Molecular and cellular actions of platelet activating factor in rat heart cells

Molecular and cellular actions of

platelet-activating factor in rat heart cells.

C V Massey, … , W J Lederer, T B Rogers

J Clin Invest.

1991;

88(6)

:2106-2116.

https://doi.org/10.1172/JCI115540

.

Platelet-activating factor (PAF) is a phospholipid with cardiovascular actions at low

concentrations (1-100 nM) but with uncertain direct myocardial actions. We investigated the

cellular and molecular effects of PAF on heart cells using isolated adult and neonatal rat

myocytes. Addition of PAF, in the superfusion solution, decreased twitch amplitude and

contractile velocity in both systems. Concentrations of PAF below 1 nM stimulated

reproducible responses with maximal effects seen at 100 nM. These functional actions of

PAF could be blocked by the known PAF antagonist, BN 50739, in a dose-dependent

manner. Parallel biochemical studies showed that nanomolar PAF rapidly stimulated the

phosphoinositide pathway in cultured myocytes, evidenced by the accumulation of

[3H]inositol phosphates in prelabeled cultured myocytes. The potency and specificity of

PAF, as well as the time course, for the response were nearly identical in the biochemical

and functional assays. PAF produced no functional changes in protein kinase C-depleted

myocytes, but it did stimulate inositol trisphosphate accumulation in such cells. We

conclude that: (a) PAF exerts a direct negative inotropic effect on myocardial tissue; (b) the

effects of PAF are mediated by a specific, high affinity cardiac receptor; (c) an underlying

biochemical mechanism for the action of PAF includes the activation of the phospholipase

C/phosphatidylinositol intracellular signaling pathway, which leads to activation of protein

kinase C.

Research Article

Molecular

and

Cellular Actions

of

Platelet-activating Factor

in Rat Heart Cells

Clara V. Massey,* Trudy A.Kohout,*Shirley T. Gaa,*W. J.Lederer,'andTerry B.Rogersl"

Departments of *Medicine,Division ofCardiology, *BiologicalChemistry,Physiology,and Medical BiotechnologyCenter, University ofMaryland School ofMedicine,Baltimore, Maryland21201

Abstract

Platelet-activating factor (PAF) isaphospholipidwith

cardio-vascular actions at low concentrations (1-100 nM) but with uncertain direct myocardial actions. We investigated the cellu-lar and molecucellu-lar effects of PAFonheart cells using isolated adultandneonatalrat myocytes.Addition of PAF, in the super-fusion solution, decreased twitch amplitude and contractile ve-locity in both systems. Concentrations of PAF below 1 nM

stimulatedreproducibleresponseswith maximal effectsseenat

100 nM. These functional actions of PAF could be blockedby the known PAF antagonist, BN 50739, inadose-dependent manner. Parallel biochemical studiesshowed thatnanomolar PAFrapidly stimulated thephosphoinositidepathway in cul-turedmyocytes,evidenced by theaccumulationof

_VIHlinositol

phosphates in prelabeled culturedmyocytes.Thepotencyand specificity of PAF,aswellasthe timecourse,for theresponse werenearly identical in thebiochemicaland functionalassays. PAFproducednofunctional changes in protein kinase

C-de-pletedmyocytes,but it did stimulate inositol trisphosphate ac-cumulationinsuch cells. We conclude that: (a) PAFexertsa

direct negative inotropic effect onmyocardial tissue; (b) the

effects of PAFaremediatedbyaspecific, high affinity cardiac receptor,(c)anunderlyingbiochemicalmechanism for the

ac-tion ofPAF includes the activation of thephospholipase

C/

phosphatidylinositol intracellular signaling pathway, which leadstoactivation ofprotein kinase C. (J. Clin. Invest. 1991.

88:2106-2116.) Key words: cardiac*myocytes*inotrope-

re-ceptor-phosphoinositide

Introduction

Platelet-activating factor

(PAF',

I-O-alkyl-2-acetyl-sn-glycero-3-phosphorylcholine)isapotentsignaling moleculewhose bio-logical activitywasinitially associated with platelet activation

(1, 2). Thisphospholipidisnowknown to have profound

ef-Portions of this workwerepresentedattheAmericanHeart Associa-tion ScientificMeetings, 12-15 November1990, Dallas,TX.

Addresscorrespondence to Dr. Terry B. Rogers, Department of

BiologicalChemistry, University ofMarylandSchool ofMedicine,660 WestRedwoodStreet, Baltimore,MD21201.

Received for publication6November 1990 andinrevisedform8

July1991.

1. Abbreviations used in thispaper: Ins(1,4,5)P3,

inositol-1,4,5-tris-phosphate; InsP2, inositol bisinositol-1,4,5-tris-phosphate; InsP, inositol monophos-phate; InsP3, inositol trisphosphate; PAF,platelet-activating factor; TPA,12-O-tetradecanoylphorbol-I3-acetatephorbolester.

fects

on manytissues including muscle cells, vasculartissue, and neurons (3-5). PAF can be synthesized in and released frommanytissues including kidney,lung, and heart, as well as

fromcellsinthecirculation inresponse toappropriatestimuli

(3, 6-8). Biochemical

investigation

into the biosynthesis of

PAFindicatesthat heart musclecouldbeexposedtothis

mole-cule ina

physiological

setting. In this regard, recent studies

havereportedthatsignificant quantities ofPAF weremeasured in coronary effluent afteranischemic intervention(9, 10).

Despite this evidence,there has beenconfusion concerning theaction ofPAF oncardiac tissue.Inrecentstudies,PAFhas

beenreportedto havelittleornodirect actionon the

myocar-dium(10, 11). It wasproposedthat thecardioactive effects of

PAF arise secondarily to platelet-dependent release of cardioactive agents (thromboxane A2, histamine, and leuko-trienes). Incontrast, inearlier studies

using

electrically paced isolated cardiac tissue,acase wasmade for direct action ofPAF ontheheart(12, 13). Detailed information concerning the

ac-tion ofthis

phospholipid

incardiac cells hasbeenlimited bythe

complex tissue and cellular systemsused in these

experiments.

