Protein Kinase C, Mediates Phospholipase D Activation by ...

Vol. 269,No. 14,Issue of April 8,pp. 10511-10516, 1994 Printed in U.S.A.

THEJOURNALOF BIOUCICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Protein Kinase C, Mediates PhospholipaseD Activation by Nucleotides and Phorbol Ester in Madin-Darby Canine Kidney Cells STIMULATION OF PHOSPHOLIPASE D IS INDEPENDENT OF ACTIVATION OF POLYPHOSPHOINOSITIDE-SPECIFICPHOSPHOLIPASE C AND PHOSPHOLIPASE &* (Received forpublication, December 20, 1993, and in revised form, January 24, 1994)

Maria A. Balboa, Bonnie L. Firestein, Catherine Godson*, Kelly S . Bell, and PaulA. Insel§ From the Department of Pharmacology 0636, University of California at Sun Diego, La Jolla, California 92093-0636

lipids, producing phosphatidic acid and a free polar headgroup (1).Phosphatidic acid has been suggested to act as a second messenger in certain cell types (2, 3) and to induce nuclear events and synthesisof cytokines (4-6). In addition, phosphatidic acid has been shown to specifically stimulate phospholipase C,1 (PLC,l) (7) and phosphatidylinositol-4-phosphate kinase (81, and in liver cells it is thought to activatea protein kinase (9). In the presence of a primary alcohol, PLD catalyzes a transphosphatidylation reaction in which the phosphatidyl moiety of the substratephospholipid is transferredto the alcohol so as to produce the corresponding phosphatidylalcohol(1, 10, 11).This unique property hasbeen widely used to identify PLD activity in manycell systems (12). Considerablebiochemical evidence suggests the existence of multiple PLD isoenzymes, each oneprobablyplayingdifferentroles insignal transduction (6, 13-15). However, the mechanisms by which receptor agonistsactivate PLD remain poorly defined.Evidence for the involvement of GTP-binding proteins includes the inhibitory effects of pertussis toxin treatment of intact cells, and the formation of phosphatidylethanol in response t o nonhydrolyzable GTP analogues in cell-free systems (16-18). Arole of tyrosine phosphorylation has also beenpostulated, since peroxides of vanadate, which inhibit tyrosine phosphatases, can induce activation of PLD (19, 20). Addition of phorbol esters to intact cells increases PLD activity, suggesting thatprotein kinase C (PKC) isinvolved in the regulation of that activity(21).Down-regulation of PKC, caused by prolonged exposure of cells to phorbol esters, abolishes PLD activation by phorbol esters and various agonists (22-24). However, PKC is not a unique protein. At least eight different subtypeshave been cloned andcharacterized biochemically (25). As no isoform-specific inhibitors of PKC are Phospholipase D (PLD)’catalyzes thehydrolysis of phospho- currently available, it hasbeen difflcult to demonstrate specificity of function for a given isoform in vivo. * This work has been supported by National Institutes of Health Recent work in Swiss/3T3 cells and rat fibroblasts overexGrants GM 31487 and HL 35018 and by a postdoctoral fellowship from pressing PKC, and PKC, has shown enhanced PLD response to Fulbright CommisionlMinisterio de Educacih y Ciencia, Spain (to platelet-derived M. A. B.),a predoctoral fellowship from Pharmaceutical Manufacturersphorbol esters,endothelin,thrombin,and Association Foundation (toB. L. F.), anda postdoctoral fellowship from growth factor, suggesting a role for different PKC isozymes in American Heart Association, California affiliate (to C. G.). Thecosts of regulating PLD activity (26-28). Unfortunately, no firm conclupublication of this article were defrayedin part by the payment of page sions are possible because the PKC subtype implicated in difcharges. This articlemust therefore be hereby marked “aduertisernent” ferent systems seems to be different. Moreover, substrate availin accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ability, receptor number andotherregulatory factorsmay This paper is dedicated to the memory of the father of one of us critically limit responsiveness in these systems. Also, overex(M. A.B.). f Current address: Fondation pour Recherches Medicales, Universitepression of an enzyme likePKC has the potentialof producing de Geneve, Faculte de Medecine, 64,Avenue de la Rosarie, 1211 Geneva effects secondary to indirectconsequences resulting from phos4, Switzerland. 5 To whom correspondence should be addressed. Tel.: 619-534-2295; phorylation of substrates that impact on signal transduction pathways. An alternative approach to assess the role of an Fax: 619-534-6833. The abbreviations used are: PLD, phospholipase D; PLC, phospho- enzyme isoform is to “knock out” this particular isoform. Our lipase C; P q ,phospholipase 4; A A , arachidonic acid; CPK,antisense protein kinase C cDNA IP,, inositol 1,4,5-trisphosphate; MDCK, Madin-Darby canine kidney; NDGA, nordihydroguaiaretic acid; 4aPDD, C; PMA, phorbol 12-myristate, 13-acetate; PI,phosphoinositide; DAG, 4a-phorbol didecanoate; PEt, phosphatidylethanol;PKC, protein kinase diacylglycerol; MeSATP, methylthio-ATP.

