Signal transduction by neutrophil immunoglobulin G Fc receptors. Dissociation of intracytoplasmic calcium concentration rise from inositol 1,4,5-trisphosphate.

The signal transduction mechanisms involved in the regulation of phagocytosis are largely unknown. We have recently shown that in neutrophils, when IgG-mediated phagocytosis is stimulated by formyl-methionyl-leucyl-phenyl-alanine (fMLP), the enhanced ingestion is dependent on the increase in [Ca2+]i which results from ligation of Fc receptors by the IgG-coated target (Rosales, C., and Brown, E. (1991) J. Immunol. 146, 3937-3944). Now, we have studied the mechanism by which this rise in [Ca2+]i occurs. Aggregated IgG, the monoclonal antibody 3G8 (which recognizes Fc receptor type III), and insoluble immune complexes caused an increase in [Ca2+]i. The rise in [Ca2+]i induced by Fc receptor ligation was resistant to pertussis toxin. In contrast, fMLP induced a rise in [Ca2+]i which was inhibited by pertussis toxin. fMLP-induced [Ca2+]i was accompanied by an accumulation of inositol 1,4,5-trisphosphate (IP3) which peaked by 15 s, and which was also abolished by pertussis toxin. IP3 accumulation after aggregated IgG, 3G8, or insoluble immune complexes was much less than after fMLP. Unlike [Ca2+]i rise induced by Fc receptor ligation, this small increase in IP3 was inhibited by pertussis toxin. These data demonstrated that the [Ca2+]i increase induced by Fc receptor ligation is not mediated by IP3. Immediate pretreatment of human polymorphonuclear neutrophils with optimal doses of fMLP also reduced subsequent increase in [Ca2+]i rise from thapsigargin, a sesquiterpene lactone tumor promoter that releases intracellular Ca2+ from IP3-sensitive stores without IP3 turnover. Similarly, to its effects on thapsigargin, fMLP inhibited the [Ca2+]i rise upon subsequent immune complex binding. Pretreatment of cells with immune complexes also prevented subsequent [Ca2+]i rise from thapsigargin and fMLP. These data demonstrate that IgG Fc receptor ligation and fMLP activation of human polymorphonuclear neutrophils use distinct signal transduction mechanisms to release Ca2+ from the same thapsigargin-sensitive intracellular pool. In contrast to fMLP, signal transduction for increased [Ca2+]i after Fc receptor stimulation does not involve a pertussis toxin-sensitive G protein, and is independent of IP3.

The signal transduction mechanisms involved in the regulation of phagocytosis are largely unknown. We have recently shown that in neutrophils, when IgGmediated phagocytosis is stimulated by formyl-methionyl-leucyl-phenyl-alanine (fMLP), the enhanced ingestion is dependent on the increase in [Ca2+]i which results from ligation of Fc receptors by the IgG-coated target (Rosales, C., and Brown, E. (1991) J. Immunol. 146,[3937][3938][3939][3940][3941][3942][3943][3944]. Now, we have studied the mechanism by which this rise in [Ca2+]i occurs. Aggregated IgG, the monoclonal antibody 3G8 (which recognizes Fc receptor type 111) , and insoluble immune complexes caused an increase in [Ca2+]i. The rise in [Ca2+Ii induced by Fc receptor ligation was resistant to pertussis toxin. In contrast, fMLP induced a rise in [Ca2+Ii which was inhibited by pertussis toxin. fMLP-induced [Ca2+Ii was accompanied by an accumulation of inositol 1,4,5-trisphosphate (IPS) which peaked by 15 8, and which was also abolished by pertussis toxin. IPS accumulation after aggregated IgG, 3G8, or insoluble immune complexes was much less than after fMLP. Unlike [Ca2+]i rise induced by Fc receptor ligation, this small increase in IP3 was inhibited by pertussis toxin. These data demonstrated that the [Ca2+Ii increase induced by Fc receptor ligation is not mediated by IP3. Immediate pretreatment of human polymorphonuclear neutrophils with optimal doses of fMLP also reduced subsequent increase in [Caz+Ii rise from thapsigargin, a sesquiterpene lactone tumor promoter that releases intracellular Ca2+ from IP3-sensitive stores without IP3 turnover. Similarly, to its effects on thapsigargin, fMLP inhibited the [Ca2+]i rise upon subsequent immune complex binding. Pretreatment of cells with immune complexes also prevented subsequent [Ca2+Ii rise from thapsigargin and fMLP. These data demonstrate that IgG Fc receptor ligation and fMLP activation of human polymorphonuclear neutrophils use distinct signal transduction mechanisms to release Ca2* from the same thapsigargin-sensitive intracellular pool. In contrast to fMLP, signal transduction for increased [Ca2+Ii after Fc receptor stimulation does not involve a pertussis toxin-sensitive G protein, and is independent Of Ips.
