Immunophwmacologv, 23 (1992) 191-197 Q 1992 Elsevier Science Publishers B.V. All

rights reserved. 0162-3109/92/$05.O0

IMPHAR 00589

EMD 52692 (bimak attenuates lumin superoxide anion r polyrnorphonuclear leubcytes Galen M. Pieper and Garrett J. Gross Department of Pharmacdogv ond Toxiiologv, Medicul Cokge of Wircorspin. MirwCukee, WI S3226. U.S.A.

(Received 15 March

1991;

accepted 10 January 1992)

Abstract: We kestigated the relationship of potassium channel activation on modulation of oxidative respiratory bursts in canhte neutrophils. Generation of superoxide anion radicals in opsonized zymosan-astivated cells w-as determined using the technique of ferricytochrome c reduction. Preincubation of cells with the selective potassiuln channel opener, EMD 52692 (I-IOOpM). attenuated superoxide anion radical prwhction. Furthermore, EMD 52692 also prcduced a concentrationdependent inhibition of luminol-enhanced chemilumiaescence by activated wutrophils. Giyburide, a selective antagonist of ATFsensitive potassium channels, prevented the modulatory effect of EMD 52692 on both superoxide anion generation and cant role in luminol-enhanced chemiluminescence. The results suggest that ATP-sensitive potassium channels may play a si regulating oxygen-derived ftee radical production in neutrophiknduced tissue injury. Key nordr: Oxygen-derived free radical; Chemiluminescencc; Neutrophil; Potassium channel; Glyburide (glibenclamide); Superoxide anion

Introduction

Whole cell patch-clamp techniques have now identified potassium currents in a variety of immunoresponsive cells (Ypey and Clapham, 1984; Gallin, 1986; Choquet and Kom, 1988; Gardner, 1990) including neutrophils (Krause and Welsh, 1990). A precise role for potassium

Correspondetrce to: G.M. Pieper, Department of Phannacology and Toxicology, Medical College of Wisconsin, 870t Watertown Plank Road. Milwaukee WI 53226, U.S.A. Abbrcviatkns: EMD 52692. 6-oyano-2,2dimethyl4[2-oxo1,2-dihydro-l-pyridyl]-2H-l-benzopyran.

channels in neutropbil functions such as chemotaxis, respiratory oxidative bursts, or protease release has not been ascertained. Yet, it is assumed that the potassium channels in neutrophils may pIay a role in stimulus-response coupling. Our interest in potassium channel openers stems from originai observations that the nitrovasodilator, nicorandil, produces a signticant improvement in postischemic contractile recovery in a canine model of reversible, ischemia-reperfusion injury (i.e. the stunned myocardium) unlike that seen with traditional nitrovasodilators (Pieper and Gross, 1987). Nicorandil is d&rent from traditional nitrovasodilators in that it pos-

sesses the ability to hyperpolarize vascular smooth muscle by promoting potassium conductance (Kinoshita and Sakai, 1990). We subsequently showed that nicorandil also suppressed superoxide anion free radical generation by isolated neutrophils (Gross et al., 1989). It has been well-established that the etiology of the stunned myocardium is related to increased oxygenderived free radical production particularly during early reperfusion (Bolli, 1990). Some investigators have indicated that neutrophils may be one primary source of these oxygen-derived free radicals (Engler and Covell, 1987; Westlin and Mullane, 1989). With the discovery that nicorandil relaxed vascular smooth muscle by promoting potassium conductance, several selective potassium channel openers have been developed such as the novel benzopyran derivative, EMD 52692 (bimakalim). We hypothesized that certain of these po:assiam char& openers may act upon other cells such as neutrophils to attenuate the production of toxic superoxide anion radicals. Accordingly, we examined the inhibitory effect of the newly developed potassium channel opener, EMD 52692 (De Peyer et al., 1989) on superoxide anion radical production and liminol-enhanced chemiluminescence in canine neutrophils which were activated with opsonized zymosan.