Protein kinase C (PKC) has been implicated in the activation of phospholipase D (PLD)in a number of systems. By antisense technology, we have “knocked out” a and p isoforms of PKC to study the role of these isoforms in PLD activation in Madin-Darby canine kidney (MDCK) cells. To this end, we have studied PLD activation by phorbol12-myristate13-acetate (PMA),ATP, UTP, and 2-methylthio-ATP in cells labeled with [3H]palmiticacid.[3H]Phosphatidylethanol (PEt) production catalyzed by PLDin thepresence of ethanol was time- and concentration-dependentin PMA- and nucleotide-stimulated cells. In Ca2+-free medium, [‘HIPEt accumulationwasdiminishedforallstimuliassayed. Treatment of cells with chelerythrine, an inhibitor of PKC, and phorbol ester down-regulation of PKC inhibited [‘HIPEt production by both PMA and nucleotides. In cells transfected with antisense PKC, or both PKC, and PKC,,PLD activation was inhibited by both PMA and nucleotides, whereas in cells transfected with antisense PKC,, PLD activation was similar to that of control cells. Moreover,inhibition of polyphosphoinositidespecific PLC (by neomycin) or of release of arachidonic acid and arachidonic acid metabolites (bynordihydroguaiaretic acid or by indomethacin) failed to decrease [‘HIPEt accumulation in PMA- and nucleotide-stimulated MDCK-Dl cells.From these data, we conclude that in MDCK-Dl cells PMA and nucleotide receptors utilize PKC, to regulate PLD activity and that PLD activation is independent of the activation of polyphosphoinositidespecific PLC and phospholipase &,-mediatedrelease of arachidonic acid or arachidonic acid metabolites.

10511

10512

Phospholipase D Regulation by Protein Kinase C,

laboratory has developed strategies to isolate stably transfected cells deficient in specific PKC isoforms. This methodology has allowed us to demonstratethe involvement of PKC, in arachidonic acid (AA) release in Madin-Darby canine kidney (MDCK) cells(29). In the current work, we have taken advantage of this methodology to study whether a specific PKC isoform is involved in PLD activation in MDCK cells. Our findings suggest that PKC, regulates PLD activity in MDCK stimulated with phorbol esters, ATP, UTP, or 2-MeSATP. Furthermore, our data demonstrate that PLD activation takes place independently of both polyphosphoinositide-specific PLC (PI-PLC) and AA release.

3.0

-

2.5 -

1.5 1.0 -

2.0

0 755025

100 125 150

Phorbol ester (nM)