* This work was supported by Public Health Service Grant GM 38330. 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.
$ To whom reprint requests should be addressed Infectious Diseases Div., Washington University, School of Medicine, Box 8051, 660 S. Euclid Ave., St. Louis, MO 63110.
Antibodies have two major functions: the binding to antigen via their antigen-combining sites and the activation of defense mechanisms via their carboxyl termini, the Fc region. Ligation of the Fc portion to specific receptors on many cells of the immune system triggers several functions including phagocytosis, antibody dependent cell-mediated cytotoxicity, secretion of inflammatory mediators, generation of the respiratory burst, and clearance of immune complexes (1,2). Because the Fc receptors (FcR)' mediate these important defense functions of the immune system, understanding signal transduction from receptor-ligand interaction is a central question in phagocyte biology.
Human phagocytic cells bear at least three distinct types of FcR, all members of the Ig gene superfamily. Knowledge of their structures and gene organization has progressed rapidly in the past few years (3, 4). However, the signals transduced by each of the FcR to activate particular cell responses are not completely elucidated.
Human neutrophils (PMN) express two FcR FcRII (CD32) which has a traditional membrane spanning domain and a cytoplasmic tail (5, 6) and FcRIII (CD16) which has a glycosylphosphatidylinositol anchor (7). FcRIII exists as a transmembrane anchored form in NK cells (8). Each of these receptors has been shown to express structural polymorphisms that might be functionally significant (9, 10). FcRIII, with a higher density on the cell membrane than FcRII, has been thought to be primarily a binding molecule, used to present ligand to FcRII, which in turn will mediate signaling across the membrane (11,12). This idea is attractive because the glycosyl-phosphatidylinositol linkage of FcRIII has no obvious signaling mechanism. However, recent reports (13, 14) and our own results (15) indicate that FcRIII is able to transduce the signal for several cell responses, including an increase in cytoplasmic free Ca2+ concentration ([Ca2+Ii).
Previously, we found that FcR-mediated phagocytosis stimulated by W L P , but not by phorbol esters or platelet-activating factor, is dependent on an increase in [Ca2+Ii (15) 5265 been extensively studied in PMN stimulated with the chemoattractant peptide fMLP. fMLP receptor signal transduction involves a pertussis toxin-sensitive G protein and activation of a phospholipase C, resulting in the production of inositol 1,4,5-trisphosphate (IP,) (16), which leads to the increase in [Ca2+Ii (17). IP3 is thought to release Ca2+ directly from intracellular stores and indirectly open membrane calcium channels either through a metabolite or via increased [Ca2+Ii itself (18)(19)(20).
Results presented in this report indicate that, in contrast to fMLP receptor stimulation, FcR ligation causes a [Ca2+Ii rise via a pertussis toxin-insensitive pathway which is independent of IPS generation. FcR ligation and fMLP binding apparently release Ca2+ to the cytosol from the same intracellular pool which is sensitive to the sesquiterpene lactone thapsigargin. Thus, signal transduction for increased [Ca2+Ji in PMN initiated by different ligand-receptor interactions can occur by at least two distinct pathways.

