Regulation of Capacitative Ca2+ Influx in Human Neutrophil Granulocytes

Ca2+ entry through the capacitative (store-regulated) pathway was shown to be inhibited in neutrophil granulocytes by the protein kinase C activator phorbol 12-myristate 13-acetate and the chemoattractantN-formyl-methionyl-leucyl-phenylalanine (fMLP) by a hitherto unknown mechanism. Measuring both Ca2+ and Mn2+ entry into store-depleted cells we show in the present study that inhibition of the capacitative pathway is absent in various forms of chronic granulomatous disease. To establish the possible relationship between inhibition of the capacitative pathway and ability of O·̄2 production and consequent membrane depolarization, gradual changes of the membrane potential were evoked in neutrophils of healthy individuals. This was accomplished by pharmacological manipulation of the membrane potential and by variations of the concentration and type of the stimulant. Close relationship was observed between membrane depolarization and inhibition of Mn2+ entry through the capacitative transport route. Our results provide an explanation for the inhibitory action of fMLP and phorbol 12-myristate 13-acetate on capacitative cation influx and reveal that upon physiological stimulation, Ca2+ entry into neutrophils is restricted by the depolarization accompanying O·̄2 production.

Ca 2؉ entry through the capacitative (store-regulated) pathway was shown to be inhibited in neutrophil granulocytes by the protein kinase C activator phorbol 12myristate 13-acetate and the chemoattractant N-formylmethionyl-leucyl-phenylalanine (fMLP) by a hitherto unknown mechanism. Measuring both Ca 2؉ and Mn 2؉ entry into store-depleted cells we show in the present study that inhibition of the capacitative pathway is absent in various forms of chronic granulomatous disease. To establish the possible relationship between inhibition of the capacitative pathway and ability of O 2 . production and consequent membrane depolarization, gradual changes of the membrane potential were evoked in neutrophils of healthy individuals. This was accomplished by pharmacological manipulation of the membrane potential and by variations of the concentration and type of the stimulant. Close relationship was observed between membrane depolarization and inhibition of Mn 2؉ entry through the capacitative transport route. Our results provide an explanation for the inhibitory action of fMLP and phorbol 12-myristate 13-acetate on capacitative cation influx and reveal that upon physiological stimulation, Ca 2؉ entry into neutrophils is restricted by the depolarization accompanying O 2 . production.
Stimulation of neutrophil granulocytes by various receptor agonists is accompanied by an increase of intracellular Ca 2ϩ concentration. This Ca 2ϩ transient is a crucial element in the transduction of extracellular signals, leading to the activation of different effector responses (e.g. degranulation, superoxide production) of the cells (1)(2)(3). The Ca 2ϩ signal of neutrophils consists of two components: (i) a rapid, transient increase of [Ca 2ϩ ] i 1 due to Ca 2ϩ release from the internal stores induced by inositol 1,4,5-trisphosphate and (ii) a sustained elevation of [Ca 2ϩ ] i due to Ca 2ϩ influx from the extracellular space (1).
Neutrophil granulocytes do not possess voltage-dependent Ca 2ϩ channels, but there is growing evidence that emptying of the internal Ca 2ϩ stores initiates Ca 2ϩ influx from the extracellular milieu, i.e. the Ca 2ϩ content of the stores regulates Ca 2ϩ -conducting channels in the plasma membrane (4 -7). This Ca 2ϩ entry is often referred to as capacitative or store-operated Ca 2ϩ influx (8).