Inthis study isolated adult and neonatalrat heartcellswere

usedto

investigate

theactions ofPAFon acardiac tissue

un-complicated

by other celltypes. We reporthere thatPAFexerts a direct

negative inotropic

effecton such cells and thatthis effect is mediated bya

specific

cardiacPAFreceptor.

Further-more, we suggest thatthe functional changes inheartcells

pro-ducedbyPAFarise from the activation of

protein

kinase C via the

phospholipase

C/phosphatidylinositol

intracellular

signal-ing pathway.

Methods

Preparation of cardiac myocytes. Isolatedsinglecells from adult rat cardiac ventricular tissuewereprepared byenzymatic disaggregationof the intact tissues. The cellswereplacedinafour-well 1.7-cm culture dish. After 30 min the cells settledtothe bottom of theplateandwere

usedthesameday (14). Spontaneously beating primary cultures of neonatalratcardiac ventricular myocyteswerepreparedfroml-dold Sprague-Dawleyrats(15).Thecellswereplatedinfour-well 1.7-cm tissue culture dishes and grown in culture for 4-5 d(15-17).

Measurement of contractile properties ofadult orneonatalcells. Contractions ofsinglecardiac myocytes, in four-well culturedishes,

were measured usinganoptical video dimensionanalysissystemas

previously described

(15).

Briefly, the four-well culture dish was

mountedonthe stage ofaninvertedphase microscopeand the cells ina

singlewellweresuperfusedat1ml/minwith control medium

(DME,

25 mMHepes, 0.25%BSA,pH7.3) equilibratedwith 95% 2-5%C02

at

320C.

Aftera10-min equilibration period,the cellswere_superfused

with control mediumcontainingvariouscompoundsasindicated in Results. When the cellswereelectricallydriven 40-Vpulsesof 10ms durationwereused. PAFwasaddedtothesuperfusionbuffers from stock solutions indimethylsulfoxide,such that the medium contained

afinaldimethylsulfoxide concentrationof0.1%.Controlexperiments

indicated that thisconcentrationofvehicle hadnoeffectoncontractile behavior. Theimageof thebeatingcellswasmonitoredusingavideo J.Clin.Invest.

©TheAmericanSocietyforClinicalInvestigation,Inc. 0021-9738/91/12/2106/11 $2.00

camera(CohuInc., SanDiego, CA)and thevideosignalwasrecorded on tape. The motion of the cell wallwasmeasuredwith theuseofa

video dimension analyzer (Instrumentation forPhysiology& Medi-cine; San Diego,CA)duringtheexperimentandoff-line. The video dimension analyzer identifies the cell edge bycontrasttransition and trackstheedgeasitspositionchangesduringcontraction. A voltage proportional to cellshorteningis produced by theanalyzer,recorded, and thenanalyzedusinganIBM PCcompatiblecomputer. The con-traction and relaxation velocitieswereobtainedmathematically by dif-ferentiatingtheposition versustime output of the video dimension analyzer.

Loading conditions forcontraction caused by attachment ofthe cell

tothecoverslipor toother cells may vary from celltocell.However,by expressingthe contractile response of the cellasapercentage of the basal response, these differences inloadingconditionsaretaken into

accountand eachcellcan serve asitsowncontrol.

Measurementofinositolphosphates. Cells in culture(4-5 dold) werelabeledwith 10MACi/mlof[3H]myo-inositolfor 24hinserum-free DME-Hepes,0.25%BSA, unless otherwise indicated. Pilot studies

re-vealed that the absence of serum during thelabeling phase hadno

qualitative effectsonthebiochemicalorbiologicalresponsesreported in Results. However, serum-free media greatlyimprovedthesignalin the assay.Atthetimeof theexperimentthe cellswerewashedwith 4 x 0.5mlDME-Hepes,0.25% BSA, and treatedasindicated in Results. After aspirating the supernatant, the reactionwasstoppedbyrapidly adding0.5 mlof 6%perchloricacidat0C. The acidextract was

neu-tralized and the[3H]inositolphosphateswereresolvedusingion

ex-changechromatographyasdescribedpreviously (15).Total

[3H]inosi-tol phosphateswerecollectedbyelutingthemfrom the columns with 3 ml of 2 Mammoniumformate/0.1 Mformic acid. In some

experi-mentsonly the[3H]inositol bisphosphate(InsP2) fractionwascollected since itssignal provided the largest accumulationoverbasal.

Materials. The following substances wereobtained from Sigma Chemical Co., St. Louis, MO: DME-Hepes; PAF; lyso-PAF (L-a-lysophosphatidylcholine, -y-O-alkyl); formic acid; ammonium for-mate; TPA(12-O-tetradecanoylphorbol-13-acetatephorbolester). Per-chloricacidwasobtainedfrom J. T. Baker Chemical Co.,Phillipsburg, NJ).Myo-[2-3H(N)]-inositoland[3H]Ins(1,4,5)P3werefrom NEN Re-searchProducts,Wilmington,DE.DowexAGI-X8(formateform)was

from Bio-Rad Laboratories, Rockville Center, NY. The following com-panies kindly supplied these PAF antagonists: BN52021 and BN

50739,Institut Henri Beaufour, Le Plessis Robinson, France; Brotizo-lam and WEB 2086, Boehringer Ingelheim Pharmaceuticals Inc.,

Ridgefield, CT; Ro 19-3704/001, Hoffmann-LaRoche Inc., Nutley, NJ; SRI 63-441, Sandoz Research Institute, E. Hanover, NJ.

Results

PAFhas diverseactionsoncardiacand smooth muscleaswell

as onnerves.It has been showntoreducemyocardial contrac-tile force

(12, 18-20)

and has beenproposedas apromotor of

ischemia-reperfusion injury by

both direct and indirect action

(21-23).

Itisnotclear ifPAF-mediatedcardiacabnormalities result from direct actions of this substance on heart cellsor

from indirectactionssuchasreduced blood

flow,

altered

neuro-nalinputtothemyocardium,orplatelet activationwith subse-quent release of

secondary

mediators. Initial experimentswere

designedtodetermineifPAF hadanydirect effectsoncardiac tissue free of these potentially confounding factors.

