Differential effects of the protein kinase C activator phorbol 12-myristate 13-acetate on calcium responses and secretion in adherent and suspended RBL-2H3 mucosal mast cells.

Adhesion of RBL-2H3 mucosal mast cells to fibronectin-coated surfaces has been linked to changes in secretion and tyrosine kinase activity. We now show that adhesion affects the sensitivity of RBL cells to the protein kinase C activator phorbol 12-myristate 13-acetate (PMA). In suspended cells, PMA inhibited antigen-induced calcium influx (as measured by manganese influx) and changes in intracellular free calcium and had complex effects on antigen-stimulated secretion. However, in adherent cells PMA had little effect on these responses. Suspended cells only secreted in response to thapsigargin if they were co-treated with PMA, while adherent cells secreted in response to thapsigargin alone. The thapsigargin-induced secretion in adherent cells was inhibited by protein kinase C down-regulation and by the protein kinase C inhibitor GF 109203X, but not by calphostin C. We suggest that protein kinase C is constitutively activated in adherent cells, possibly due to modification of the regulatory domain of the enzyme.

Adhesion of RBL-2H3 mucosal mast cells to fibronectin-coated surfaces has been linked to changes in secretion and tyrosine kinase activity. We now show that adhesion affects the sensitivity of RBL cells to the protein kinase C activator phorbol 12-myristate 13-acetate (PMA). In suspended cells, PMA inhibited antigen-induced calcium influx (as measured by manganese influx) and changes in intracellular free calcium and had complex effects on antigen-stimulated secretion. However, in adherent cells PMA had little effect on these responses. Suspended cells only secreted in response to thapsigargin if they were co-treated with PMA, while adherent cells secreted in response to thapsigargin alone. The thapsigargin-induced secretion in adherent cells was inhibited by protein kinase C down-regulation and by the protein kinase C inhibitor GF 109203X, but not by calphostin C. We suggest that protein kinase C is constitutively activated in adherent cells, possibly due to modification of the regulatory domain of the enzyme.
The RBL-2H3 mucosal mast cell line has been used extensively as a model of stimulus secretion coupling (1). Activation of these cells by antigen leads to a complex series of events including tyrosine phosphorylation of various proteins (2,3), including the receptor for immunoglobulin E (IgE) 1 (4), phosphoinositide breakdown (5) leading to activation of protein kinase C (6), emptying of intracellular calcium stores by inositol 1,4,5-trisphosphate (IP 3 ) (7,8), and influx of calcium across the plasma membrane (9 -11). These events culminate in the secretion of various mediators of the inflammatory response (1,8). It is clear that both the increase in intracellular calcium and protein kinase C activation are important steps in the signaling pathway and that these two signals act synergistically to promote secretion (12,13).
Activation of protein kinase C with the phorbol ester phorbol 12-myristate 13-acetate (PMA), alone, does not induce secre-tion in rat basophilic leukemia (RBL) cells (12)(13)(14). Some laboratories have reported that PMA potentiates antigen-induced secretion at concentrations below 15 nM (12,15), but other reports do not support this finding (13,14). Nevertheless, there is general agreement that PMA markedly potentiates secretion in response to calcium ionophore (12)(13)(14). A similar synergism has been seen when protein kinase C is activated by PMA while intracellular calcium is increased by treatment with the endoplasmic reticulum Ca 2ϩ -ATPase inhibitors thapsigargin 2 or cyclopiazonic acid (16). Additionally, the protein kinase C inhibitors staurosporine, Ro31-7549, and calphostin C have been shown to inhibit antigen-stimulated secretion (17). In general, it appears that the combination of protein kinase C activation and increases in intracellular calcium are sufficient to induce secretion.
In addition to promoting secretion, activation protein kinase C by PMA has a second, inhibitory effect on RBL cells in suspension (12,13,15). Increases in intracellular Ca 2ϩ are inhibited at concentrations above 10 nM (12,13,15), possibly by the inhibition of phospholipase C-␥ (13,18), thus preventing phosphoinositide breakdown. Some groups have also shown that antigen-stimulated secretion is inhibited by high concentrations of PMA (12,15), presumably due to the inhibition of the Ca 2ϩ response.
