoinositides (16, 17). In several cells including human lysoPC and alkyl-['':P]lysoPC generate ['Hlalkyl-DG neutrophils (6, 8), hepatocytes (18), 3T3 fibroblasts (19, 20), and [32P]orthophosplhate (['"P]PO,) with superimposREF52 rat embryo cells (21, 22) and MDCK-DI cells (23), able kinetics, indicating degradationof PA by a phos- certain stimulants and growth factors cause PC hydrolysis phohydrolase. Generation of ['Hlalkyl-DG and [""PI and DG accumulation, suggesting PC as a source of DG. PO4 lags behindP A formation and parallels the decline Although activation of a PC-specific phospholipase C has in PA accumulation,. In addition, generation of both been proposed (18), an alternative pathway involving sequen['Hlalkyl-PA and ['EIIalkyl-DG requires extracellular Ca2+and cytochalasin B. Furthermore, the phospho- tial action of PLD and PA phosphohydrolase (21, 22, 24) hydrolase inhibitor, propranolol, decreases both ["HI would also result in PC hydrolysis and DG formation. Recent studies from our laboratory have firmly established alkyl-DG and [32P]Pj04while increasing ["Hlalkyl-PA that upon stimulation with M L P , HL-60 granulocytes proand not altering [S'H]alkyl-PEt. Moreover, thedecreases in DG are accounted for by increases in PA. duce PA via PLD-catalyzed hydrolysis of PC (5, 6). We now These results demonstratethat PLD-derived alkyl-PA demonstrate that, in cytochalasin B-treated neutrophils stimis degraded by a phosphohydrolase to produce alkyl- ulated with fMLP, PLD is activated to selectively hydrolyze DG. DG formed during stimulation contains both ester PC generating PA and that thisP A is then degraded rapidly and ether-linked species and this DG formation is in- by PA phosphohydrolase to produce DG. We further conclude hibited completely by propranolol. Upon stimulation, that, under these conditions, PC is the major source of both alkyl-[32P]PC-labeled neutrophilsdo not produce [""PI PA and DG produced during neutrophil stimulationand that phosphocholine, suggesting that PC is not hydrolyzed the previously reported sustainedDG formation occurs largely by phospholipase C. In addition,PA is formed in by sequential actions of PLD and PA phosphohydrolase on amounts sufficient to account for of all the DG formed PC andnot by phospholipase C. during stimulation. is It concluded that theDG formed during fMLP stimulationis derived almost exclusively The abbreviations used are: DG, diglyceride; PA, phosphatidic from PC via thePLDI/PA phosphohydrolase pathway. acid; alkyl-PA, l-O-alkyl-PA; PLD, phospholipase D; PC, phospha* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

tidylcholine; alkyl-PC, l-O-alkyl-PC: alkyl-lysoPC, l-O-alkyl-lysoPC; fMLP, N-formyl-methionyl-leucyl-phenylalanine;PEt, phosphatidylethanol; alkyl-PEt,l-O-alkyl-PEt;Hepes, N-2-hydroxyethylpiperazine-N-2-ethanesulfonicacid; EGTA, [ethylenebis(oxyethylenenitri1o)ltetraacetic acid; TLC, thin-layer chromatography.