Accord-ingly,

weexaminedtheresponsesofisolatedcardiac myocytes afterapplication ofPAF. We used bothacutely isolatedadult cardiac cells and cultured neonatal myocytes. Myocytes, acutely dissociated from adultratventricle,wereplacedin a

superfusion chamber, driven electricallyat90beatsper

min,

and thenexposedto200nMPAF.Asshown inFig. 1,

applica-tion of PAF resulted inadecreaseinbeating amplitudeanda

decreaseinthe maximum velocitiesof contraction and

relax-ation. In thesignal-averagedresponses,the overall shape ofthe twitch and waveform for thecontraction/relaxation velocities

were notalteredby PAFtreatment.This result isfroma repre-sentative experiment. The results from several experiments withseparatepreparations indicate thatPAFdecreasedtwitch amplitude by 22±15% (n=3).These resultsdemonstratethat PAFcandecreasecontractility inacardiac cell preparationthat

isnotcomplicated byneuronal,platelet,orhemodynamic

ef-fects.

We complemented these studies by examining the effects of PAF in cultures ofspontaneously beating neonatalrat

ventricu-CONTROL

200nM

PAF

AVERAGE

w

r--)

.D

E--q

0-4

.4

Q--,

I

2 sec

I

lium

0.2 sec

E--U

v C)

04

₄

₄₀

sec

larmyocytes. This well defined cardiac preparation allows for molecular and functional studies to be performed in parallel (15-17).Fig.2shows that spontaneously contracting heart cells

superfusedwith80 nM PAF exhibited a rapid increase in the beating rate (66% over 2 min). Decreases inamplitude (50%)

and in maximum contraction/relaxation velocities were also observed in this representative experiment. The addition of PAFin several experiments showedan average maximal in-crease in spontaneous beating rate of 72±5.4% above control values (mean±SEM, n = 7). The average decrease in amplitude was39±4.6% below control values (mean±SEM, n = 7).

Since it ispossiblethat thePAF-mediated decreases in both velocity andamplituderesult from the increases in contractile

frequency, the action ofPAFon myocytes at afixed frequency

wasinvestigated. Fig.3 showsthat when myocytes were

electri-cally drivenat90beats permin, application of 20nMand 80 nM PAFresultedin arapid decreaseintwitch amplitude. The

resultsfromaseries of experimentsrevealed that 20 nM and 80 nM PAF decreased twitch amplitude by 32±6.3% and 24±8.8%,respectively (mean±SEM,n=5-6),valueswhich are not statistically different. Fig. 3 also shows that the overall shape ofthe twitchisnotalteredby 80nMPAF.These results demonstrate that theeffects ofPAF onelectrically driven neo-natal myocytes aresimilartotheresponsesobservedinadult heart cells.However,thenegative inotropic effects ofPAF are

reducedinpaced neonatalcellscomparedtospontaneous

beat-ing neonatalmyocytes.Takentogether, these datasuggest that some of the action ofPAF israte related, but the negative inotropic effects ofPAF cannot beexplained on the basis of ratechange alone.

Toinvestigate the possibility that the functionalresponses we have observed are receptormediated, thepotencyand

speci-ficity of thePAFeffectswerecharacterized.The effects of PAF onboth frequency and amplitude were dose dependent. As shownin Fig. 4, PAF was very potent in evoking functional responses, with 1 nM PAF increasing frequency by > 20% over control. The dose-response relation was biphasic, with an

ini-tial phase between 0.5 nM and 20 nM and a second phase

between 20nMand100nM.Contractile amplitudein the sec-ond phase decreased by up to 65%. Fig. 5 shows that the

biologi-cally inactive analogue ofPAF, lyso-PAF, had no effect on

contractile behavior,even atconcentrationsas high as100nM. ThespecificityofthePAF-evokedeffectswasfurther

exam-ined by the use of a potent PAF receptor antagonist,BN50739 (24).Before application of 20 nM PAF, cells were pretreated

with various concentrationsof BN 50739 for 10min.Asshown

inFig.6, BN 50739 was very potent in blocking the functional responsesofthe cells to PAF, with half-maximal doses in the10

nMrange. Iftheinhibitoryeffects of BN 50739 are due to its

actionatthelevelofthe receptor, then theinhibitionshould be

competitive in character.Totestthis hypothesis,a PAF dose-responsecurve was generated in the presence ofa half-maximal

doseofBN50739(10 nM). As showninFig.7,addition ofBN

50739caused adisplacement of thePAFdose-response curve totheright byabout oneorderof magnitude, without altering

themaximalresponseofthe cells athighdosesofPAF. These resultssupport the view that PAF and BN 50739 compete for a commonreceptorsite.Takentogether, they indicate thatthere

isa specific cardiac PAFreceptor thatmediatesthenegative inotropic action of this phospholipid.

Theseresultssuggest that there are specific PAF-receptor

mediated intracellularmolecularmechanismsthatunderliethe potent biological effects of PAF on cardiac myocytes. The

functionalresponsesevokedby PAF, decreases in twitch

am-CONTROL

2

sec

80nm

PAF

AVERAGE

0.2

sec

20,um/sec

Figure2. Effects of PAFoncontractilebehaviorofspontaneously beatingcultured neonatal rat ventricularmyocytes.Cultured neonatal myo-cytesweresuperfusedwith control mediumandthespontaneous beatingbehaviorwasrecorded. Thesuperfusionsolution was switched to

con-trolmediumcontaining80 nM PAF andthechangesinspontaneous beatingbehaviorweremonitored. Shown aretheresponsesin atypical ex-perimentundercontrol conditionsand 3minafterexposuretoPAF.Ontherightarethe_{signal-averaged}twitches and cellshorteningvelocity

curvesunder controlconditions(solid line)andafterexposuretoPAF(dotted line). H

2--mD

1|lm

A

CONTROL

2OnM PAF

2 sec

B

CONTROL

8OnM

PAF

AVERAGE

2

sec

0.2

sec

Figure 3. Effects ofPAFonthecontractile behaviorof electrically driven cultured neonatalratventricular myocytes. Cultured cells(intrinsic spontaneousbeatingrate40-70 beats permin)werefield stimulated with 40-Vpulsesof 10msdurationat a constant rateof 90pulsesper min. After an equilibration period either 20nM PAF(A) or 80nM PAF(B)wereaddedtothesuperfusate.Shownaretypicalresponses after 3-min exposurestoPAF. On the right side ofBthesignal-averagedtwitches under control conditions(solid line)andafterexposureto80nM PAF

(dotted line)areshown.