EXPERIMENTALPROCEDURES M~terials-[~HIPalmiticacid (specific activity,54 Cilmmol), [,H]arachidonicacid(specificactivity, 100 Cilmmol), and [3H]inositol1,4,5trisphosphate (specific activity, 10 Ci/mmol) were purchased from DuPont NEN. PMA, 4a-phorbol didecanoate (4aPDD), quinYAM,indomethacin, nordihydroguaiaretic acid (NDGA), neomycin, ATP, UTP, and phosphatidic acid werepurchased from Sigma, 2-MeSATPwas obtained from Research Biochemicals Incorporated (Natick, MA). Phosphatidylethanol (PEt) was obtained from Biomol (Plymouth Meeting, PA). Chelerythrine was fromL. C. Services Corporation (Woburn,MA). G60 2-' 0.50 thin layer chromatography plates were purchased from Whatman. The 0.25 organic solvents were from Fisher. Cell Culture-MDCK-Dl cells, a clone isolated from the parental 0.00 MDCK cell line (30), were grown in Dulbecco's modified Eagle'smedium 0 0.1 with 7.5% heat-inactivated horse serum, 2.5% heat-inactivated fetal Agonist (pM) calf serum, 15 m~ HEPES, and 3.7 gfiiter sodium bicarbonate, pH 7.4 (31).For experiments, cells were treated with trypsin-EDTA solution FIG.1. Accumulation of ['HIPEt in stimulated MDCK cells. and diluted subsequently into six-well plates. Cells were used after MDCK cells, prelabeled with [SHlpalmiticacid, were stimulated in the growth for 3 days. MDCK cells transfected with pCMV expression vec- presence of 1% ethanol by (A) PMA (0)or 4aPDD (0)for 30 min, or by or 2-MeSATP (A) for 15 min. ARer extraction, tor containing antisense PKC, (CKPJ andor PKC,(CKp,J plus the ( B )ATP (a),UTP (0) neomycin resistance gene fRc-CMVNeo, gen), and the control cells lipids were separated by thin-layer chromatography, and radioactivity transfected with Rc-CMVNeo only (Neo'), were grown as previously content in PEtwas quantified by liquid scintillation counting. Data are given as a percentage of radioactivity in PEt with respect to total ce1described (29). lular radioactivity. Values are shown as means f S.E. of triplicate dePLL)Aetiuation-Wild-type or transfected MDCK-Dl cells were la- terminations in a representative of five m e r e n t experiments. beled the day before confluency with f3H]palmiticacid (2.25 pCilml). M e r removal of Iabefing medium, cultures were rinsed twice with phosphate-buffered saline and equilibrated with serum-free medium RESULTS (Dulbecco's modified Eagle'smedium) containing 1mg/ml bovineserum PEt Production in Stimulated MDCK-Dl Cells-PLD-cataalbumin at 37 "C for 45 min. Cells were then incubated at 37 "C with the specified concentra~onsof stimuli for indicated times in the pres- lyzed transphosphatidyiation is the only known mechanism by ence of 1%ethanol. Reactions were stopped by aspirating the medium which cells can form phosphatidylalcohols.Therefore, PEt forand by adding 0.5 mi of methanol containing 1% HCI, and lipids were mation catalyzed by PLD in the presence of ethanol is a sensiextracted according to Bligh and Dyer (32).rHlPEt was resolved from tive and speciik assay for PLD activation (36).As observed total lipids by TLC on Silica Gel G plates using the upper phase of a with parental MDCK cells (15,3'71,PMA increased 13H1PEt system consisting of ethyl a c e ~ t ~ i s ~ t a n ~ acidlwater a~tic accumulation in MDCK-Dl cells is a con~ntration-de~ndent (130:20:30:100, vlv)(33).The lipids were identified by eo-migration with manner, while 4aPDD, an inactive phorbol ester, was ineffecauthentic standards and visualized by iodine staining. ~ ~ o a c t i v i t ytive (Fig. lA),We next determined if PLD could be activated by was determined in the silica gel scrapings by liquid scintillation specpurinergic receptor agonists in MDCK cells. As shown in Fig. trometry. B ,ATP, and 2-MeSATP increased E3HfPEtaccumulation Ca2+-depletedcells were prepared by treatment with 40 p~ q u i n ~ A M Z plus 1 m~ EGTA in Ca2+-freeHanks' salt solution (Sigma) a t 37 "C for in a concentration-de~ndentmanner, reaching maximal activ60 min (34), washed twice,and stimulated as appropriate, either in the ity at 100 p~ ATP, 300 p~ UT", and 1000 p~ 2-MeSATP, represence or absence of extracellular Ca". spectively,with EC, values of approximately 10,3, and 150 p~ Acid Relea~e-[~H]AArelease was for ATP, UTP, and 2-MeSATP, respectively. The time course of ~ e a s u ~ eof~ ~~JAraehidonic n t measured as described previously (31). Cells were labeled with 0.33 PLD activation was different for different agonists: response to pCi/ml t3HlAA for 16 h. At the end of the labeling period, cells were ATP, UTP, and 2-MeSATP was rapid (within 3 m i d (Fig. 2 8 1 , washed four times with phosphate-buffered saline. One ml of serumfree medium containing 1mg/ml bovineserum albumin and the appro- while PMA response was more delayed (Fig. 2A). In addition, priate concentration of stimulus was added. M e r 30 min at 37 "C, response to PMA continued for at least 50 min, while response C3H]AA release into the extracellular media was determined by liquid to the nucleotides was almost hlly achieved within 5-10 min for all three nucleotides. This may reflect a rapid desensitizascintillation counting. lnositol 1,4,5-Bisphosphate Assay-MDCK-Dl cells were washed tion of the signaling pathways for the nucleotides. twice with phosphate-bufferedsaline and incubated with the appropriRole of Ca2+in Receptor Stimulation of PLD in MDCK-Dl ate stimulus in serum-free medium at 37 "C for the indicated times. Cells-One way to identify the possible involvement of differReactions were stopped by aspirating the medium and by adding 1ml ent PLD subtypes is by assessment of requirements for Ca2+for ofice-cold 20% trichloroacetic acid. The plates were scraped, the samples transferred to Eppendorf tubes, and protein was quantified PLD activity. Ca2+has been proposed to be a key regulator for and particularly for according to the Bradford procedure (35). The trichloroacetic acid was PLD activity in many cell types (6,3'7-40), extracted using diethyl ether, and IP, was determined by a specific phosphatidylinositol-hydrolyzingPLD in MDCK cells (15).AS shown in Fig. 3, removing extracellular Ca2+from MDCK-Dl radioreceptor assay kit (DuPont NEN).

U'I'P,

Phospholipase D Regulation by Protein Kinase C, 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0

o 10

0

20 30 Time (min)

40

~

0

50

"

fO

"

10513

"

'

~

l

'

20 30 40 50 60 70 EO

Chele~hnne@M) I

B

1.25

1

1

1.oo

0.75 0.50 0.25 0.00

0

5

10

15 20

25

30

Time (min) RG.2. Time course of ImPEt accumulation in stimulated MDCR cells. Cells, prelabeled with [3Hlpalm.itic acid for 24 h, were challenged withLA)medium alone0 , PMA(32 ru4, e),or ( B )ATP (300 p,e),UTP (300 w, 0) or 2-MeSATP (300ptd, A)for the indicated time points, inthe presenee of 1%ethanol. Dataare shown as means t S.E. of triplicate determinations in a representative experiment that was

replicated five Merent times.