Materials am! Methods Chemicals, reageflts

Chemicals of the highest purity grade including: zymosan A, cytochrome C, superoxide dismutase, xanthine, xanthine oxidase, glyburide and aaamin were purchased from Sigma Chemical Company (St. Louis, MO). Luminol from Pierce Chemicals (Rockford, IL) was dissolved in silylation grade dimethyl sulfoxide and diluted in buffer for the chemiluminescence assay (final assay concentration of dimethyl sulfoxide was 0.05%). Hanks buffered saline solution (HBS S) with and without Mg*+ and Ca* + was purchased from Gibco BRL (Grand Island, NY). For the

neutrophil isolation, HESPAN (DuPont Pharmaceuticals, Wilmington, DE) and Percoll (Pharma&, Fiscataway, NJ) were utilized. EMD 52692, from E. Merck (Darmstadt, Germany), was dissolved in a small volume of propylene glycol and diluted with HBSS for the stock solution. Preparation of polymorphonuciear leukocytes

Canine neutrophils were isolated by the procedure outlined previously in our laboratory (Pieper and Gross, 1990). Briefly, 40 ml of canine whole blood was collected in heparinized (200 U) syringes containing 20 ml of hydroxyethyl starch (HESPAN). The mixed blood samples were allowed to sediment for 45 min at room temperature. The supernatant of this sediment was centrifuged at 500 x g for 7 min and the pellet resuspended in 15 ml of water at 4 “C for 20 s to lyse residual erythrocytes. Isotonicity was achieved by reconstitution with the addition of 5 ml 3.6% NaCI. This suspension was centrifuged at SO0x g for 7 min at 4 “C. The resulting pellet was resuspended in 5 ml of HBSS (pH 7.4) and layered upon Percoll (density = 1.075). Upon centrifugation at 20 000 x g for 30 min at 4 ’ C, the resulting neutrophil fraction was counted on a Model Z, Coulter counter (Coulter Company, Hialeah, FL) and adjusted to the desired number of cells in HBSS containing calcium and magnesium. Using this procedure, these fractions contained cells which were > 95% neutrophils with < 1y. platelet contamination. Trypan bitrueexclusion tests indicated viability > 95% Opsoniration of zymosan

Gpsonized zymosan was prepared by the procedure outlined by Simpson et al. (1988). Briefly, 10 mg/ml of zymosan A was incubated with normal canine serum for 30 mm at 37 “C and washed twice with sterile normal saline. Opsonized zymosan was resuspended in stock solutions of 10 mg/ml containing normal saline and frozen at - 70 “C for later use.

Supemxide union radicd prdiic?ion

Superoxide anion radicals were measured by the superoxide dismutase (SOD)-inhibitable reduction of 75 ,uM fe~cytochrome C as previously described (Fantone and Kinnes, 1983). In a 37 *C water bath, 2 x 10’ ceils/tube in HBSS (plus Ca’ + and Mg2 + ) were incubated with or without EMD 52692 for IS min. After preincubation, cells were stimulated with 1 mg/ml opsonized zymosan, The reaction was terminated after 30 min by placing on ice and immediately pelleting the cells. ARer removal of the su~mat~t, each sample was diluted with HBSS. Samples were run in triplicate against duplicate samples which were stimulated in the presence of 90 pg/ml SOD. Control (unstimulated) samples were atso run with each assay to verify the lack of nonspecific stimulation of celIs during processing. Absorbance of samples was measured at 550 nm using a Gilfbrd 250 s~~~ophotomet~r. Chemilutninescettce measurements

was chemiluminescence Luminol-enhanced measured by the method of ~eChatelet et al. (1982) as modified by Da@egriet al. (1989) using a Model 560 Luminometer (Chronolog Instruments, Havertown, PA). Luminol-dependent ~hemiiumi~escen~e measles reactive oxygeh species and neutrophil degranulation which is specific and azurophilic (DeChatelet et al., 1982). The reaction mixture (total vohme = 1.0 ml) consisted of 2.5 x IO5cells in HBSS with Caz + and Mg+ , 5 FM luminol and I m&/ml opsonized rymosan. To facilitate stable sequential control responses over a period of time, the HBSS for these studies also contained 100 mg/dl glucose and 0.1% fatty acid-free bovine serum atbumin. After 15 min of incubation in the presence or absence of EMD 52692 (l-100 FM), a 10 yl aliquot of ~umino1was added to the cuvette, foilow~ 2 min later by the addition of 10 ~1 opsonized zymosan to initiate chemiluminescence. The maximum chemiluminescence response in mV was measured at the plateau of the curve. Inhibition was calculated as a % of the SODinhibitable chemiluminescence response to con-

trol ceih which were stimulated in the absence of EMD 52692. The SOD-inhibitable portion of the chemiluminescence was 90 + 1Yb of the control response. Characreriwiott of &te of po~assitmt chnttnels

Additional studies were performed in the ferri~~tochrome C and chemiluminesce~ce assays to determine the type of potassium channel mediating the modulatory effects of EMD 52692. Accordingly, different antagonists of potassium channels (Cook, 1988) such as g~yburide (an ATP-sensitive potassium channel antagonist), agamin (a selective antagonist of low-conductance Ca’+ -activated potassium channels), or ~~inopy~dine (a nons~~tive ant~onist of potassium channels), was given 5 min prior to the addition of EMD 52692, Finally. other studies were performed in a celC free system to determine that EMD 52692 did not produce diiect scavenging of superoxide anion free radicals. In these experiments, the luminolenhanced chemiluminescence produced by the suboxide anion radical gen~a~n~ system of xanthine (6.7 mM) plus xanthine oxidase (0.3 U/ml) in HBSS (pH 7.4) was performed in the presence or absence of EMD 52692.