In human neutrophil granulocytes the regulation of capacitative Ca 2ϩ influx is more complex: agonists of several chemotactic receptors, N-formyl-methionyl-leucyl-phenylalanine (fMLP) and platelet-activating factor, initiate Ca 2ϩ influx by emptying the internal stores in an inositol 1,4,5-trisphosphatedependent manner while fMLP, but not platelet activating factor, was also shown to inhibit capacitative Ca 2ϩ influx (9). A similar action of the protein kinase C activator phorbol 12myristate 13-acetate (PMA) has been reported, too (9 -13). The inhibitory effect of both fMLP and PMA is absent in undifferentiated HL-60 cells and develops gradually during the differentiation induced by dimethyl sulfoxide (13). Despite the similarity in the basic feature, i.e. blocking the capacitative Ca 2ϩ influx, several differences were also described in the characteristics of the inhibition brought about by the two compounds. (i) The fMLP-induced inhibition is transient while the effect of PMA is sustained (9). (ii) Staurosporine, an inhibitor of protein kinase C prevents the inhibitory effect of fMLP only partially but blocks the effect of PMA completely (9). (iii) The inhibitory action of fMLP is sensitive to pertussis toxin whereas the effect of PMA is not influenced by the toxin (14). (iv) Okadaic acid, an inhibitor of protein phosphatases prolongs the inhibition elicited by fMLP but it does not affect the PMA-induced response (12). On the basis of these observations a role of protein phosphorylation was suggested but the intervening kinase(s) remained unidentified (9,(12)(13)(14).
The characteristics of Ca 2ϩ influx inhibition by fMLP and PMA are remarkably similar to the features of superoxide production, evoked by the same agents (15,16). Generation of superoxide by the NADPH oxidase induces rapid depolarization of the cells mainly due to electron transfer from the intracellular milieu to the extracellular environment (17)(18)(19)(20)(21). Depolarization itself was shown to inhibit the agonist-induced Ca 2ϩ influx in several cell types (22)(23)(24). However, the possibility of a causal relationship between inhibition of capacitative Ca 2ϩ influx by fMLP and PMA and depolarization coupled to superoxide generation elicited by the same agents has never been raised.
In the present study two approaches were used to test this hypothesis. First the effect of fMLP and PMA on capacitative Ca 2ϩ influx was investigated in neutrophil granulocytes of patients suffering from chronic granulomatous disease (CGD), i.e. in cells which are unable to produce superoxide, but possess the capacitative Ca 2ϩ entry mechanism. The second approach consisted of testing the inhibitory action of fMLP and PMA on Ca 2ϩ influx in healthy neutrophils, where O 2 . production or membrane potential have been modulated by pharmacological means. The obtained results reveal a close relationship between plasma membrane depolarization and inhibition of Ca 2ϩ influx into neutrophil granulocytes. The suggested mechanism may play a significant autoregulatory role in healthy granulocytes, whereas in CGD cells the lack of inhibition of capacitative Ca 2ϩ entry may represent an alteration of cellular Ca 2ϩ metabolism. To our knowledge this is the first report showing disturbed regulation of capacitative Ca 2ϩ entry in a human disease.

EXPERIMENTAL PROCEDURES
Materials-Fura-2/AM was obtained from Calbiochem; Percoll from Pharmacia; dimethyl sulfoxide, PMA, fMLP, cytochrome c, and thapsigargin were from Sigma and di-O-C 5 (3) from Molecular Probes. Diphenylene iodonium (DPI) was the kind gift of Dr. Arvind Nanda (Hospital for Sick Children, Toronto, Canada). All the other reagents were of research grade.
The routinely used medium (referred to as H medium) contained in mM: NaCl 145, KCl 5, MgCl 2 1, CaCl 2 0.8, HEPES 10, glucose 5, pH 7.4. The Ca 2ϩ -free medium consisted of the same constituents except for CaCl 2 and was supplemented with 0.5 mM EGTA. In the KCl based H medium NaCl was replaced by KCl.
Cell Isolation-Human neutrophils were prepared from blood of healthy volunteers by dextran sedimentation followed by Percoll gradient centrifugation according to the procedure described in Ref. 25. Contaminating red cells were removed by hypotonic lysis. Cells were finally suspended in H medium and kept at room temperature. Preparations contained more than 95% neutrophils, viability as determined by erythrosin B dye exclusion exceeded 97%.
