Transient Inhibition by Chemotactic Peptide of a Store-operated Ca2’ Entry Pathway in Human Neutrophils*

Emptying the intracellular calcium stores of fura-2-loaded human neutrophils by treatment with the en-domembrane ATPase inhibitor thapsigargin leads to a maintained increase of [Ca2+], by Ca2+ entry through a store-operated Caz+ entry pathway. Under these conditions, [Ca2+li was reduced transiently by N-formyl-methionyl-leucyl-phenylalanine (fMLP) and perma-nently by phorbol 12,13-dibutyrate (PDB). Platelet-activating factor (PAF) had no effect. The fMLP- and PDB-induced [Ca2+Ii decreases were not due to stimulated Ca2+ efflux but to inhibition of store-operated Ca2’ entry pathway. PDB and fMLP, but not PAF, inhibited the entry of Ca”, Mn2+, and Baa+ in thapsi-gargin-treated cells. This inhibition was dependent on [Ca2+]~, barely detectable at [Ca2+]i of 50 nM and in-creasingly strong and fast to appear at 170 and 630 nM. Inhibition of entry by fMLP was complete within 5-10 8, disappeared within 2-3 min, and was partially prevented by staurosporin (100 nM). Inhibition by PDB was equally fast, but no recovery was detected within 6

The agonist-induced increase of cytosolic Ca2+ concentration ([Ca2+]J in human neutrophils is often composed of two phases: (i) an early and transient [Ca2+]i peak due to inositol 1,4,5-trisphosphate-mediated Ca2+ release from the intracellular Ca2+ stores, and (ii) a sustained [Ca2+]; increase due to increased Ca" entry through the plasma membrane (1)(2)(3)(4)(5). The mechanism for the late Ca2+ entry is still obscure, although there is strong evidence that the opening of this plasma membrane Ca2+ entry pathway is signaled by the emptying of the intracellular Ca2+ stores (5,6). We shall use * This work was supported by the Spanish Direcci6n General de Investigacih Cientifica y TCcnica Grant PB89-0359. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
In addition to the agonist-induced rise in [Ca2+];, there is evidence that some agonists also induce homeostatic processes, which tend to restore [Ca"]; to the basal level. An agonist-induced [CaZ+li decrease has been described in thrombin-stimulated platelets (ll), vasopressin-stimulated hepatocytes (12), thrombin-stimulated endothelial cells (13), and fMLP-stimulated neutrophils (14,15). It has also been shown that fMLP reduces [Ca2+Ii in neutrophils permeabilized to Ca" with ionomycin (14, 15). A similar effect has been obtained with phorbol esters (14, 15), suggesting that the [Ca2+]; decrease could be mediated by protein kinase C. The fMLP-induced reduction of [Ca2+Ii in ionomycin-treated human neutrophils has been attributed to a stimulation of the plasma membrane Ca2+ pump, leading to increased Ca" efflux (14, 15). Stimulation of the Ca2+ pump by phorbol esters has been demonstrated in neutrophil's inside-out plasma membrane vesicles (16) and cytoplasts (17), although this effect was evident only about 2 min after the addition of the phorbol ester. Additionally, it has been shown that phorbol esters inhibit fh4LP-induced MnZ+ entry, although the physiological significance of this effect was not clear (15).
We have used here thapsigargin-treated human neutrophils to investigate the mechanism of the fMLP-induced decrease in [Ca2+Ii. Thapsigargin is a very specific inhibitor of the sarcoplasmic and endoplasmic reticulum Ca2+ pump (18), which induces a permanent elevation of [Ca2+Ii in intact cells. This increase of [Ca2+], is initiated by slow release of Ca2+ from the intracellular Ca2+ stores, and it is maintained by a stimulation of Ca2+ entry due to activation of SOCP once the stores are emptied of Ca2+ (19). We show here that, in thapsigargin-treated human neutrophils, fMLP and phorbol ester, but not platelet activating factor (PAF), decrease [Ca2+Ii by The abbreviations used are: SOCP, store-operated Caz+ entry pathway; PAF, platelet-activating factor(s) (1-0-alkyl-2-sn-glycero-3-phosphorylcholine); N L P , N-formyl-methionyl-leucyl-phenylalanine; PDB, phorbol 12,13-dibutyrate.