17069

17070

Diglyceride Production by Phospholipase DIPhosphatidate Phosphohydroke

separation of PA, PEt, monoglyceride, DG, and triglyceride. Major EXPERIMENTAL PROCEDURES phospholipids were analyzed using chloroform/methanol/acetic acid/ Materials - 1-0-[9,10-3H]Octadecyl-2-lyso-sn-glycero-3-phosph~choline ([3H]alkyl-lysoPC,90 Ci/mmol) was purchased from Amer- water (503083, by volume) (solvent system IV) (28). The lipids were sham COT. [*4C]Phosphocholinewas from Du Pont-New England located by autoradiography or by staining with iodine vapor and the Nuclear. [Y-~'P]ATP(4500 Ci/mmol) and [32P]orthophosphate(car- silica gel areas containing individual lipids were scraped and quantified by liquid scintillation spectrometry. rier-free) obtained from ICN, Imine, CA. l-O-Hexadecyl-2-acetyl-snAnalysis of Water-soluble Products by TLC-To analyze waterglycerol was from Nova Biochem, Switzerland and Escherichia coli soluble products, neutrophils labeled in alkyl-PC with 32Pwere inDG kinase was from Lipidex, Westfield, NJ. Precoated Silica Gel-G thin layer plates (0.25-mm thick) from Whatman and precoated cubated, as appropriate. Samples containing unlabeled cells were also cellulose thin layer plates were from Analtech, Newark, DE. Choline included. Immediately after stopping the reactions, standard samples tosylate was from Synthon, Houston, TX, and 2,4,6-triisopropylben- of [32P]P04and ["C]phosphocholine were added to those samples zenesulfonyl chloride was from Aldrich. Unlabeled lipids including that contained unlabeled cells. Following extraction as outlined above (26), aliquots (0.5 ml) of the aqueous phase were dried under a flow PEt were obtained from Avanti Polar Lipids, Birmingham, AL. All other reagents including fatty acid-free bovine serum albumin, cyto- of nitrogen. The residues were dissolved in 50%ethanol in water and spotted on cellulose thin layer plates. The plates were developedusing chalasin B, WLP, andpropranolol were purchased from Sigma. All methanol, 10% trichloroacetic acid, NH,OH (653510, by volume) solutions and cell suspensions were prepared in Hepes/saline buffer containing Hepes (25 mM, pH 7.4), NaCl (125 mM), M&12 (0.7 mM), (29). In this solvent system, PO4 ( R p = 0.2) was well separated from phosphocholine (RF= 0.7) and phospholipids (RF= 0). Cellular ["PI EGTA (0.5 mM), glucose (10 mM), and fatty acid-free bovine serum Po4 and [32P]phosphocholinewere localized by autoradiography and albumin (1mg/ml). comparison with standards. Preparation of Alkyl-r'P]lysoPC-This radiolabeled phospholipid Determination of Phospholipid Mass-Triplicate samples containwas prepared in a series of enzymatic and chemical reactions as ing 5 X lo7 double (3H and 32P)-labeledcells in 5 ml were stimulated was for 30 s with fMLP in theabsence and thepresence of 0.5% ethanol. detailed elsewhere (5). Briefly, l-0-alkyl-2-acetyl-sn-glycerol phosphorylated by E. coli DG kinase in the presence of [32P]ATP to Appropriate control samples that received no fMLP were also inThe lattercompound was reacted cluded. The lipid extracts equivalent to 2 X IO6cells were analyzed produce l-O-alkyl-2-a~etyl-[~~P]PA. chemicallywith choline tosylate and 2,4,6-triisopropylbenzenesulfon- by TLC for PC using solvent system IV. The remainder of the extracts whichwas then was used to separate PA and PEt by TLC using solvent system I. ylchloride to produce l-O-alkyl-2-a~etyl-[~~P]PC, hydrolyzed to alkyl-[3zP]lysoPCby treatment with methanolic KOH. The silica areas containing either PA, PEt, or PC were scraped and Neutrophil Zsohtion and Preparation of Labeled Neutrophiksuspended in 10 mM EGTA solution. The suspension was eluted twice Venous blood from healthy male donors was drawn in acid/citrate/ by the procedure of Bligh and Dyer (26) and purified lipid samples dextrose buffer. Polymorphonuclear leukocytes (neutrophils) were (PA, PEt, PC) were digested with 70% perchloric acid (0.25 ml). The isolated by dextran sedimentation, Ficoll/Hypaque Gradient centrif- resultant inorganic phosphate was determined colorimetrically (30) ugation and hypotonic lysis to remove erythrocytes (25). These in a final assay volume of 1.25 ml. The 3H and 32Pwere quantified in washed cells were suspended in Hepes/saline buffer a t a concentration the silica suspensions and in perchloric acid digest (for 32Ponly). The of 2 X lo' cells. Alkyl-[32P]lysoPC(2-4 pCi/ml) or [3H]alkyl-lysoPC 32Pvalues for perchloric acid digests were compared with those for (2-5 pCi/ml) was added to thesuspension, and in some experiments, initial silica suspensions to determine the percent recovery. Values both alkyl-[32P]lysoPCand [3H]alkyl-lysoPCwere added simultane- presented were corrected for the recovery losses, which were within ously. These suspensions were then incubated for 75 min at 37 "C. 15-30%. Unincorporated radiolabels were removed by washing cells twice in Determination of DG Mass-DG mass was determined enzymatiHepes/saline buffer. Washed neutrophils were finally suspended in cally as described (31) except that cardiolipin was replaced with 20 the Hepes/saline buffer (2 X 10' cells/ml). Cell viability was assessed mM dioleoyl-phosphatidyl-glyceroland that the octyl-@-glucoside/ by trypan blue exclusion and was consistently found to be over 95%. phosphatidylglycerol mixture contained 1% fatty acid-free bovine Eighty % of the added alkyl-[3zP]lysoPCwas taken up by the cells. serum albumin. Briefly, lipid extracts from 5 X lo6 cells obtained Ninety-eight % of this cell-associated 32Pwas recovered in the lipid before and after stimulation were suspended in octyl-8-glucoside/ fraction and the remainder in the aqueous phase. Analysis of the phosphatidylglycerol mixture and thentreated with E. coli DG kinase lipids by thin layer chromatography showed that 95% of the "P was in the presence of [32P]ATPof known specific activity (100,000incorporated into alkyl-PC and theremainder was found with alkyl- 300,000 dpm/nmol). The resultant [32P]PAwas separated by TLC lysoPC (3%) and other phospholipids (2%). High performance liquid using solvent system I1 and quantified by scintillation counting. The chromatographic analysis of the aqueous phase obtained after lipid mass of [32P]PAwas determined from the known specific activity of extraction (5) revealed that cellular ATP contained little or no 32P. the [32P]ATP. Therefore, this procedure allowed rapid labeling of cellular alkyl-PC Subspecies Analysis of Various Lipids-Distribution of the sn-1 with 32Pwithout labeling cellular ATP. [3H]Alkyl-lysoPC, either bonds in various lipids was analyzed by selective hydrolysis with alone or in combination with alkyl-[32P]lysoPC,was similarly taken NaOH and HCl. Alkali and acid selectively hydrolyze ester and up by neutrophils with subsequent incorporation into alkyl-PC. alkenyl bonds, respectively, whereas alkyl bonds are resistant to both Incubation Conditions and Lipid Extraction-Appropriately labeled treatments. Large samples (5.0-7.5 X 10' cells) of 3H-labeledneutroneutrophils (5 X lo6) were suspended in 0.45 ml of Hepes/saline phils containing lo' cells/ml, 1.5 mM CaCIZ,and 5 p M cytochalasin B buffer containing 1.5 mM CaC12 and 5 p~ cytochalasin B and then were incubated for 5 min before adding either buffer or 100 nM fMLP incubated at 37 "C for 5 min. Various compounds were dissolved in for an additional 30 s in the presence of 0.5% ethanol. PA and PEt dimethyl sulfoxide and diluted into buffer before use. The final were separated by TLC, andsilica gel areas corresponding to PA and concentration of dimethylsulfoxide in the assays did not exceed 0.2%. PEt were extracted (26). PC was similarly isolated from unstimulated The reaction was initiated by adding either vehicle or fMLP in the neutrophils. Aliquots of these lipid samples were treated separately absence or presence of 0.5% ethanol in a volume of 50 pl. After with either buffer, 0.5 N NaOH in methanol for 5 min (32)or 0.5 N appropriate times, the reaction was terminated by the addition of NaOH in methanol (5 min) followed by 5 N HCl for 30 min (33). chloroform/methanol/acetic acid (100:200:4, byvolume) and standard Samples treated with NaOH or HCl were brought to neutrality by samples (10 pg each) of PA, PEt, andDG (or other lipids as needed) dropwise addition of 0.5 N formic acid or 5 N NaOH, respectively. were added. The phases were then separated (26). The lipids were extracted, digested with 70%perchloric acid, and the Analysis of Lipids by TLC-The lipids in the lower chloroform resultant inorganic phosphate was determined (30). Lipid recovery phase were spotted on Silica Gel G plates. The plates were developed was monitored by 3zP.Phospholipid phosphorus remaining as lysousing the organic phase of ethylacetate/isooctane/acetic acid/water phospholipids after alkali treatment was taken to represent the (110:5020:100,by volume) (solvent system I) (7). In this system, PA amount of ether-linked (1-0-alkyl- and 1-0-alkenyl-) phospholipid (RF= 0.1) and PEt ( R F = 0.33) were separated from each other and species. Subtraction of this value from that of total phospholipid content gave the amount of ester-linked (diacyl-) phospholipid spefrom major phospholipids (RF= 0) and neutral lipids (RF= 0.7-0.95). In some experiments, PA was also separated using chloroform/meth- cies. Phospholipid phosphorus remaining after the sequential treatanol/acetic acid (65:15:5, byvolume) (solvent system 11).When it was ment with NaOH and HCl was taken to represent the amount of necessary to analyze neutral lipids, the plates were first developed alkyl-linked phospholipid species. Data Presentation-Assays were performed in duplicate, unless half-way in solvent system I, dried, and developed again using hexane/ diethylether/methanol/acetic acid (90203:2, by volume) (solvent sys- mentioned otherwise. Data points arethe mean of two determinations tem 111) (27). This double-developmentprotocol allowed simultaneous which were within 5% of the mean. Each set of experiments was

17071

Diglyceride Production by Phospholipase DIPhosphutidate Phosphohydrolase 50000 -

performed at least two times and the representative experiment was presented.

