D HL-60

Lipopeptides activate G,-proteins in dibutyryl cyclic AMP-differentiated HL-60 cells

Jan F. KLINKER, Ariane HOER, Ingo SCHWANER, Stefan OFFERMANNS, Katharina WENZEL-SEIFERT and Roland SEIFERT*

Institutfur Pharmakologie, Freie Universitat Berlin, Thielallee 69-73, D-14195 Berlin, Federal Republic of Germany

Synthetic lipopeptides activatesuperoxide-anion (02-) formation in human neutrophils in a pertussis-toxin (PTX)-sensitive manner, suggesting the involvement of G-proteins of the

G,

family in the signal-transduction pathway. We compared G- protein activation by lipopeptides and the chemotactic peptide N-formylmethionyl-leucyl-phenylalanine (fMLP) in dibutyryl- cyclic-AMP-differentiated HL-60 cells. Thelipopeptide (2S)-2- palmitoylamino-6-palmitoyloxymethyl-7-palmitoyloxyhepta- noyl-SK4 (Pam3AhhSK4) and fMLP activated high-affinity GTPase, i.e. the enzymic activity of G-protein a-subunits, in HL- 60 membranes in a time- and protein-dependent manner, but they had no effect on Mg2+-ATPase and Na+/K+-ATPase.

Pam3AhhSK4 and fMLP increased

Vmax

of GTP hydrolysis.

Pam3AhhSK4 activated GTP hydrolysis with half-maximal and maximal effects at about

2,uM

and 10,uM respectively. Other lipopeptides activated GTPhydrolysis as well. Lipopeptides were less effective than fMLP toactivate GTPase. In membranes from PTX-treated cells, the stimulatory effects of lipopeptides and

INTRODUCTION

Human neutrophils play an important role in host defence against bacterial infections and are activated by the bacterial chemotacticpeptide fMLP (for review see Rossi, 1986; Seifert andSchultz, 1991). fMLP, after bindingto specific heptahelical membrane receptors, activates the PTX-sensitive G-proteins,Gi2 and

G,3

(Gierschik etal., 1989; Offermanns et al., 1990). This

process leads to the activation of phospholipase C and ofnon-

selective cation channels (resulting inanincrease in[Ca2+] ) and ofphospholipaseD(resulting in PA formation) (Paietal., 1988;

Bauldry etal., 1991; Kessels et al., 1991; Seifert etal., 1992a;

Krautwurst et al., 1992). Stimulation by fMLP of neutrophils and differentiated HL-60 leukaemic cells cumulates in the ac-

tivation of the 02-forming NADPH oxidase (for review see

Rossi, 1986; Seifert and Schultz, 1991).

The outercell wall ofGram-negative bacteria contains lipo- protein (for review seeBraun, 1975). Lipoprotein and synthetic lipopeptides are effective activators of B-lymphocytes (Resch andBessler, 1981; Bessleretal., 1985). We have shown recently thatlipopeptides activate02-formation in humanneutrophils in

a PTX-sensitive manner, suggesting the involvement of

G,-

proteins in thesignal-transduction pathway (Seifertetal., 1990).

As there isnoknownheptahelicalreceptorforlipopeptides,and

as only lipopeptides bearing positive charges activate 02-

fMLP on GTPase were abolished. InN-ethylmaleimide-treated membranes, the relative stimulatory effect ofPam3AhhSK4 on GTP hydrolysis was enhanced, whereas that of fMLP was diminished. fMLP and Pam3AhhSK4 activated GTPase in an over-additive manner in N-ethylmaleimide-treated membranes.

UnlikefMLP, Pam3AhhSK4 did not enhanceincorporationof GTP azidoanilide into, and cholera-toxin-catalysed ADP-ribo- sylation of

G,-protein

oc-subunits in,HL-60membranes and did not induce rises in cytosolicCa2+ concentration. Pam3AhhSK4 and fMLP stimulated phosphatidic acid formation in a PTX- sensitive manner. Pam3AhhSK4 itself did *not activate 02- formation, but potentiatedthestimulatory effects offMLP. Our data suggest that (i) lipopeptides activate the GTPase of

G,-

proteins, (ii) lipopeptides and fMLP activate

Gi-proteins

differently, (iii) lipopeptides stimulate phospholipaseD via

G,-

proteins, and (iv)phosphatidic acid formation is not sufficient foractivationof

02-

formation.

formation, we put forward the hypothesis that lipopeptides activate

Ga-proteins

directly, i.e. in a manner similar to thatof othercationic-amphiphilicpeptidessuch as substance Pand the wasp venom mastoparan(Higashijimaetal., 1988,1990;Mousli etal., 1990; Seifertetal., 1990; Tomita etal., 1991).However, the cationic-amphiphilic lipopeptides Pam3CSK4 and Pam3AdhSK4donot stimulateGTPhydrolysis by G-proteinsin HL-60membranes(Seifertetal., 1992b).Additionally,wehave shown that lipopeptides, i.e. Pam3CSK4 and Pam3AhhSK4, activate tyrosine phosphorylation in Bt2cAMP-differentiated HL-60cells(Offermannsetal., 1992). The latter findingprompted ustoexamine theeffects ofPam3AhhSK4onG-protein activation in membranes fromBt2cAMP-differentiated HL-60cells and to studyitseffects on

[Ca2+]1,

PAformation and02- formation in thesecells.