Characterization of CGD Patients-The diagnosis of CGD was established on the basis of serious reduction of O 2 . production and by detection of the molecular defect using Western blot analysis. Qualitative investigation of O 2 . generation was carried out by the nitro blue tetrazolium slide test as described in Ref. 26. Quantitative measurements consisted of determination of chemiluminescence (27) and cytochrome c reduction (28) following stimulation of the cells by 300 nM PMA. The amount of cytochrome b 558 in the separated membrane fraction was determined on the basis of the heme spectrum as detailed in Ref. 29. Expression of the different subunits of NADPH oxidase was assessed by Western blot analysis using monoclonal antibodies against gp91 phox and p22 phox and polyclonal antibodies against p47 phox . All the applied antibodies were developed and kindly provided by Professor Dirk Roos (Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam, Netherlands). Genetic analysis of the mutation underlying CGD was carried out by Professor Dirk Roos (CLB, Amsterdam, Netherlands) and the results were reported in Ref. 30. Four patients (N. M., L. F., A. R., and G. B.) investigated in the present study had a mutation on the cytochrome b␤ (CYBB) gene and expressed neither gp91 phox nor p22 phox in their neutrophils. One patient (I. B.) with p47 phox deficiency had a GT deletion on the gene neutrophil cytosol factor 1 and the cytosolic protein p47 phox was absent in the granulocytes of this patient. Further data on the patients investigated in the present study are detailed in Ref. 30.
Measurement of [Ca 2ϩ ] i -For loading with the fluorescent dye, cells (5 ϫ 10 7 /ml) were incubated in the presence of 4 M Fura-2/AM for 30 min at 37°C. Thereafter cells were washed to remove the extracellular dye and resuspended in H medium at a density of 5 ϫ 10 7 /ml. Fura-2loaded cells were stored at room temperature.
For measurement of [Ca 2ϩ ] i , Fura-2-loaded cells (10 6 ) were suspended in 3 ml of H medium and preincubated at 37°C for 5 min prior to addition of any agent. Changes of the fluorescence were recorded in a Deltascan dual-wavelength spectrofluorimeter (Photon Technology International, South Brunswick NJ) using wavelengths 340 and 380 nm for excitation and 505 nm for emission. Measurements were performed at 37°C with continuous stirring. [Ca 2ϩ ] i was calculated from the ratio of fluorescence excited at 340 and 380 nm, following the method detailed in Ref. 31. Data were analyzed with the Felix software (PTI).
Measurement of Mn 2ϩ Influx-Mn 2ϩ influx measurements were performed under the same conditions as described for [Ca 2ϩ ] i except that the excitation wavelength was 360 nm. At this wavelength, fluorescence of Fura-2 is quenched by Mn 2ϩ but it is not influenced by changes of [Ca 2ϩ ] i (31). Unless otherwise stated, Mn 2ϩ influx was initiated by adding 300 M MnCl 2 .
Measurement of Membrane Potential Changes-Changes of the membrane potential were monitored by the potential sensitive fluorescent dye 3-3Ј-dipentyloxacarbocyanine (di-O-C 5 (3)) as desribed in Refs. 17 and 32. Cells (10 6 ) were suspended in 3 ml of H medium in the presence of 100 nM di-O-C 5 (3) and stimuli were added after binding of the probe reached equilibrium and the fluorescence had stabilized. Changes of the fluorescence were recorded in the spectrofluorimeter used for [Ca 2ϩ ] i measurements or in a Perkin-Elmer 3000 spectrofluorimeter with wavelengths 484 and 510 nm for excitation and emission, respectively. Calibration of the fluorescent signal was carried out in the presence of 2 g/ml valinomycin and varying K ϩ concentration in the external medium, as described previously (32,33). In accordance with earlier reports (17,32)

The Effect of fMLP and PMA on the Capacitative Ca 2ϩ Influx into Neutrophils of Healthy Individuals and CGD Patients-