20f]%

4

8-'N

80 80

60-6

T

40

i-40

±

zo~~~~~~~~~~~~~~~~~~~~~~~~~z

20

T

20

T~~~~

1

1

0

100

log [PAF]

nM

Figure4.Dose-response relation for PAF effects on contractile behavior of spontaneously beating cultured neonatal rat ventricularmyocytes. The initial beating behavior of cultured cells was assessed in cells that were superfused with control medium. Thesuperfusatewas switched tocontrol medium supplemented with a range ofPAF concentrations and the resulting beating behavior was recorded. The results are reported as percent change relative to the initial, pre-PAF, values for frequency (-) and amplitude(n). The data points are themeans±SEMfrom four to nine

CONTROL

1OOnM

LYSO-PAF

1,unm

md

2

see

Figure 5. Effects of lyso-PAF on contractile behavior of spontaneously beating cultured neonatal rat ventricular myocytes. After the cells were equilibratedin control medium, the superfusate was switched to control medium supplemented with 100 nM lyso-PAF. Shown is a representative beatingrecordofacell under control conditions and4min after application of lyso-PAF. This agent produced no changes in contractile behavior in three separate experiments.

t~7

50

.I

50

~>

ON

40

'

U

|-

n

30

-3

So

20

L

Um

10

[

z

1

10 100

log [BN 50739]

nM

Figure6.Effectof PAFantagonist,BN50739,onthe PAF-induced functionalresponsesof cultured neonatal rat ventricularmyocytes. Sponta-neously contractingculturesweresuperfusedwithmediumcontainingvarious concentrations of BN 50739 for 6min.Thesuperfusatewas switchedtomediumcontainingthesameconcentrationofBN50739supplementedwith25nM PAF. Thechangesinbeatingfrequency (-)and

twitchamplitude (o)aftera3-minexposureto PAFareshownrelative to the initial values.Thedatapointsarethe means+SEM from four to eightseparateexperiments.

II

Q-)

0

v-To

40pnm

sec

40

30

20

10

U4

>--<

V

Z4

Lo

X4

l'

:

V

;D

z C

80

CfZ

I I

X~

Z;

60

40

20

0

1000

log [PAF] nM

Figure7.Effects ofBN50739onthe PAFdose-responserelation inspontaneously

beating

cultured neonatalratventricular myocytes. A dose-response curve for the functionaleffectsof PAFonspontaneousbeatingbehaviorwasmeasuredasdescribed in

Fig.

4 except that thesuperfusate

wasswitched to control medium withPAFplus10 nM BN 50739. Shownarethe PAFdose-responsecurvesfor mediacontainingPAF(-)or mediacontaining PAFsupplementedwith 10nM BN50739

(i).

The dataarethemeansofthreeto

eight

separateexperiments±SEM. plitude and increases in

frequency,

are

analogous

tothose

ob-served for

angiotensin

II. This hormone stimulatesthe phos-phoinositidesignaling

pathway

in these cultures

(15).

Accord-ingly,

experiments

were

designed

that wouldtestthehypothesis thatPAFactivates thesame

signaling pathway.

Myocytesthat hadbeenprelabeled with[3H]inositol wereexposedto25 nM PAFand the resulting

production

of

[3H]inositol

phosphates

wasquantitated. Asshown in Fig. 8, timecourse studies

re-vealed thatapplication

of

PAFresulted in the

rapid

accumula-tion of

[3H]inositol

monophosphate

(InsP),

[3H]InsP2,

and [3H]inositol

trisphosphate

(InsP3). Maximal values for

[3H]-InsP2and

[3H]InsP3

of1.7-fold stimulationoverbasal and

3-fold

stimulationoverbasal,

respectively,

werenotedat 2min. Thus thetime coursefor

phosphoinositide

turnoveris

com-parabletotherateofonsetof the functionalresponsesevoked byPAF.

Taken together, these results suggest that activation of

phosphoinositide

turnoverisanunderlying mechanism for the actionofPAF oncardiac cells. This hypothesiswastestedby

examining

the potencyand

specificity

ofPAFin stimulating this

signaling

pathway.PAFevokedadose-dependent

accumu-lation of

[3H]inositol

phosphates over a 2-min interval as shownin

Fig.

9. PAF was very potentin stimulating

[3H]InsP

and

[3H]InsP2

accumulation withhalf-maximalconcentrations in the rangeof5 nM. Maximal responses were obtained at a

dose of 20 nM. Thus, the threshold and effective

concentra-tions ofPAFin boththebiochemical and functional experi-ments are verysimilar.

Otherexperiments wereperformedto further explore the

specificity

of the action of PAF. The biologically inactive

ana-logue, lyso-PAF, did notincrease the accumulation of

[3H]-inositol phosphates at a concentration of 25 nM (data not

shown). Since thespecificPAFantagonist,BN 50739, is a po-tentinhibitor of the functional responses of cardiac myocytes toPAF,its effects on PAF-evokedphosphoinositideturnover wereexamined.AsshowninFig. 10,when cells were preincu-bated with BN 50739 before theapplicationofPAF,the

antago-nist blocked the subsequent PAF-evoked accumulation of

[3H]InsP2 in adose-dependentmanner. This agentwas very potent,with anICW in the range of 5 nM. Thus, theantagonist

potency of BN 50739 wascomparablein both the biochemical and functional assays.

These data support theview that PAF-evokedbiochemical

andphysiological responses are mediatedby aspecific, high

affinityreceptor.Since thepossibility of multiplereceptors has

beenraised (25),an

important

issue arises

concerning

the phar-macological characteristics of the cardiac receptor. Measure-mentsofphosphoinositideturnover wereemployedto

charac-terize the

pharmacology

of the cardiacPAF receptorby using theinhibitory potencies of differentPAFantagonistsas an in-dex oftheiraffinities for thePAFreceptor. As shown inFig.10, BN 50739 was most potentand SRI 63-441wasabout10-fold lesspotent

(ICv-

50

nM).