MATERIALS

AND

METHODS Materials

The lipopeptides Pam3AhhSK4 Pam3CSK4RPQASGVYMGN- LTAQ and Pam3CSK4RPQASVYMNLTAQ were kindly pro- videdby Dr. J. Metzger,Dr.K.-H.WiesmiullerandDr.G.Jung, Institut fur Organische Chemie der Universitiit

Tiibingen,

Germany. Thelipopeptide Pam3CSK4YGGFLwas

kindly

pro- videdbyDr.C.Sakarellos,Departmentof

Chemistry, University

Abbreviations used: Bt2cAMP, dibutyryl cAMP;

[Ca2+]i,

cytosolic Ca2+ concentration; CTX, cholera toxin; G,-proteins, family of highly similar G-proteins

(Gil-G3);

fMLP, N-formyl-L-methionyl-L-leucyl-L-phenylalanine; NEM, N-ethylmaleimide; 02-, superoxide anion; PA, phosphatidic acid, Pam3, N-palmitoyl-S-[2,3-bis(palmitoyloxy)-(2RS)-propyl]; Pam3Adh, (2S)-palmitoylamino-6,7-bis(palmitoyloxy)heptanoyl; Pam3Ahh, (2S)-palmitoyl- amino-6-palmitoyloxymethyl-7-palmitoyloxyheptanoyl; PTX,pertussistoxin.PeptidechainslinkedtoPam3, Pam3Adhand Pam3Ahharegiven in the one-letter code.

* To whomcorrespondence shouldbeaddressed.

3-phosphocholine (60Ci/mmol) was from Dupont/New Eng- land Nuclear (Bad Homburg, Germany). NEM was from Sigma Chemie (Deisenhofen,Germany).Stocksolutionsoflipopeptides (1 mM each) anddilutions were prepared in distilled water and storedat -20 'C.Sources ofothermaterials havebeendescribed elsewhere (Ebel et al., 1976; Rosenthal et al., 1986; Seifert and Schultz, 1987; Seifertet al., 1990, 1992a,b; Offermanns et al., 1990, 1991; Krautwurst et al., 1992; Wenzel-Seifert and Seifert, 1993).

Cell culture and membrane preparatlon

HL-60 cells were grown in suspension culture at 37 'C and were differentiatedtowards neutrophil-like cells with Bt2cAMP (0.2 mM) for48 h(Wenzel-Seifert andSeifert, 1993). For deter- mination of the activities ofhigh-affinity GTPase,Mg2+-ATPase and Na+/K+-ATPase, andforphotolabelling, HL-60 membranes wereprepared asdescribed bySeifertand Schultz (1987). PTX (100ng/ml) or itscarrier (control) was added to cell cultures 24h before measurement of PAformation ormembrane prep- aration. Under theseconditions, more than 95 % of

G,-protein

a-subunits wereADP-ribosylated (results not shown).

GTPase assay

GTPhydrolysiswasdetermined as describedbyWenzel-Seifert and Seifert (1993). Reaction mixtures (100,d) contained membranes from Bt2cAMP-differentiated HL-60 cells (3.0- 7.0,ug of protein/tube), 0.5

1sM

[y-32P]GTP (0.1,Ci/tube), 0.5 mMMgCl2, 0.1mMEGTA, 0.1mMATP, 1 mMadenosine

5'-[Ly-imido]triphosphate,

5 mM

phosphocreatine,

40

jug

of creatine kinase, 1mM dithiothreitol and 0.2% (w/v) BSA in 50 mM

triethanolamine/HCl,

pH 7.4. Reaction mixtures con- tained substances at various concentrations and were pre- incubated for3min at 25'C.Reactionswereinitiated byaddition of

[y-32P]GTP

andwereconductedfor 10-20min, unlessstated otherwise.

Low-affinity

GTPaseactivity was determined in the presenceof50

,sM

GTP and amountedto < 5%oftotalGTPase activity.

Treatment of HL-60 membranes

with NEM

HL-60 membranes were treated with NEM as described by McLeish et al., (1989) with modifications. Briefly, membranes (5.0,Cg of protein/tube) were incubated for 10min in reaction mixtures containing0.1 mM NEMorsolvent(control), 0.5,uM

[y-32P]GTP

(0.1

sCi/tube),

0.5 mM MgCl2, 0.1mM EGTA, 0.1 mM ATP, 1 mM adenosine

5'-[fi,y-imido]triphosphate,

5 mMphosphocreatine,40

,ug

of creatine kinase and 0.2 % BSA in 50 mM

triethanolamine/HCl,

pH 7.4.Thereafter, dithiothrei- tol (5mM) was added, and reaction mixtures were incubated forafurther10min.The membraneswerethen used

immediately

forthe GTPaseassay.

Mg -ATPase and Na+/K+-ATPase assays

Theactivities of

Mg2+-ATPase

and

Na+/K+-ATPase

in HL-60 membranesweredetermined as described by Ebeletal. (1976).

determined asdescribedbyAmes(1966).

Assay for photolabelling of membrane proteins

HL-60membranes

(50,Cg

of protein in60,ul)wereincubatedat 30°C in a buffer consisting of0.1 mM EDTA, 5 mM MgCl2,

1mM benzamidine, 10 ,M GDP and 30 mM Hepes/NaOH, pH7.4. After exposure to various substances, samples were

incubated foranother 3 min with10nM[a-32P]GTP azidoanilide (1 uCi/tube). Stopping of reactions and irradiation of samples

wereperformed asdescribedbyOffermannset al.(1990).

Assay

for

CTX-catalysed ADP-ribosylatIon of

membrane proteins HL-60 cellswerehomogenized by 20passesthrougha26-gauge needle in a buffer consisting of 150 mM NaCl, 3 mM MgCl2,

1 mM EDTA, 1 mM dithiothreitol and 20 mM Hepes/NaOH, pH 7.4. After centrifugation at 1000g for 10min at 4°C, the supernatantsuspensionwasincubated for 60 minat30°C in the above buffer supplemented with 1mM guanosine

5'-[/Jy- imido]triphosphate.

Thereafter, the suspensionwas centrifuged at 30000g for 15min at 4 'C. The pellet, referred to as

membranes, wasdissolved in abuffercontaining 1 mMEDTA and 10 mM Tris/HCl, pH 7.4. CTX was activated by mixing toxin stock solution (2mg/ml) withanequal volume of 40 mM dithiothreitol and subsequent incubation for 10 min at 30'C.

