The protein kinase C inhibitor H-7 activates human neutrophils: effect on

shape, actin polymerization, fluid pinocytosis and locomotion

H. U. KELLER, V. NIGGLI, A. ZIMMERMANN

Institute of Pathology, University of Bern, Freiburgstrasse 30, CH-3010 Bern, Switzerland

and R. PORTMANN

NC-Laboratory Spiez, CH-3700 Spiez, Switzerland

Summary

The present study demonstrates new properties ofH-7. The protein kinase inhibitor H-7 is a potentactivator of several neutrophil functions. Stimu-lation of initially spherical nonmotile neutrophilselicits vigorous shape changes within a few seconds,increases in cytoskeletal actin, altered F-actin distri-bution, increased adhesiveness and a relatively smallincrease in pinocytic activity. H-7 has also chemoki-netic activities. Depending on the experimentalcondition, H-7 may elicit or inhibit neutrophil loco-motion. It failed to induce chemotaxis. Thus, theresponse pattern elicited by H-7 is different from that

of other leukocyte activators such as chemotacticpeptides, PMA or diacylglycerols. The finding thatH-7 can elicit shape changes, actin polymerizationand pinocytosis suggests that these events can occurwithout activation of protein kinase C (PKC). PMA-induced shape changes and stimulation of pino-cytosis were not inhibited by H-7.

Key words: H-7, protein kinase C, neutrophil granulocytes,shape, actin, pinocytosis.

Introduction

Activators of protein kinase C (PKC), such as phorbolesters or diacylglycerols, stimulate several neutrophilfunctions including the metabolic burst, exocytosis (Smithet al. 1988), pinocytosis (Keller and Zimmermann, 1987;Robinson et al. 1987), shape changes (Roos et al. 1987;Zimmermann etal. 1988), actin polymerization (Rao, 1985;Roos et al. 1987; Zimmermann et al. 1988; Sha'afi et al.1983), surface adhesion and others. Other neutrophilfunctions as, for example, directional locomotion (Gallinand Wright, 1978), polarization or locomotor activity (Rooset al. 1987) are suppressed by these PKC activators. Thefindings suggest that the diacylglycerol-protein kinase Cpathway is involved in signal transduction of these par-ticular functions. Therefore, one expects that inhibitors ofprotein kinase C may block these responses. H-7 was foundto inhibit PKC and other kinases (Hidaka et al. 1984).Several previous studies demonstrated inhibitory effectson neutrophil responses, including chemotactic peptide-induced chemotaxis and locomotion (Gaudry et al. 1988),phosphorylation of the 50000Mr protein kinase C sub-strate stimulated by PMA (Sha'afi et al. 1986) and neutro-phil superoxide anion release in response to PMA and thecalcium ionophore A 23187 (Berkow et al. 1987).

The present work shows new properties of H-7. H-7 is byitself an activator of several leukocyte responses includingshape changes, actin polymerization, fluid pinocytosis andchemokinesis.Journal of Cell Science 96, 99-106 (1990)Printed in Great Britain © The Company of Biologists Limited 1990

Materials and methods

MaterialsReagents and suppliers were: iV-formyl-L-norleucyl-L-leucyl-L-phenylalanyl-L-norleucyl-L-tyrosyl-L-lysine (fNLPNTL), Bachem,Bubendorf/Switzerland; human serum albumin (HSA), Behr-ingwerke, Marburg/FRG; NBD-phallacidin (Molecular Probes,Inc., Junction City, Oregon/USA; phorbol-12-myristate-13-acet-ate (PMA), FITC-dextran FD-70, lysolecithin (L-cMyaophosphati-dylcholine, palmitoyl), H-7 (l-(5-isoquinolinesulfonyl)-2-methylpiperazine), paraformaldehyde, EGTA, Sigma Chemical Corp., StLouis, MO/USA; glutaraldehyde, N-2-hydroxyethylpiperazine-iV'-2-ethanesulfonic acid (Hepes), Serva Feinbiochemica,Heidelberg/FRG; neutrophil isolation medium (NIM), PackardInstrument International SA; diisopropylfluorophosphate (DFP,95.7% pure), Fluka AG, Buchs/Switzerland. Water-insolublecompounds were dissolved in dimethylsulfoxide (DMSO), FlukaAG, Buchs/Switzerland. Triton X-100 was obtained as a 10%solution" in water from Pierce (Oud-Beijerland/Netherlands) andstored under nitrogen. Actin was prepared from rabbit skeletalmuscle as described by Pardee and Spudich (1982).