40000

ALKYL-PA

-ETHANOL

-

RESULTS

Direct Proof for PLD ‘4ctivation in fMLP-stimulated Human Neutrophils-Two immediate products of PLD action on PC are PA and choline. These products could also be formed indirectly via phospholipase C action on phosphatidylcholine. This latterreaction generates DG and phosphocholine, which could subsequently be converted to PA and choline by kinase and phosphatase activities, respectively. Thus, although radiolabeled choline or PA could be conveniently measured in intact cells prelabeled in cellular phospholipids with radiolaa 0 40000 beledglycerol, fatty acids, or choline, such measurements ALKYL-PA +ETHANOL would not distinguish PLD from phospholipase C. Because E 30000 formation of [32P]PAfrom 32P-labeledphospholipids can ocP cur only by PLD, a conclusive demonstration of PLD activity 6 20000 in intact cells requires specific labeling of phospholipids with I P 32Pwithout labeling cellular ATP. u) 10000 When 32P-labeledhuman neutrophils are stimulated with ?! % 0fMLP in the presence ofCa2+,a l k ~ l - [ ~ ~ P ]isPformed A rapidly, a reaching a maximum within 60 s and remaining unchanged for 5 min thereafter (Fig. 1).Pretreatment of neutrophils with +ETHANOL cytochalasin B greatly augments fMLP-induced generation of ALKYL-PE1 alkyl-[32P]PA.Because, under the presentlabeling conditions, cellular ATP has not been labeled with 32P,a l k ~ l - [ ~ ~ P ] P A 20000 must be formedby the action of PLD on alkyl-[32P]PC.These 10000 4 4 I data, therefore, conclusively link alkyl-PA formation to PLD activation in fMLP-stimulated neutrophils. Previous studies have shown that in HL-60 granulocytes, PLD-mediated hydrolysis and transphosphatidylation activitiesare closely associated (5, 6). Therefore, to determine 0 60 120 180 240 300 360 whether PLD activation in human neutrophils is associated TIME, seconds with transphosphatidylation, 32P-labeledcells have been stimulated with fMLP in the presence of ethanol. As shown in FIG. 1. Time-course of alkyl-[SaP]PAand alkyl-SaP]PEtforFig. 1, in the presence of ethanol, stimulated neutrophils mation by fMLP-stimulated neutrophils in the absence and of cytochalasin B. Washed human neutrophils were produce, in addition to a l k ~ l - [ ~ ~ P ] P A , a l k ~ l - [ ~ ~ This P ] P Epresence t. a l k ~ l - [ ~ ~ P ] Pformation, Et like a l k ~ l - [ ~ ~ P ] formation, PA is labeled in alkyl-PC with 32Pby incubation with alkyl-[32P]lysoPC. samples (0.5 ml) containing 5 X lo6 labeled neutrophils greatly augmented by pretreatment of neutrophils with cyto- Duplicate and 1.5 mM CaClz were incubated with (0,0 ) or without (M) 5 HM chalasin B and occurs rapidly, reaching a maximum within cytochalasin B before adding either buffer (0)or 100 nM fMLP (0, 30 s. However, in contrast t o a l k ~ l - [ ~ ~ P ] P A , a l k ~ l - [ ~ ~B) P ]for P Evarious t times (0-5 min) in the absence (upper panel) or the levels continue to increase, albeit at a slower rate, after the presence (middle and lower panels) of 0.5% ethanol. Alkyl-PA and initial peak. These data suggest that, in neutrophils, alkyl- alkyl-PEt were separated by TLC using solvent system I and the PA is rapidly converted to secondary products (see below) radioactivity quantified by liquid scintillation spectrometry. Data are the average of two determinations, which were within 5% whereas alkyl-PEt is metabolically stable. Cytochalasin B by points of the mean. For additional details, see “Experimental Procedures.” itself induces production of neither a l k ~ l - [ ~ ~ P ]nor P Aalkyl[32P]PEt. Furthermore, whenaddedindividually, neither adyl-PC) account for half of the total cellular phospholipid fMLP nor ethanol indu.ces alkyl-[32P]PEtformation. Furthermore, alkyl-[j2P]PEt formation is accompanied by content and consist of ether-linked (alkyl-PC and alkenylreduced accumulation of a l k ~ l - [ ~ ~ P ] PHowever, A. the com- PC) andester-linked (diacyl-PC) species (33,34).Of the total bined accumulation of a l k ~ l - [ ~ ~ P ]and P A a l k ~ l - [ ~ ~ P ] PinE t diradyl-PC, alkyl-PC and alkenyl-PC constitute 45 and 5%, Alkali ethanol-treated cells actually exceeds the amount of alkyl- respectively, and the remainder is diacyl-PC (33, 34). [32P]PAdetected in the absence of ethanol and is equivalent treatment completely deacylates ester-linked (diacyl-) phosto about 4% of the total [32P]PC content. Thus,the reduced pholipids to produce water-soluble phosphorus-containing accumulation of a l k ~ l - [ ~ ~ P is ] PaAreflection of redirection of products whereas ether-linked (1-0-alkyl- and 1-0-alkenyl-) are deacylated only at the sn-2 position to the putative phosphatidyl-PLD intermediate to a l k ~ l - [ ~ ~ Pphospholipids ] To determine whether esterPEt rather than inhibition of PLD. By several other criteria producelysophospholipids. including fMLP dose rlesponse, Ca2+requirements and inhi- linked PC is also hydrolyzed by PLD, PA, PEt, and PC have bition by an fMLP receptor antagonist (data not shown), been isolated from neutrophils before and after fMLP stimPhospholipid masses before (total) and after (etheralkyl-[32P]PEtsynthesis is closely correlated with a l k ~ l - [ ~ ~ Pulation. ] PA formation, as shown previously for HL-60 granulocytes linked) NaOH treatment have then been quantified by phos(5, 6). Together these results demonstrate that neutrophil phorus estimation. As reported previously (33,34),neutrophil PLD is capable of catalyzing the transphosphatidylation re- PC is composed of diacyl-PC (52%)and ether-linked PC action. (48%) in equal amounts (Table I). This distribution remains Choline-containingPhosphoglyceridesAre the Preferred unchanged during fMLP stimulation. PA and PEt present in Substrates for PLD in fMLP-stimulated Neutrophils-In hu- unstimulated cells are primarily of the diacyl-type (80-90%). man neutrophils, choline-containing phosphoglycerides (dir- However, net PA formed after stimulation (0.71 nmol/107

.r;

I .