MATERIALS AND METHODS
Reagents-A 10 times concentrated stock solution of Hanks' balanced salt solution (HBSS) was obtained from GIBCO. fMLP, phor-bo1 dibutyrate, lithium chloride, mannitol, EGTA rabbit anti-BSA antibodies, and BSA were purchased from Sigma. fMLP and phorbol dibutyrate were stored as stock solutions in dimethyl sulfoxide from Aldrich Chemical Co. at -70 "C, and diluted into buffer just before use. Thapsigargin was purchased from LC Services Co. (Woburn, MA). Fura-2-AM ester was from Molecular Probes (Eugene, OR). 25% human serum albumin was from the New York Blood Center (New York, NY). RPMI-1640 tissue culture media was from Whittaker Bioproducts (Walkersville, MD); fetal calf serum was from Hyclone (Logan, UT). Hybridoma cells producing anti-Fc receptor 111 mAb 3G8 were the kind gift of Dr. Jay Unkeless, Mt. Sinai Medical Center, New York, NY. 3G8 IgG was purified from ascites by octanoic acid precipitation, and DEAE-Sephacel chromatography (21). Pertussis toxin was from List Biological Laboratories Inc. (Campbell, CA). The IP3-specific radioimmunoassay kit, [3H]inositol, and a mixture of tritiated inositol mono-, bis-, and trisphosphate were purchased from Amersham Corp. Scintillation liquid CytoScint ES* was from ICN Biomedicals Inc. (Cleveland, OH). All other chemicals of analytical grade purity were from Sigma.
Preparation of Neutrophil Suspensions-PMN were obtained from heparinized venous blood from healthy adult donors. PMN were purified by standard techniques as previously described (22). Generally cells were suspended in HBSS, containing 10 mM HEPES, without Ca2+ or M e . When necessary, 1.5 mM Ca2+ and 1.5 mM M e were added to this buffer (HBSS++).
Aggregated ZgG-IgG purified from human serum by octanoic acid precipitation (21) was heated at 60 "C for 30 min, chilled on ice, centrifuged at 12,000 rpm for 20 min, and chromatographed on a sizing column packed with Bio-Gel A-1.5 m, Bio-Rad. Aggregated IgG (agg-IgG) collected with the void volume was concentrated and stored at 4 'C.
Insoluble Immune Complexes-10 pl of bovine serum albumin (10 mg/ml) were mixed with 300 pl of rabbit anti-BSA antibody (2.6 mg/ ml), incubated at 37 "C for 60 min, and then chilled on ice. The insoluble immune complexes formed were washed and resuspended in 300 pl of HBSS. A 25-fold final dilution of these preparation was used to stimulate FcR. This concentration was found to give in PMN a [Ca2+Ii rise similar in magnitude to that elicited by lo-' M fMLP.
Fluorescence Calcium Measurements-PMN were loaded with fura-2-AM as previously described (23). Fluorescence changes of a 2-ml stirred PMN suspension kept at 37 "C were monitored with a F-2000 Hitachi Instruments (Danbury, CT) spectrofluorimeter, using 340 and 380 nm excitation wavelengths and 510 nm emission wavelength. Calcium concentrations were calculated as described by Grynkiewicz et al. (24).
Treatment of Neutrophils with Pertussis Toxin-PMN at lo7 cell/ ml in HBSS++ were treated with pertussis toxin (PT) (0-10 pg/ml) at 37 "C for 75 min. After this time, cells were washed twice and resuspended in fresh buffer for [Ca2+li and IP3 measurements. 2 pg/ ml pertussis toxin caused a complete inhibition of degranulation, as measured by release of myeloperoxidase in response to fMLP (data not shown).
Measurement of Inositol 1,4,5-Trisphosphate by RIA-Neutrophils (3 X 107/ml) in buffer containing 20 mM LiCl were stimulated with IO-' M fMLP or 300 pg/ml aggregated IgG. After various times an equal volume of 20% trichloroacetic acid was added. Cells were vortexed and let stand at 4 "C for 10 min. After centrifugation, supernatants were recovered, treated with 10 mM mannitol, and extracted four times with 3 ml of ether. The aqueous phase was neutralized with 7.5% NaHC03 and used to measure IP3 with a commercial RIA kit following the manufacturer's instructions.