L h

B

none PMA ATP UTP 2-MeSATP

FIG. 4. Effect of chelerytkine on ['HIPEtaccumulation in MDCK cells. Cells prelabeled with [3H]palmitic acid were preincubated with the indicated concentrations of chelerythrine for 30 min prior to stimulation with 300 ATP in the presence of I% ethanol for 15 min (A). Cells were preincubated with 50 p chelerythrine (black bars)or vehicle(open bars) and stimulated with ATP (300p), UTP (300 p), or 2-MeSATP (300p)in the presence of I%ethanol for 15min ( B ). Data are shown as means t S.E. of two experiments with triplicate determinations.

to be activated by the nucleotides. PKC Requirement for PHIPEt Aecumulatiolt in MDCK"D1 CeEls--In several cell systems, PLD is modulated by PKC activation (33,40421. Increase in PEtformation promoted by PMA 0.75 (Figs. IA and 2) suggested that PKC activation enhances PLD s activation in MDCK-Dl cells.In order to determine further the i3 role of PKCin PEt production in MDCK cells, cells were treated 0.50 with chelerythrine, a powefil PKC inhibitor (IC, 0.66 p d (43). The drug strongly inhibited t3H]PEtaccumulation in ATPstimulated cells (Fig. 441, with an IC,of about 10 piv. Chel0.25 erythrine also inhibited E3HJPEt production by2-MeSATP, VIT,and PMA (Fig. 4B 1, suggesting a role for PKCin receptormediated PLD activation in MDCK-Dl cells. The smaller ex0.00 tent of inhibition of PLD activation by PMA compared to that 2-MeATP UTP ATP PMA none by the nucleotides may reflect the differing kinetics (Fig. 21, FIG. 3. Effect of Ca* on PEt production by stimulated MDCK and perhaps mechanisms, of activation of the two classes of [~~IP~~mitic a ~ d - ~ a ~MDCK l e d cells were treated with 32 rn activators. Other i ~ i b i t o r s (e.g. , staurosporine, sph~ngosine) PMA, 300 w ATP, 300 w UTIP, or 300 p 2-MeSATP in medium with 1.3 m~ CaCI, (closed bars)or in calcium-free medium withf m EGTA were not effective in decreasing [3H]PEt accumulation in our e, has been fopen bars) for 15 min. Resdta, given as percent release with respect tosystem. Moreover, wefound that s ~ u r o s p o ~ nwhich cellular radioactivity, are themean * S.E.of three different determina- described as an activator of PLD in rabbit peritoneal macrotions in a presentative experiment that was replicated four times. phage (441, st~mulated[3H]PEt production in MD~K-Dlcells (data not shown). cells nearly abolished the responses to ATP, UTP, and 2-MePKCa Is Inuo~ued in PLD A c ~ i u i tin ~ M ~ C K - D lCellsSATP, while the response to PMA was inhibited by 50%. PEt Prolonged treatment with phorbol esters leads to down-reguproduction was also examined in cells that were depleted of lation of PKC because phorbol ester promotes translocation of PKC to the membrane facilitatesproteolysis of the membrane their intracellular Ca2" stores by treating them with 40 quin2tAM plus 1 n m EGTA. Under these conditions, PEt pro- associated form of the enzyme (45). PKC isoforms have differ[data ential s u s c e p ~ b ~ ito t yproteolysis and to phorbol ester-induced duction did not occur in theabsence of extracellular ea2+ not shown). These data suggest that a threshold concentration d o ~ - r e ~ l a t i (46, o n 47). We have previously described that of Ca2+is required to sustain receptor-mediated PEt production PKCa is preferentially down-regulated (relative to PKC,) in in MDCK cells, and that a ea2+ dependent form of PLD is likely phorbol ester-treated MDCK-Dl cells (48). Thus, we measured 1O .O

h

0 .

3

cells.

-


10514

2.0

-l5 $?

"

0.75

t

1.5

a 1.0 0.5

0.0

CKPaP ControlCKPp

2-MeSATP UTPATP PMA FIG.5. Effect of PKC down-regulationon ['HIPEt production in MDCK cells. Cells were treated with 80 m PMA (black bars) or vehicle (open bars) for 20 h. Then the cells were stimulated with PMA (32 m),ATP (300 w), UTP (300 w) or 2-MeSATP (300 w) in the presence of 1% ethanol for 15 min. Data are shown as means ? S.E. of two experiments w t h triplicate determinations. Unstimulated PEt production has been substrated. PEt values in control and PKCdownregulated cells were 0.02 ? 0.01,and 0.20 ? 0.04, respectively.

CKPa

FIG.6. PEt production in stimulated CPK transfected MDCK cells. CPK,,CPKCPKm+,, and Neo' transfectants were prelabeled with [3H]palmitich d and stimulated with PMA (32 m,gray bars), ATP (300 w, striped bars), UTP (300 w, closed bars), 2-MeSATP (300 w, open bars), or medium alone(dotted bars) in thepresence of 1% ethanol for 15 min. Results are shown as means ? S.E. fromtriplicate determinations in a representative of three different experiments. 260