I

Diglyceride Production by Phospholipase DIPhosphatidate Phosphohydrolase

17072

TABLE I Distribution of sn-1 bondsin PC, PA,and PEt in humanneutrophils Samples (75 ml) containing 3H-labeledneutrophils (7.5 X 10’ cells), 1.5 mMCaC12, and 5 p~ cytochalasin B were incubated for 5 min before adding buffer (control) or 100 nM fMLP for an additional 30 s in the presence of 0.5% ethanol. Following separation by TLC, PC, PA, and PEt were eluted from silica and analyzed for the distribution of the sn-1 bonds (ester versus ether) by alkali hydrolysis and phosphorus estimation. Values wereobtained by subtracting silica controls and therefore, represent the total amounts of individual phospholipids. For further details, see “Experimental Procedures.” EtOH + fMLP

EtOH

Phospholipid

TotalEther-linked

Diacyl- TotalEther-linked

nmolll0’ cells

PC

PA PEt

31.7 0.15 0.07

15.0 0.03 0.01

Diacyl-

nmoll10’ cells

16.7 29.6 0.12 0.86 0.06 0.72

14.2 0.31 0.31

15.4 0.55 0.41

TABLEI1 Contents, specific activities and 3H/32Pratios of double-labeled PC, PA, and P E t in human neutrophils Washed human neutrophils were double-labeled in alkyl-PC with 3H and 32P by incubation with [3H]alkyl-lysoPC and alkyl-[32P]lysoPC. Triplicate samples (5 ml) containing these double-labeled cells (5 X lo7 cells) and 1.5 mM CaC12were incubated with 5 p M cytochalasin B for 5 min before adding buffer (control) or 100 nM fMLP for 30 s in the absence or the presence of 0.5% ethanol. PC, PA, and PEt containing 3H and 32Pwere separated by TLC and analyzed for mass and radioactivities as described under “Experimental Procedures.” The PA and PEtvalues for mass and radioactivities of unstimulated samples were subtracted from the values for the appropriate parametersof stimulated samples to obtain the tabulated net values (& S.D.). Parameters

Vehicle PC

fMLP PA

fMLP + EtOH PA

PEt

Mass (nmol/107 cells) 30.0 & 0.6 1.07 & 0.01 0.71 & 0.03 0.37 2 0.04 3H (dpm/nmol) X 22.6 & 0.220.8 & 1.4 26.8 & 2.620.0 & 0.4 10-3 32P(dpm/nmol) X 16.8 & 0.718.0 & 0.1 19.6 k 1.1 15.0 f 0.4 10-3 1.33 3H/32Pratio 1.37 1.16 1.35

cells) is comprised of 60% diacyl-PA (0.43 nmol/107 cells) and 40% ether-linked PA (0.28 nmol/107 cells), whereas net PEt formed after stimulation (0.65 nmol/107 cells) contains 52% diacyl-PEt (0.35 nmol/107 cells) and 48% ether-linked PEt (0.30 nmol/107 cells) (Table I). Amounts of phospholipid phosphorus for PC, PA, and PEt remaining after NaOH alkyl-PEt formed are derived from alkyl-PC exclusively via treatment do not differ from those obtained after sequential PLD. Because alkyl-PA and alkyl-PEt generation represents treatments with NaOH and HCl (data not shown),suggesting diradyl-PA and diradyl-PEtformation (Tables and I 11),these that theether-linked species is virtually devoid of 1-0-alkenyl results further suggest that diacyl-PA and diacyl-PEt are also bond. These results demonstratethat fMLP-stimulated PLD formed primarily via PLD action on diacyl-PC. hydrolyzes both diacyl-PC and alkyl-PC to generate the rePLD-derived Alkyl-PA Is Degraded by PA Phosphohydrospective PA and PEt species and that the relative distribu- lase to Produce Alkyl-DG-Neutrophils stimulated with fMLP tions of sn-1 ester- and ether-bonds in PA and PEt formed generate DG in large quantities (8, 13-15, 17), and this bulk during stimulation are very similar to thatin PC. DG comes from sources other thanphosphatidylinositol (13PLD-catalyzed Hydrolysis of PC Is the Primary Route to PA 17). Recent studies have suggested PC as the origin of this and PEtin fMLP-stimulated Neutrophils-In stimulated neu- DG in fMLP-stimulated human neutrophils(6,8). DG can be trophils, PA and PEtcould be formed by pathways other than derived from PC either by phospholipase C or by sequential PLD action onPC. These alternative pathwaysmight include actions of PLD andPA phosphohydrolase. To evaluate these phospholipase C followed by DG kinase and/or de mvo syn- alternatives, endogenous alkyl-PC has been labeled with 3H thesis from glycerol. In neutrophils, fMLP causes increased by preincubation of cells with [3H]alkyl-lysoPC. Upon stimincorporation of exogenous [32P]P04 intoPA via DG kinase ulation with fMLP, these 3H-labeled neutrophilsgenerate (9, 16). To determine the relative contribution of PLD acti- [3H]alkyl-PA (Fig. 2). As observed with [32P]alkyl-PAgenervation to PA and PEt formation, neutrophils double-labeled ation (Fig. I), formation of [3H]alkyl-PAis rapid and is greatly by incubation with [3H]alkyl-lysoPC and alkyl-[32P]lys~PC enhanced by cytochalasin B pretreatment (Fig. 2). In the were stimulated with fMLP for 30 s. Mass measurements of absence of cytochalasin B, stimulated neutrophils produce double-labeled PC, PA, and PEt show that the specific ac- little or no [3H]alkyl-DG. In the presence of cytochalasin B, tivities of 3H and 32Pin the total PA and PEtformed during fMLP also induces formation of [3H]alkyl-DG. Within 30 s stimulation are very similar to each other and tothose of PC after fMLP addition to cytochalasin B-treated neutrophils, (Table 11).These data furtherconfirm PC as the major source the [3H]alkyl-PAlevel reaches a maximum, but no increase of PA and PEt and negate a major contribution of de novo in [3H]alkyl-DG is detected. Recoverable [3H]alkyl-PA then synthesis and/or phospholipase C utilizing phospholipids declines and this decline parallels generation of [3H]alkylother thanPC. Significant contributions by these mechanisms DG. These results demonstratethat infMLP-stimulated neuwould have lowered the specific activities of PA and PEt trophils, alkyl-DG formation from alkyl-PC occurs subsebelow that of PC. Furthermore, both by mass measurements quent to alkyl-PA generation. This DG formation correlates and by quantitation of 3H and 32P,the combined formation with the decline in [3H]alkyl-PAand, like alkyl-PA formation, of P A and PEtis equivalent to 3-4% of the total PCcontent, requires cytochalasin B. Formation of [3H]alkyl-PAand [3H] suggesting that radiolabeled PA and PEtformation represents alkyl-DG shows identical dependence on fMLP concentrathe net mass accumulation of these products in fMLP-stim- tions and is similarly inhibited by the fMLP receptor antagonist, t-BocMet-Leu-Phe(Fig. 3). In addition, inthe presence ulated neutrophils. Although alkyl-[32P]PAis formed exclusively byPLD, [3H] of EGTA and no added Ca2+,cytochalasin B-treated neutroalkyl-PA could also come from DG kinase-catalyzed phos- phils stimulated with fMLP produce neither [3H]alkyl-PA phorylation of [3H]alkyl-DG generated by phospholipase C nor [3H]alkyl-DG (data not shown). Both [3H]alkyl-PA and action on [3H]alkyl-PC.If the DG kinase pathway were to be [3H]alkyl-DGare formed simultaneously when Ca2+is added contributed significantly to the formation of VHIalkyl-PA, in excess of EGTA, although Ca2+ by itself is ineffective. the 3H/32P ratio of alkyl-PA in double-labeled neutrophils Formation of these products shows identical dependence on would be higher than thatof alkyl-PC. As shown in Table 11, Ca2+ concentrations. Maximum formation occurs at 100 FM the 3H/32P ratios of alkyl-PA and alkyl-PEt formed 30 s after of free Ca2+and remains unchanged at least up to 1 mM of stimulation with fMLP are identical to that of alkyl-PC. free Ca2+. These observations suggest aprecursor-product These data suggest that, at early times, the alkyl-PA and relationship between alkyl-PA and alkyl-DG involving PA