Separation of Inositol Phosphates by HPLC-PMN were incubated in inositol-free media with 5.5 mM glucose for 4 h in the presence of 50 pCi/ml [3H]inositol. Pertussis toxin (2 pg/ml) was added to the cells at the same time that the labeled inositol. Pertussis toxin did not affect the level of incorporation of radioactivity into the cells. 10 million PMN in HBSS++ containing 20 mM LiCl were treated with fMLP for 15 s, or agg-IgG or mAb 3G8 for 30 s. For these experiments the concentration of fMLP and agg-IgG was increased to lo" and 500 pg/ml, respectively, to induce maximum stimulation and to promote the highest production of inositol phosphates. Inositol phosphates were then isolated by trichloroacetic acid precipitation and ether extraction as described above. Samples were neutralized and applied to a S5 SAX Spherisorb HPLC column, from Phase Separations Inc. (Norwalk, CT). Inositol phosphates were eluted with an ammonium formate gradient from 0 to 1 M in 0.5 h. Reproducibility of the gradient was monitored by the resolution of a mixture of AMP, ADP, and ATP added as an internal standard to each sample. Fractions (0.5 ml) were collected and counted in 10 ml of scintillation liquid. Peaks corresponding to inositol mono-(IP,), bis-(IP2), and trisphosphate (  I  & '  ,  ) were identified by the retention time of known radioactive standards.  (Table I). As shown previously (15) the [Ca2+Ii increase induced by agg-IgG and 3G8 came exclusively from intracellular stores. While extracellular EGTA decreased the extent (Fig. 1) and duration (not shown) of the fMLP-induced [Ca2+Ii rise, EGTA had no effect on the increase in [Ca2+]i induced by agg-IgG or 3G8 (Fig. 1). Moreover neither extracellular Mn2+ or Ni2+, which competitively inhibit extracellular Ca2+ entry induced by fMLP (25), affected the extent or duration of agg-IgG or 3G8-induced were not statistically different from baseline. ND, not done. [Ca2+Ii rise (not shown). Pretreatment of PMN with pertussis toxin caused a dose-dependent inhibition of the fMLP-induced [Ca2+Ii rise. Maximal inhibition (85%) was observed at 2 pg/ml pertussis toxin (Fig. 2 A ) . Surprisingly 3, and Table I). This [Ca2+]; rise from IIC was more prolonged than the one from f"LP or agg-IgG.

Signal Transductic
Despite the marked increase in [Ca2+Ii rise compared to agg-IgG or 3G8, pertussis toxin treatment still had no effect on the response induced by IIC (Fig. 3). These results indicated that there is not a pertussis toxin-sensitive G protein involved in the mechanism of [Ca2+]i release induced by FcR ligation in PMN. These differences in the generation of an increase in [Ca2+Ii by fMLP and FcR ligands led us to test the hypothesis that the molecular mechanisms involved in the increase were distinct for the two ligands.
IP3 Is Not the Second Messenger for FcR-mediated [Ca2+]; Rise-Since inositol 1,4,5-trisphosphate produced after stimulation of several types of receptors is known to mediate Ca2+ release from intracellular stores (18, 19), we examined whether ligation of FcR and fMLP receptors also resulted in formation of IP3. PMN were stimulated with fMLP, agg-IgG, 3G8, or IIC for different periods of time and then inositol phosphates were isolated by trichloroacetic acid precipitation and ether extraction. IP3 was quantitated in the neutralized cell extracts using an IP3-specific radioimmunoassay kit. Basal IP3 measured in this way was 1.38 k 0.24 pmol/106 PMN (mean f S.E., n = 12). The relatively high levels of basal IP3 measured by RIA, may reflect a partially activated state generated during the purification of PMN. fh4LP induced an increase in formation of IP3 by 15 s after stimulation (Table I, Fig. 4). 1 nM fMLP produced a smaller but still significant and easily detectable peak of IP3 with the same kinetics as 10 nM fMLP (Fig. 4). The effect of fMLP on both IP3 and [Ca2+]; was dose dependent, since 1 nM fMLP led to approximately 50% of the increases in [Ca2+Ii and IP3 seen with 10 nM fMLP (Table I, Fig. 4). Agg-IgG gave a [Ca2+]; rise comparable in magnitude to that elicited by 1 nM fMLP (Table I) but caused only very small IP3 increases over base line (Fig. 4). In fact at no time was the measured IP3 in response to either agg-IgG or 3G8 statistically different from base line. As expected, IP3 formed after 10 nM fMLP stimulation was eliminated by pertussis toxin (Table I). Importantly, the small amount of IP3 generated by FcR ligation by either agg-IgG or 3G8 also was completely eliminated by  (Table I). Treatment with pertussis toxin had no significant effect on the basal IP,. Because pertussis toxin had no effect on FcR-dependent increase in [Ca2+]i (Table I), these data suggested that the small rise in [Ca'+li caused by agg-IgG or 3G8 was unrelated to IP3 generation.