For the selective study of the capacitative Ca 2ϩ influx, neutrophil granulocytes were treated with the microsomal Ca 2ϩ -ATPase inhibitor thapsigargin (TG). In this way the agonistresponsive internal Ca 2ϩ stores could be depleted even in the absence of increased inositol 1,4,5-trisphosphate generation and full activation of the capacitative pathway could be achieved both in the absence and presence of extracellular Ca 2ϩ (34). Comparative experiments were carried out on neutrophil granulocytes of healthy individuals which produced high amounts of O 2 . upon stimulation by fMLP and PMA, and on cells of CGD patients lacking a functional oxidase. The active O 2 . producing enzyme is a multicomponent structure and genetic defects have been described in both subunits of the membrane-bound cytochrome b 558 (gp91 phox and p21 phox ) and in the two phagocyte-specific cytosolic proteins (p47 phox and p67 phox ) (30,35,36). We carried out measurements on granulocytes of patients with verified defects in the gene of gp91 phox or p47 phox . ] i after addition of CaCl 2 to store-depleted cells reflects the sum of two opposing processes: Ca 2ϩ influx through the capacitative pathway and Ca 2ϩ extrusion via the plasma membrane Ca 2ϩ pump. The capacitative pathway of many different cell types including human neutrophils is also permeable for Mn 2ϩ , whereas Mn 2ϩ is not a substrate for the Ca 2ϩ pump. Thus, for isolated study of the capacitative influx Mn 2ϩ is a better indicator than Ca 2ϩ itself. As the reaction of Fura-2 with Mn 2ϩ results in quenching of the fluorescence of the dye, Mn 2ϩ entry into the cells is represented by and proportional to the detected decrease of Fura-2 fluorescence (5).
In the following experiments we examined the effect of fMLP and PMA on the TG-evoked Mn 2ϩ entry. Cells were exposed to 100 nM TG in the presence of extracellular Ca 2ϩ . After a 10-min incubation, 300 M MnCl 2 was added. As shown in Fig. 2 (traces 1) the addition of MnCl 2 induced a rapid quenching of fluorescence representing Mn 2ϩ entry via the capacitative pathway opened by the store depletion. A rapid fall in fluorescence occurred in the first 30 s (about 80% of total decrease of fluorescence), followed by a slower phase of quenching. In control experiments, where the cells were not treated with TG, MnCl 2 caused only negligible decrease in fluorescence (data not shown), proving that the plasma membrane of resting neutrophils is relatively impermeable for Mn 2ϩ and the dye is localized almost exclusively in the cytosol. When normal cells were stimulated by 10 nM PMA ( Fig. 2A, trace 2) or 1 M fMLP ( Fig.  2A, trace 3) 1 min before the addition of MnCl 2 , the quenching by Mn 2ϩ was strongly reduced (69.2 Ϯ 8.9% inhibition by PMA, n ϭ 4 and 66 Ϯ 3.2% by fMLP n ϭ 4) indicating that both fMLP and PMA inhibited Mn 2ϩ influx into the cells. Fig. 2, B and C (traces 2 and 3), summarizes the effect of the same agonists on the TG-induced Mn 2ϩ entry in gp91 phox (B) and p47 phox -deficient (C) CGD cells. As the nearly parallel experimental recordings show, PMA and fMLP were both ineffective in reducing the TG-induced Mn 2ϩ entry into the cells. In four experiments carried out on gp91 phox -deficient CGD patients the rate of Mn 2ϩ entry in the presence of fMLP or PMA was even slightly higher (by 6.7 Ϯ 7 and 6.2 Ϯ 4.5% respectively) than in the absence of any stimulator. In the experiment carried out on p47 phox -deficient neutrophils, fMLP induced 8% and PMA effected 14% inhibition of Mn 2ϩ entry.

Control of O 2 . Production and Membrane Potential Changes
in CGD Neutrophils-Both methods, widely used for investigation of Ca 2ϩ entry through the capacitative pathway provided similar results: neither fMLP nor PMA inhibits this transport route in CGD cells. We wished to establish whether the altered regulation of capacitative Ca 2ϩ entry into CGD cells might be related to the characteristic functional defects observed in this disease. To this end O 2 . production and the consequent membrane potential changes had to be checked in the cells of the CGD patients investigated in this study. When stimulated with PMA, superoxide production was not detectable in the granulocytes of three gp91 phox -deficient patients (N. M., A. R., and G. B.). The neutrophils of one gp91 phox patient (L. F.) produced less than 10%, whereas the cells of the p47 phox -deficient patient (I. B.) produced less than 5% of the control value.