Reaction mixtures (50 ul) contained 3 ,iM [32P]NAD'

(5,c1Ci/

tube), 3 mM MgCl2, 1mM ATP, 10mM thymidine, 0.2 % BSA, 2 ,g of activated CTX and 0.1 M potassium phosphate, pH7.4, and various substances. After preincubation for 3 min, reactionswere initiated by addition of membranes from about

1 x107 HL-60 cells. After incubation for 60 min at 30°C, reactionswereterminatedby addition of 20 mMHepes/NaOH, pH 7.4 (4'C),andcentrifugationat12000gfor10minat4'C.

Assay

for PA formation

PAformationwasdetermined asdescribedby Paiet al. (1988) andBourgoin and Grinstein (1992), with modifications. In brief, HL-60 cells were centrifuged at 250gfor 10minat 20'C and

were suspended at 1 x106 cells/ml in a buffer consisting of 125 mMNaCl, 0.7 mM MgCl2, 0.5 mMEGTA, 10 mM glucose, 0.1% fatty-acid-free BSA and 25mM Hepes/NaOH, pH 7.2.

Centrifugation was repeated. HL-60 cells were suspended at 2x107 cells/ml in the above buffer and incubated for 90 minat 37'C in the presence of 10-15 fiCi/ml 1-0-[3H]hexadecyl-2- acetyl-sn-glycero-3-phosphocholine. Thereafter, cells were di- lutedto 1 x106cells/ml in the above buffer andwerecentrifuged at250gfor 10minat20'C. HL-60 cells[(0.5-1.0)x10"cellsin

100

fil]

^were incubated for 6min at 37'C in the presence of

1 mMCaCl2 and1 fg/ml cytochalasin B. Cellswereexposedto solvent (control),

Pam3AhhSK4

(10 fiM) or fMLP (IfM) for 15min. Reactionswereterminatedby the addition of 500,l of CHCl3/CH3OH/conc. HCl (400:200:1, by vol.). Phase sep-

arationwasachievedby additionof 250,ulofwaterand 150flof CHC13. The sampleswerecentrifuged at 12000gfor 10minat 4'C. The upperphasewasremoved, and 400,ulof the organic phasewas dried undernitrogenat 30'C. Lipidsweredissolved in 50jl of

CHC13

^and spotted on t.l.c. plates (Whatman LK 6 D). Non-radioactive lipid standards were added, and the plates were developed in a solvent system consisting of CHCl3/CH3OH/acetic acid (13:3:3, by vol). Lipid standards Reaction mixtures

(100 ,1d)

contained

6.0,ug

of

protein

and were detected

by

exposure to iodine vapour, and the areas

1200

E

0E

C1

.aa) 0 I- 0

10 15 20 25 30 0 2 4 6 8 10

lime(min) Protein(i/tube)

E 300 0 ._n 0

200 I- 100

Fire1 Time-andprotein-dependenceof

high-affinity

GTPhydrolyski InmembranesfromBt2cAMP-dlfferentlated HL-60

cells

High-affinity GTP hydrolysiswasdetermined as described in the Materials and methods sectioninthe presenceofPam3AhhSK4 (10jiM)(*),fMLP(10jiM)(-),orwateradded instead of stimulus (control) (-). (a) GTPhydrolysisas afunction of time. Reactionswereconducted for the indicatedperiodsof time. Reaction mixtures contained 5.7 g ofprotein (mainpanel)and 6.2jigof protein (inset). (b) GTP hydrolysisasafunction of theamountofprotein. Reaction mixtures contained the indicatedamountsofprotein,and reactionswereconducted for 15 min.

The open circlesindicate that at the designated incubation time and amount of protein no GTP was hydrolysed. Thestimulatory effects ofPam3AhhSK4onGTPhydrolysisweresignificantversus control (P<0.05,Wilcoxontest) under allconditions studied.

corresponding to PA were scraped off after sublimation of iodine. Lipids were eluted from the silica gel with 1 M HC1/CH30H(1/1,v/v), andradioactivitywasdetermined ina liquid-scintillation counter.

Assay for 02- formatlon

2- formationwasmonitoredat550nm

by

continuousmeasure- ment offerricytochrome c reduction inhibitable

by superoxide

dismutase,by usinganUvikon810dual-beam

spectrophotometer

(Kontron, Eching, Germany) (Seifert et

al.,

1990). Bt2cAMP- differentiatedHL-60cells(2.5x106cells in 500

,ul)

wereincubated for 3min at 37°C before addition of stimuli. The absolute amountsof 2-generatedwithin 10 min were calculated.

Miscellaneous

Proteinwas determined

by

the method of

Lowry

etal.

(1951).

[y-32P]GTP

was preparedasdescribed

by

Walsethetal.

(1991).

[c_-32P]GTP

azidoanilide was

prepared

as described

by

Offermannsetal.

(1991). [32P]NAD

was

synthesized

asdescribed by Cassel and Pfeuffer(1978).

SDS/PAGE

and

autoradiography

wereperformedas described

by

Rosenthal etal.

(1986). [Ca2+]J

wasdetermined

by using

thefluorescent

dye fura-2,

asdescribed by Seifertetal.

(1992a).

Data

reproducibility

Datashown inFigures 1-3and 5 and Tables 1-3are themeans of assay

quadruplicates.

Unless

shown,

the S.D. values were

generally less than 5%

(GTP hydrolysis)

and

10% (02-

formation) ofthemeans. Similar results were obtainedwith at least threedifferent

preparations

ofHL-60membranesorintact HL-60 cells. Basal GTP

hydrolysis

and the extent of GTPase stimulation caused

by

fMLP and

Pam3AhhSK4 (10 j/M each)

varied to some extentamong different membrane

preparations

from

Bt2cAMP-differentiated

HL-60cells

(see Figures

1-3 and Tables 1and

2).

Similar

findings

have beendocumented forbasal and fMLP-stimulated GTP

hydrolysis

in membranes from dimethyl

sulphoxide-differentiated

HL-60cells

(McLeish

et

al.,

1989). Theautoradiographsshownin Figure 4 are representative of at least threeindependentexperiments.