Brotizolam,

RO

19-3704/001,

and WEB 2086 were also less potentwith

_IC~v's

inthe range of

100-300nM.Theginkgolide

antagonist,

BN 52021, was not aneffective blocker,even atdosesof1

_gM

(data notshown).

The rankorderof potencyofthese drugs is consistent with the

relativeaffinities oftheseantagonists for the PAF receptor in

platelets (26). This demonstratesthat: (a) there are potent

antag-onists ofthe cardiacPAF receptor;and(b) the pharmacology of

the cardiac PAF receptor issimilar to that characterized in

platelets.

C/D

Ed--CID

0

(:Il

~-CID

0

zH

~-C/D)

M

02

3

2

0-1

2

4

6

8

10

TIME

(m)in

Figure 8. Time course of PAF-stimulated release of inositol phosphates in cultured neonatal rat ventricular myocytes. Cultured myocytes were prelabeled with[3H]inositoland then incubated with 25 nM PAF for increasing time periods. After incubation, the level of intracellular

[3Hjinosi-tolphosphates was determined from cell extracts. Shown are the time course curves for[3H]InsP

(i),

[3H]InsP2

(.),and

[3H]InsP3

(A).The results

arereportedasfold-stimulation over basal. Basal values were obtained in parallel cultures incubated in the absence ofPAF.These values were:

[3H]InsP,200 cpm;

[3HJInsP2,

30 cpm;and

[3HJInsP3,

65cpm. The data points are the means of three experiments, each performed in duplicate, ±SEM.

functionalresponsesofPAF.Further,the responses to PAF are

similartothosewithagentsthatareknowntostimulate protein kinase C inthese cultures(15,16). Toexaminethehypothesis thatprotein kinase C mediates the effects ofPAF,experiments

were designed usingcellsthatwere depleted of this enzyme. Myocytes weretreated overnight withthephorbolester, TPA

(1

gM).

Wehavepreviously shownthatin neonatalmyocytes undertheseconditionsthereis>96%lossofenzymeactivity (17).Thesecultureswereexposedto25 nMPAF, a dosethat maximally stimulated

phosphoinositide

turnover. Asshown in Fig. 11,PAFdidnot evoke functional changesin thesecells;

application

of25 nM PAF had noeffectoneitherthe

frequency

(34beats permin)orthe

amplitude

ofcontraction.The signal-averaged twitches underscore the lack ofresponseof

protein

kinase C-depletedcells to PAF. The failure ofsuch cells to

respondto PAFmightresultfromalack ofa

phosphoinositide

response. However, asshownin Fig. 12, PAFstimulatedthe

production ofinositol phosphatesin

protein

kinaseC-depleted

cellsinamanneranalogoustothat in normalcells.These data support the conclusion that

protein

kinase

C,

rather than Ins(1,4,5)P3, isanimportant mediator of the effects ofPAFon

cardiaccells.

Discussion

The

ability

ofPAF to

adversely

affect cardiac functionhas been

well documentedin theliterature

(3, 27).

Yet the

physiological

importance ofthisphospholipid in cardiac disease has been

difficulttoestablishdue to the complexexperimentalmodels

used.Forexample, administration ofexogenous PAF to labora-tory animals leadstocardiac dysfunction (18-20, 28). How-ever,thisagent alters vascular tone,activatesthesympathetic

nervous system, andstimulatesmost cellsofthesystemic

circu-lation (3, 7, 8, 27). Thus, it is difficult to delineate specific cardiaceffects in suchmodels.

Otherexperimental evidence implicatesPAFalterations in cardiac function. PAF receptor antagonists are protective in models ofmyocardial ischemia(4, 22, 23). More focused

ex-periments with paced isolatedhumanatrialtissueandguinea pig papillary muscle suggest that PAF has a direct negative inotropic effect (12, 13). Yet, other recentexperimentshave been interpreted to suggest that theeffects ofPAF on heart

dependonplatelet activationand releaseof secondary

media-tors

(10,

11). Thus, it isclear that other

approaches

are

required

inordertodefinethe roleofPAFincardiac functionand

dis-ease.

It waswithinthiscontext thatweinitiatedstudiesto

exam-ine the effects ofPAFin isolated ratventricularmyocytes: a

simplified

cell system thatcontainsnovascular,

neuronal,

or blood components. The

principal findings

ofthisreportare:

(a)

Cl)

2

cn

m

3

2

1

1

10

100

log [PAF]

nM

Figure9.Dose-responsecurvesfor the PAF-stimulatedrelease of inositolphosphates. [3H]Inositol-prelabeledcultureswereincubated with a

rangeof PAF concentrations for2-minintervals. After theincubations,the levels of[3H]InsP(A)and[3HJInsP2 (v)weredetermined fromcell

extracts.The datapointsrepresent themeansof threeexperiments,eachperformedinduplicate,±SEM. When theerrorbarsarenotseenthey

donotextendbeyondthesize ofthesymbol.

biochemical mechanism for the action of PAF includes the activation of the phospholipase

C/phosphatidylinositol/pro-tein kinase C intracellular signaling pathway;and(d) cultured neonatalratmyocytesrepresentavalidandusefulmodel for

investigating the myocardial actions of PAF and its antago-nists.

Negative inotropiceffects ofPAF. Theinitialstudieswere

performedinelectricallystimulated dissociated adult rat

myo-cytes. PAF evokedadecreaseinmyocardial contractility char-acterized byadecreaseinbeating amplitude, velocityof

con-traction, and velocity of relaxation. Havingdocumenteda

di-rect negative inotropic effect in adult cells, subsequent experimentswereperformedin cultures ofspontaneously beat-ing neonatal rat myocytes. Here the functional studies

pro-duced results analogous to the responses observed in adult

cells; decreases in beating amplitude, velocity of contraction, andvelocity of relaxation.