The medium was usually prepared as follows: 138 mM NaCl,6mMKCl, 1.1 mM EGTA, 1 mM Na2HPO<p 5min NaHCO3,1.1 mMCaCl2, lmM MgSO4l 5.5mM glucose, 20mM Hepes, 2% (w/v)HSA. If so indicated, the medium was modified by (1) replacing2 % HSA with 0.1 % HSA (actin assays); or (2) by omitting CaCl2and MgS04, and/or by adding 10 mM EDTA. In most assays (seeFig. 4, below; pinocytosis), or shape changes (data not shown), theresults with or without divalent cations were similar. SincefNLPNTL-stimulated cells showed increased adhesion at the tailend and, therefore, failed to migrate in the absence of Ca2+ and

99

Mg2"1", locomotion assays were performed in the presence ofdivalent cations.

Cell preparationNeutrophils were isolated from heparinized (10 units ml"1) hu-man blood withdrawn from healthy volunteers. In the first step,red cells were removed by separation over Isopaque-Methocel(Bo'yum, 1968). In the second step, neutrophils and mononuclearcells were usually separated by means of NIM (Ferrante andThong, 1980). The white cells in these preparations consisted of93-99.5 % neutrophils. There were between 2 and 17 platelets perneutrophil. Cells used in locomotion assays were separated byIsopaque-Methocel followed by a discontinuous Ficoll/Isopaquegradient (Bo'yum, 1968).

Cell shape assayCells (108ml~1) were preincubated in a reciprocating water bathat 37 °C for 10 min, and then incubated with or without stimuli atuniform concentrations for another 30 min. Cells were fixed in 1 %glutaraldehyde (final concentration) at 37 °C for 30 min, washedand resuspended in 0.9% NaCl, containing 1 mgml"1 NaN3. Theshape of fixed cells was analyzed by means of differential inter-ference contrast microscopy (DIC, Nomarsky optics; Zeiss EM 35microscope, 100X objective, NA 1.25). Shape changes in livingcells were analyzed using the same microscope, the stage beingheated to approx. 37 °C with an airstream stage incubator (Nichol-son Precision Instruments Inc., Bethesda, MD).

Basically, neutrophils were classified by shape as previouslydescribed (Roos et al. 1987) with one additional category ((4) typeB polarity). We distinguished between the following categories.

(1) Spherical cells with a smooth surface lacking surfaceprojections. These cells are characteristic of unstimulated con-trols (Fig. 2A).

(2) Spherical smooth cells with small unifocal projections thatare shorter than the diameter of the respective cell body.

(3) Front-tail polarity (type A polarity), i.e. cells with contrac-ted rear end (tail) and ruffles at the expanding front (see Fig. 2C,120 s).

(4) Type B polarity: these cells show unifocal projections thatare longer than the diameter of the body of the respective cell,which lacks the tail knob. The cell body at the rear end tends to bespherical rather than triangular (Fig. 2D, 0 s, below).

(5) Nonpolar cells with surface projections. These cells mayproduce protrusions at one or several sites, without developingclear-cut polarity (Fig. 2C, 80 s and Fig. 2E, below).

Neutrophil locomotion, chemokinesis and cell-substratum contactsIn the first series of experiments (Table 1) the locomotor behaviorof neutrophils was studied in slide-coverslip preparations, whichare narrow enough to prevent floating of neutrophils. Cells(0.75xl06/6 jil) were placed on a slide, a round coverslip (25 mmdiameter) was placed on top and the preparation was sealed withparaffin. In another series of experiments locomotion was deter-mined in Sykes-Moore chambers with an approximate depth of3 mm. The locomotor behavior of cells was then recorded at 37 °C

for 10 min using videomicroscopy. The proportion of locomotingcells and the speed was determined by morphometry. Also thecontact areas between coverslip and cells was assessed usingreflexion contrast microscopy (Keller et al. 1983). Directionallocomotion was determined using a two-filter assay as previouslydescribed (Keller et al. 1983).

Fluid pinocytosis of FITC-dextranNet uptake of FTTC-dextran was used as a measure for fluidpinocytosis (Davis et al. 1982). Neutrophils (106 cells ml"1) werepreincubated in plain medium at 37 °C for 10 min. FTTC-dextran(5 mg ml"1 final concentration) was added with or without stimuliat the concentrations indicated. After 30 min of incubation thereaction was stopped by fixation with 1 % glutaraldehyde (finalconcentration) at 37 °C for 10 min. The cells were washed fivetimes in phosphate-buffered saline (pH7.4), and the relativefluorescence intensity was determined immediately by flow cyto-metry (Ortho cytofluorograph-50, Ortho Instruments, Raritan,New Jersey /USA), using a 5 W argon laser, an excitation wave-length of 488 nm, and a long-pass filter giving an emissionwavelength between 515 and 555 nm. The increase in the netuptake of FITC-dextran was determined as follows: the differ-ence in the median channel number of matched samples incu-bated with and without FITC-dextran was calculated. Themedian value for the net uptake of FITC-dextran by unstimu-lated cells was set as zero and the shift in the median channelnumber following stimulation is shown.