To further compare FcR ligation with fMLP, IIC were used as an FcR ligand, at a concentration which gave an equivalent increase in [Ca'+li as M fMLP (Fig. 3, Table I). Similar to agg-IgG and 3G8, the IPS produced after stimulation with IIC was minimal compared to fMLP and not statistically different from base line. The small nonsignificant rises in response to IIC were completely eliminated by pertussis toxin treatment (Table I). Kinetic experiments showed that there was no significant increase in IP, at any time from 5 to 30 s after IIC addition. These data demonstrate that even with an FcR ligand that gives equivalent [Ca'+]i rises as 10 nM fMLP, no significant rise in IP, could be found. These data are consistent with the hypothesis that the increase in [Ca'+Ii caused by FcR ligation is unrelated to IP3 generation.
To verify and extend the results obtained by RIA, we measured inositol phosphate turnover directly. In order to do this, PMN were labeled with [3H]inositol and treated with pertussis toxin or buffer for 75 min before stimulation. Cells were incubated with fMLP for 15 s and agg-IgG or 3G8 for 30 s, and then inositol phosphates were isolated and separated by HPLC. Pertussis toxin did not affect the level of incorporation of radioactivity into PMN or the base-line (unstimulated) levels of inositol phosphates. fMLP induced phosphoinositol turnover, manifested by the appearance of clear IP1, IP2, and IPS peaks above the nonstimulated level of inositol phosphates. Treatment with pertussis toxin completely prevented IP3, IP', and IP1 formation (Fig. 5). In agreement with the RIA, stimulation with agg-IgG caused a much smaller increase of inositol phosphates over nonstimulated conditions. In this assay, the amount of inositol phosphates generated by agg-IgG was statistically different from base line. Importantly, for all these inositol phosphates, this small rise was eliminated by pertussis toxin. Quantitatively similar results were obtained for mAb 3G8 stimulation, but the magnitude of the inositol phosphate accumulation was smaller (Fig. 5). These results confirmed the previous observations. These data supported the hypothesis that although a small .."" . amount of IPS can be detected after FcR stimulation, it is not coupled to the generation of the [Ca2+]i rise, since IP3 is completely eliminated by treatment with pertussis toxin, but the increase in [Ca'+li is unaffected. fMLP and Insoluble Immune Complexes Rebase [Ca"li from the Same Pool-Studies in other cell types that compared Ca'+ release by IP3 and other substances such as GTP, caffeine, and ryanodine have suggested that there are at least two intracellular Ca" pools that can be independently regulated (26,27). Since increments of IP3 in PMN after agg-IgG, immune complexes, or 3G8 treatment were minimal and no significantly different from resting IP, levels, we examined the possibility that the Ca'+ mobilized after FcR ligation was from a different pool than the one that responds to IP3 (17). To address this question, PMN were treated with thapsigargin, a tumor promoting sesquiterpene lactone that releases Ca" from the IP3-sensitive pool by blocking the Ca2+-ATPase required for reuptake (28). Exposure of PMN to 200 nM thapsigargin caused a rapid release of Ca'+ from intracellular stores and also entry of Ca2+ from the extracellular media (not shown). In the presence of external 3 mM EGTA, thapsigargin induced a [Ca2+Ii rise that slowly returned to lower levels (Figs. 6A and 7A), leaving the cells with this Ca'+ pool empty. Fig. 6 shows the [Ca2+Ii changes which occurred when fura-2-loaded PMN were treated with thapsigargin before and after stimulation with fMLP. Thapsigargin reduced almost completely the [Ca2+ji rise induced by fMLP (Fig. 6A) indicating that the fMLP-sensitive pools of ca'+ were depleted by thapsigargin. Thapsigargin also eliminated the [Ca'+]i rise from stimulation with IIC (Fig. 7A) compared to 225 f 25 nM (mean f S.E., n = 8) from IIC with no pretreatment. Thus, both fMLP and IIC release Ca2+ from a thapsigargin-sensitive intracellular pool. While further doses of fMLP after an initial dose led to much smaller increases in [Ca2+li, we could not distinguish whether this inhibition on repeated agonist exposure was due to receptor occupancy, receptor desensitization, or depletion of the relevant intracellular pool. Therefore, we examined the size of the fMLP-releasable intracellular pool by adding thapsigargin after fMLP stimulation. Pretreatment of PMN with IO-' M fMLP inhibited a subsequent increase in [Ca2+]; in response to thapsigargin (Fig. 6B). The maximum [Ca2+]; after thapsigargin in the absence of fMLP was 71.4 f 2.1 nM (mean f S.E., n = 5), while after fMLP, it was 24 f 3.2 nM (mean f S.E., n = 3), suggesting that -69% of this intracellular pool had been depleted by prior exposure to fMLP. Thus, we concluded that lo-' M fMLP effectively depleted this pool.