In the spontaneously beating neonatal cell preparation, PAF also evokedarapidincreasein the beatingfrequency.It couldbe argued that in this cellsystemthe PAF-mediated fall in myocardial contractilitywasthe resultof increased

contrac-tile frequency. However, when neonatal cellswereelectrically drivenatafixedrate,decreases in amplitude, velocity of con-traction, and velocity of relaxation comparableto those ob-servedinthe adult cell preparationwerestill noted. Although

partof theresponsein neonatal cellsmayberaterelated,these results reveal that PAF is a negative inotropicagent in both isolated adult and neonatalratventricular cells.

Characterization ofacardiac PAFreceptor. Aconclusion

from the datareportedhere is that the effects of PAFare

me-diatedbyaspecific, high affinityreceptor locatedoncultured

myocytes. PAF is potent in evoking cardiac responses, with threshold doses in the 1 nMrange.Further,the effects of PAF

on contractilebehavioraredose-dependentand saturable, in respect toboth frequency and amplitude. Thedose-response

relationwasbiphasic withaninitial phase between 0.5 and 20 nM anda second phase between 20 and 100 nM. Thedoses seen in the initialphaseareconsistent with recent

measure-ments ofPAF in the circulation. PAF concentrations have beenseentoincrease from undetectable levelsto0.6-2 nM in twoseparatemodels: (a) incoronaryeffluent of isolated rabbit heart during initial reperfusion followinganischemic

interven-tion; (b) incoronarysinus samples of patients withcoronary arterydiseaseundergoingatrial pacing (9, 10). These data

un-derscore the physiological relevance of the PAF-evoked re-sponsesreported here.

Examinationofthe specificity of the PAF effects provided further evidence foracardiacreceptor. Thebiologically inac-tive analogue of PAF,lyso-PAF, had noeffectoncontractile behavior. BN 50739,apotentsynthetic PAFreceptor

antago-nist(24), blocked the effects of PAF inadose-dependent

man-ner,with half maximal doses in the nanomolarrange.Finally,

experiments performed in thepresenceofahalf-maximal dose

of BN 50739 causedadisplacementof the PAFdose-response

curvetothe right, without altering the maximalresponse to

agonist. These resultssupporttheview that PAF and BN 50739

C/D

CI)

CO

Ed--0

100

80

60

40

20

0

100

80

60

40

20

0

1 10 100 1000 1 10 100

1000

log

[ANTAGONIST]

nM

Figure10.Dose-responsecurvesforPAF antagonists. [3H]Inositol-prelabeled cultureswereincubated for 6 minat37°Cwith the PAF

antago-nists.After this preincubation,25 nM PAFwasappliedtothe culturesfor2minand the levelsof[3HJInsP2weremeasured. The resultsare

reportedaspercentmaximalvalueswhichwereobtainedin parallel experiments in which antagonistswerenotincluded.Thecurvefor themost

potentantagonist,BN50739 (-), is shownineach panel for comparison.Thecurvesforthe other antagonists(x)arelabeledineach panel. The

datapointsarethemeansfromatleast three experiments±SEM.

competefora commonreceptorsite. Taken together, they indi-catethat thenegative inotropic effects of PAFaremediated by aspecific cardiac PAFreceptor.

Molecularmechanism foraction ofPAF. A major goal of this investigation was to explore the underlying molecular mechanisms responsible for the effects of PAFonmyocardial

cells. Inmultipleexperimentsreported here, application of low doses ofPAFresultedinarapid accumulation ofinositol phos-phates. Thus, PAF stimulates thephosphoinositidepathway in cardiac cells. Thisfinding is consistent with previousreportsof PAF-evokedphosphoinositideactivation in platelets and other tissues(8, 29).

Itwasimportanttodetermine if the PAF-evoked activation ofthephosphoinositidepathway underlies the observed func-tionalresponses. Indirectevidencestronglysupportsthis

con-clusion.First,PAFevokedadose-dependentaccumulation of inositol phosphates with threshold and maximal concentra-tions comparable tothose seenin the high affinity response

observedinthefunctional studies.Second,the timecoursefor

PAF-mediatedinositolphosphates accumulation is

compara-blewiththerateofonsetof PAF-evoked functionalresponses.

Finally, the potent PAF receptor antagonist, BN 50739, blockedPAF-stimulated inositolphosphates accumulationin a

dose-dependentmanner,withahalf maximalinhibitorydose comparable tothat observed inthe functional experiments. Takentogether, these resultsareconsistent with the viewthat

thehigh affinity PAF-evokedresponsesaremediated,in part,

by thephosphoinositide pathway. Thisalsosuggests that the responsesofmyocytestodoses ofPAF inthe50-100nMrange

aremediatedby other additionalmechanisms.

Thereare twoimportant second messengersproduced in the phosphoinositide pathway, Ins(1,4,5)P3, and 1,2-diacyl-glycerol. The latter activates protein kinase C. Ins(1,4,5)P3

mo-bilizes

Ca2"

inmanycells butthere is disagreement concerning itsrole incardiac cells (30-32). Incontrast,thereis consider-able evidence thatprotein kinase C activation results in nega-tiveinotropicresponsesinavariety of cardiac tissues,

includ-ing human heart (16, 33-35). The functional studies reported here suggestthatthe effects of PAFare mediated by protein

kinaseC because the effects notedareidenticaltothoseseenin

previous studies withangiotensinIIandphorbolesters(15, 16). Both of theseagentsareknownactivators ofproteinkinaseC.

More direct studies with protein kinase C-depleted cells

re-vealed that PAF failedtoproduce functional changes in the absenceofthisenzyme.It isnotlikely that the depletion proto-coldamagedthecells since several results thatwehave previ-ously reportedindicate that these myocytes functionnormally. ThephosphoinositideresponsestoangiotensinIIand

a-adren-ergicreceptor stimulationarecompletelynormal in these cells

(36). Further,theexpressionandfunctionofsarcoplasmic retic-ulum,atarget forproteinkinaseC,isnotaltered when

com-paredtocontrol cells(17). Finally, in thepresent studyPAF

stillactivatedthereceptor-mediated phospholipaseCpathway indepleted cells, demonstrated bythe accumulation of inositol phosphates, including Ins(1,4,5)P3. These data dissociate Ins(1,4,5)P3 production fromthe PAF-evoked functional ef-fectsanddefineanimportantroleforproteinkinase C in

me-diatingtheeffects of PAFonheartcells.