Diglyceride ProductionPhospholipase by

17073

DIPhosphatidute Phosphohydrolase 50000

15000

ALKYL-PA

ALKYL-PA 40000

-

30000

-

10000

20000 -

5

P 0

5000 10000

i k > F 0

a 0 0 a r FJ

-

0

30000

IOWO 20000

5000

10000

o

!

.

, 0

0 0

GO

120

180

240

300

TIME, seconds

FIG. 2. Time course osf ['Hlalkyl-PA and ['Hlalkyl-DG formation by fMLP-stimulated neutrophils in the absence and the presence of cytochalasin B. Human neutrophils were labeled

in alkyl-PC with 3H by incubation with [3H]alkyl-lysoPC.Duplicate samples containing 5 X lo6 [3H]alkyl-PC-labeledcells and 1.5 mM CaC12were incubatedin.theabsence (0)or presence (0)of 5 p M cytochalasin B for 5 min before adding 100 nM fMLP for the additional times (0-5 min). ['HIAlkyl-PA and [3H]alkyl-DG levels remained unchanged throughout the incubation inthe absence of either cytochalasin B or fMLP.

phosphohydrolase. Con.sistent with alkyl-PA being the immediate precursor of alkyl-DG is the observation that fMLP stimulates ethanol-treated cells to produce alkyl-PEt with a concomitant reduction of both alkyl-PA (Fig. 1 and Table 11) and alkyl-DG (data not shown). To furtherevaluate the role of PA phosphohydrolase in the generation of alkyl-DG, the effects of propranolol, a phosphohydrolase inhibitor (35, 36), on the formation of various products of fMLP-stimulated neutrophilshave been examined (Fig 4). In theabsence of fMLP, propranolol has no effect on basal levels of [3H]alk,yl-DGbut causes small increases in [3H]alkyl-PA. In fMLP-stimulated neutrophils, propranolol inhibits the formation 'of [3H]alkyl-DG in a dose-dependent manner (ICsovalue = 70 pM) with concomitant increments in [3H]alkyl-PA accumulation. Propranolol concentrations (150-200 p ~ that ) cause virtually complete inhibition of [3H] alkyl-DG formation, inhibit [3H]alkyl-PEt synthesis by less than 10%. Therefore, propranolol increases PA accumulation in fMLP-stimulated neutrophils by inhibiting the phosphohydrolase action on PL:D-derived PA. PC Is Not Hydrolyzed by Phospholipase C-Although propranolol inhibits neither PLD (Fig. 5) nor phosphoinositidespecific phospholipase C (37), a putative PC-specific phospholipase C (38,39) could still be inhibited by propranolol. It has, therefore, been essential to identify and measure various

.

, 20

.

, 40

.

, 60

.

, 80

.

,

.

100

l 120

fMLP, nM

FIG. 3. Dose-response of ['Hlalkyl-PA and ['Hlalkyl-DG formation by fMLP-stimulated neutrophils inthe absence and the presence of t-BocMet-Leu-Phe. Duplicate samples containing 5 X lo6 [3H]alkyl-PC-labeledcells, 1.5 mM CaC12,and 5 pM cytochalasin B were incubated with buffer (0)or with 10 pM t-BocMet-LeuPhe (0)for 5 min before stimulating with various concentrations of fMLP for 3 min.

putative products of phospholipase C and PLD-phosphohydrolase pathways. The PA degradation by phosphohydrolase should resultinaconcurrent generation ofDG and Po4 whereas PC hydrolysis by a phospholipase C would form DG and phosphocholine in parallel. In order to simultaneously measure these products, experiments have been performed using human neutrophils that have been double-labeled in alkyl-PC by incubationwith [3H]alkyl-lysoPC and alkyl[32P]-lysoPC.When stimulated withfMLP in thepresence of cytochalasin B, these double-labeled neutrophils produce [3H] alkyl-DG with a concomitant increase in 32Pradioactivity in the aqueous phase (Fig. 5 ) . Analysis of the aqueous phase by TLC hasrevealed that the [32P]P04 content increases with a time course that is virtually identical to that of total 32P increase in theaqueous phase and of [3H]alkyl-DGformation. The [32P]phosphocholinecontent increases only slightly upon stimulation. Thus, in stimulated neutrophils, [3'P]P04 is by far the major water-soluble product (>go%) with [32P]phosphocholine accounting for a relatively minor fraction(40%). Propranolol inhibitsincreases in [3H]alkyl-DG,[32P]P04, and total aqueous 32P.The [32P]phosphocholinecontent, however, is not affected by propranolol. The fact that propranolol inhibits [32P]POr formation without increasing the levels of [32P]phosphocholinerules out the possibility that [32P]phosphocholine formed during stimulation might have been degraded rapidly by a putative phosphocholine phosphatase to produce choline and [32P]P04. The 3H$zP ratio of PA formed during fMLP stimulation increases with time (Fig. 6, lower

Diglyceride Production by Phospholipase DIPhosphatidatePhosphohydrohe

1

12000

i

ALKYL-DG

0

60

0

120

180

360 0

2300 40

ALKYL-PA

/” I f oc 1

lom:

5000

-

loax,

1

.

.

0

‘Oo0

ALKYL-PEt

.

.

.

.

.

.

.

.

.

.