RESULTS

First, the time- and protein-dependence ofhigh-affinity GTP hydrolysisinHL-60 membranes was studied. In the presence of 5.7jig of protein per tube, Pam3AhhSK4 and fMLP (10 ,M each) stimulated GTP hydrolysis in a linear manner for up to 30min(Figure 1).Activation ofGTPase by thelipopeptideand fMLPoccurred withoutmeasurable delay (see Figure la, inset).

At anincubation time of15 min,basal, lipopeptide- and fMLP- stimulated GTP hydrolyseswerelinear up to 10jigofprotein per tube.

Figure 2 shows typical Lineweaver-Burk plots of basal, Pam3AhhSK4-andfMLP-stimulatedGTPhydrolyses in HL-60 membranes. TheKmofbasal high-affinity GTPase in HL-60 mem- branes was 0.45+0.12,M (mean+S.D., n=6). This value corresponds to the Km ofhigh-affinity GTPase in human and rabbitneutrophil membranes(Feltner et al., 1986; Kupper et al., 1992). Thelipopeptide and the formyl-peptideincreased

VmJ..

^of

GTPhydrolysis in HL-60membranes withoutaffecting

K,.

Theinfluence of

Pam.AhhSK4and

fMLP onMg2+-ATPaseand Na+/K+-ATPase wasstudied.Theactivities of these enzymes in HL-60 membranes were 0.26 + 0.06 and 0.15 + 0.02

j#mol/min

permg

respectively.

Pam3AhhSK4 and fMLP(10,M each) did notalter theseenzymeactivities (results not shown).

Concentration/responsecurvesfor the stimulatory effectsof Pam3AhhSK4 and fMLP on high-affinity GTPase in HL-60 membranes are shown inFigure3.

Pam.AhhSK4

activated GTP hydrolysis with an

EC50

^{of about 2,M} and a maximum at

10,uM.

fMLP activated GTPase with an

EC50

of 0.5

jiM

and a plateau at 30-100I M. The effectiveness of Pam3AhhSK4 (100 ,M)to activateGTPase was about 30 % of that of fMLP (100jiM).

Theeffectsoffourlipopeptides(10

jiM

each)on GTPhydro-

lysis

were compared (see Figure 3, inset). Pam3CSK4RPQA-

SGVYMGNLTAQ

contains theepitope of lymphocytic chorio- meningitisvirusnucleoprotein, RPQASGVYMGNLTAQ, used for induction ofcytotoxic

T-lymphocytes

in vivo(Schulzet al.,

Treatments with carrier(control) and PTX were performed as described in the Materials and methods section. High-affinity GTPase activity in HL-60 membranes was determined as described in the Materials and methods section. For determination of basal GTP hydrolysis, water(solvent) was added instead of stimulus.

GTPhydrolysis (pmol/min per mg)

Stimulus Control PTX

Water(solvent) fMLP(10 ,uM) Pam3AhhSK4(10 1sM)

26.0 + 0.5 61.4 +2.1 35.2 +1.0

15.4+0.7 14.8 + 0.5 15.5+0.6

Figure 2 KInetIc analysis ^of hgh-affinity GTP hydrolysis in membranes from Bt2cAMP-dNfferentlated HL-60 cells

GTPhydrolysiswasdeterminedasdescribedintheMaterials andmethods section with GTP (0.07-1.0 1sM)inthepresenceofPam3AhhSK4(10,PM)(*),fMLP(10 ,uM) (M),orwater addedinstead ofstimulus(control) (M). Lineweaver-Burk plotsoftypical experimentsare shown. Thestimulatory effectsof Pam3AhhSK4on GTP hydrolysisweresignificant versus control (P<0.05,Wilcoxontest)underallconditionsstudied.

15 -

E~ ^9-

13-

,~ ~ ~ ~~0_ I I-*I E

.Co 11 -o

CL~ ~ ~ ~ ~

0.001 0.01 0.1 1 10 100

[Substance](pM)

Figure 3 ConcentratIon/respons curvesfor effects ^of Pam,AhhSK and fMLP on high-affinity GTP hydrolysis In membranes from Bt24MP- dMifrentiated HL-60 elis: comparlsonof effctsofvariousllpopeptides High-affinityGTPaseactivityinHL-60 membraneswasdeterminedasdescribed in the Materials andmethods section. Main panel:effects ofPam3AhhSK4 (*)and fMLP(U)atvarious concentrations on GTP hydrolysis. Inset shows effects of various lipopeptides (101sM each) on GTP hydrolysis: Pam3AhhSK4 (LP1), Pam3CSK4RPOASVYMNLTAQ (LP2), Pam3CSK4RPQASGVYMGNLTAQ (LP3), Pam3CSK4YGGFL (LP4).

1991). The peptide RPQASVYMNLTAQ, contained in Pam3CSK4RPQASVYMNLTAQ,lackstwoglycineresidues and

waspreparedascontrolsubstance for induction ofcytotoxicT- lymphocytes(Schulzetal., 1991). Pam3CSK4YGGFLcontains the Leu-enkephalin pentapeptide, YGGFL. Pam3CSK4RPQ-

ASGVYMGNLTAQ and

PamsCSK4RPQASVYMNLTAQ

Table 2 Effectof NEM onhigh-affinityGTPhydrolysisinmembranes from

Bt2cAMP-dlfferentiated

HL-60cells:

interactIon

of PamAhhSK4 withfMLP

Treatments with solvent (control) and NEM were performed as described in the Materials and methods section. High-affinity GTPase activity in HL-60 membranes was determined as described in the Materials and methods section. For determination of basal GTP hydrolysis, water(solvent) was added instead of stimulus.

GTPhydrolysis (pmol/min per mg)

Stimulus Control NEM

Water(solvent) fMLP(10,uM) Pam3AhhSK4 (10 1M) fMLP+Pam3AhhSK4

12.8+0.2 21.4 +0.8 17.0+0.5 23.3 +0.4

6.5+0.1 9.1 +0.2 9.5+0.2 13.3+0.4

weresimilarly effective as Pam3AhhSK4 to stimulate GTPase, whereasPam3CSK4YGGFLwasbyabout 50% less effective.