Stimulation of intact neutrophils by fMLP or PMA was shown to induce rapid depolarization of the cells, the main component of which seems to be the electrogenic operation of the oxidase. In accordance with this suggestion, CGD cells were reported to undergo only marginal depolarization upon fMLP or PMA treatment (17-21, 37).
In our experiments addition of PMA or fMLP to neutrophils obtained from healthy individuals resulted in a sudden depo- larization of similar amplitude but with different kinetics (Fig.  3). Whereas PMA induced a stable depolarization lasting over 5 min (Fig. 3A), the fMLP-evoked depression of the membrane potential showed a partial recovery after 2 min (Fig. 3B). Both PMA and fMLP had only a marginal effect on the membrane potential of neutrophil granulocytes obtained either from gp91 phox -deficient (Fig. 3) or p47 phox -deficient (not shown) patients. Our measurements carried out on healthy and CGD cells are thus in full agreement with previous reports (17,18,37).

Correlation between Membrane Potential and Inhibition of the Capacitative Influx in Healthy Neutrophil Granulocytes-
The experiments carried out on neutrophil granulocytes of CGD patients revealed that both fMLP and PMA failed to inhibit capacitative Ca 2ϩ influx in these defective cells and they both failed to induce any significant depolarization. In our next experiments we wanted to test the correlation of these two parameters in healthy neutrophils under conditions where gradual changes of the membrane potential could be evoked.
As a first approach we tried to reduce the agonist induced depolarization by pharmacological means and study the effect of fMLP and PMA on the Ca 2ϩ influx under these conditions. For this purpose we applied a combined pretreatment using two drugs which affect charge movements in two different ways. The NADPH-oxidase inhibitor DPI reduces the rate of electron efflux through the oxidase (38) whereas the K ϩ -selective ionophore, valinomycin, allows charge compensation through outward movement of K ϩ . These two agents with different sites of action were previously described to clamp efficiently the PMA-induced depolarization (33). The addition of 10 nM PMA after 10 min incubation with TG induced a rapid depolarization response (Fig. 4A, trace b) whereas pretreatment of the cells with valinomycin and DPI largely diminished the depolarizing effect of PMA (Fig. 4A, trace a). Clamping of the membrane potential by these agents was similarly effective when fMLP was used as a stimulus (Fig. 4B). Both DPI and valinomycin, when applied separately, were able to reduce only partially the membrane potential change induced by either stimulating agent (data not shown).
The effect of pretreatment with valinomycin and DPI on the PMA-induced inhibition of the capacitative influx is demonstrated in Fig. 4C. Addition of CaCl 2 to store-depleted control cells resulted in an increase of [Ca 2ϩ ] i from 100 nM to approximately 250 nM (trace 1), whereas in PMA-stimulated cells the same treatment induced a rise of [Ca 2ϩ ] i only to approximately 130 nM (trace 2). Pretreatment of the cells with valinomycin and DPI (trace 3) eliminated the inhibitory effect of PMA and the increase of [Ca 2ϩ ] i was similar to the value observed without PMA treatment. The inhibitory effect of fMLP on the capacitative Ca 2ϩ influx was also prevented by DPI plus valinomycin (Fig. 4D). When DPI or valinomycin was used alone to attenuate the fMLP-or PMA-induced membrane potential changes the inhibition of the influx was only partially antagonized (traces not shown). Prevention of membrane potential changes also effectively counteracted the inhibitory effect of both fMLP and PMA on Mn 2ϩ entry into TG-pretreated cells (Fig. 5).
A parallel increase was observed in membrane depolarization and inhibition of Mn 2ϩ entry when the concentration of the receptor agonist fMLP was varied between 1 nM and 1 M, i.e. in the range where O 2 . production is augmented from just detectable to maximal intensity. Also platelet-activating factor, another chemotactic receptor agonist which evokes a similar Ca 2ϩ signal as fMLP but has only slight effect on O 2 . production (39) induced partial depolarization and reduced the rate of Mn 2ϩ entry into TG-treated cells moderately (data not shown). The close relationship between membrane potential changes and inhibition of the capacitative Mn 2ϩ influx observed under these remarkably different experimental conditions suggests a causative relation.