The name SOCP was chosen by analogy to voltage-operated (VOCC), receptor-operated (ROCC), or second messenger-operated (SMOCC) Ca2+ channels. Although SOCP could be a subclass of SMOCC, we want to stress with this name that this Ca2+ entry pathway opens as a consequence of the emptying of the intracellular calcium stores rather than by the action of a second messenger generated directly on interaction of an agonist with its membrane receptor. SOCP is preferred to the term "capacitative Ca2+ entry" originally proposed by Putney (7), as it is generally thought at present that the entry from the extracellular medium to the cytoplasm does not take place "capacitatively" across the calcium stores (8). The term store-operated Ca2+ channel was not used, as the data presented here cannot distinguish between a channel and some other kind of entry mechanism, even though available evidence on the effects of membrane potential (9) and electrophysiological recordings (10) fits better to a "channel" than to a "carrier" mechanism.

13055
inhibition of SOCP and not by activation of the Ca2+ pump. Inhibition of SOCP by fMLP is transient and precedes the better known fMLP-induced activation of this entry pathway (4, 5, 15). Similarities with the effect of phorbol ester suggest that this inhibition may be mediated by protein kinase C, although a clear difference in sensitivity to staurosporin between fMLP-and PDB-induced decrease of [Ca2+Ii was apparent.

MATERIALS AND METHODS
Human neutrophils were obtained from blood of healthy volunteers anticoagulated by mixing 6:l (v/v) with acid citrate-dextrose. Dextran (T500, Pharmacia LKB Biotechnology Inc.) was then added to obtain a final concentration of 1.3%. After 45 min at room temperature, the upper phase containing no red cells was removed and centrifuged (300 X g, 10 rnin). The cell pellet was resuspended, layered on a Ficoll gradient (lymphocyte separation medium; Flow Laboratories, Irvine, Scotland), and centrifuged for 20 min at 400 X g. The cells were resuspended, and contaminating red cells were disrupted by hypotonic lysis (20). Neutrophils were finally suspended at 1-2% cytocrit in standard medium containing (in mM): NaCl, 145; KCl, 5; MgCI2, 1; CaCI,, 0.2; sodium-HEPES, 10; glucose, 10 (pH 7.4).
Neutrophils were loaded with fura-2 by incubation with 2-4 p~ fura-Z/AM for 30 min at room temperature in standard incubation medium. Cells were then washed twice and resuspended at 2% cytocrit in nominally Ca2+-free standard medium. [Ca2+Ii was measured in 0.5-ml aliquots of this cell suspension and kept at 37 "C under magnetic stirring using a fluorescence spectrophotometer constructed by Cairn Research Ltd. (Newnham, Sittinbourne, Kent, U.K.), which allows simultaneous excitation of fluorescence at 340, 360, and 380 nm. Fluorescence emission was set at 530 nm. Fluorescence readings were integrated at 1-s intervals, and [Ca2+]; was calculated from the ratio of the fluorescences excited at 340 and 380 nm (21). Mn2+ uptake was followed from the quenching of the fura-2 fluorescence excited at 360 nm wavelength, which is insensitive to variations in [Ca2+Ii. Ba2+ uptake was followed from the increase of fura-2 fluorescence excited at 360 nm wavelength. Fura-2 fluorescence excited at 360 nm increases by about 50% after binding to Ba2+ (22) (see spectrum of fura-2-Ba2+ complex in Ref. 23).