Distribution of cellular actin between the cytosol andcytoskeletonTo prevent proteolytic degradation of cytoskeletal proteins(Amrein and Stossel, 1980), neutrophils were incubated with5 mM DFP on ice for 5 min, and resuspended in medium with 0.1 %HSA, 10 mM EDTA without divalent cations before stimulation.Cytoskeletal structures were subsequently isolated as describedby White et al. (1982), with small modifications described else-where (Keller et al. 1989a). Pellets and supernatants togetherwith known amounts of purified rabbit skeletal muscle actin weresubsequently electrophoresed through a 7.5% to 15% gradientpolyacrylamide slab gel (Laemmli, 1970), stained with CoomassieBrilliant Blue R, and scanned at 590 nm with a Camag TLCscanner. From comparison with known amounts of purified actin,the amount of actm in the cytoskeleton and the cytosol could bedetermined. F-actin was localized by NBD-phallacidin stainingas previously described (Roos et al. 1987).

Results

H-l'-induced shape changesShape changes occur within 10 s after addition of H-7(Fig. 1). At this time point, the shape changes induced byH-7 are similar to those elicited by 10~ 9 M fNLPNTL, i.e.almost all cells (97-100%) show generalized surface ruf-fling without substantial elongation of the cells. Between

Table 1. Effect of H-7 on locomotion, adhesion and cell polarity of human neutrophils

StimuluB

Cellsmigrated

(%)

8.7±4.10

59 3 ±7.81.7±1.0

Locomotion

Mean speed(/on min"1)

Locomotingcells

4.3±10

9.4±0.41.8±1.0

All cells

0.45±0.270

5.5±0.50.06±0 03

Adhesioncontact area

0.5±0.3161±8106±3161±24

Shape

TypeApolarity

(%)

2±26±2

69±36±2

in suspension

TypeBpolarity

(%)

07±21±13±1

NoneH-7 (300 JIM)fNLPNTL (10"9M)H-7 (300 j(M)

+ £ N L P N T L ( H T 9 M )

Cells were separated with Isopaque-Methocel followed by a Ficoll gradient and suspended in 2 % HSA-Gey's-Hepes with Caa+ and Mg2*. Neutrophillocomotion was measured in slide-coverslip preparations. Adhesion was determined in Sykes-Moore chambers. Mean of three experiments±s.D.M.

100 H. U. Keller et al.

Fig. 1. Shape changes and altered F-actin distribution in human neutrophils stimulated by H-7. A. Unstimulated controls, 30min;B, 300/JM H-7, 30 s; C, 300 /an H-7, 30min. Neutrophils in 2 % HSA-GeyV-Hepes without added Ca2+ and Mg24" were preincubatedat 37 °C for lOmin in plain medium followed by incubation with or without H-7 for the time indicated. NBD-phallacidin staining ofparaformaldehyde-fixed cells. Photographs were taken with DIC optics (top) and fluorescence microscopy (bottom). Bar, 10 /<m.

1 and 5 min after stimulation with 300 JXM H-7 or 10 9 MfNLPNTL, neutrophils develop different shapes. WhereasfNLPNTL-treated cells gradually acquire type A polarity,H-7-treated neutrophils show preferentially type B po-larity or nonpolar cells with surface projections. Cellstreated with H-7 for longer than 5 min often show a moreor less spherical cell body with one or few fairly largeprojections (Figs 1 and 2). Many of these cells, in particu-lar those that show a single large surface projection, maysuperficially resemble polarized cells as induced by chemo-tactic peptides (type A polarity). However, the shapediffers from classical front-tail polarity (type A polarity)observed after stimulation with chemotactic peptides inthat H-7-treated cells show a fairly rounded cell bodywithout a persisting contracted area (tail knob, triangularshape) at the rear end of the cell. The one-sided pro-trusions showing ruffles at the front end are often ex-tremely long (type B polarity; Fig. 2C, 0 s; Fig. 2D, 0 s and20 s). In some cells the one-sided protrusion may persist fora fairly long time (Fig. 2D, upper cell), in others it is lessstable (Fig. 2D, lower cell). Large single protrusions maybe withdrawn and finally formed anew at the opposite sideof the cell (Fig. 2B). Or, they may split and temporarilyform two very large protrusions on opposite sides of thecell (Fig. 2C). Cells with structures resembling a tail occuras well (Fig. 2B, 0 s, or C, 120 s or 160 s) and were classifiedas polarized (type A polarity) in the shape analysis.Usually, they develop when one of the large pseudopodscontracts, as seen in Fig. 2C 80 s (cell with lamellipodia onopposite sides) versus Fig. 2C 120 s or 160 s (cell withcontracting area on one side and lamellipodia on theopposite side), similar to lateral pseudopods in polarizedcells, or they appear transiently at the rear end of cells.However, they do not usually persist (Fig. 2C), like thetails in cells polarized by chemotactic peptides.H-7-stimulated cells may also simultaneously form pro-trusions into different directions (Fig. 2E). The nonpolarcells with surface projections elicited by H-7 differ fromnonpolar neutrophils induced by PMA, diacylglycerols or2H2O in that the protrusions are larger and a spherical cell

body of reduced size is often clearly visible. The sequencesshow that the movements are quite vigorous (Fig. 2). Thedose-response curve of neutrophils stimulated with H-7for 30 min is shown in Fig. 3.