The Effect of Superoxide Anions and Oxidase-independent Depolarization on the Capacitative Mn 2ϩ Entry-In the above described experiments attenuation of the agonist-induced depolarization was accompanied by a decrease in the superoxide production. To test the possible role of superoxide anions themselves in the inhibition of Ca 2ϩ influx via the capacitative pathway, we examined the TG-induced Mn 2ϩ influx under conditions where superoxide was generated by xanthine oxidase in the presence of xanthine. The concentration of the enzyme and its substrate were chosen so that a similar amount of O 2 . was formed as in the case of stimulation of the cells by fMLP. Superoxide production by the xanthine/xanthine oxidase system did not affect the membrane potential of the cells (not shown) and did not inhibit the TG-induced Mn 2ϩ influx (Fig.  6A). Apparently O 2 . anions themselves are not responsible for the inhibitory action of fMLP and PMA on capacitative Ca 2ϩ influx, supporting our view that the change of the membrane potential resulting from the electrogenic operation of the enzyme may be the decisive factor. To further substantiate the role of depolarization in the agonist-induced inhibition of the store-regulated influx, we examined the effect of oxidase-independent depolarization on the capacitative Mn 2ϩ influx. We followed the TG-induced Mn 2ϩ entry into cells, suspended in a KCl-based medium supple- mented with valinomycin. In the presence of the potassium ionophore the membrane potential is close to the potassium equilibrium potential, which, in the case of 150 mM external K ϩ , is presumed to be around 0 mV. Fig. 6B shows, that under these conditions the TG-induced Mn 2ϩ influx is markedly inhibited. In two parallel experiments, the Mn 2ϩ -induced fluorescence change detected in cells suspended in KCl medium containing also valinomycin (trace 1 in Fig. 6B) was 52 and 40% of the value observed in control cells in NaCl medium (trace 2 in Fig. 6B). This observation indicates that depolarization induced by other mechanisms than activation of the NADPH oxidase, is also effective in the inhibition of the capacitative influx. The inhibition was, however, more pronounced when the cells were stimulated by PMA or fMLP (compare Figs. 2A  and 6B). This is in accordance with the results of the membrane potential measurements where PMA and fMLP were found to depolarize the cells over 0 mV (Figs. 3 and 4). The difference in the extent of depolarization attained in the presence of high [K ϩ ] o and valinomycin or by stimulation of the NADPH oxidase is in accordance with and may provide an explanation for the detected difference in the effectivity of inhibition of capacitative Ca 2ϩ influx.

Does Membrane Depolarization Interfere with Divalent Cation Entry during Physiological Stimulation of Neutrophil
Granulocytes?-In all the previous experiments, Ca 2ϩ or Mn 2ϩ entry was investigated in TG-pretreated cells, i.e. under conditions where the capacitative pathway has been opened by extensive depletion of the agonist-responsive internal Ca 2ϩ stores. In this experimental setup a clear correlation was revealed between membrane depolarization (induced by various mechanisms) and inhibition of the capacitative Ca 2ϩ pathway. The question arises whether this mechanism plays any detectable role during activation of intact (non-depleted, non-predepolarized) cells. To address this problem, Mn 2ϩ entry was followed in neutrophils incubated in H medium containing 1 mM CaCl 2 . In the experiment demonstrated in Fig. 7, fMLP was added after Mn 2ϩ , so that the entry of divalent cation(s) occurring in the very early phase of cell stimulation could also be detected. As shown in trace 2, in healthy neutrophils the decrease of fluorescence indicating Mn 2ϩ entry was delayed for almost 2 min after fMLP addition. This lag phase corresponds to the period where the majority of fMLP-induced O 2 . production takes place and membrane depolarization reaches its maximum (17,40). In sharp contrast to the normal cells, addition of fMLP was followed by instantaneous Mn 2ϩ influx in the neutrophils of gp91 phox -deficient CGD patients (trace 1). The result shown in Fig. 7 clearly indicates that in healthy cells the decrease of the plasma membrane potential that accompanies O 2 . production prevents the entry of divalent cations during the early phase of stimulation. This preventive mechanism is absent in the granulocytes of CGD patients.