Fura-Z/AM was obtained from Molecular Probes, Eugene, OR. Thapsigargin, PAF, and ionomycin were purchased from Calbiochem. fMLP, PDB, and staurosporin were obtained from Sigma, Madrid. Other chemicals were obtained either from Sigma or E. Merck, Darmstadt. Fig. 1 (left panels) shows the effects of fMLP (1 p~) , PAF (36 nM), and PDB (100 nM) on [Ca2+]i of human neutrophils. The first two produced an increase of [Ca2+]i, and PDB did not modify [Ca2+Ii. fMLP produced an early and transient (30-9 duration) [Ca2+Ii peak due to Ca2+ release from the stores followed by a sustained plateau due to increased Ca2+ entry (5). The [Ca2+Ii increase produced by PAF was composed of a wider early [Ca2+Ii peak followed by a sustained plateau. Fig. 1  The decrease of [Ca2+Ji elicited by fMLP and PDB ( Fig. 1, right panels) could, in principle, be due either to stimulation of Ca2+ efflux or to inhibition of Ca2+ entry. The possible stimulation of a Ca2+ efflux mechanism by fMLP or PDB was investigated in two ways. In a first series of experiments, we  5 mM Ni2+ was used instead of EGTA to inhibit Ca2+ entry (results not shown). A second approach to this problem is shown in Fig. 3. Cells with filled Ca" stores (not treated with thapsigargin) were incubated in EGTA-containing medium, and the content of the intracellular Ca2+ stores was released to the cytoplasm with either ionomycin alone, ionomycin + fMLP, or ionomycin + PDB. In the three cases, [Ca2+Ii rose to a similar level, and the subsequent rate of decrease to the basal level was similar. The rate of decrease of [Ca2+]i should reflect the activity of the Ca2+ extrusion mechani~rns.~ Fig. 3 shows that no significant differences could be observed among the rates of [Ca2+Ii decrease in the three cases. This implies that at least in the first 30 s after the addition of fMLP or PDB, there is no significant activation of Ca" efflux.

RESULTS
If the decrease of [Ca2+]i induced by fMLP and PDB in thapsigargin-treated cells is not due to stimulated Ca2+ efflux, then it should be due to inhibition of Ca2+ entry. We have used Mn2+ and Ba2+ as Ca2+ surrogates for the Ca" entry pathway in order to trace unidirectional divalent cation movements. Fig. 4 shows that fMLP inhibited transiently Mn2+ entry in thapsigargin-treated cells (as reflected by the slowing of the quenching of F360), whereas PDB produced a long lasting inhibition of Mn2+ entry, Additionally, Fig. 4 shows that the inhibition of Mn2+ entry by both fMLP and PDB was strongly dependent on the [Ca2+], at the moment they were added (see also Table I Mn2+ has been shown to be a good Ca2+ surrogate in human neutrophils (5,6) but, in other cells, the adequacy of Mn2+ to follow unidirectional Ca2+ movements has been questioned (24)(25)(26)(27). We have also used Ba2+ to trace the Ca2+ entry pathway opened by emptying the intracellular Ca2+ stores with thapsigargin. As with Mn2+, Ba2+ is not pumped out by the Ca2+ pump, and its uptake can be followed by the increase of the fluorescence of fura-2 at the Ca2+-insensitive excitation wavelength of 360 nm (22)  The time course of the inhibition of Mn2+ entry by fMLP was studied by adding fMLP at different times before starting Mn2+ uptake measurements. The results of such experiments are shown in Fig. 6. Inhibition was maximum 10 s after the addition of fMLP, and then it decreased with time, 50% recovery being reached between 1 and 2 min after the addition of fMLP. The time course of the inhibitory effect of fMLP was also tested directly on Ca2+ entry. Fig. 7 shows the effect of fMLP on the increase in [Ca2+Ii induced by the addition of 1 mM Ca" to thapsigargin-treated cells incubated previously in 0.1 mM Ca2+. ca2+ entry was completely prevented 5-10 s after the addition of fMLP, and then it recovered with time. In similar experiments (not shown), PDB also inhibited the [Ca2+Ii rise induced by addition of 1 mM Ca2+ to thapsigargintreated cells. Inhibition by PDB was complete 10 s after its addition, and no recovery was observed for at least 5 min. PAF did not reveal any inhibitory effect in similar experiments (results not shown). Fig. 8 shows the concentration dependence of the inhibition by fMLP and PDB of the MnZ+ entry. With fMLP, the inhibition was complete at 100 nM and half-maximal at about 10 nM. With PDB, inhibition was complete at 10 nM and halfmaximal at about 1 nM. When [Ca2+Ii was kept at lower levels (thapsigargin-treated cells suspended in 0.1 mM Ca", [Ca2+Ii = 170-230 nM), the inhibitory effects were smaller, but the concentration dependence was similar (results not shown).