In the past, experiments on RBL cells have been performed interchangeably with cells in suspension or with adherent cells. However, recent experiments have shown that adhesion itself affects RBL cell responses. Adhesion of RBL cells results in the tyrosine phosphorylation of several proteins including pp125 FAK (19). In addition, antigen-stimulated secretion is enhanced in adherent RBL cells (20). In studying the effects of protein kinase C activation on secretion and calcium handling, we have discovered another effect of adhesion on RBL cell responses, namely a loss of sensitivity to the effects of the protein kinase C activator, PMA.

EXPERIMENTAL PROCEDURES
Sensitized RBL Cells-All experiments were performed with the secreting subline 2H3 of rat basophilic leukemia cells (21) maintained in monolayer culture in Eagle's minimum essential medium containing 10% fetal bovine serum, 8% newborn bovine serum, and antibiotics as described (22). For secretion experiments in adherent cells, 0.4 ϫ 10 6 cells in 0.3 ml of culture medium containing 0.18 g of mouse monoclonal IgE anti-dinitrophenyl (mIgE␣DNP) for sensitization were plated into each well of 24-well multiwell plates (Falcon, Oxnard, CA) and incubated overnight at 37°C in a humid atmosphere containing 5% CO 2 . For fura-2 experiments with adherent cells, aliquots of 5 ϫ 10 6 cells in 2.5 ml of culture medium containing 1 g/ml mIgE␣DNP were added to 35-mm plastic culture dishes, each of which contained a 22 ϫ 22-mm glass coverslip that had been scored down the middle. The cells were incubated overnight at 37°C in a humid atmosphere containing * This work was funded in part by Grant AI 19910 from the National Institutes of Health and National Science Foundation Grant DCB-9105361. 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.
5% CO 2 . For secretion and fura-2 experiments with cell suspensions, cells grown to confluence in a 75-cm 2 tissue culture flask were incubated overnight at 37°C in 10 ml of culture medium containing 6 g of mIgE␣DNP.
Solutions-The standard saline solution used was a modified Tyrode's solution composed of 135 mM NaCl, 5 mM KCl, 1.8 mM CaCl 2 , 1 mM MgCl 2 , 5.6 mM glucose, 0.05% gelatin, and 10 mM HEPES adjusted to pH 7.4 with NaOH. Ice-cold quenching solution for secretion experiments contained 135 mM NaCl, 5 mM KCl, and 10 mM Na-HEPES (pH 7.4). For loading cells with fura-2, the saline solution contained 250 M sulfinpyrazone (Sigma) and 0.1% bovine serum albumin instead of gelatin to maximize uptake and retention of the dye (23,24).
Reagents-Fura-2 acetoxymethyl ester (fura-2/AM) was purchased from Molecular Probes (Junction City, OR). Calphostin C and GF 109203X were purchased from Calbiochem-Novabiochem International (San Diego, CA). PMA, thapsigargin, and 4-methylumbelliferyl-Nacetyl ␤-D-glucosaminide were purchased from Sigma. Stock solutions of fura-2/AM, PMA, thapsigargin, 4-methylumbelliferyl-N-acetyl ␤-Dglucosaminide, calphostin C, and GF 109203X were prepared in dry dimethyl sulfoxide. Cells were never exposed to Ͼ0.2% dimethyl sulfoxide, and at this or lower concentrations the solvent did not affect the responses of RBL cells. Purified mIgE␣DNP (25) was a gift from Barbara Baird and David Holowka, Department of Chemistry, Cornell University. The antigen used was bovine ␥-globulin to which an average of 10 dinitrophenyl groups/molecule had been coupled (26), except for the measurements of translocation and tyrosine phosphorylation of protein kinase C, where the dinitrophenyl groups were coupled to bovine serum albumin.