Table 1 compares theeffects ofPam3AhhSK4 and fMLP on

GTPhydrolysis in control membranes and in membranes from PTX-treated cells. In membranes from PTX-treated cells, the stimulatory effects offMLP andPam3AhhSK4onGTPhydrolysis

were abolished. Similarly to

Pam.AhhSK4.

Pam

CSK4RPQ-

ASGVYMGNLTAQ, Pam3CSK4RPQASVYMNLTAQ and Pam3CSK4YGGFL (10 uM each) had no effect on GTP hydrolysis in membranes from PTX-treated cells (results not shown).

NEM alkylates

G,-protein

a-subunits, thereby uncoupling receptors fromG-proteins ina mannersimilar to thatof PTX (Jakobsetal., 1982; McLeishetal., 1989). In control and NEM- treated membranes, fMLP stimulated GTPase by 67 % and 40% respectively (Table 2). The corresponding values for Pam3AhhSK4 were 33 % and 46 % respectively. In control membranes,Pam3AhhSK4andfMLPactivated GTPhydrolysis in ^a sub-additive manner (82% stimulation), and in NEM- treated membranesthey interacted inan over-additivemanner

(105% stimulation).

In addition to the determination of GTP hydrolysis, photolabellingwith the reactive GTPanalogue GTP azidoanilide is an established method to assess activation ofG1-protein a-

subunitsbyformyl-peptidereceptors. Asreportedfor membranes fromdimethylsulphoxide-differentiatedHL-60cells(Offermanns et al., 1990), fMLP (10 M) increased incorporation of GTP azidoanilide into 40/41 kDa proteins, corresponding tothe a-

subunits of

Gi2

and GI3, in membranes from Bt2cAMP- differentiated HL-60 cells, but Pam3AhhSK4 (10 M) had no

stimulatory effect (Figure 4).

7 0.20 E XI 0.15

.E

E 0.10 -0.05

3-

(a) (b)

2

. . #.. . .

0.1 1 10 100

[Lipopeptidel(pM)

43*

43010-~~~~~~~~~~~~~~* ^{:.: :}

3.*... ..o 30*;^

20 >

..

1 2 3

0.001 0.01 0.1

[fMLPI(WM)

Figure

5 Effects ofPam,AhhSK on02- formation In Bt2cAMP-dlfferentlated HL-60 cells: synergism with fMLP

02-

formationwasdetermined asdescribed in the Materials and methods section. (a)Cells wereexposed toPam3AhhSK4atvariousconcentrationsinthe presence of fMLP(1,M) (U) orwater instead of fMLP(*).(b) CellswereexposedtofMLPatvariousconcentrations in thepresence of Pam3AhhSK4 (10FM)(*)orwaterinstead ofPam3AhhSK4 (-).

Figure 4 Effects of

Pam3AhhSK4

and fMLP on incorporafon of GTP

azidoanhlide

Into, and CTX-catalysed ADP-rlbosylation of

G,-proteln

ac- subunits Inmembranesfrom Bt2cAMP-dlfferentlated HL-60 cells (a) Photolabelling wasperformed as described in the Materials and methods section. The autoradiogram ofanSDSgelcontaining4Mureaand9%(w/v)acrylamideis shown.Lane 1, fMLP(101M); lane2,wateradded instead of stimulus (control); lane3, Pam3AhhSK4 (10,M).(b)ADP-ribosylationwasperformedasdescribed intheMaterials andmethods section.

The autoradiogram of an SDS gel containing 10% (w/v) acrylamide is shown. Lane 1, Pam3AhhSK4 (10,M); lane 2, wateradded instead of stimulus (control); lane 3,fMLP (10

#M).

Numbersontheleftaremolecularmassesof markerproteins(kDa). DF,dyefront.

Table 3 StUmulatlon by

Pam,AhhSK4

and fMLP of PA formaffon in Bt2cAMP-dlfferentlated HL-60

cells:

effectofPTX

Treatments with carrier(control) and PTXwereperformedasdescribed in the Materials and methods section. PAformation wasdetermined asdescribedin the Materials and methods section. The concentration of fMLPwas1,uM and that of Pam3AhhSK4was10,uM. For determination of basal PAformation,water(solvent)wasaddedinstead of stimulus.

PAformation (d.p.m.)

Stimulus Control PTX

Water(solvent fMLP Pam3AhhSK4

652+36 924+66 799+ 33

458 +6 492 +27 497+21

Moreover, CTX-catalysed ADP-ribosylation of

G,-protein

a- subunits was studied. By analogy to dimethyl sulphoxide- differentiatedHL-60cells (Gierschiket al.,1989), fMLP (10

FM)

enhanced CTX-catalysed ADP-ribosylation of

G,-protein

a- subunits in membranes from Bt2cAMP-differentiated HL-60 cells(seeFigure 4).Unlike fMLP,

Pam3AhhSK4

(10 uM) did not enhance ADP-ribosylation of

G,-protein

a-subunits.

We investigated the effects of

Pam3AhhSK4

on

[Ca2+]1

in

Bt2cAMP-differentiated

HL-60 cells. The lipopeptide (0.3- 10

FM)

did notinducerises in

[Ca2+]1

in HL-60 cells (results not shown). By contrast, fMLP potently and effectively increases

[Ca2+1]

in Bt2cAMP-differentiated HL-60 cells (Seifert et al., 1992a,b).

The effects of

lipopeptides

and fMLP on PA formationare shown in Table 3.

Pam3AhhSK4 (10

M) and

fMLP (1 FM)

increased PAformation by 23 % and 42 %respectively. InPTX- treatedcells, thestimulatory effects of Pam3AhhSK4 and fMLP onPA formation were greatly diminished. In the presence of ethanol (0.5 %, v/v),Pam3AhhSK4 (10

FM)

and fMLP(1 FM) stimulatedphosphatidylethanol formation to a similar extent as PAformation(results not shown).