To test for the possible involvement of protein kinase C in the inhibition of SOCP, we have investigated the effects of the protein kinase C inhibitor staurosporin on the fMLP-and PDB-induced inhibition of the increase in [Ca2+Ji in experiments similar to those of Fig. 7. Fig. 9 shows that 100 nM staurosporin prevented the inhibition of the [Ca2+]i increase by PDB but blocked only partially the inhibition induced by fMLP. Similarly, 100 nM staurosporin prevented the inhibition of Mn2+ entry in thapsigargin-treated cells by PDB but had only a partial effect on the inhibition produced by fMLP (results not shown).
Inhibition of SOCP by fMLP in cells not treated with thapsigargin was investigated in the experiments shown in Fig. 10. Mn2+ was used to trace activation of SOCP, which was achieved by PAF. PAF produces a Ca2+ release from the intracellular Ca2+ stores, which is followed by an increased plasma membrane permeability ( 5 ) . PAF Table I). 1 mM Ca" (5). The addition of fMLP together with PAF produced an additional delay of the start of the PAF-induced acceleration of Mn" entry. This effect of fMLP was concentration-dependent, half-maximal inhibition being reached between 4 and 10 nM. Since fMLP itself mobilizes Ca2+ and accelerates MnZ+ entry (4, 5), it should in principle cooperate with and make faster the effect of PAP. Therefore, the results shown in Fig. 10 strongly support that fMLP induces an early and transient inhibition of SOCP, which prevents the effects of PAF during the first 1-2 min of incubation.

TABLE I Effect
Further evidence for this hypothesis was provided by the experiment shown in Fig. 11. We show that PAF induces a long lasting increase of [Ca2+] (upper panel), which typically shows secondary peaks dependent on the entry of extracellular [Ca"'] (4, 5). Simultaneous addition of fMLP and PAF induced a [Ca2+Ii transient whose width was less than half that obtained with PAF alone. This inhibitory effect was observed

DISCUSSION
It is widely known that fMLP activates Ca2+ influx from the extracellular medium in human neutrophils, and there is strong evidence that this activated Ca2+ influx is a consequence of the emptying of the intracellular Ca2+ stores (5, 6).
We report here that the fMLP-induced activation of Ca2+ influx through SOCP is preceded by a transient inhibition of this pathway, which contributes to delay the onset of activated Ca2+ entry. This inhibition is maximum 10 s after the addition of 1 p~ fMLP and gradually decreases with a half-time of about 1 min. fMLP-induced stimulation of Ca2' (Mn") entry (as a consequence of the emptying of the Ca2+ stores) becomes gradually apparent as inhibition is relieved, allowing Ca2+ to enter and refill the stores. This is why fMLP is able to induce Caz+ (Mn2+) entry on its own but transiently inhibits PAFinduced Caz+ (Mn") entry (see Figs. 10 and 11). Refilling of the stores then finally blocks Ca2+ (Mn2+) entry, leading the cell back to the resting situation.