All statistical analysis was performed on a Macintosh computer with statistical software, The IC, (the bunt of drug prancing a 50% inhibition) was calculated. Statistical analysis between any two group means was determined by Student’s t-test for unpaired observations. Rest&s were consider si~~c~t at p < 0.05. Results

Superoxide anion radical formation was 1.2 + 0.3 and 23.2 + 1.3 ,uM in unstimulated and opsonized zymosan-stimulated cells, respectiveIy (n = 28 each). EMT) 52692 produced a concentration-dependent inhibition of superoxide anion formation with a maximal response occurring as

EMD 52692 (LOG Ml Fig. I. Concentration-dependent inhibition of superoxide anion radical formation hy EMD 52692in canine neutrophils which were stimulated with 1 m&ml apsonized zymosan, Results show the mean f SEM cf 17 det~inants pertbrmed on hatches of neutr~~ls from 6 ind~vida~ animals.

min (superoxide anion concentration = 24.2 + 3.0 and 23.4 4 2.0 &M for untreated and apamintreated cells, respectively; R = 9 detentions). Lmrdnol-enhanced ch~ilumines~nce was atso measured in zymosan-activated cells. EMD 52692 produced a concentration-dependent attenuation of luminol-enhanced chemiluminescence (Fig. 3). The threshold inhibition of chemi~umin~c~ce was seeu at 2 pM EMD 52692 (not shown) with an EC,, of 12 + 1 FM (n = 7 determinations). Control samples with vehicle concentrations equal to that used to dissolve lOOtiM EMD 52692 did not alter the chemiluminescence response of control cells (i.e, 100% of control response). Addition ofglyburide, but not apamin, reversed the effects of EMD 52692 on hnninol-

low as 30 ,uM EMD 52692 (Fig. 1). Vehicle control samples produced no inhibition of superoxide anion fo~ation (not shown). ~etrea~ent of ceils with glyburide (10 FM}, but not apamin (20 nM), completely blocked the inhibitory action of EMD 52692 on zymosan-activated superoxide anion radical production (Fig. 2). Glyburide, in the absence of EMD 52692, had no effect on superoxide anion formation by activated cells (Fig. 21. Similar results were obtained with apa-

1w

EMD 52692

WY)

so

0j

Qpsdnizsd fymosan

Fig. 2. ln~bitio~ of superoxide anion radical fo~ation by EMD 52692in zymosan-activated canine neufrophiis and its blockade by the ATP-sensitive potassium channel blocker, glyburide. Results show the mean 1 SEM of I5 determinations performed on batches of neutrophils from 5 individual animals. *p c 0.05 vs. control.

2 Min

Fig. 3. Concentration-dependant inhibifion of rymosanactivated, IuminoLenhanced chemiluminescence of canine neutrophils by END 52692.The example (repeated on seven different occasion) shows individual tracings from the same batch of cells using the same gain settings.