CharacterizationofcardioactivePAFantagonists.There is evidence formultiple receptorsfor PAF in several cells(25).

;-

VV-CONTROL

PKC

DEPLETED

lpM

Em

isec

SIGNAL

AVERAGE

aO.5pLM

O.2sec

Figure 11. Functional effects ofPAF onprotein kinase C-depleted cells. Cultured neonatalratventricular myocyteswereincubated with 1 4M TPAovernight to reduceproteinkinase C activity. The enzyme-depleted culturesweresuperfused with control medium and the spontaneous beatingbehaviorwasrecorded. Thesuperfusatewasswitchedtocontrolmediumcontaining25nM PAFand theresultingbeatingbehavior was recorded.Shownis thebeating record ofacellunder controlconditions and 2 min after exposuretoPAF.Displayed belowarethe signal-averaged twitches derived from the same data under control conditions(solid line)andafter exposure to PAF(dotted line).

IP1

IP3

Figure12. PAF-stimulated release of inositol phosphates in protein

ki-naseC-depleted cells. Protein kinase C-depletedcultures(hatched bars) orcontrolcells(openbars)were

ex-posedto25 nM PAF for 2min. The

levels of[3H]InsPand[3H]InsP3

weredeterminedasdescribed inFig.

8. Thedata pointsrepresentthe

meansof three experiments, each

performed in duplicate, ±SEM.

00

~-m)

0-2.0

1.5

1.0

0.5

0.0

Cf)

M

Cf

Ed--Ct)

0

Ez

U/)

TO

Thus, it was important to further characterize the

pharmacolog-ical features ofthe cardiac PAF receptor. Measurements of

phosphoinositideturnover were used to assess the potency of

sixknown PAFantagonistsas an index oftheiraffinitiesfor the receptor. The rank order of potency we noted in cardiac tissue is consistent with the relative affinities of these drugs for the PAF receptor in platelets, demonstratingthatthe cardiac re-ceptoris similar pharmacologicallyto that observed in platelets (26). Thus,thiscellsystem can be used to evaluate the relative potencyand potential cardiac efficacyofavariety of PAFantag-onists.

Insummary, the present resultsdemonstratethat PAF de-creases contractile behavior and increases phosphoinositide

turnoverinisolatedratventricular myocytes.Theseeffectsare

mediatedby aspecific cardiacPAFreceptor which ultimately

leads to the activation of protein kinase C. The receptor ap-pears to besimilartothat whichhas beenidentified inother tissues. Finally, withapotentialroleforPAFantagonists inthe treatmentof ischemia-reperfusion injury, thiscultured cell sys-temprovidesavaluable tool for assessingtherelative efficacy of these antagonists.

Acknowledgments

Wethank Dr. Giora Feuerstein, Smith Kline Beecham, King of Prus-sia, PA, for many conversations and helpful suggestions throughout the

courseof this work.

This work wassupported in part by the National Institutes of Health Grants HL 28131 (T. B. Rogers) and HL 25675 and HL36974 (W. J. Lederer). Dr. Rogers isarecipientofa Research Career Develop-ment Award from the National Institutes of Health.

References

1.Benveniste, J.,P. M.Henson, and C. G. Cochrane. 1972. Leukocyte-depen-denthistamine release fromrabbitplatelets: the role ofIgE,basophils, anda

platelet-activatingfactor. J. Exp.Med.136:1356-1377.

2. Hanahan,D.J., and R. Kumar. 1987.Platelet-activating factor, chemical andbiochemical characteristics.Prog.LipidRes. 26:1-28.

3. Braquet, P., L.Touqui,T. Y.Shen,andB.B.Vargaftig.1987.Perspectives

inplatelet-activatingfactor research.Pharmacol.Rev.39:97-145.

4.Koltai,M.,A.Tosaki,J.-M.Guillon,D.Hosford,and P.Braquet. 1989.

PAFantagonistsaspotential therapeuticagents in cardiacanaphylaxisand myo-cardial ischemia.Cardiovasc. DrugRev.7:177-198.

5. Braquet, P., M. Paubert-Braquet, M.Koltai,R.Bourgain,F.Bussolino,and

D.Hosford. 1989. Is thereacasefor PAFantagonistsin thetreatmentofischemic states? TIPS(Trends Pharmacol. Sci.).10:23-30.

6.Handley,D. A.1990.Preclinicalandclinicalpharmacologyof

platelet-acti-vating factor receptorantagonists. Med.Res. Rev. 10:351-370.

7.Feuerstein,G. Z.1989.PAFand thecardiovascularsystem. In Platelet-acti-vating Factor and HumanDisease.E.Barnes,editor.BlackwellScientific Publi-cations, Cambridge.138-157.

8.Prescott, S.M.,G.A.Zimmerman,andT.M.McIntyre.1990. Platelet-ac-tivating factor.J.Biol. Chem.265:17381-17384.

9.Montrucchio, G., G.Camussi,C.Tetta, G.Emanuelli,F.Orzan,L.Libero,

andA. Brusca. 1986. Intravascular release ofplatelet-activatingfactorduring

atrialpacing.Lancet.ii:293.

10.Montrucchio,G., G.Alloatti,C.Tetta,R. DeLuca,R. N.Saunders,G. Emanuelli,and G.Camussi.1989. Releaseofplatelet-activatingfactor from

isch-emic-reperfusedrabbitheart.Am.J.Physiol.256:H1236-1246.

11.Alloatti, G.,G.Montrucchio,and G.Camussi.1990.Prostacyclininhibits theplatelet-dependent effects ofplatelet-activatingfactor in the rabbitisolated heart. J.Cardiovasc. Pharmacol. 15:745-751.

12.Robertson,D.A.,A.Genovese, and R. Levi. 1987.Negative inotropic

effectofplatelet-activating factoronhumanmyocardium: apharmacological

study.J.Pharmacol.Exp. Ther.243:834-839.

13.Robertson,D.A., D.-Y. Wang, C. 0.Lee,andR.Levi.1988.Negative

inotropic effect of platelet-activatingfactor:association with a decrease in intra-cellular sodium activity. J.Pharmacol.Exp. Ther. 245:124-128.