Actin reorganization in response to H-7H-7 produced a dose-dependent increase in cytoskeleton-associated actin and, at somewhat lower concentrations, afairly parallel increase in the proportion of cells undergo-ing shape changes including polarized cells (type A and B)and nonpolar cells with surface projections (Fig- 3). At10~ 5 M H-7 there was an increase in the proportion ofspherical cells with a single small projection to about 40 %,which was not associated with a measurable increase inpolymerized actin. Thus, the shape change response is inthis case a slightly more sensitive parameter for neutro-phil activation than the measures for the amount ofcytoskeletal actin. The increase in cytoskeletal actinobserved following stimulation with H-7 was in a rangesimilar to that following stimulation with the chemotacticpeptide fNLPNTL (Fig. 3). F-actin as detected byNBD-phallacidin binding showed a fairly homogeneousdistribution in unstimulated spherical cells (Fig. 1). At30-60 s following exposure to 300 /.IM H-7 the cells showeda marked increase in F-actin associated all along the cellmembrane. At 30 min, neutrophils had developed longprojections and F-actin was found preferentially in projec-tions, whereas the cell-body contained less F-actin than inunstimulated controls.

Effect of H-7 on fluid pinocytosis of FITC-dextranH-7 alone has a small dose-dependent effect on the netuptake of FITC-dextran by human neutrophils (Fig. 4A).Also the small stimulating effect of 10~ 9 M fNLPNTL onfluid pinocytosis is slightly increased rather than in-hibited by simultaneous addition of H-7, indicating anadditive effect (Fig. 4B). PMA alone has a very strongstimulating effect on pinocytosis of FITC-dextran, whichis not inhibited by simultaneous addition of the protein

Protein kinase C inhibitor H-7 101

Fig. 2. Dynamics of shape changes of individual living neutrophils stimulated with or without H-7. Unstimulated controls (A); H-7,300 ;IM (B,E); H-7, 100/IM (C,D). Neutrophils in 2 % HSA-Gey's-Hepes without Ca2+ und Mg2"1" but with 10 HIM EDTA werepreincubated at 37°C for 10 min and then incubated with or without H-7 for 30 min. Sequential photographs of living cells were thentaken with DIC optics at the time intervals (seconds) indicated. Bar, 10 fan.

102 H. U. Keller et al.

100-

90-

80-

70-

E 60-&5O-TO

30-

20-

10-

0- :-{ ' > .

20

18

16

14

• 1 2

•10

• 8

/—\auca

~a

toi

o

tin

(9i

1 ac

l:le

ta

in

mtv

DC

•6

14-

12-

10-

8-

6-

4 •

2-

0-

0 10"7 5.KT7 10"6 5.10~6 10"3 5.10"5 10"4 3.10"4

H-7 concentration (M)

Fig. 3. H-7-induced shape changes and cytoskeletal actin:dose-response relationship. The proportion of all neutrophilswith marked shape changes (nonpolar cells with surfaceprojections, type A and type B polarity) (A A), of cellsshowing type B polarity only (D • ) and of spherical cellswith a short unifocal protrusion ( • - — • ) was determined.Cells were preincubated in 2 % HSA-Ge/s-Hepes with Ca2+

and Mg^+ at 37°C for lOmin and then incubated with H-7.After 30 min of incubation the cells were either fixed withglutaraldehyde for microscopic examination, using DIC optics,or processed for assessment of the amount of cytoskeletal actin(* *) as described in Materials and methods. The value forthe percentage of cytoskeletal actin obtained with cells exposedto 10~9M flSTLPNTL for lmin was 17.3±1.5% in the sameseries of experiments. Mean of three experiments±s.D.M.

kinase inhibitor H-7 (Fig. 4C). Basically similar resultswere obtained when neutrophils were first incubated withH-7 alone for 30 min and then with H-7 and PMA foranother 30 min (data not shown).

Adhesiveness, chemokinesis and directional locomotionOne series of experiments with Sykes-Moore chamberswas conducted after stimulation with 100 ^M H-7 for30 min. In contrast to neutrophils stimulated with H-7,control cells were mostly spherical, non-motile and non-adherent. The percentage of migrating cells was 1 ± 1 % inthe controls, 77±10% after stimulation with 100/XM H-7and 4±3% with 300 /ZM H-7. The mean speed of themigrating cells stimulated with 100//M H-7 was 7±l/miper minute (mean of 3 experiments±S.D.M.). Thus, H-7 iscapable of stimulating locomotion (Fig. 5) under appropri-ate conditions. The locomotor activity decreases withprolonged incubation in the Sykes-Moore chambers. It isalso lower in slide-coverslip preparations than in Sykes-Moore chambers, probably because the contact areas arelarger (data not shown).