The question arises whether the alteration of Ca 2ϩ entry detected in CGD cells has any influence on the Ca 2ϩ metabolism of these cells. In accordance with previous results obtained with Quin-2 as Ca 2ϩ -sensitive dye (19), the Ca 2ϩ signal induced by fMLP was basically similar in granulocytes from healthy or CGD individuals also under our experimental conditions, using Fura-2 as fluorescent probe (Fig. 8, A and C). However, investigation of the ionomycin-releasable Ca 2ϩ pool revealed remarkable difference between the two cell populations. In the experiment showed in Fig. 8 external Ca 2ϩ was chelated either before or 5 min after fMLP stimulation and ionomycin was added thereafter. In healthy cells the ionomycin-induced signal was only slightly different under the two conditions whereas in CGD cells the ionomycin-releasable Ca 2ϩ pool was definitely larger when chelation of external Ca 2ϩ occurred 5 min after fMLP addition. Apparently in healthy cells blockade of Ca 2ϩ entry during fMLP induced O 2 .
production impairs the refilling of intracellular Ca 2ϩ stores, whereas in CGD cells the absence of membrane depolarization allows rapid Ca 2ϩ entry and earlier and more complete restoration of the Ca 2ϩ pools.

DISCUSSION
Summing up the presented experimental data, we propose that the previously described inhibition of the capacitative Ca 2ϩ pathway of neutrophil granulocytes by the stimulators fMLP and PMA is mediated via primary changes of the membrane potential. Our hypothesis is supported by the following Our suggestion that the change of the membrane potential is the common denominator provides a suitable explanation for the previously reported characteristics of inhibition of Ca 2ϩ entry into neutrophils. Differences in the kinetics, sensitivity to staurosporine and pertussis toxin described for the inhibitory action of fMLP or PMA on Ca 2ϩ import reflect the properties of O 2 . production evoked by the two stimulants (15,16). Appearance of Ca 2ϩ -influx inhibition during differentiation of HL-60 cells corresponds to the expression of oxidase components and increase of O 2 . production (41). Finally, prolongation of the inhibitory effect of fMLP on capacitative Ca 2ϩ entry by various phosphatase inhibitors can be ascribed to increased phosphorylation of the oxidase components with consequent prolongation of O 2 . production and related depolarization (42,43). It should be noted that the mechanism proposed on the basis of our experiments does not necessitate the phosphorylation of the capacitative Ca 2ϩ pathway itself or of any intermediary factor transmitting the signal from the intracellular stores to the transport protein located in the plasma membrane. In addition to a drastic drop of the membrane potential, initiation of O 2 . production also induces characteristic changes of the intracellular pH which could modify Ca 2ϩ entry, too. However, there are two observations which argue against this possibility. Stimulation by PMA evokes an immediate acidification which is followed after approximately 2 min by a remarkable alkalinization (44). In contrast to this biphasic response of pH i , both membrane depolarization and inhibition of Mn 2ϩ influx induced by PMA were found to be stable for more than 5 min (Figs. 2 and 3). On the other hand, addition of the K ϩ -ionophore valinomycin to neutrophil cytoplasts suspended in a Na ϩ -based medium significantly augmented the acidification following PMA stimulation (45), whereas in our experiments valinomycin partially antagonized both the inhibition of Ca 2ϩ and Mn 2ϩ entry and membrane depolarization brought about by PMA or fMLP. Apparently, changes of pH i do not occur in parallel with inhibition of Ca 2ϩ influx and in view of the above arguments we consider the alteration of the membrane potential as the decisive factor in determining Ca 2ϩ entry through the capacitative pathway.