195

* ~~~ Y

l-l

2Mim

Fig. 4. Reversal by glyburide {IO pM), but not spamin (20 nM), of the inhibitory effect of 20 CM END (52692 on luminolenhanced chemiluminescence ofzymosan-activated canine neutrophils. Examples shown for glyburide and apamin were repeated on four and six occasions, respectively.

enhanced c~e~luminesc~ce {Fig. 4). Glyburide or apamin alone had no effect on luminolenhanced chemiluminescence in the absence of EMD 52692 (i.e. 98.4 & 4% and 96 f 2% of the control response, respectively). Pretreatment with SOQ~M ~~~opy~dine also si~~~~y (p < 0.05) prevented ~hibition of lumin& enhanced chemiluminescence by EMD 52692 (i.e., responses were 107 + 6% of control responses, n = 4 determinations). In cell-free experiments, luminol-enhanced che~um~escence was measured in presence of the superoxide anion ranch-gen~ating system of

xantiine plus xanthine oxidase. EMD 52692 did not quench the chemiluminescence induced by this system. Additions of either 30,50 or 100 PM EMD 56292 (n = 4-6 experiments each) produced responses which were 106 + 7, 106 f 4 and 99 + 8% of control (p r 0.05, not sign& cant).

Discussion

To our ~owledge, the present results are the &st to show that potassium channel openers have a

1% modulatory effect on the function of immunoresponsive cells. Specifically, our results indicate that the selective potassium channel opener, EMD 52692, inhibited both superoxide anion radical production and luminol-enhanced chemiluminescence in opsonized zymosan-activated canine neutrophils. This was not due to a direct scavenging effect of EMD 52692 on superoxide anion radicals but rather to an action on a selective type of potassium channel located in the neutrophil. Membrane potential changes, most likely induced by transmembrane potassium ion movement, are known to occur in neutrophils subsequent to activation by various stimuli (Seligmann and Gallin, 1983). Activation of neutrophils sets in motion a cascade of events. It is wellknown that changes in membrane potential occur before release of oxygen&rived free radicals (Korchak and Weismann, 1978; Kitagawa and Johnston, 1985). Some indirect studies show that an impairment of membrane depolarization due to dietary manipulation is coupled with reduced superoxide anion production in rat neutrophils (Gyllenhammar and Palmblad, 1989). Unfortunately, it has been difficult to document a causal relationship between membrane potential changes and the release of superoxide anion by polymorphonuclear leukocytes. Indeed, different investigators have suggested that the early event of membrane depolarization during neutrophil activation may, but not necessarily, be sufficient to trigger oxidative respiratory bursts suggesting that this may operate via different pathways (Kitagawa and Johnston, 1985; Seeds et al., 1985; Fletcher and Seligmann, 1986). This could explain why EMD 52692 did not elicit total inhibition of either superoxide anion production or luminol-enhanced chemiluminescence. In human neutrophils, membrane depolarization is followed by a subsequent repolarization (Kuroki et al., 1982; Seligmann et al., 1980). Recent whole cell patch-clamp studies have documented a voltage-dependent (depolarizationdependent), inward rectifying potassium channel and a Caa+ -activated, voltage-independent po-

tassium channel which was outward rectifying (Krause and Welsh, 1990). The function of these potassium channels is not yet known with certainty, It is possible that these currents turn the respiratory burst on and off. Indeed, previous observations using guinea pig neutrophils show that addition of positively charged alkylamines to the surface membrane inhibited while negatively charged agents reversed superoxide anion generation to various types of stimuli (Miyahara et al., 1987). Our observations suggest that EMD 52692 produced modulatory effects on superoxide anion production and chemiluminescence by activating APT-sensitive potassium channels since the actions of EMD 52692 were completely blocked by glyburide, the ATP-sensitive potassium chan-

nel blocker. The action of glyburide on tie EMD 526924nduced response could not be explained by an enhanced production of superoxide anion or chemiluminescence by glyburide per se since the same concentration of glyburide given alone had no effect on either assay. The suggestion that EMD 52692 acted speciflcahy via ATP-sensitive potassium channels was supported by additional studies showing that apamin, a selective blocker of low-conductance Ca2+ activated potassium channels, did not alter the responses to EMD 52692. Therefore, the present studies using neutrophils are in agreement with the reports wherein EMD 52692 has been shown to activate ATP-sensitive potassium channels in smooth muscle cells (De Peyer et al., 1989). Whether EMD 52692 alters other aspects of neutrophil function in vitro is not entirely known at this time. Recent in vivo observations in our laboratories have shown that EMD 52692 sign& candy reduces myocardial infarct size in the dog and that this is associated with a marked decrease in neutrophil infiltration into the ischemic area f&owing 5 h of reperfusion relatively to untreated animals (Gross and Auchampach, 1992). These data suggest that EMD 52692 may have other actions on neutrophil recruitment but the precise mechanism is not yet defined. In conclusion, we suggest the possibility that