The chemokinetic effect of 300 /JM H-7 has also beentested in slide-coverslip preparations, using either un-stimulated control cells, which show relatively little loco-motor activity, or chemotactic peptide-stimulated cells,which show stimulated locomotor activity. Under theseparticular conditions H-7 inhibits locomotion of controlcells or cells stimulated with 1 0 " 9 M fNLPNTL almostcompletely (Table 1), i.e. it has a negative chemokineticeffect. Thus, the shape changes elicited by H-7 at thisparticular stage (mainly nonpolar cells with surface pro-jections) are probably not instrumental in promotinglocomotion. The inhibition of locomotion associated with

14-

12-

e

2 6

4

2

0

70

60

50

40

30

20

10

0

A. H-7 only

**rX-..-\.\B. H-7+fNLPNTL

C. H-7+PMA

o icr7 5 to-7 ICT4 5 ur" KT' 5.HT! io-4 3.10-1

H-7 concentration (M)

Fig. 4. Effect of H-7 on fluid pinocytosis. Uptake ofFITC-dextran by neutrophils stimulated by H-7 alone (A), orsimultaneously by 10~9 M fNLPNTL and increasingconcentrations of H-7 (B), or 10~ 9 M PMA and increasingconcentrations of H-7 (C). The net uptake of FTTC-dextran isexpressed as shift in the median channel number (uptake bystimulated neutrophils minus uptake by unsthnulated cells).The net uptake measured in unstimulated neutrophils in amedium with or without divalent cations was 4.1 and 6.8,respectively. Note the difference in scale in A and B versus C.Neutrophils in medium containing 2 % HSA with Caa+ andMgf+ ( • • ) or 2 % HSA and 10 mM EDTA without addedOf?"1" and Mg2* (A A) were preincubated at 37°C for10 min, then exposed to FITC-dextran and 10~ 9 M fNPNTL or10"9 M PMA with or without H-7 for 30 min. Mean of threeexperiments±s.D.M.

these shape changes could be due to suppression of front-tail polarity by H-7 as well as to increased cell-substratum adhesiveness (Table 1). H-7 alone increasedthe contact areas substantially. Combined stimulationwith H-7 and fNLPNTL produced a further increase.

In contrast to 20% activated serum (26+6% cellsmigrated) or 10~ 9 M fNLPNTL (9±4%), H-7 at concen-trations ranging from 1/IM (0.1±0.1%) to 300 ^M(1.5±1.4%) did not induce significant directional loco-

Protein kinase C inhibitor H-7 103

We suspect that these morphological differences are as-sociated with differences in the locomotor behaviour (e.g.persistence). However, we do not yet know how useful thisclassification will be in the long run.

Both chemotactic peptides (Keller et al. 1983) and H-7(Fig. 3) elicit a dose-dependent increase, followed by adecrease, in the proportion of polarized cells. At 300 /IMH-7 (Table 1; Fig. 3) polarized cells are to a large extentreplaced by nonpolar cells with surface projections. Thismay explain the finding that H-7, an agent capable ofpolarizing neutrophils, is also capable of suppressingchemotactic peptide-induced polarity (Table 1). We havenot tested whether H-7 and fNLPNTL also have synergis-tic or additive effects on polarity because H-7 polarizesneutrophils only within a very narrow concentrationrange.

B

100 pm

Fig. 5. Stimulation of neutrophil locomotion by H-7. Cells inplain medium (A, top) or with 100/iM H-7 (B, bottom) werepreincubated in Sykes-Moore chambers on a heated (37 °C)microscope stage for 10 min. Then, the initial position of theneutrophils (•) and the tracks ( ) produced during anobservation time of 10 min were recorded on videotape anddrawn on the TV monitor. Bar, 100 /an.

motion (filter assay) as compared to plain medium(0.2±0.1%). Thus, there is no evidence of chemotacticactivity.

Comparison of responses induced by H-7 versusfNLPNTLThe early morphological responses to H-7 are very similarto those elicited by the chemotactic peptide fNLPNTL.Both agents elicit immediate formation of cells ruffling allover the surface within one minute. At this stage theresponses to these agents cannot be distinguished on thebasis of DIC morphology, actin distribution or increase incytoskeletal actin (Figs 1 and 3). Dissimilarities developafter 5 min of stimulation. At 30 min the shape changesproduced by H-7 are sufficiently characteristic to allowmorphological diagnosis of H-7 stimulation as opposed tostimulation with chemotactic peptides, PMA, diacylgly-cerols, 2H2O or microtubule disassembling agents. Wedistinguish between type A and B polarity because thishelps us to discriminate between responses to H-7 (mainlytype B) and those to chemotactic peptides (mainly type A).