It had been previously proposed that fMLP and PDB activate Ca2+ efflux (14-17). We do not find any evidence for such activation, at least within the initial 30 s after the addition of any of these agents. Previous evidence for activation of CaZ+ efflux by fMLP relies on the stimulation by fMLP of '%a loss in cells preequilibrated with 45Ca (17) and in the reduction by fMLP of the increase in [Ca2+Ii induced by ionomycin (14, 15). Stimulation of 46Ca loss by fMLP could also be explained as a consequence of the fMLP-induced release of tracer from the stores and the [Ca2+]i increase, which activates the plasma membrane Ca2+ pump. The effects reported for fMLP in ionomycin-treated neutrophils (14, 15) cannot be explained by inhibition of the ionophoric ionomycin-mediated Ca2+ influx. However, we have reported previously that low concentrations of ionomycin induce a large Ca2+ influx, which is not directly mediated by the ionophore acting at the plasma membrane but secondary to the ionophore-mediated emptying of the Ca2+ stores and activation of the SOCP (5). Inhibition of this Cazc entry pathway by fMLP may therefore reduce the [Ca2+Ii increase in the ionophoretreated cells without interfering with the ionomycin-mediated fluxes. Activation of "Ca loss by phorbol esters has also been reported (16,17), even though phorbol esters do not increase [Ca2+]i. This effect, however, was only evident about 2 min after the addition of the phorbol ester, whereas PDB-induced inhibition of Ca2+/Mn2+ uptake starts in less than 10 s (see Fig. 4. and Table I).
The fMLP-induced early inhibition of the SOCP is very much paralleled by the phorbol ester PDB, suggesting the involvement of protein kinase C in the mechanism of the inhibition. Both agents, fMLP and PDB, inhibited Ca2+ entry with a delay of 10 s or less (Figs. 2 and 4 and Table I), in both cases the inhibition was strongly dependent on [Ca2+Ii (Fig. 4 and Table I), and the decrease of [Ca2+Ii in thapsigargintreated cells took place at the same rate for both agents (see Fig. 2). Inhibition by fMLP was transient, while that induced by PDB was permanent, as it would be expected for a nonmetabolizable protein kinase C activator. Another important difference was the smaller sensitivity to staurosporin of the fMLP-induced inhibition, as compared with that induced by PDB. A similar discrepancy had been reported for the decrease of [Ca2+Ii induced by fMLP and PDB in ionomycintreated neutrophils (14, 15). The different sensitivity to this inhibitor may be due to a different mechanism for activation of protein kinase C in both cases. If, as suggested above, NLP-induced inhibition of SOCP was mediated by protein kinase C, the inability of PAF to produce such inhibition might seem rather puzzling. PAF is able to produce inositol 1,4,5-trisphosphate (28) and to release Ca2+ from the intracellular stores more strongly than fMLP does (5).4 Then, if activation of protein kinase C were due to diacylglycerol derived from phophatidylinositol4,5-bisphosphate by activation of phospholipase C, PAF should be at least as strong as fMLP to inhibit SOCP. An alternative view would be that the activation of protein kinase C produced by fMLP in human neutrophils took place by a different pathway, perhaps through phospholipase D (29) and that PAF were unable to activate this pathway.
It has been recently reported that thrombin inhibits Ca2+ entry induced by emptying the Ca2+ stores of endothelial cells with histamine or ionomycin (13). This effect seems rather similar to the fMLP-induced inhibition of SOCP reported here. However, several differences exist that make it difficult to decide at present whether both effects share a common mechanism. In the first place, phorbol esters do not modify Ca2+ entry in endothelial cells, and staurosporin does not affect the thrombin-induced inhibition of Ca2+ entry (13). Additionally, the thrombin-induced effect is much more long lasting than the effect of fMLP; no signs of recovery were still detected 5 min after the addition of thrombin (13). Despite these differences, phenomenological similarities between the effects of thrombin in endothelial cells and of fMLP in neutrophils suggests that agonist-induced inhibition of Ca2+ entry may be a more general phenomenon involved in Ca2+ homeostasis.
The physiological significance of the fMLP-induced inhibition of Ca2+ entry could be to prevent coincidence between the Ca2+ mobilization and the Ca2+ influx induced by the emptying of the stores, allowing for a fast return of [Ca2+]i to basal levels. The subsequent delayed activation of Ca2+ influx would then allow refilling of the intracellular Ca2+ stores, leaving the cell ready to respond to a new stimulus. Alternatively, inhibition of SOCP might allow the cell to distinguish between agonists that induce initially only mobilization of stored Ca2+ and a late Ca2+ influx, such as fMLP, and agonists M. Montero