The presented experiments allow some speculation about the possible mechanism of inhibition of capacitative Ca 2ϩ entry by membrane depolarization. In resting neutrophil granulocytes suspended in a medium containing 1 mM Ca 2ϩ , the applied fluorescent dye detected an internal [Ca 2ϩ ] of about 100 nM, suggesting that a 10 4 -fold concentration gradient (equalling 120 mV membrane potential change) could drive Ca 2ϩ inward. However, in view of recent reports demonstrating a significantly higher [Ca 2ϩ ] in the subplasmalemmal space than in the averaged cytosol (46,47), the true concentration gradient is at least 1 order of magnitude smaller. The membrane potential of resting neutrophils is suggested to be around Ϫ60 mV but upon stimulation with PMA it was reported to overshoot up to positive values (20,33). A drop of 60 mV or more in the driving force for Ca 2ϩ could certainly slow down the inward movement of the divalent cation considerably. This suggestion is supported by our control experiment where a decrease in the concentration of external Ca 2ϩ added to TG-pretreated cells (in the same experimental setting as shown in Fig. 1) from 500 to 50 M resulted in a measurable impairment of Ca 2ϩ entry. Thus in the case of human neutrophil granulocytes, the decrease in the driving force for Ca 2ϩ ions due to the extensive depolarization following fMLP or PMA stimulation is certainly sufficient to cause detectable diminution of Ca 2ϩ influx by itself. Whether in addition to the drop in the driving force depolarization also affects the conductivity of the capacitative Fura-2loaded cells suspended in H medium were exposed to 1 M fMLP at the time point indicated by the arrow. External Ca 2ϩ was chelated by the addition of 3 mM EGTA 5 min after (A and C) or immediately before (B and D) the stimulation by fMLP. Cells were exposed to 1 M ionomycin at the time point indicated by the asterisk. Traces are representative of two similar experiments performed with healthy cells and show the result of one experiment with gp91 phox -deficient CGD cells.
cation pathway remains to be elucidated by electrophysiological measurements.
Irrespectively of the mechanism of action, the phenomenon that Ca 2ϩ entry is blocked by depolarization seems not to be restricted to neutrophil granulocytes. Various experimental means of depolarization were reported to inhibit agonist-induced Ca 2ϩ influx in T-lymphocytes stimulated via the antigen receptor (23), in platelets upon thrombin stimulation (22), and in parotid acinar cells upon carbachol stimulation (24). Inhibition of voltage-dependent K ϩ -channels decreased Ca 2ϩ import induced by anti-CD2 and anti-CD3 antibodies in T lymphocytes. This effect is probably mediated by membrane depolarization, too (48). This apparently general mechanism may have a significant biological role in phagocytic cells where physiological stimulation results in drastic drop of the membrane potential. Inhibition of Ca 2ϩ entry by membrane depolarization may represent an autoregulatory mechanism limiting Ca 2ϩ uptake following cell stimulation. According to the result shown in Fig.  7, this mechanism is operating during physiological stimulation of granulocytes: detectable influx of divalent cations begins only after the rate of O 2 . production declines and other compensatory mechanisms (e.g. opening of the conductive H ϩ transport pathway (49,50)) contribute to the restoration of the membrane potential. Investigation of neutrophil granulocytes obtained from CGD patients revealed that independently of the genetic form of the disease, neither fMLP nor PMA is able to restrict Ca 2ϩ influx via the capacitative pathway and stimulation of CGD neutrophils by fMLP induces intensive flow of Ca 2ϩ ions from the extracellular space into the cell at an earlier time point than in normal granulocytes. The difference in Ca 2ϩ entry is not reflected in the fluorescent signal detectable with Quin-2 (19) or Fura-2 following fMLP stimulation. In evaluating this apparent contradiction it should be recalled that both Ca 2ϩ -sensitive dyes are inefficient in detecting the correct [Ca 2ϩ ] i at specialized cell regions like the subplasmalemmal space where the [Ca 2ϩ ] i can exceed the cellular average by 1 or 2 orders of magnitude (46,47). In view of these considerations, our observation on the absence of inhibition of Ca 2ϩ entry and faster restoration of the internal Ca 2ϩ stores following fMLP stimulation may represent a biologically important alteration of Ca 2ϩ metabolism of these defective cells. The contribution of this alteration to the pathomechanism of CGD has to be clarified in future investigations.