Discussion

Remodeling of membrane lipids and activation of proteinkinase C are believed to be involved in the transmembranesignaling of neutrophils. Formation of diacylglycerols hasbeen reported to occur in neutrophils stimulated bychemotactic peptides, PMA or opsonized particles (Eeib-mann et al. 1988; Fallman et al. 1989; Dougherty et al.1989). The dynamics of the responses may be controlled bysequential generation of diacylglycerols, which activatePKC, and l-O-alkyl-2-acylglycerols (AAG), which inhibitPKC (Dougherty et al. 1989). One current conception isthat kinase inhibitors like H-7 can be used to blockcellular responses that involve protein kinase-dependenttransduction mechanisms. Hence, inhibition by H-7 orlack of inhibition is considered as evidence for or againstthe involvement of protein kinase-dependent pathways.For instance, the finding that H-7 inhibits neutrophillocomotion has been interpreted to mean that activation ofprotein kinase C regulates spontaneous migration and thedirectional locomotion to chemotactic peptides (Gaudry etal. 1988). Our interpretation of similar findings (Table 1)is different. H-7 alone is a neutrophil activator thatmodifies rather than suppresses responses to chemotacticpeptides. A mere inhibition would result in spherical cells,which is not the case. H-7 produces shape changes andincreased adhesiveness. It can thereby suppress polarityinduced by chemotactic peptides. This suffices to explaininhibition of locomotion. Since protein kinase inhibitorssuch as H-7 can function as agonists, they may not bereliable tools for distinguishing between PKC-dependentand PKC-independent responses. A new approach to theuse of H-7 and possibly other kinase inhibitors is needed.Their capacity to activate cells and modulate responses toother agents should be taken into account before onepostulates that they act exclusively as inhibitors.

The available data suggest that the protein kinaseinhibitor H-7 has diverse and sometimes even opposingeffects (e.g. positive and negative chemokinesis) on vari-ous leukocyte functions.

(1) H-7 has been reported to inhibit certain leukocytefunctions including the basal release of arachidonic acid(Matsumoto et al. 1988), spontaneous neutrophil mi-gration (Gaudry et al. 1988) and chemotactic peptide-induced directional locomotion (Gaudry et al. 1988). Fur-thermore, H-7 was found to interfere with several effects ofPMA on neutrophils, including PMA-stimulated phos-phorylation of the 50000Afr protein kinase substrate(Sha'afi et al. 1986), the PMA-induced oxidative burst

104 H. U. Keller et al.

(Gaudry et al. 1988; Berkow et al. 1987), and various effectsof PMA or diacylglycerols on secretion and intracellularcalcium (Berkow et al. 1987; Sha'afi et al. 1986; Smith et al.1988).

(2) H-7 had no effect on the NADPH oxidase activity ofneutrophils, the superoxide release stimulated by thechemotactic peptide FMLP (Gaudry et al. 1988; Berkow etal. 1987; Seifert and Schultz, 1987; Wright and Hoffman,1986), and the Quin 2 and secretory responses of neutro-phils to fMLP and leukotriene B4 (Sha'afi et al. 1986;Berkow et al. 1987).

(3) The present results show that H-7 can also functionas a potent activator of certain neutrophil functions. H-7has been found to induce shape changes, increased associ-ation of actin with the cytoskeleton, alterations in intra-cellular F-actin distribution, chemokinesis (stimulation orsuppression of locomotion depending on the conditions)and a small pinocytic response.

H-7 is a new type of leukocyte activator drug and theresponse pattern produced differs to some extent fromthose elicited by chemotactic peptides, PMA or diacylgly-cerols. In the early phase (up to 1 min) responses to H-7 orfNLPNTL were even very similar. Later (5-30 min) theydiffered in that fNLPNTL-stimulated neutrophils developmainly type A polarity. H-7-stimulated cells developvarious shapes but mainly nonpolar cells with surfaceprojections and/or type B polarity, whereas type A po-larity is only seen in a small proportion of cells. Both typesof polarized cells (A and B) may show locomotor activity.At high H-7 concentrations and with time the proportionof polarized cells decreases and the proportion of nonpolarcells with surface projections increases. H-7 can alsoabrogate front-tail polarity produced by chemotactic pep-tides and this is associated with the development ofnonpolar cells with surface projection (Table 1). H-7 differsfrom chemotactic peptides in that it has only chemokineticbut no chemotactic properties.

PMA and diacylglycerols, which activate PKC as well asthe PKC inhibitors H-7 (Fig. 3) and staurosporine (V.Niggli, personal communication), can elicit nonpolar cellswith surface projections (though the projections formed inresponse to H-7 tend to be larger) and increased amountsof F-actin or cytoskeleton-associated actin. However, H-7differs from PMA or diacylglycerols in that it can, within alimited concentration range, elicit polarity and stimulatelocomotion. Furthermore, H-7 has only a small effect onthe net pinocytic uptake, whereas PMA and diacylgly-cerols are very potent stimulators of pinocytosis andvacuolization. H-7 failed to inhibit the pinocytic responsesinduced by PMA or fNLPNTL (Fig. 4), indicating thatthese effects may not be mediated through activation ofPKC. In Walker carcinosarcoma cells PMA and H-7 wereeven found to exert additive inhibitory effects on loco-motion (Keller and Zimmermann, 1989).