Regulation bylipopeptidesof02-formation in HL-60 cells is depicted inFigure5. Pam3AhhSK4itselfup to 100

FM

did not activate02-formation.However,Pam3AhhSK4 potentiated

O2

formationinducedby fMLP (1FM)with an

EC50

ofabout 2

FM

and a maximum at 10

FM.

fMLPactivated02-formation with an

EC50

of 15 nM and a maximum at 100 nM. Pam3AhhSK4 (10FM) substantially enhanced the effect of fMLP at sub- maximally andmaximally effective concentrations.

DISCUSSION

Lipopeptides activate 02- formationinhuman neutrophils in a PTX-sensitivemanner,suggestingtheinvolvementofG1-proteins in thesignal-transduction pathway (Seifertetal., 1990). Sofar, however, a heptahelical receptor for lipopeptides has not yet been identified. In addition, lipopeptides, unlike chemo- attractants, do not induce rises in

[Ca2+],

in HL-60 cells (Wenzel-SeifertandSeifert, 1993; Seifertetal., 1992a,b). Thusit is unlikely that lipopeptides activate Bt2cAMP-differentiated HL-60 cells through a chemoattractant receptor with known ligands orthrough thenewly clonedformyl-peptide-related re- ceptor with as-yet unknown ligands (Murphy et al., 1992).

Thereforewehaveputforward thehypothesisthatlipopeptides activate G1-proteins directly, i.e. in a receptor-independent manner(Seifertetal., 1990).Thisassumptionissupportedbythe finding thatonly

lipopeptides

bearing positive chargesactivate 02- formation in human neutrophils (Seifert et al.,

1990).

By analogytolipopeptides,thestimulatory effects ofvarious direct G-proteinactivators inneutrophilsare

inhibited,

atleast inpart, by PTX (Serra et al., 1988; Perianin and

Snyderman,

1989;

Seifertetal., 1990, 1992b;Norgaueretal., 1992;Kanahoetal., 1992). To corroborate our

suggestion further,

we have now studiedtheeffects of

lipopeptides

inBt2cAMP-differentiatedHL- 60cells,asthese cellsare auseful modelsystemfor the

analysis

of

Ga-protein-mediated signal-transduction

processes at the

940,,

670'

94 >, 670

-o

UE

0

'._E

C C 0

0

.:.9,...

....x..'..<..

30>

1 2 3

1

Wefound that the lipopeptide Pam3AhhSK4 activated high- affinity GTP hydrolysis in HL-60 membranes in a time-, membrane-protein- and stimulus-concentration-dependent manner(see Figures 1 and 3). Most importantly, the stimulatory effects of lipopeptidesonGTPase were inhibited by PTX, as were those of fMLP(seeTable 1). These data indicate thatlipopeptides increase the GTPase activity of

G1-proteins.

Pam3AhhSK4 increased

Vm..

of GTP hydrolysis, suggesting that it stimulated the catalytic rate ofGTP turnover (see Figure 2). By analogy with lipopeptides, formyl-peptides also increase

VmJ-

of GTP hydrolysis (seeFigure 2) (Feltner etal., 1986). The stimulatory effect of Pam3AhhSK4 on high-affinity GTPase was not unspecific, inasmuchasthelipopeptide didnotaffect theactivities of othernucleotide-metabolizingenzymes,i.e.Mg2+-ATPase and Na+/K+-ATPase, inHL-60membranes.

NEM, through alkylation of

G,-protein

a-subunits, disrupts the interaction ofheptahelical receptors with G-proteins in a manner similar to that of PTX-catalysed ADP-ribosylation (Jakobs et al., 1982; McLeish et al., 1989). With respect to the interaction of formyl-peptide receptors with

G,-protein

a-subunits, alkylation and ADP-ribosylation have similar consequences, i.e they are inhibitory (see Tables 1 and 2) (McLeishetal., 1989). Concerning the interaction of lipopeptides with

G,-proteins,

ADP-ribosylation is alsoinhibitory(seeTable 1). Unexpectedly, alkylationwasfoundto enhance the relative extent oflipopeptide-stimulated GTP hydrolysis (see Table 2).

These datasuggest thatalkylation and ADP-ribosylation ofa- subunits may affect G-protein function in different manners.

Alkylation of

G(-protein

a-subunits, unlike ADP-ribosylation, mayincrease theirlipophilicity, and therebymayfacilitate their interaction with lipophilic portions of lipopeptides. It is also possible that alkylation and ADP-ribosylation induce different conformations of

G,-protein

a-subunits, resulting in opposite effects concerning their interaction with lipopeptides, but not withregardtoreceptors.-Additionally, Pam3AhhSK4 and fMLP activated GTPasein NEM-treated membranesin a synergistic manner. Takentogether,these results suggestthatlipopeptides and formyl-peptides differently activate the GTPase of

G1-

proteins.

Certainlipopeptides, i.e.

Pam3CSK4

and

Pam3AdhSK4,

donot stimulate GTP hydrolysis in membranes from Bt2cAMP- differentiatedHL-60cells(Seifertetal., 1992b).Itispossible that the extent of activation of

G,-proteins

induced by these lipopeptides is too small to be detected in the GTPase assay.

Pam3CSK4,

Pam3AdhSK4 and

Pam.AhhSK4

have in common the positively charged peptide chain, but they differ in the structureof thelipid moiety (Seifertetal., 1992b). Withrespect toGTPhydrolysis, only

Pam.AhhSK4

^isstimulatory (see Figure

3) (Seifert

etal.,

1992b).

Thesedifferential effects of

lipopeptides

onGTPhydrolysissupportthe view that these substances donot activate the GTPase of

G,-proteins

in an unspecific manner.

Intriguingly, the resultsofa recentstudy indicatethat

lipophilic

amino acids in the third cytoplasmic loop of heptahelical receptors may be moreimportant for G-protein activation than are

positively

charged amino acids (Cheung et

al., 1992).

By analogy, the structure of the lipid moiety of

lipopeptides

is important for their G-protein-activating

properties

(see

Figure 3)

(Seifert et al., 1992b). However, the peptide chain is also of relevance for G-protein activation by

lipopeptides.