-/

T

In

A [,H]PEt in cells treated with PMA for 16-20 h. Fig. 5 shows 240 . h that PKC down-regulationabolished [3H]PEt generation by 2 220 c PMA and by nucleotides. 8 200 We next conducted experiments in MDCK-Dl cells transfected with PKC antisense for isoforms a,p, and a + p (29), in 180 , -In order todefine more precisely the role of these isoforms in PLD . activation. The results, summarized in Fig. 6 , showed that -$0 ) 160 . PKC, antisense-transfected cells had a diminished ability to a" 140 produce L3H1PEt in response to PMA and to nucleotides, while 120 PKC, antisense transfectant cells behaved as did control cells 100 (cells transfected with Rc-CMVNeo). In double transfectants 0 5 10 15 20 (PKCdBantisense-transfected cells), we observed the same response found with PKC, antisense-transfected cells. Time (s) The inhibition of [3H]PEt accumulation might be due to a delayed activation of PLD in transfectedcells. To rule out this 1.75 possibility, we carried outkinetic experiments to study whether T B L3H1PEt accumulation would occur at longer periods of stimulation. Neither PKC, or PKCdBtransfected cells stimulated for g 1.25 as long as 50 min could produce [,H]PEt at levels similar to those seen in control cells (data not shown). PHlPEt Accumulation Is Not Inhibited by Neomycin in MDCK-Dl Cells-It is well known that stimulationof ATP and 0.75 2-MeSATP receptors increases PLC activity in a variety of cell 0.50 types (49). Thus we reasoned that PLD activation by nucleotides might be a consequence of PKC activation throughPI-PLC. To 0.25 assess thishypothesis, we performed experiments withneomy0.00 cin. This drug inhibits PLC activation (50,51) and been has used none PMA ATP UTP 2-MeSATP in ourlaboratory to ascertain the relationship between PI-PLC activity and AA release (52). We found that neomycin strongly FIG.7.Effect of neomycin on I P S and PEt productionin stimuinhibited IP, formation in ATP-stimulated MDCK-Dl cells (Fig. lated MDCK cells. A, MDCK cells werestimulated with ATP (300 w, 7A), but exerted no effect on [,H]PEt accumulation in cells A, 0 ) or vehicle ( 0 )in the presence (A,0 ) or absence (0)of 1 m treated either with PMA, ATP, UTP, or 2-MeSATP (Fig. 7B). neomycin for the indicated times. These data are given as mean 2 S.E. from triplicate determinations in a representative experiment, and are These results strongly suggest that a PI-PLC and PLD are in- expressed as percent increase with respect to control. 100% value was dependently activated inMDCK-Dl cells. 15.2 pmol of IP, per mg of protein. B , t3H1palmitic acidprelabeled cells or PHlPEt Accumulation Is Independent of AA Release in were stimulated with PMA (32 m),ATP (300 phi), UTP (300 y), MDCK-Dl Cells-MDCK-Dl cells stimulated by ATP,UTP, and 2-MeSATP (300 w) in the presence of 1% ethanol, and in the presence (closed bars) or absence (open bars) of 1 m neomycin. These data are 2-MeSATP release AA and AA metabolites to the extracellular given as means ? S.E. of triplicate determinations in a representative medium (53). To investigate the relationship, if any, between experiment. AA release and[,H]PEt generation, we used NDGA. NDGAis a well recognized inhibitor of lipoxygenase but, in addition, has role in receptor-mediated AA release in a variety of cell types been found to inhibitP& in vitro (54,551 and PW-mediated (57). As shown in Fig. €A,we found that L3HIAA release in AA release in intactcells (56). Indeed, NDGA acts as a potent response to ATP, UTP, and 2-MeSATP was strongly inhibitedby inhibitor of the high molecular weight, AA-specific, Ca2+-de- NDGA in MDCK-Dl cells in a concentration-dependent manNDGA, a concentration that totally inhibited AA pendent P& (55), a n enzyme that ispresumed to play a major ner. At 25

-

'

'

'

'