The mechanisms involved in neutrophil activation byH-7 may be rather complex. One problem is that theability of various compounds to inhibit PKC in vitro doesnot correlate with their inhibitory effects in intact cells(Schachtele et al. 1988). Furthermore, the view that PMAor diacylglycerols elicit their effects exclusively by acti-vation of PKC and that kinase inhibitors exert theireffects exclusively by inhibition of protein kinase C havebeen challenged recently in several respects (Schachtele etal. 1988; Love et al. 1989; Hockberger et al. 1989; Krishna-murthi and Joseph, 1989; Sha'afi, 1989). More work isrequired to define the biochemical mechanisms involved inneutrophil activation by H-7. The role of cyclic nucleotide-

dependent kinases (Hidaka et al. 1984), phospholipase A2(Matsumoto et al. 1988) or other yet unidentified targets(Love et al. 1989) should be considered.

The work was supported by the Swiss National Science Foun-dation. The technical assistance of M. Kilchenmann, D. Meier,and P. Kirschner is gratefully acknowledged.

References

AMHBIN, P. C. AND STOSSBL, T. P. (1980). Prevention of degradation ofhuman polymorphonuclear leukocyte proteins bydiisopropylfluorophosphate. Blood 56, 442—447.

BERKOW, R. L., DODSON, R. W. AND KRAFT, A. S. (1987). The effect of aprotein kinase C inhibitor, H-7, on human neutrophil oxidative burstand degranulation J. Leuk. Biol. 41, 441-446

BOYUM, A. (1968). Separation of leucocytes from blood and bone marrow.Scand. J. clin. Lab. Invest. 21, 1-89.

DAVIS, B. H., WALTER, R. J., PEARSON, C. B., BECKER, E. L. AND OLIVER,J. M. (1982). Membrane activity and topography of f-Met-Leu-Phe-treated polymorphonuclear leukocytes. Acute and sustained responsesto chemotactic peptide. Am. J. Path. 108, 206-216.

DOUGHERTY, R. W., DUBAY, G. R. AND NEEDEL, J. E. (1989). Dynamics ofthe diradylglycerol responses of stimulated phagocytes. J. biol. Chem.284, 11263-11269.

FALLMAN, M., LEW, D P., STENDAHL, O. AND ANDERSSON, T. (1989).Receptor-mediated phagocytosis in human neutrophils is associatedwith increased formation of inositol phosphates and diacylglycerol.J. clin Invest. 84, 886-891.

FERRANTE, A. AND THONGH, Y. H. (1980). Optimal conditions forsimultaneous purification of mononuclear and polymorphonuclearleucocytes from human blood by the Hypaque-Ficoll method. J.Immun. Meth. 36, 109-117.

GAT.I.IN, J. I. AND WRIGHT, D. G. (1978). Role of secretory events inmodulating human neutrophil chemotaxis. J clin. Invest. 62,1364-1374.

GAUDHY, M., PERIANIN, A., MARQUBTTY, C. AND HAKIM, J. (1988).Negative effect of a protein kinase C inhibitor (H-7) on humanpolymorphonuclear neutrophil locomotion. Immunology 63, 715-719.

HIDAKA, H., INAQAKI, M., KAWAMOTO, S. AND SASAKI, Y. (1984)Isoquinolinesulfonamides, novel and potent inhibitors of cyclicnucleotide dependent protein kinase and protein kinase C.Biochemistry 23, 5036-5041.

HOCKBERGER, P., TOSELU, M., SWANDULLA, D. AND LUX, H. D. (1989). Adiacylglycerol analogue reduces neuronal calcium currentsindependently of protein kinase C activation. Nature 338, 340-342.

KELLER, H. U., NIGGLI, V. AND ZIMMERMANN, A. (1989a). Diacylglycerolsand PMA induce actin polymerization and distinct shape changes inlymphocytes: relation to fluid pinocytosis and locomotion. J. Cell Sci.93, 457-465.

KELLER, H. U. AND ZIMMERMANN, A (1987). Shape, movement andfunction of neutrophil granulocytes. Biomed. Pharmac. 41, 285-289.

KELLER, H. U., ZIMMERMANN, A. AND COTTIER, H. (1983). Crawling-likemovements, adhesion to solid substrata and chemokinesis ofneutrophil granulocytea. J. Cell Sci. 64, 89-106.

KELLER, H. U., ZIMMERMANN, A. AND NIGGLI, V. (19896). Diacylglycerolsas well as the protein kinase inhibitor H-7 suppress cell polarity andlocomotion of Walker 266 carcinosarcoma cells. Int. J. Cancer 44,934-939.