.

e

PHOSPHATE

50

350 150

8000

-

4000

-

250

n PROPRANOLOL, uM

FIG. 4. Effects of propranolol on the formation of [‘HlalkylPA, [‘Hlalkyl-DG, and [‘Hlalkyl-PEt by fMLP-stimulated neutrophils. Duplicate samples (0.5 ml) containing 5 X lo6 [3H] alkyl-PC-labeled neutrophils, 1.5 mM CaC12,and 5 p~ cytochalasin B were incubated for 5 min with the indicated concentrations of propranolol before adding either buffer (0)or 100 nM W L P (0)for 3 min in the absence (for alkyl-DG and alkyl-PA) or the presence of 0.5% ethanol (for alkyl-PEt). Propranolol was dissolved in dimethyl sulfoxide and then diluted into Hepes-saline buffer. This diluted solution was sonicated in a bath sonicator before transferring to the assay tubes. [3H]Alkyl-DG, [3H]alkyl-PA, and [3H]alkyl-PEt were separated by TLC as described under “Experimental Procedures.”

panel), demonstrating that at later times, [3H]alkyl-PA is formed from [3H]alkyl-DGvia the DG kinase pathway. However, the increases in the 3H/32P ratio are completely prevented by propranolol, suggesting that the[3H]alkyl-DGgenerated by phosphohydrolase actionon [3H]alkyl-PA is rephosphorylated by DG kinase. These results clearly demonstrate that [3H]alkyl-DG is formed almost exclusively by sequential actions of PLD and phosphohydrolase on [3H] alkyl-PC and thatin fMLP-stimulated neutrophils, alkyl-PC is not hydrolyzed by phospholipase C. Phosphohydrohe Activity Appearsafter PLD ActivationThe onset of PO, and DG formation lags behind that of PA formation in fMLP-stimulated neutrophils (Fig. 5). This lag may be due either to inadequate phosphohydrolase activity to convert the PA or to rapid clearance of POr andDG. To test these possibilities, the time courses of PA formation by fMLPstimulated cells have been followed in the absence and the presence of 200 PM propranolol which completely inhibits the accumulation of [3zP]P04 and[3H]alkyl-DG (Fig. 5). During

. . _ . _ . . _ . . . . . PHOSPHOCHOUNE

0

60

120

180

240

300

-I 360

~

TIME, seconds

FIG. 5. Time courses for formation of [‘Hlalkyl-DG, [“PI PO4, and [‘zP]phosphocholine by fMLP-stimulated neutrophils. Duplicate samples containing double-labeled neutrophils, 1.5 mM CaClz and 5 P M cytochalasin B were incubated with (0,W) or without (0)200 p~ propranolol for 5 min before adding either buffer (0)or 100 nM W L P , (0,W) for various times (0-5 min). PA and DG were separated by TLC using the double-development method. Aliquots of the aqueous phase were counted for the total water-soluble 32Pproducts. The aqueous phase (500 pl) was further analyzed for [3zP]P04 and[3P]phosphocholine by TLC as described under “Experimental Procedures.” For other conditions, see Fig. 1 and “Experimental Procedures.”

the lag period (first 30 s), propranolol treatment causes no additional increments in PA (Fig. 6), suggesting lack of substantial phosphohydrolase activity during this time, and not rapid clearance of the products. Only at later times, when DG and PO, formation occurs in detectable amounts (Fig. 5), does propranolol treatment increase P A accumulation (Fig. 6). Theseresults suggest that phosphohydrolase activity lags behind PLD activation leading to theinitial build-up of PA. PLD-phosphohydrohe Is the Predominant Route to DG Formation in fMLP-stimulatedNeutrophils-The relative amounts of alkyl-DG and diacyl-DG formed during fMLP-

Diglyceride Production by Phospholipase DIPhosphatidate Phosphohydrolase 16000

17075

500

1

400

4mi t

3

-I

W

0 r-

z

300

0

E

”n

200

3

PROPRANOLOL, ut4

FIG.7. Effect of propranolol on total DG formation by fMLP-stimulated neutrophils. Duplicate samples (0.5 ml) containing 5 X lo6 neutrophils, 1.5 m M CaC12,and 5 p~ cytochalasin B were incubated with variousconcentrations of propranolol for 5 min before addingeither buffer (0)or 100 nM fMLP (0)for an additional 3 min. DG mass was determined by enzymatic conversionto [32P]PA, as described under “ExperimentalProcedures.” mass measurements clearly demonstrate that generation of alkyl-PA, alkyl-PEt, andalkyl-DG represents that of diradylPA, diradyl-PEt, and diradyl-DG, respectively (Tables I and 11).Furthermore, activitiesof PLD andphosphohydrolase are TIME, seconds not dictated by the nature of bonds at the sn-1 position of FIG.6. Time course of double-labeled PA formation by phospholipid substrates (see below) These conclusions valifMLP-stimulated neutrophils in the absence and the presence date the selective labeling of alkyl-PC as an approach to of propranolol. Duplicatesamples (0.5 ml) containing double-labeled neutrophils(5 X lo6cells), 1.5 mM CaC12,and 5 pM cytochalasin investigating PLD-mediated lipid metabolism in stimulated B were incubated with (0,m) or (0,O) without 200 p~ propranolol neutrophils. PLD activation and subsequent DG generation require for 5 min before adding either buffer ( 0 , O ) or 100 nM fMLP (0,m) for various times (0-5 rnin.).Duplicate determinations were within + cytochalasin B (Figs. 1 and 2). Several other neutrophil re2% of the mean. sponses including degranulation (40), respiratory burst (1417), phospholipase AP activation (41), and protein kinase C stimulation of human neutrophils are very similar to that of translocation (42) are also greatly potentiated by cytochalasin alkyl-PC and diacyl-PC, respectively (8), suggesting PC as a B. These responses are clearly elicited in vivo possibly due to source of DG. To determine the relative contribution of PLD- exposure of neutrophils to multiple agonists either sequenphosphohydrolase pathway to the total DG generation by tially or simultaneously. In vitro,such heterologous exposures fMLP-stimulated neutrophils, the diradyl-DG mass has been might obviate the need for cytochalasin B. Indeed, 5-hydroxquantified in the absence and the presence of propranolol. yeicosatetraenoic acid, a naturallyoccurring mediator, mimics Propranolol, by itself, has no effects on basal levels of DG cytochalasin B in augmenting agonist-induced degranulation mass (Fig. 7). However, propranolol causes inhibition of (43) and phospholipase Az activation (44). Thus, cytochalasin fMLP-induced increases in DG mass at concentrations that B-treated neutrophils could serve as a model for the study of also inhibit conversion of radiolabeled PA to DG and PO, receptor-mediated changes in lipid metabolism and their rel(Figs. 4 and 5). These results suggest that notonly alkyl-DG evance to neutrophil functions. Although cytochalasin B is but also diacyl-DG is generated by the sequential actions of known to perturb thecytoskeletal organization in neutrophils PLD andphosphohydrolase and that [32P]P04 and [3H]alkyl- (45), the actual mechanism of its priming effects on neutrophil DG represent net increases in total DG mass. Also note that responses has yet to be identified. the net accumulation of PA mass (Table 11) far exceeds the The present results extend our previous studies in HL-60 net increases in DG production (Fig. 7). granulocytes (5, 6) and demonstrate that human peripheral neutrophils, like HL-60 granulocytes, can be selectively laDISCUSSION beled in PC with 32Pin the absence of [32P]ATP.Using this Ether-linked lysophospholipids are metabolically more sta- labeling approach, it is conclusively demonstrated that, in ble than theirester-linked counterparts because the latter are cytochalasin B-treated neutrophils, fMLP stimulates a PLD readily hydrolyzed by intracellular lysophospholipases. Thus, activity that hydrolyzes PC to produce PA (Fig. 1, Tables I the use of labeled alkyl-lysoPC provides an efficient means of and 11). The data further demonstrate that this receptorlabeling endogenous PC. Specific activity determinations and mediated PLD, like PLD activities from other sources (46,