14.Cohen, N. M., and W. J. Lederer. 1988. Changes in the calcium current of rat heartventricularmyocytes during development. J. Physiol. (Lond.). 406:115-146.

15. Allen,I.S., N. M. Cohen, R. S. Dhallan, S. T. Gaa, W. J. Lederer, and T. B.Rogers. 1988. Angiotensin II increases spontaneous contractile frequency andstimulates calcium current in cultured neonatal rat heart myocytes: insights intounderlying biochemical mechanisms. Circ. Res. 62:524-534.

16.Dosemeci,A., R. S. Dhallan, N. M. Cohen, W. J. Lederer, and T. B. Rogers. 1988. Phorbol ester increases calcium current and simulates the effects of angiotensinII oncultured neonatal rat heart myocytes. Circ. Res. 62:347-357.

17. Rogers, T. B., S. T. Gaa, C. Massey, and A.Dosemeci. 1990. Protein kinase CinhibitsCa2+accumulation in cardiac sarcoplasmic reticulum. J. Biol. Chem.265:4302-4308.

18. Kenzora, J. L., J. E. Perez, S. R. Bergmann, and L. G.Lange.1984. Effects ofacetyl glyceryl ether ofphosphorylcholine (platelet-activating factor) on ventric-ularpreload, afterload,andcontractility in dogs. J.Clin.Invest. 74:1193-1203. 19. Levi, R., J. A. Burke, Z.-G. Guo, Y. Hattori, C. M. Hoppens, L. M. McManus, D. J. Hanahan, and R. N. Pinckard. 1984. Acetyl glyceryl ether phos-phorylcholine(AGEPC)aputativemediatorof cardiacanaphylaxis in the guinea pig. Circ. Res. 54:117-124.

20.Benveniste,J., C. Boullet, C. Brink, and C. Labat. 1983. The actions of Paf-acether(platelet-activatingfactor) on guinea-pig isolated heart preparations.

Br.J.Pharmacol.80:81-83.

21. Lepran, I., and A. M.Lefer.1985.Ischemiaaggravatingeffects ofplatelet-activating factorinacutemyocardial ischemia.Basic Res.Cardiol.80:135-141. 22.Stahl, G. L.,Z.-I.Terashita,and A. M.Lefer. 1988. Role of platelet-acti-vating factor in propagation of cardiac damage during myocardialischemia.J. Pharmacol.Exp.Ther.244:898-904.

23.Montrucchio,G., G. Alloatti,F.Mariano,R.dePaulis,A.Comino, G.

Emanuelli,andG. Camussi. 1990. Role ofplatelet-activating factorin the reper-fusioninjury of rabbit ischemicheart.Am.J.Pathol.137:71-83.

24.Yue, T.-L., M. Farhat, R. Rabinovici, P. Y. Perera, S.N.Vogel, and G. Feuerstein. 1990. Protectiveeffect ofBN50739,anewplatelet-activatingfactor antagonist,inendotoxin-treated rabbits. J.Pharmacol.Exp. Ther. 254:976-981. 25. Hwang, S.-B. 1988.Identification ofasecondputativereceptor ofplatelet-activatingfactorfrom humanpolymorphonuclear leukocytes.J. Biol. Chem. 263:3225-3233.

26.Tahraoui,L.,A.Floch, S. Mondot, and I. Cavero. 1988.High affinity specific bindingsites fortritiatedplatelet-activating factor in canineplatelet membranes: counterpartsofplatelet-activating factorreceptorsmediating plate-letaggregation.Mol.Pharmacol.34:145-151.

27.Feuerstein,G., andJ.M. Hallenbeck. 1987.Prostaglandins, leukotrienes, andplatelet-activatingfactor in shock.Annu.Rev.Pharmacol.Toxicol. 27:301-313.

28.Sybertz,E.J., R. W.Watkins,T.Baum, K.Pula,and M.Rivelli. 1985. Cardiac,coronary andperipheralvascular effectsof acetylglycerylether phos-phoryl choline in theanesthetizeddog. J.Pharmacol. Exp.Ther. 232:156-162. 29.Salari,H., V.Duronio,S.Howard,M.Demos,andS.L.Pelech. 1990. Translocation-independentactivation ofproteinkinase Cbyplatelet-activating

factor, thrombin,andprostacyclin. Biochem.J.267:689-696.

30.Vites, A.-M.,andA.Pappano.1990. Inositol1,4,5-trisphosphatereleases intracellularCa2l inpermeabilizedchickatria.Am.J.Physiol.(HeartCirc. Phys-iol.).258:H1745-H1752.

31.Movsesian,M.A.,A.P.Thomas,M.Selak,andJ. R.Williamson. 1985. InositoltrisphosphatedoesnotreleaseCa2'frompermeabilizedcardiac myocytes andsarcoplasmicreticulum. FEBS(Fed.Eur.Biochem.Soc.)Lett.185:328-332. 32.Nosek,T.N.,M.F.Williams,S.T.Zeigler,and R.E.Godt. 1986.Inositol

trisphosphateenhancescalciumrelease in skinned cardiac and skeletal muscle.

Am.J.Physiol. 250:C807-C811.

33.Capogrossi,M.C.,T.Kaku,C.R.Filburn,D. J.Pelto,R.G.Hansford, H. A.Spurgeon,andE.G.Lakatta. 1990.Phorbolesteranddioctanoylglycerol

stimulate membrane association ofproteinkinaseC and havea_negative

inotro-piceffect mediatedby changesincytosolic Ca2+in adultratcardiac myocytes. Circ.Res.66:1143-1155.

34. Yuan,S.,F. A.Sunahara,andA.K.Sen. 1987.Tumor-promoting

phor-bolestersinhibit cardiac functions and induceredistributionofproteinkinase C inperfused beatingratheart.Circ.Res.61:372-378.

35.Gwathmey,J.K., andR.J._Hajjar. 1990. Effect ofproteinkinaseC activationonsarcoplasmicreticulum function andapparentmyofibrillarCa2' sensitivityin intact and skinned muscles from normal anddiseasedhuman myo-cardium. Circ. Res. 67:744-752.