Next, we stimulated PMN with IIC at a dose that gave a [Ca2+Ii rise similar to that from lo-' M fMLP. The [Ca2+]i rise after thapsigargin was almost completely eliminated when IIC were given before thapsigargin (Fig. 7B). The thapsigargininduced [Ca2+]; rise was 25 f 1.4 nM (mean f S.E., n = 3) after IIC compared to 71.4 f 2.1 nM (mean f S.E., n = 5) when thapsigargin was given first, an extent of depletion similar to that seen with fMLP. Most significantly, IIC and fMLP were able to inhibit the [Ca2+]; rise due to subsequent stimulation with the other agonist (Figs. 6C and 7C). The maximum [Ca2+]; rise induced by IIC after fMLP was 69 f 2.8 nM (mean f S.E., n = 3) compared to 225 f 25 nM (mean f S.E., n = 8) in the absence of fMLP, a 69% inhibition, identical to that seen when thapsigargin was added after fMLP. Thus both fMLP and IIC could effectively deplete Ca2+ from a thapsigargin-sensitive pool, and each inhibited a subsequent [Ca2+]; rise from the other. We concluded that, despite the differences in mechanism of release of Ca2+ from intracellular stores, fMLP and Fc receptor ligation utilized the same intracellular Ca2+ pool. Interestingly, prior incubation with fMLP did not depress the [Ca2+Ii rise induced by agg-IgG. In these experiments agg-IgG induced an increase in [Ca2+Ii of 116 f 8 nM (mean f S.E., n = 8) when incubated with PMN prior to fMLP; agg-IgG induced an increase of 104 f 2 nM when incubated after fMLP. Similarly, agg-IgG did not alter the fMLP-induced increase in [Ca2+]i. The fMLPinduced increase in [Ca2+Ii was 303 f 7 nM before agg-IgG and 291 f 8 nM after agg-IgG. Nonetheless, prior treatment with thapsigargin eliminated the [Ca2+Ii rise on subsequent exposure to agg-IgG. These data suggest that there was sufficient Ca2+ left in the thapsigargin-releasable pool after fMLP to support a normal submaximal FcR stimulation.

An increase in [Ca2+]i occurs as a result of interaction of
IgG with PMN (29,30). This increase in [Ca2+], is thought to regulate multiple events which occur on ligation of Fc receptors, including degranulation (31), activation of arachidonate metabolism (32,33), and diapedesis (34). Recently, we have shown that the increase in [Ca2+], which takes place after Fc receptor ligation is necessary for fMLP induction of the highly phagocytic state characteristic of PMN at inflammatory sites (15). Therefore, we have investigated the mechanism by which [Ca2+Ii is elevated after engagement of Fc receptors by model immune complexes.
Recent data from several laboratories have demonstrated that the IgG-mediated increase in [Ca2+Ii comes entirely from release of intracellular stores of Ca2+ (14,15). The best understood mechanism by which Ca2+ is released is by production of IP3, which acts to open a channel in the membrane of the intracellular pool of Ca2+ (17,35). IPS receptors have been purified and cloned from several tissues (36)(37)(38)(39)(40)(41). These cDNAs predict a molecule with many of the characteristics expected of a ligand-gated channel. Therefore, we examined the hypothesis that the IgG-mediated increase in [Ca2+Ii occurs because of generation of IPS. Our data strongly suggest that this is not the case. First, although increases in [Ca2+]; in response to agg-IgG were about half of t,hat achieved with an optimal concentration of fMLP, very little IP3 gen-eration could be detected from Fc receptor ligation, by either of two assays. Using insoluble immune complexes, which are a more physiological and potent stimulator for FcR, we found that the [Ca2+Ii rise was as large as the one induced by fMLP. These immune complexes also generated very little Ips. Moreover, there was no clear relationship between the amount of Ips generated and the [Ca2+Ii rise induced by three different FcR ligands (3G8, agg-IgG, and IIC). In contrast, there was a correlation between Ips generation and [Ca2+], rise for fMLP, a ligand which has been shown previously to activate the pathway leading to Ips generation (16). We considered that the Ips might be metabolized extremely rapidly to IP2 and IP1 after receptor ligation. Both of these potential metabolites of Ips were elevated upon IgG stimulation of PMN, but neither was elevated strikingly more than Ips in proportion to fMLP stimulation, suggesting that increased metabolism of Ips was not the explanation for the small amount of Ips observed.