This as-

sumption

is

supported by

the

findings

that the

lipopeptides Pam3CSK4, Pam3CSK4RPQASGVYMGNLTAQ, Pam8CSK4-

RPQASVYMNLTAQandPam3CSK.YGGFLhaveincommon

The microbialalkaloidstaurosporine isa potentinhibitor of protein kinases, in particular of protein kinase C (Tamaokietal., 1986).Inaddition, staurosporine stimulates phospholipase D in rabbitneutrophils, presumablyvia directactivation of

Gi,

butit does not activate phospholipase C (Kanaho et al., 1992). By analogy, positively charged lipopeptides inhibit protein kinase C anddonotinduce rises in

[Ca2+]i

inHL-60 cells (Bessler, 1990;

Seifert et al., 1992b). These findings prompted us to study the effects of

Pam.AhhSK4

^on phospholipase D in HL-60 cells. Pam3AhhSK4 activated the formation of PA and of phosphatidylethanol, a transphosphatidylation product specificallyformedby phospholipase D (see Table 3) (Paietal., 1988; BourgoinandGrinstein, 1992). Additionally, stimulation by

PamkAhhSK4

of PAformationwasPTX-sensitive(see Table 3). These data indicate that the lipopeptide activates phospholipaseDvia

G,-proteins.

In this context, the

question

arises whether the effects of Pam3AhhSK4^on phospholipaseD weremediated via Gi2, Gi3

or both Ga-proteins. Therefore we studied the effects of Pam3AhhSK4 ^on photolabelling and CTX-catalysed ADP- ribosylationof

G,-protein

a-subunitsinHL-60membranes. Both methods have been shown to be useful forthe analysis of G-

protein activation by formyl-peptide receptors in membranes fromdimethyl sulphoxide-differentiated HL-60 cells (Gierschik et al., 1990; Offermanns et al., 1990). In agreement with the aforementionedstudies,wefoundprominent stimulatory effects of fMLP on photolabelling and CTX-catalysed ADP- ribosylation of Ga-protein a-subunits in membranes from Bt2cAMP-differentiated HL-60 cells (see Figure 4). However, Pam3AhhSK4 ^was devoid of any stimulatory effect in these regards.Thuswecannotyetanswerthequestionastowhich

G,-

protein subtypesareinvolved in theactivation ofphospholipase D by lipopeptides. Interestingly, some lipopeptides do not measurably stimulate theGTPase of

G,-proteins

(Seifert etal., 1992b). Moreover, even the most effective lipopeptides with regard to GTPase activation, i.e. Pam3AhhSK4, Pam3CSK4- RPQASGVYMGNLTAQand Pam3CSK4RPQASVYLMNLT- AQ,arestill muchlesseffective than fMLP(see Figure 3). Thus the sensitivityofphotolabellingand CTX-catalysed ADP-ribo- sylationmay betoolow todetect therelatively smallextentof

Gi-protein

activation by Pam3AhhSK4. However, itcannot be excluded that the lack ofstimulatory effect of lipopeptides on

photolabellingandCTX-catalysed ADP-ribosylation reflects also qualitative differences in activation of

G,-proteins

by these substances andby formyl-peptides (McLeishet al., 1989).

The dataobtained withPam3AhhSK4 concerning stimulation of PAformationimplythatanincrease in[Ca2+]1 viaactivation ofphospholipaseCand/ornon-selectivecation channels isnot

aprerequisite forstimulation ofphospholipaseD(see Table 3) (Seifert et al., 1992b). In agreement therewith, Rosoff et al.

(1988), Bourgoinetal.(1990)and Kanahoetal.(1992) suggested thatphospholipase Dis activated in a Ca2+/phospholipase C- independentmanner. However, in human neutrophils, rises in [Ca2+]1 may play apart in the activation ofphospholipase D (Kessels etal., 1991; Bauldryet al., 1992).

The effectivenessofPam3AhhSK4 to activate PAformation

wasabout 50%of thatof fMLP(seeTable3). Nonetheless,the lipopeptideitself didnotactivate 2-formation(see Figure 5).

Thus stimulation ofphospholipase D is not sufficient for the activation of NADPH oxidase.Our dataareinagreementwith thoseobtainedbyKesselsetal.(1991)andBauldryet al.(1992).

By analogy to thecytokines, granulocyte/macrophage-colony- the

lipid portion,

but differ in the

peptide

chain and their

stimulating

factor and^tumouir necrosis

factor-a,

activation of

PAformationby lipopeptidesmay beimportantforpotentiation of fMLP-induced 02- formation (see Table 3 and Figure 5) (Bourgoin etal., 1990; Bauldryetal., 1990).

In conclusion, we have shown thatlipopeptides activate the GTPase of G1-proteins and that lipopeptides and fMLP activate

G,-proteins

differently. Lipopeptides activate phospholipase D via

G,-proteins,

but PA formation is not sufficientforthe activationof02-formation.

J.F. K.is a recipientofapredoctoralfellowship of the FreieUniversitat Berlin. We appreciatethesupplyoflipopeptides byDr.J.Metzger,Dr. K.-H.Wiesmullerand Dr.

G. Jung,Institut furOrganische ChemiederUniversitatTubingen,Germany,andby Dr. C.Sakarellos, Department of Chemistry, University of loannina, Greece.

Stimulatingdiscussions with Dr. J. Metzger, Dr. K.-H.Wiesmuller and Dr. G.Jung arealso acknowledged. We also thank Dr. G.Schultz, Institut fur Pharmakologie, Freie UniversitatBerlin,forhelpfuldiscussion. We aregratefultoMiss E. Bombien and Mrs. E. Glass for expert technical assistance. This work wassupported bygrants of the Deutsche Forschungsgemeinschaft.