1

d


10515

pancy of receptors by ATP, UTP, and 2-MeSATP leads to the activation of a Ca2+-dependentPLD. This is supported by the fact that no response is seen in Ca2+-depletedcells. Moreover, strong inhibitionof since removingextracellular Ca2+ leads a to the response,PLD activation inMDCK-Dl cells likely requires a sustained elevation of intracellular Ca2+ levels,a result consistent with findings in other systems (6,3840). Inasmuch as PLD is mostly a cytosolic enzyme, it is not yet clear whether the or translocation increase incalcium acts to facilitate activation of PLD. Both ATP and 2-MeSATP stimulate rapid polyphosphoinosiNDGA (pM) tide hydrolysis, resulting in increasedIP, production (Fig. 7A) (59). Because previous studies by our laboratory have indicated 1.00 that these analogs can increase AA release inMDCK cells (531, we believe that the same typeof P, receptors activate PI-PLC, 8 0.75 between PI-PLC PLD, and PL4. To determine the relationship z 0.50 and PLD, we took advantage of the inhibitory action of neomycin on PI-PLC, which we have previously found fails to blunt I 0, 0.25 AA release in MDCK-Dl cells (52). Neomycin inhibits formation of IP, by binding topolyphosphoinositides, thereby making 0.00 them unavailable toPLC attack (60). Because neomycin inhibnone P A U 2M ited ATP-induced IP, formation but hadno effect on PEt accumulation stimulated either by PMA, ATP, UTP, or 2-MeSATP (Fig. 71, IP, formation and theIP,-mediated discharge of intra1.50 cellular Ca2+ stores are not prerequisites for ATP, UTP, and s 1.25 2-MeSATP to activate PLD in MDCK-Dl cells. Similar conclu1.oo sions have also been drawn by others using platelet-activating z 0.75 factor-stimulated humanU937 cells (40) and chemotactic pepI 0.50 tide-stimulated human neutrophils(61). Previous work by our laboratory has strongly suggested that 0.25 AArelease inMDCK cells is the consequence of P& activation 0.00 and that other pathways, such as the DAG deacylation pathnone P A U 2M way or the inhibitionof AA reacylation, are probably of limited FIG.8. Effect OfNDGAonAArelease and ['HIPEt accumulation in stimulated MDCK cells. A, [3HlAA-prelabeled cells were stimu- significance in this particular cell type (62). Thus, in our system, AA release appears to be a useful index of P q activity lated with ATP (300 p,O),UTP (300 p,0),2-MeSATP(300 p,A),or vehicle (0) in the presence of the indicated NDGAconcentrationsfor 30 (62). Using NDGA as a P q inhibitor (54, 55) we found that min. B , [3H]palmitic acid-labeled cells were stimulated withPMA(P,32 this agent did not block PLD activation, thus strongly suggestm,),ATP (A, 300 p,), UTP (U, 300 p,) or 2-MeSATP(2M, 300 p,) in to PLD activation inMDCK-Dl the presence (closedbars) or absence (open bars) of 25 p NDGA, or C, ing that AArelease is unrelated in the presence (closedbars) or absence (open bars) of 1 p indometha- cells. cin. These data are given as means f S.E. of triplicate determinations If PLD activation requires PKC, in particular PKC, activafrom a representative of three different experiments. tion, how might PKC, be activated prior to its effects on PLD stimulation? Since PI-PLC hydrolysisand the subsequentgenrelease, we found no inhibition of [,H]PEt accumulation with eration of DAG and IP, seem not to be required, we hypothesize any of the stimuli tested(Fig. 8B). that otherphospholipases could be involved. In this regard,we We also carried out experiments with indomethacin, a well have found that a,-adrenergic receptor activation produces a established cyclooxygenase inhibitor (521, and did not see any rapid formationof phosphocholine and DAG in MDCK-Dl cells, effect on [,H]PEt accumulation in MDCK-Dl-stimulated cells presumably by activation of a PLC that promotes rapid hy(Fig. 8C).These data suggest that PLD activation is related drolysis of phosphatidylcholine (63). Thus, if the nucleotide neither toAA release norto eicosanoid production in MDCK-Dl receptors are coupled to PLC activitythat actson phospholipids cells. others than polyphosphoinositides, this could induce a rapid generation of DAG, and thus a stimulation of PKC,. DISCUSSION Recent studies by Liscovitch and colleagues (28) usingPKC,The data presented in this study indicate that ATP, UTP, and overexpressing cells have suggested that this PKC isoenzyme 2-MeSATP, which are likely to interact withP, and P,, recep- regulates the levels of PLD expression rather than PLD actitors expressed on MDCK-Dl cells (53, 581, can stimulate PLD vation per se. We have found that ethanol only modestly elactivity. Our observations suggest that PLD activation is not evates L3H1PEt accumulation (around 20%) in MDCK-Dl cells, secondary to the two phospholipases previously noted to be indicating that basalPLD activity is very low in MDCK cells. activated upon P,, and P,, receptor occupancy, i.e. PI-PLC and Because we have found that basal PLD levels in unstimulated P w (53). Moreover, using MDCK-Dl cells with antisense in- cells were very similar in all of the transfected cells, PKC, in hibited expressionof a, p, or a + p PKC isoforms (29), we have particular PKC,, although implicated in PLD activation, isnot demonstrated that PKC, is implicated in PLD activation in likely to be critical for PLD expression in MDCK-Dl cells, in MDCK cells by PMA and by the receptors activated by the contrast with results obtained in Swisd3T3 fibroblasts (28). nucleotides. Other overexpression studies inrat fibroblasts have implicated Recently, Huang et al. (15) haveidentified two different PLD the p, isoform of PKC in PLD activation (26), but our results activity isoforms in MDCK cells. One of these isoforms does not using antisensetechnology to decreaseexpression of PKC, fail require extracellular Ca2+, while the other ishighly dependent to implicate this isoform in our system.We do not know if this of extracellular Ca2+ (15). Our studies suggest that the occu- reflects speciesor cell-type differences in PLD regulation or if it