Secretion-This was determined from the release of the granuleassociated enzyme, ␤-hexosaminidase. Secretion was carried out in 24-well plates in which cells had been plated overnight and washed with saline solution, or in polystyrene tubes containing 0.5 ϫ 10 6 cells in saline solution. Secretion was initiated by antigen (1 g/ml) added directly to the wells or tubes, and was terminated by adding ice-cold quenching solution to each well after 30 min for adherent cells or 60 min for suspended cells. These incubation times resulted in maximal secretion. An aliquot from each supernatant was assayed fluorimetrically for ␤-hexosaminidase (excitation 360, emission 450) using 4-methylumbelliferyl-N-acetyl ␤-D-glucosaminide as the substrate. Secretion is expressed as a percent of the ␤-hexosaminidase content of the cells prior to stimulation.
Fura-2 Measurements in Cell Suspensions-For measurements of free ionized calcium, sensitized cells (10 6 /ml) were incubated with 0.5 M fura-2/AM for 1 h at 37°C. In manganese quench experiments, the fura-2/AM concentration was 5 M. The cells were then washed and resuspended in saline solution containing 250 M sulfinpyrazone and 0.05% gelatin. Three-ml aliquots of cell suspension (10 6 cells/ml) were added to acrylic cuvettes maintained at 37°C and constantly stirred. Fura-2 fluorescence at 510 nm was monitored with a Perkin-Elmer LS-5 fluorescence spectrophotometer. Fura-2 was excited at 334 nm for measurements of free ionized calcium, or at 360 nm for manganese quench experiments.
Fura-2 Measurements in Adherent Cells-Sensitized cells on coverslips were washed twice and incubated for 45 min with 1 M fura-2/AM for measurements of free ionized calcium, or 5 M fura-2/AM for manganese quench experiments. After loading, the cells were washed twice and each coverslip half was placed in a holder made from a 1.5-ml centrifuge tube, which was then inserted into a 3-ml acrylic cuvette containing 2.5 ml of saline solution containing 250 M sulfinpyrazone and 0.05% gelatin. The temperature was maintained at 37°C, and fluorescence was monitored as described above for suspended cells.

Measurement of Translocation and Tyrosine Phosphorylation of Protein Kinase C Isozymes Derived from Suspended and Adherent Cells-
Suspended or adherent cells were sensitized and activated with antigen essentially as described above, with activation times of 1 min. In some experiments cells were treated with either 50 or 100 nM PMA for 2 min, followed by the addition, or not, of antigen for 1 min. In experiments using thapsigargin, cells were incubated with 500 nM thapsigargin for 5 min in order to achieve a maximal increase in intracellular calcium. All incubations were in 0.9 ml of saline solution. Following activation, 0.1 ml of a 10 ϫ sonication buffer (27) was added and samples were immediately sonicated at 4°C. Suspended cells were treated as described (27), while adherent cells were activated and sonicated directly in the 25-cm 2 flasks in which the cells were cultured. A 0.1-ml sample of the nuclei-free sonicate was mixed with an equal volume of 2 ϫ Tris-glycine SDS sample buffer for determination of the amount of protein kinase C isozymes present in the cells. The soluble and pelleted fractions were then recovered from the remaining nuclei-free sonicate as described (27). Pelleted fractions were resuspended to 0.9 ml, and 0.1-ml aliquots were removed from both soluble and pelleted fractions and mixed with the 2 ϫ Tris-glycine SDS sample buffer as above. Proteins derived from the soluble and particulate fractions were resolved by SDS-PAGE (8%) and transferred to nitrocellulose for analysis of the relative amounts of protein kinase C isozymes present in each fraction. Analysis was by Western blots using the polyclonal or monoclonal antibodies described previously (28), except for the antibody to protein kinase C-, which was obtained from Transduction Laboratories, Lexington, KY.