17076

Diglyceride Production by Phospholipase DIPhosphutidate Phosphohydrolase

47), is also capable of catalyzing transphosphatidylation between PC and ethanol toform PEt (Fig. 1, Tables I and 11). Similar conclusions have been drawn for fMLP-stimulated PLD activity inHL-60 granulocytes (5), emphasizing the utility of HL-60 cells as a granulocyte model for the study of PLD. Recent studies in HL-60 granulocytes demonstrate that phorbol diesters, synthetic diglycerides, and the Ca2+-ionophore A23187 activate PLD activities that act on endogenous alkyl-PC to generate both alkyl-PA and alkyl-PEt (7). However, NG108-15 neuroblastoma cells treated with phorbol diesters produce only PEt and not PA (48), raising the possibility that hydrolysis (PA formation) and transphosphatidylation (PEt formation) might be catalyzed by distinct PLD activities. Phospholipid analysis (Table I) and comparative specific activity determinations (Table 11) reveal that alkyl- and diacyl-linked PC species are hydrolyzed equally by PLD. Neutrophil PLD might also act on phosphatidylinositol (49) and/ or phosphatidylethanolamine (50). However, the distribution of sn-1 bonds in PA and PEt resembles that of PC (Table I) and not phosphatidylinositol or phosphatidylethanolamine (33,34). Moreover, previous studies in neutrophilshave demonstrated that the fatty acid composition of PA formed during fMLP-stimulation differs substantially from that of phosphatidylinositol (13) but is very similar to thatof PC (51). Thus, phospholipids other than PC are degraded only poorly, if at all, by PLD in fMLP-stimulatedneutrophils. Apparently, the substrate specificity for fMLP-stimulated PLD is determined not by the composition of the sn-1 bonds but by the nature of the phospholipid base. PLD activities from other sources including rat brain microsomes (52, 53) and rat liver membranes (10) also prefer endogenous PC as substrate. Immediately following stimulation, both PA and PEt are formed almost exclusively by PLD (Table I1 and Fig. 6). A t later times, up to about 20% of the PA is generated by DG kinase-catalyzed phosphorylation of DG, as indicated by the increments in the 3H/32P ratio (Fig. 6). However, even this DG is derived by phosphohydrolase from PLD-derived PA because inhibition of the phosphohydrolase by propranolol blocks the increments in 3H/32Pratios (Fig. 6). Therefore, at no time following stimulation does the phospholipase C/DG kinase pathway contribute noticeably to theincrease in PA. Both by fatty acid distribution (13) and by distribution of sn-1 bonds (8), the bulk DG produced during fMLP stimulation resembles PC more closely than itdoes phosphoinositides or phosphatidylethanolamine (33, 34). Evidently, PC is the major source of DG formed during neutrophil stimulation. However, there is noevidence for PC hydrolysis by phospholipase C in fMLP-stimulated cells (Fig. 6). On the contrary, our present data clearly establish that analternative pathway involving sequential breakdown of PC by PLD andPA phosphohydrolase functions in stimulated cells. PC hydrolysis by PLD generates PA inamounts (Table 11)sufficient to account for all of the DG formed (Fig. 7). Furthermore, the phosphohydrolase inhibitor propranolol does not inhibit phosphoinositide-specific phospholipase C stimulated by fMLP2 andyet propranolol causes virtually complete inhibition of DG accumulation (Fig. 7). Therefore, it must be concluded that the combined activities of PLD andphosphohydrolase constitute the major pathway for DG formation in fMLP-stimulated neutrophils. Such a mechanism may also be functional in other cell systems where PC has been identified as the source of DG (18-23). In rathepatocytes, vasopressin which activates PLD M. M. Billah, S. Eckel, T.J. Mullmann, R. W. Egan, and M. I. Siegel, unpublished observations.

(lo), causes both translocation and activation of phosphohydrolase (54). In various cells, unsaturated fattyacids activate phosphodrolase by translocating the enzyme to themembrane (54, 55). The possibility that, in neutrophils, fatty acids released during stimulation (41) as well as PLD-derived PA itself might trigger phosphohydrolase translocation awaits further research. In summary, selective 32P-labeling of endogenous PC of human neutrophilshas unequivocally demonstrated receptorlinked activation of PLD, which catalyzes both hydrolysis and transphosphatidylation to generate PA and PEt, respectively. PC is the cellular substrate for neutrophil PLD, and in fMLP-stimulatedneutrophilsPA and PEt are formed almost exclusively byPLD. ThePA is then dephosphorylated to DG byPA phosphohydrolase. This pathway involving PLD and phosphohydrolase is responsible for the large, sustained accumulation of DG by stimulated neutrophils,and under the presentexperimental conditions, PC is not hydrolyzed by phospholipase C. Both PA and DG have been implicated in important neutrophil functions such as degranulation and superoxide anion formation (13,17, 56,57). While the central role of phosphohydrolase in de mu0 lipid synthesis is well established (12), our presentdata are the first evidence for its major involvement in signal transduction mechanisms. Acknowledgments-We thank Dr. W. R. Bishop for helping us with DG determinations, Dr. Jim Pai for discussions, and Lisa Ramirez for typing the manuscript. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22, 23. 24. 25.