REFERENCES

Ames, B. N.(1966) Methods Enzymol.8,115-118

Bauldry, S.A., Bass, D. A., Cousart, S. L. and McCall, C. E. (1991) J. Biol. Chem. 266, 4173-4179

Bauldry, S.A., Elsey, K.L. andBass, D. A.(1992) J. Biol. Chem.267, 25141-25152 Bessler, W. G. (1990)bga-Schriften 1, 196-204

Bessler,W.G., Cox, M., Lex, A., Suhr, B.,Wiesmuliler,K.-H.andJung, G. (1985) J. Immunol.135,1900-1905

Bourgoin, S. andGrinstein, S. (1992) J. Biol. Chem.267,11908-11916 Bourgoin,S., Plante, E., Gaudry, M., Naccache, P. H., Borgeat, P. and Poubelle,P.E.

(1990) J.Exp. Med. 172, 767-777

Braun, V.(1975) Biochim. Biophys. Acta415,335-377

Cassel, D. andPfeuffer,T.(1978) Proc. Natl. Acad. Sci. U.S.A. 75, 2669-2673 Cheung,A.H., Huang, R.-R. C. andStrader,C.D.(1992) Mol. Pharmacol. 41, 1061-1065 Ebel, H.,Aulbert,E. and Merker, H. J. (1976) Biochim. Biophys. Acta433,531-546 Feltner, D.E.,Smith, R. H. and Marasco,W. A.(1986) J. Immunol. 137, 1961-1970 Gierschik, P.,Sidiropoulos, D. and Jakobs,K.H.(1989)J.Biol. Chem.264, 21470-21473 Higashijima,T., Uzu, S., Nakajima, T. and Ross, E. M. (1988) J. Biol. Chem. 263,

6491-6494

Higashijima,T., Burnier, J. and Ross, E. M.(1990) J. Biol. Chem.265, 14176-14186 Jakobs, K.H., Lasch, P., Minuth, M.,Aktories,K.andSchultz,G.(1982) J. Biol. Chem.

257,2829-2833

Kanaho,Y., Takahashi, K.,Tomita,U., liri, T., Katada, T., Ui, M. and Nozawa, Y. (1992) J. Biol.Chem.267, 23554-23559

Kessels, G. C. R., Roos, D. and Verhoeven, A. J. (1991) J. Biol.Chem.266, 23152-23156 Krautwurst, D.,Seifert, R., Hescheler, J. and Schultz, G. (1992) Biochem. J. 288,

1025-1035

Kupper, R. W.,Dewald, B., Jakobs, K. H., Baggiolini, M. andGierschik, P. (1992) Biochem. J.282, 429-434

Lowry,0. H., Rosebrough, N. J., Farr,A.L.andRandall, R. J. (1951) J. Biol. Chem.193, 265-275

McLeish,K.R.,Gierschik, P., Schepers, T.,Sidiropoulos, D. and Jakobs,K.H. (1989) Biochem. J.260,427-434

Mousli, M.,Bronner, C., Landry, J.,Bockaert, J. and Rouot, B. (1990)FEBS Lett.259, 260-262

Murphy,P.M.,Ozcelik,T., Kenney, R. T.,Tiffany, H. L.,McDermott, D.and Franck, U.

(1992) J. Biol.Chem.267, 7637-7643

Norgauer,J., Eberle, M., Lemke, H.-D. andAktories, K. (1992) Biochem. J.282, 393-397 Offermanns, S., Schafer, R., Hoffman, B., Bombien, E., Spicher, K., Hinsch, K.-D.,Schultz,

G. andRosenthal,W. (1990)FEBS Left.260,14-18

Offermanns, S.,Schultz,G.andRosenthal, W. (1991) Methods Enzymol. 195,286-301 Offermanns, S.,Seifert,R., Metzger, J. W., Jung,G., Lieberknecht, A.,Schmidt, U. and

Schultz, G. (1992) Biochem. J.282, 551-557

Pai, J.-K.,Siegel, I.,Egan, R. W. and Billah, M. (1988) J. Biol. Chem. 263, 12472-12477 Perianin, A. andSnyderman, R. (1989) J. Immunol.143, 1669-1673

Resch,K.andBessler, W. (1981) Eur. J. Biochem. 115, 247-252

Rosenthal, W.,Koesling, D., Rudolph, U., Kleuss, C., Pallast, M., Yajima, M. and Schultz, G.(1986) Eur. J. Biochem.158, 255-263

Rossi, F. (1986) Biochim. Biophys. Acta853, 65-89

Rosoff, P. M.,Savage, N. and Dinarello, C.A.(1988)Cell 54, 73-81

Schulz, M.,Zinkernagel,R. M.andHengartner,H.(1991) Proc. Natl. Acad.Sci. U.S.A. 88, 991-993

Seifert, R. andSchultz, G.(1987) Biochem. Biophys. Res. Commun.146,1296-1302 Seifert, R. andSchultz,G.(1991) Rev. Physiol. Biochem. Pharmacol. 117,1-338 Seifert, R., Schultz,G.,Richter-Freund, M., Metzger, J.,WiesmUller, K.-H., Bessler, W. G.

and Hauschildt, S. (1990) Biochem. J.267, 795-802

Seifert, R., Hoer A.,Offermanns,S., Buschauer, A. and Schunack,W. (1992a) Mol. Pharmacol.42, 227-234

Seifert, R., Serke,S.,Huhn, D., Bessler, W. G., Hauschildt, S., Metzger, J., Wiesmuller, K.-H.andJung, G. (1992b). Eur. J. Biochem. 203,143-151

Serra, M. C., Bazzoni, F., Della Bianca, V., Grzeskowiak, M. and Rossi, F. (1988) J.Immunol. 141, 2118-2124

Tamaoki, T., Nomoto, H., Takahashi, I., Kato, Y., Morimoto,M.andTomita, F. (1986) Biochem.Biophys. Res. Commun. 135, 397-402

Tomita, U., Inanobe,A., Kobayashi, I.,Takahashi, K., Ui, M. and Katada,T.(1991) J.Biochem. (Tokyo) 109, 184-189

Walseth, T. F.,Yuen, P. S.T. and Moos, M. C. Jr. (1991) MethodsEnzymol. 195, 29-44 Wenzel-Seifert, K.and Seifert,R.(1993) J.Immunol.150, 4591-4599

Received 20May 1993/26 July 1993; accepted 30 July1993