197 the beneficial effects of certain potassium channel openers on ischemia-reperfusion injury might be explained, in part, by suppressing the adverse actions of circulating neutrophils. Acknowledgements The authors extend appreciation to Ms. Anna Hsu for technical assistance. This work was supported, in part, by funds provided from E. Merck (Darmstadt, Germany) and NIH grant HL 08311. References Bolli R. Mechanism of myocardial stunning. Circulation 1990; 82: 723-738. Choquet D, Kern H. Modulation of voltage-dependent potassium channels in 3 lymphocytes. Biochem Pharmacol 1988; 37: 3797-3802. Cook NS. The pharmacology of potassium channels and their therapeutic potential. Trends Physiol Sci 1988: 9: 21-28. Dallegri F, Ballestrero A, OnonelloL, Patrope F. Platelet as scavengers of neutrophilderived oxidants: a possible defense mechanism at sites of vascular injury. Thromb Haemost 1989; 61: 415-418. DeChatelet LR, Long GD. Shirley PS, Bass DA, Thomas MJ, Henderson FW, Cohen MS. Mechanism of the luminol-dependent chemiluminescence of human neutrophils. J lmmunol 1982: 129: 1589-1593. De Peyer JE, Lues I, Gericke R. Haeusler G. Characterization of a novel K+ channel activaror. EMD 52692, in electrophysiological and pharmacological experiments. PBiQers Arch 1989; 414 (Sup@. 1): Sl9l. En&r R, Cove8 JW. Granulocytes cause reperfusion ventricular dysfunction after 1%min &hernia in dogs. Circ Res 1987: 61: 20-28. Fantone JC, Kinnes DA. Prostaglandin E, andprostaglandin I2 modulation of superoxide production by human neutrophils. Biochem Biophys Res Common 1983; 113; 506-512. Fletcher MP. Seligmann BE. PMN heterogeneity: long-term stability offluorescent membrane potential response to the chemoattractant N-formyl-methionyl-leucyl-phenylalanine in healthy adults and correlation with respiratory burst activity. Blood 1986; 68: 611-618. Gallin EK. Ionic channels in leukocytes. J Leuk Biol 1986; 39: 241-254. Gardner P. Potassium channels in immunoresponsive cells. In: Cook NS, ad., Potassium Channels: Structure, Class& cation, Function and Therapeutic Potential. Ellis Harwood, Chichester 1990, pp. 382-399.

Gross G. Auchampach J. The novel potassium channel opener. EMD 52692. reduces myocardial infarct size in dogs by multiple mechanisms. FASEB J 1992: (in press). Gross G. Pieper G, Farber NE. Warltier D. Hardman H. Effects of nicorandil on coronary circulation and myocardial ischemia. Am J Cardiol 1989: 63: I lJ-l7J. Gyllenhammnr H, Palmblad J. Linoieic acid-deficient rat neutrophils show decreased bactericidal capacity, superoxide formation and membrane depolarization. Immunology 1989: 66: 616-620. Kinoshita M. Sakai K. Pharmacology and therapeuticeffects of nicorandil. Cardiovasc Drugs Ther 1990; 4: 1075-1088. Kitagawa S. Johnston RB Jr. Relationship between membrane potential changes and superoxide-releasing capacity in resident and activated mouse peritoneal macrophages. J lmmunol 1985; 135: 3417-3423. Korchak HM, Weismann G. Changes in membrane potential of human granulocytes antecede the metabolic response to surface stimulation. Proc Natl Acd Sci USA 1978: 75 : 3818-3822. Krause K-H. Welsh MJ. Voltage-dependent and Caz + activated ion channels in human neutrophils. J Clin Invest 1990: 8.5: 491498. Kuroki M, Kamo N, Kobatake Y, Okimasu E, Utsumi K. Measurement of membrane potential in polymorphonuclear leukocytes and its changes during surface stimulation. Biochim Biophys Acta 1982: 693: 326-334 Miyahara M, Watanabe S. Okimasu E, Utsumi K. Chargedependent regulation of NADPH oxida% activity in guinea-pig polymorphonuclear leukocytes. Biochim Biophys Acta 1987; 929: 253-262. Pieper GM, Gross GJ. Priming by plalelet-acdvating factor of neutrophil-induced impairment of endotheiium-dependent relaxation. J Vast 3iol Med 1990; 3: 56-61. Pieper GM. Gross GJ. Salutary action of nicorandil. a new antianginal drug. on myocardial merabolirm during irchemia and on posdschemic function in a canine preparation of brief, repetitive coronary artery occlusions: comparison with isosorbide dinitrate. Circulation 1987; 76: 916-928. Seeds MC, Parce JW, Szejda P, Basp DA. Independent stimulation of membrane potential changes and the oxidalive metabolic burst in polymorphonuclear leukocytes. Blood 1985; 65: 233-240. Seligmann BE, Gnllin II. Comparison of indillect probes of membrane potential utilized in studies of human neutrophils. J Cell Physiol 1983; 1 IS: 105-l 15. Simpson PJ, Mickelson J, Fantone JC, Gallagher KP, Lucchesi BR. Reduction of experimental canine myocardial infarct size with prostaglandin E,: inhibition of neutrophil migration and activation. J Pharmacol Exp Ther 1988: 244: 619-624. Ypey DL, Clapham DE. Development ofa delayed outwardrectifying K + conductance in cultured mouse peritoneal macrophages. Proc Nat1 Acad Sci USA 1984; 8;: 3083-3087.