Michell, R. H. (1975) Biochim. Biophys. Acta 416,81-147 Lapetina, E. G . (1982) Trends Pharmacol Sci. 3 , 115-118 Nishizuka, Y.(1986) Science 233,305-312 Berridge, M. J. (1984) Bwchem. J. 220,345-360 Pai, J.-K.,Siegel, M. I., Egan, R. W., and Billah, M. M. (1988) J. Biol. Chem. 263,12472-12477 Pai, J.-K., Siegel, M. I., Egan, R. W., and Billah, M. M. (1988) Biochem. Biophys. Res. Commun. 160,355-364 Billah, M. M., Pai, J.-K., Mullmann, T. J., Egan, R. W., and Siegel, M. I. (1989) J. Biot. Chem. 264,9069-9076 Agwu, D. E., McPhail, L. C., Chabot, M. C., Daniel, L. W., Wykle, R. L., and McCall, C. E. (1989) J. Biol. Chem. 264,1405-1413 Cockcroft, S. (1984) Biochim. Biophys. Acta 796,37-46 Bocckino, S. B., Blackmore, P. F., Wilson, P. B., and Exton, J. H. (1987) J. Biol. Chem. 262,15309-15315 Farese, R. V., Konda, T. S., Davis, J. S., Standaert, M. L., Pollet, R. J., and Cooper, D. R. (1987) Science 236,586-589 Brindley, D. N. (1984) Prog. Lipid Res. 23, 115-133 Cockcroft, S., and Allan, D. (1984) Biochem. J. 222,557-559 Honeycutt, P. J., and Niedel, J. E. (1986) J. Biol. Chem. 2 6 1 , 15900-15905 Rider, L. G . , and Niedel, J. E. (1987) J. Bwl. Chem. 2 6 2 , 56035608 Reibman, J., Korchak, H. M., Vosshall, L. B., Haines, K. A., Rich, A. M., and Weissmann, G. (1988) J. Biol. Chem. 63226328 Truett, A. P., Verghese, M. W., Dillon, S. B., and Snyderman, R. (1989) Proc. Natl. Acad. Sci. U. S. A. 85,1549-1553 Irving, H. R., and Exton, J. H. (1987) J. Biol. Chem. 262,34403443 Besterman, J. M., Duronio, V., and Cuatrecasas, P. (1986) Proc. Natl. Acad. Sci. U. S. A. 83,6785-6789 Muir, J. G., and Murray, A. W. (1987) J . Cell. Physiol. 130,382391 Cabot, M. C., Welsh, C. J., Zhang, Z-C, Cao, H-L, Chabbott, H., and Lebowitz, M. (1988) Bwchim. Biophys. Acta 959,46-57 Welsh, C. J., Cao, H-L, Chabbott, H., and Cabot, M. C. (1988) Biochem. Bwphys. Res. Commun. 152,565-572 Slivka, S. R., Meier, K. E., and Insel, P. A. (1988) J. Biol. Chem. 263,12242-12246 Martin, T. W. (1988) Biochim. Biophys. Acta962,282-296 Boyum, A. (1968) S c a d . J. Clin. Lab. Znuest. 21 (suppl. 97) 7789

17077

Diglyceride Production by Phospholipase DIPhosphatidate Phosphohydroluse 26. Bligh, E. G., and Dyer, W. J. (1959) Can. J. Biochem. Physiol. 37,911-917 27. Brown, J. L., and Johnston, J. M., (1972) J . LipidRes. 16, 480481 28. Skipski, V. P., Peterson, R. F., and Barclay, M. (1964) Biochem. J. 90, 374-378 29. Psi, J.-K, Liebel, E. Tottenborn, c. s.9 IkeW’Jonu, F. I., and Mueller, G. C. (1987) Carcinogenesis 8, 173-178 30. Chen, p. s., Jr., Torbara, T. A., and Warner, H. (1956) A d . Chem. 28,1756-1758 31. Preiss, J., Loomis, C. R., Bishop, W. R., Stein, R., Niedel, J. E., and Bell, R. M. (1986) J. Biol. Chem. 261,8597-8600 32. Demopoulos, C. A., Pinkard, R. N., and Hanahan, D. J. (1979) J. Biol. Chem. 254,9355-9358 33. Mueller, H. W., O’Flaherty, J. T., andWykle, R. L. (1982)Lipids, 17,72-77 34. Mueller, H. W., O’Fl.aherty, J. T., Green, D. G., Samuel, M. P., and Wykle, R. L. (11984)J. Lipid Res. 2 5 , 383-388 35. Pappu, A. S., and Hauser, G. (1983) Neurochem. Res. 8, 15651575 36. Koul, O., and Houser, G. (1987) Arch. Biochem. Biophys. 253, 453-461 37. Das, I., De Belleroche, and Hirsch, S. (1988) Prog. NeuroPsychopharmacol. Biol. Psych.12, 721-726 38. Wolf, R. A., and Gross, R. W. (1985) J. Biol. Chem. 260, 72957303 39. Clark, M. A., Shorr, R. G. L., and Bomalaski, J. S., (1986) Biochem. Biophys. Res. Commun. 4 0 , 114-119 40. Hoffstein, S., Goldstein, I. M., and Weissmann, G. (1977) J. Cell Biol.73,242-256 (z.9

41. Mead, M. J., Turner, A. G., and Bateman, P. E. (1986) Biochem. J. 238,425-436 42. Horn, w.7 and Karnovsky*M. L. Iliochem. Biophys. Res. Commun. 139,86-95 43‘ o’F1aherty, J’ T‘l J.’ Hammet, M. J’’ ”’ McCall, C. E., and Wykle, M’ R. L. (1983) Biochem. Biophys. Res. Commun. 111 , l - 7 44. Siegel, M. I., McConnel, R. T., Bonser, R. w., and Cuatrecasas, P. (1982) Biochem. Biophys. Res. Commun. 104,874-881 45. White. J. R.. Naccache. P. H.. and Sha’afi. R. I. (1983) J. Biol. Cheh. 258,14041-14047 46. Heller, M. (1978)Adu. Lipid Res. 16, 267-326 47. Kanfer, J. N. (1980) Can. J. Biochem. 5 8 , 1370-1380 48. Liscovitch, M.(1989) J. Biol. Chem. 264, 1450-1456 49. Balsinde. J.. Diez. E.. and Mollinedo. F. (1988) Biochem. BiorJhvs. Res. Commun. 154,502-508 50. Kiss. Z.. and Anderson. W. B. (1989) . . J. Biol. Chem. 264. 14831487 ’ 51. Walsh, C. E., DeChatelet, L., Chilton, F. H., Wykle, R. L., and Waite, B. M. (1983) Biochim. Biophys. Acta 750,32-40 52. Chalifour, R. J., and Kanfer, J. N.(1980) Can. J. Biochem. 5 8 , 1189-1196 53. Kobayashi, M., and Kanfer, J. N. (1987) J. Neurochem. 48, 1597-1603 54. Pollard, A. D., and Brindley, D. N. (1984) Biochem. J. 217,461469 55. Hopewell, R., Martin-Sanz, P., Martin, A., Saxton, J., and Brindley, D. N. (1985) Biochem. J. 232,485-491 56. Korchak, H. M., Vosshall, L. B., Zagon, G., Ljubich, P., Rich, A. M., and Weissmann, G . (1988) J. Biol.Chem. 2 6 3 , 1109011097 57. Bellavite, P., Corso, F., Dusi, S., Grzeskowiak, M., Della-Bianca, V., and Rossi, F. (1988) J. Biol. Chem. 263, 8210-8214 ’

I~

I