KJUSHNAMURTHI, S. AND JOSEPH, S. (1989). Arachidonate release inneutrophils: does a lack of effect of protein kinase C inhibitors implyno mvolvement of protein kinase C? Biochem. J. 261, 687-688.

LAEMMLJ, U. K. (1970). Cleavage of structural proteins during theassembly of the head of bactenophage T4. Nature 227, 680-685.

LOVE, J T. JR, PADULA, S. J., LINGBNHELD, E G., AMIN, J. K., SGROI, D.C, WONG, R. L., SHA'An, R I. AND CLARK, R B. (1989). Effects of H-7are not exclusively mediated through protein kinase C or the cyclicnucleotide-dependent kinases. Biochem. biophys. Res. Commun. 162,138-143.

MATSUMOTO, T., TAO, W. AND SHA'AFI, R. I. (1988). Demonstration ofcalcium-dependent phospholipase A2 activity in membrane preparationof rabbit neutrophils. Biochem. J. 250, 343-348.

PAJIDEB, J. D. AND SPUDICH, J. A. (1982). Purification of muscle actin.Meth. Cell Biol. 24, 271-289

RAO, K. M K. (1985) Phorbol esters and retinoids induce actinpolymerization in human leukocytes. Cancer Lett. 28, 253-262.

REIBMANN, J., KOHCHAK, H. M., VOSSHALL, L. B., HAINES, K. A., RICH, A.M. AND WEISSMANN, G. (1988) Changes in diacyglycerol labeling, cellshape, and protein phosphorylation distinguish 'triggering' from'activation' of human neutrophils. J. biol. Chem. 263, 6322-6328.

ROBINSON, J. M., BATWEY, J. A., KABNOVSKY, M. L. AND KAENOVSKY, M.

Protein kinase C inhibitor H-7 105

J. (1987). Cell surface dynamics of neutrophils stimulated with phorbolestere or retinoids. J. Cell Biol. 105, 417-426.

Roos, F. J., ZIMMERMANN, A. AND KELLER, H. U. (1987). Effect of phorbolmyristate acetate and the chemotactic peptide fNLPNTL on shape andmovement of human neutrophils. J. Cell Sci. 88, 399-406.

SCHACHTKLB, C, SEIFERT, R. AND OSSWALD, H. (1988). Stimulus-dependent inhibition of platelet aggregation by the protein kinase Cinhibitors polymyxin B, H-7 and rtaurosporine. Biochem. biophys. Res.Commun. 151, 542-547.

SEIFERT, R. AND SCHULTZ, G (1987). Fatty-acid-induced activation ofNADPH oxidase in plasma membranes of human neutrophils dependson neutrophil cytosol and is potentiated by stable guanine nucleotides.Eur. J. Biochem. 162, 563-669.

SHA'AFI, R. I. (1989). Some effects of phorbol esters are not mediated byprotein kinase C. Biochem. J. 261, 688.

SHA'AFI, R I., MOLSKI, T. F. P., HUANG, C. K. AND NACCACHK, P. H.(1986). The inhibition of neutrophil responsiveness caused by phorboleaters is blocked by the protein kinase C inhibitor H7. Biochem.biophys. Res. Commun. 137, 50-60.

SHA'AFI, R. I., WHITE, J. R., MOLSKI, T. F. P., SHErcYK, J., VOLPI, M.,NACCACHE, P. H. AND FEINSTEIN, M. B. (1983). Phorbol 12-myristate

13-acetate activates rabbit neutrophils without an apparent rise in thelevel of intracellular free calcium. Biochem. biophys. Res. Commun.114, 638-645.

SMITH, R. J., SAM, L. M. AND JUSTEN, J. M. (1988). Diacylglycerolsmodulate human polymorphonuclear neutrophil responsiveness:Effects on intracellular calcium mobilization, granule exocytosis, andsuperoxide anion production. J. Leuk. Biol. 43, 411-419.

WHITE, J. R,, NACCACHE, P. H. AND SHA'AFI, R. I. (1982). The syntheticchemotactic peptide formyl-methionyl-leucyl-phenylalanine causes anincrease in actin associated with the cytoskeleton in rabbitneutrophila. Biochem. biophys. Res. Commun. 108, 1144-1149.

WRIGHT, C. D. AND HOFFMAN, M. D. (1986). The protein kinase Cinhibitors H-7 and H-9 fail to inhibit human neutrophil activation.Biochem. biophys. Res. Commun. 135, 749-755.

ZIMMERMANN, A., GEHE, P. AND KELLER, H. U. (1988). Diacylglycerol-mduced shape changes, movements and altered F-actin distribution inhuman neutrophils. J. Cell Sci. 90, 657-666.

(Received 11 January 1990 - Accepted 16 February 1990)

106 H. U. Keller et al.