-

Y

a

Y

a

1

T

10516

Phospholipase D Regulation by Protein Kinase

C,

8. Moritz, A,, De Graan, P. N., Gispen, W. H., and Wirtz, K.W. (1992)J. Biol. is a consequence of indirect effects caused by overexpression of Chem. 267, 7207-7210 the /3 isoform. In this regard we have observed that PKC, is 9. Bocckino, S. B., Wilson, P. B., and Exton, J. H. (1991)Proc. Natl. Acad. Sei. bound to the nucleus in activated MDCK-Dl cells,2 suggesting U. S. A. 88,6210-6213 that this isozyme may be implicated in nuclear rather than in 10. Yang, S. F., Freer, S., and Benson, A. A. (1967)J. B i d . Chem. 242,477484 11. Kobayashi, M., and Kanfer, J. N. (1987)J. Neurochem. 48, 1597-1603 cell surface events in thesecells. 12. Billah, M. M., a n d h t h e s , J. C. (1990jBiochem.J. 269, 281-291 In PKC, antisense transfectedcells we found that about30% 13. Agwu, D. E., McCall,C. E., and McPhail, L.C. (1991)J. Immunol. 146, 3895-3903 of the PLD activity remained (29j, while in PMA-promoted 14. Ben-Av, P.,Eli, Y., Schmidt, U. S., Tobias, K. E., andLiscovitch, M. (1993)Eur down-regulation experiments we found a total loss in PLD acJ. Biochem. 218,455-463 tivity. This suggests that there may be another PKC isoform 15. Huang, C., Wykle, R. L., Daniel, L. W., and Cabot, M. C. (1992)J. Bioi. Chem. 267, 16859-16865 involved in PLD activation. Pfeilschifter and colleagues have 16. Kusner, D. J., Schomisch, S. J., and Dubyak G . R. (1993)J. Biol. Chem. 268, shown that prolonged treatment withPMA causes depletion of 19973-19982 PKC,,PKC,, and PKC, in rat renal mesangial cells (64). In 17. Agwu, D. E., McPhail, L. C, Chabot, M. C., Daniel, L. W., Wykle, R. L., and McCall, C. E. (1989)J. B i d . Chem. 264, 1405-1413 preliminary studies, we have found that PKC, accounts for at 18. Rttenborn, C. S., and Mueller, G. C . (1988)Biochem. Biophys. Res. Commun. least 40% of total PKC activity in MDCK cells (65). Thus the 155,249-255 possibility remains that some of the residual PLD activity in 19. Bourgoin, S., and Gristein, S. (1992)J. Biol. Chem. 267, 11908-11916 20. Dubyak, G.R., Schomisch, S. J., Kusner, D. J., and Xie, M. (1993) Biochem. PKC, antisense transfected cells, results from a calcium-inde292, 121-128 pendent PKC isoform, such as PKC, or PKC,, both of which are 21. Exton, J. H. (1990)J. Bioi. Chem. 265,9761-9765 present in MDCK-Dl c e k 3 Another possibility is that recep- 22. Cook, S. J., and Wakelam, M. J . 0. (19911 Biochim. Biophys. Acta 1092, 265-272 tor-mediated PLD might occur via action of G proteins (16-18) 23. Liscovitch, M. (1989)J. Biol. Chem. 264, 1450-1456 and that theprotocol used for down-regulation of PKC inhibits 24. Martinson, E. A,, Goldstein, D., and Brown J. H. (1989)J. Biol. Chem. 264, 14748-14754 signalling by such pathways (Fig. 5). Kusner et al. (16) have 25. Nishizuka, Y. (1989)J. Am. Med. Assoc. 262,826-1833 previously described the ATP-induced potentiation of G pro- 26. Pai, J.-K., Pachter, J. A,, Weinstein, I. B., and Bishop, W. R.(1991)Proc. Natl. Acad. Sci. U. S. A. 88,598-602 tein-dependent PLD activity in a cell-free system from U937 J. A., Pai, J.-K., Mayer-Ezell,R., Petrin, J. M., Dobek, E., and Bishop, cells. Thus activationof PLD by ATP might involve direct cou- 27. Pachter, W. R. (1992)J. Biol. Chem. 267,9826-9830 pling of G proteins and PLD or perhaps a receptor-induced 28. Eldar, H., Ben-Av, P., Schmidt, U.-S., Livneh, E., and Liscovitch M. (1993)J. Bid. Chem. 268, 12560-12564 increase in calcium concentration through G proteins. An al29. Godson, C., Bell, K. S., and Insel, P. A. (1993)J. B i d . Chem. 268,11946-11950 ternative possibility is that tyrosine kinase activity, which is 30. Meier, K.E., Sperling, D. M., and Insel, P. A. (1985)Am. J. Physiol. 246, C69477 important for stimulation of PLD in many systems (16,20,19), 31. Slivka, S. R.,and Insel, P. A. (1987)J. B i d . Chem. 262,4200-4207 is involved in MDCK-Dl cells. Tyrosine kinase activity might 32. Bligh, E. G., and Dyer, W. J. (1959)J. Biochem. Physiol. 37,911-917 amplify signals transduced by heterotrimeric andor low mo- 33. Liscovitch, M., and Amsterdam, A. (1989)J. Bid. Chem. 264, 11762-11767 lecular weight ras-like G proteins, or by analogy to PLC,, a n 34. Di Virgilio, E , Lew D. P., and Pozzan, T.(1984)Nature 310,691-693 35. Bradford M. M. (1976)Anal. Biochem. 72,248-254 isoenzyme of PLD might be activated via phosphorylation on 36. Billah, M.M., Anthes, J. C., and Mullmann, T.J. t 1991) Biochem. Soc. Duns. tyrosine residues. 19,326329 Taken together, our results suggest that inMDCK-Dl cells 37. Huang, C., and Cabot, M. C. (1990)J. Biol. Chem. 265,14858-14863 38. Halenda, S. P., and Rehm, A. G. (1990)Biochem. J. 267,479483 one or more classes of nucleotide receptors act via one or more 39. Huang, R.,Kucera, G. L., and Rittenhouse, S.E. (19911J. Bid. Chem. 266, 1652-1655 G proteins to promote calcium entry and to activate a PLC Balsinde, J., and Mollinedo, E (1991)J. Bid. Chem. 266, 18726-18730 other than PI-PLC. Based on our previous data (621, we hy- 40. 41. Martinson, E. A,, Trilivas, I., and Brown, J. H. (1990)J. B i d . Chem. 266, pothesize that this is a PC-PLC. The DAG generated may act 22282.22287 along withcalcium to facilitate a rapid activation of PKC, with 42. Conricade, K. M., Brewer, K. A,, and Exton J. H. (1992)J. Biol. Chem. 267, 7199-7202 a consequent stimulation of PLD. The generationof DAG by a 43. Herbert, J. M., Augereau, J. M., Gleye, J., and Maf€rand, J. P. (1990)Biochem. Biophys. Res. Commun. 172,993-999 phosphatidic acidphosphohyd~lase acting on the phosphatidic Y., Takahashi, K., Tomita, U., h i , T., Katada, T., Ui, M., and Nozawa acid generated by PLD, could account for a further activationof 44. Kanaho, Y. (1992)J. B i d . Chem. 267,2355443559 PKC,, and of PLD activation. An alternative possibility is that 45. Young, S.,Parker, P., Ullrich, A., and Stabel, S. (1987)Biochem. J. 244,775779 a G protein is directly coupled to PLD. Experiments areneeded 46. Huang, F.L., Yoshida, Y.,Cunha-Melo,J. R., Beaven, M.A., and Huang, K:P. to ~ s t i n g u i s hbetween these alternative hypotheses. (1989)J. Biol. Chem. 264,42384243 In summa^, the current work defines certain biochemical 47. Kishimoto,A., Mikawa, K., Hashimoto, K.,Yasuda, I., Tanaka, S.-I.,Tominaga, M., Kuroda, T., and Nishizuka, Y. (1989)J. Bid. Chem. 264,4088-4092 signals involved in nucleotidereceptor and phorbol ester48. Godson, C., Weiss, B. A., and Insel, P. A. (1990)J. Biol. Chem. 265,8369-8372 stimulated activation of PLD in MDCK-Dl cells. The results 49. O'Connor, S . E., Dainty, I. A., and Leff P. (1991)Den& Pharmacol. Sei. 12, 137-141 demonstrate a key role forPKC, in thatprocess and show that Lipsky, J. J.,and Leitman, P. A. (1982)J. Pharmacol. Exp. Ther 220,287-292 PLD activation in MDCK-Dl cells is an event that takes place 50. 51. Schwertz, D. W., Kreisberg, J. I., and Venkatachalam, M.A. (19843J. Pharindependently of activation of both PI-PLC and PLA,. maeol. Exp. Ther. 231,48-55