The strongest evidence that Ips was not involved in the increase in [Ca2+Ii came from studies with pertussis toxin. While pertussis toxin dramatically decreased the rise in [Ca"] i observed in response to fMLP, it had no significant effect on the [Ca2+Ii rise after agg-IgG, even at five times the concentration which was maximally effective in inhibition of the fMLP response. Similarly pertussis toxin had no effect on the [Ca2+Ii rise induced by IIC. In contrast, pertussis toxin inhibited the increase in Ips caused by agg-IgG, 3G8, and IIC, and indeed inhibited accumulation of all the water-soluble inositol phosphates observed with both fMLP and Fc receptor ligands. Thus, there was a clear disparity between the effects of pertussis toxin on [Ca2+Ii and Ips after Fc receptor ligation with agg-IgG and IIC. Normal increases in [Ca2+]i were observed in cells in which the Ips response was completely abrogated.
In contrast, pertussis toxin treatment of PMN inhibited both the fMLP-induced Ips accumulation and fMLP-induced [Ca2+Ii. A very small [Ca2+]i increase (15% of normal) was still observed in maximally intoxicated cells, as reported by others (42), even though inositol phosphate turnover was not detected. This suggests that fMLP may also be able to release Caz+ by a mechanism other than Ips. However, unlike Fc receptor ligation, this is a minor component of the signal for Ca2+ release. Also in contrast to FcR ligation, the main signal transduction mechanism for increase in [Ca2+]i by fMLP receptors is via a pertussis toxin-sensitive G protein pathway (16, 43).
To determine whether IgG induced release of Ca2+ from the same pool as fMLP, experiments were performed with thapsigargin. Thapsigargin has been used to distinguish between the Ips-releasable intracellular pool and other Ca2+ stores (28,44). Thapsigargin treatment of PMN inhibited subsequent Ca2+ released by both fMLP (Ips-dependent) and agg-IgG (Ips-independent). Pretreatment of PMN with optimal doses of fMLP inhibited subsequent increase in [Ca2+Ii caused by thapsigargin, suggesting that fMLP had largely emptied this intracellular Ca2+ pool. Although, under the same conditions fMLP had no effect on the [Ca2+]i increase induced by FcR ligation with agg-IgG, it inhibited the [Ca2+Ii rise by IIC (Fig.  6C). Reciprocally, IIC pretreatment blocked the subsequent [Ca2+Ii increase by fMLP (Fig 7C). This suggests that fMLP and IgG may release Ca2+ from the same thapsigargin-sensitive pool. These data raise the possibility that two functionally distinct mechanisms exist in PMN to release Ca2+ from this pool. One is Ips sensitive and stimulated by fMLP. The other is Ips insensitive, and is activated by FcR stimulation.
The identity of the intracellular mediator of Ca2+ release induced by agg-IgG remains an important unsolved problem.
It is notable that in other systems, fatty acids or their metabolites have been reported to release Ca2+ from an Ips-sensitive pool, without inducing phosphoinositide turnover (45). While these experiments suggest a possible alternative intracellular Ca2+ ionophore, it is difficult in these in uitro experiments to exclude a detergent effect of the added fatty acids. In this regard our preliminary data show that low concentrations of pura-bromophenacylbromide, a phospholipase A2 inhibitor, prevent the increase in Ca2+ in PMN after addition of agg-IgG but have little effect on fMLP-induced release. This may suggest that the phospholipase A2 activity, which has been closely associated with some forms of IgG Fc receptor (46,47), is necessary for the increase in [Ca2+]i which follows recognition of both particulate and soluble immune complexes.