A c k n f f w l e ~ g ~ e ~ thank ~ s - ~ eDm. Dan Kiel and Jesus Balsinde for helpful scientific discussions. REFERENCES 1. Heller M. (1978)Adv. Lipid Res. 16,267-326

2. Kanaho, Y.,Kanoh, H., Saitoh, K., and Nozawa, Y. (19911J. Immunoi. 146, 3536-3541 3. Cockcmft, S. (1984)Biochim. Biophys. Acta 795,3746 4. Moolenaar, W, H., Kruijer, W., "illy, B. C., Verlaan, I., Bierman, A. J., and de Laat S. W. (1986)Nature 323,171-173 5. Knauss, T. C., Jaffer, F. E.. and AbboudH. E. (1990)J. Biol. Chem. 265,

-14457-14462 .- . ... . -

6. Balboa, M. A,, Balsinde, J., Aramhuru, J., Mollinedo, F., and Upez-Botet M. (1992)J. 23x0. Med. 176.9-17 7. Jones, G. A., and Carpenter G . (iY93)J. Bid. Chem. 268,20845-20850

C. Godson, E. Masliah, M. EIlisman, and P. A. Insel, manuscript submitted for publication. C. Godson and M. A. Balboa, unpublished observations.

52. Slivka, S . R., and Insel, P. A. (19881J. Biol. Chem. 263, 14640-14647 53. Firestein, B. L., Craig, D.A., and Insel, P.A. f 1991)Soc. Neumsei. Abstr 17,81 54. Foster, K. A., Buckle, D. R., Crescenzi, K. L., Fenwick, A. E., and Taylor, J. E. (1987)Biochem. Soc. ?bans. 15,418-119 55. Wijkander, J., and Sundler, R. 11989)FEES Lett. 244,5146 56. Balsinde. J.. Fernctndez. B., Solis-Herruzo,J. A., and Diez, E. (1992)Biochim. Biophys. Acta 1136, 75-82 57. Lin. L. L., Lin, A. Y., and Knopf, J. L. (1992)Proc. Natl. Acad.Sei. U. S. A. 89, 6147-6151 58. Firestein, B. L., and Insel, P.A. (19931Pharmacologist 3 6 140 59. Cooper, C . L., Moms, A. J., and Harden, T.K. (1989)J . Biol. Chem. 264, 6202-6206 60. Robison T. W., Sevarian, A., and Foman, H.J. (1990)Toxicol.Appl. Pharmacol. 106, 113-122 61. Mullmann, T. J., Cheewatrakoolpong,B., Anthes, 3. C., Siegel, M. I.. Egan, R. W., and Billah, M. (1993)J. Leukocyte Bioi. 53, 630-635 62. Weiss, B. A., and Insel, P. A. (1991)J. B i d . Chem. 266, 2126-2133 63. Slivka, S. R., Meier, K. E., and Insel, P. A. (1988jJ.Bioi. Chem. 263, 1224212245 64. Huwiler, A,, Fabbro, D., and Pfeilschifter, J. (1992)Biochem. J. 279,441445 65. Godson, C., Ramirez, M. T., Baxter, G.,Strulovici, B., and Insel, P. A. (1991) tXSEB J. 5,A834