Immunoprecipitation of protein kinase C isozymes for analysis of tyrosine phosphorylation was done as described previously (27). Triton X-100 (final concentration 0.5%) was added to the remaining volume of the particulate fraction (see above), and the detergent lysates were used for immunoprecipitation of the individual protein kinase C isozymes. Antibodies for immunoprecipitations have been described (28). Proteins were resolved and transferred to nitrocellulose as above. The tyrosine FIG. 1. PMA shows both enhancing and inhibitory effects on antigen-stimulated secretion in RBL cell suspensions, but has much less effect on secretion in adherent cells. Antigen-stimulated ␤-hexosaminidase secretion was measured in suspended (A) and adherent (B) RBL-2H3 cells in the presence of the indicated concentrations of PMA. Spontaneous secretion was subtracted from stimulated secretion at each PMA concentration. Data are expressed as a fraction of control (antigen-stimulated secretion without PMA) and represent the mean and standard deviation of four experiments. Control secretion was 31.9 Ϯ 10.5% in suspended cells and 45.2 Ϯ 12.1% in adherent cells. Inset, a single experiment with suspended cells showing the mean and range of two replicates. The antigen concentration was 1 g/ml. Spontaneous secretion was 5.5 Ϯ 1.7% in suspended cells and 8.3 Ϯ 2.0% in adherent cells; it was unaffected by PMA.
phosphorylation of protein kinase C-␣, -␦, and -⑀ derived from suspended or adherent cells was analyzed by immunoblotting with a mouse monoclonal antibody to phosphotyrosine (4G10, Upstate Biotechnology, Inc., Lake Placid, NY). Tyrosine phosphorylation of the ␤ isozyme was not assessed due to the unavailability of an immunoprecipitating antibody. Detection was by enhanced chemiluminescence, and relative quantitation of immunoblots was performed by densitometry as described (29).

RESULTS
Antigen-stimulated Secretion-We have examined the effect of the protein kinase C activator PMA on adherent and suspended RBL cells to determine whether cell adherence can explain the conflicting reports in the literature on the effects of PMA on antigen-stimulated secretion (12)(13)(14)(15). In cell suspensions, concentrations of PMA higher than about 15 nM inhibited antigen-stimulated secretion, while lower concentrations of PMA potentiated secretion somewhat (Fig. 1A). The potentiation of secretion by low concentrations of PMA varied between experiments; Fig. 1A (inset) shows an experiment in which this potentiation was especially striking. In adherent RBL cells, however, PMA had only a small effect on secretion (Fig. 1B). Fig. 2 shows that the protein kinase C inhibitor GF 109203X (30) inhibits antigen-stimulated secretion in both suspended and adherent RBL cells, thus confirming the central role of protein kinase C in secretion from RBL cells. Although high concentrations of PMA can abolish antigen-stimulated secretion from cells in suspension (Fig. 1A), while PMA has little effect on adherent cells (Fig. 1B), the results in Fig. 2 clearly demonstrate that protein kinase C activity is necessary for secretion in both adherent and suspended cells. This result supports previous studies showing that secretion can be recon- indicator fura-2 were stimulated with 1 g/ml antigen (Ag) 1 min after treatment with 50 nM PMA as indicated. 5 M GF 109203X (GF) was added 2 min before antigen in the trace indicated, and was able to reverse the effect of PMA. The quench in fluorescence during the addition of GF 109203X was due to the strong absorbance of the compound. Data show fluorescence traces from one of three representative experiments. PMA had no effect on fluorescence measurements in unstimulated cells. stituted in protein kinase C-depleted cells by the protein kinase C isozymes ␤ and ␦ (31).
Antigen-stimulated Calcium Responses-We also examined the effects of PMA on antigen-induced changes in intracellular Ca 2ϩ in both adherent cells and in cell suspensions, using the fluorescent indicator fura-2. PMA completely abolished the antigen-induced increase of intracellular Ca 2ϩ in suspended cells (Fig. 3A), as has been shown previously (13,15). The IC 50 for this inhibition was approximately 15 nM (Fig. 4A), in agreement with results from other groups (12,15). However, in adherent cells, PMA had no significant effect on the antigenstimulated Ca 2ϩ response at any of the concentrations tested (Figs. 3B and 4B). In addition, GF 109203X was able to reverse the inhibition of the Ca 2ϩ response by PMA in suspended cells (Fig. 3A), supporting the idea that this effect of PMA is due to activation of protein kinase C.
Since PMA abolished not only the initial increase but also the prolonged elevation in intracellular Ca 2ϩ in suspended cells, it should inhibit both the release of calcium from intracellular stores and calcium influx across the plasma membrane. We therefore examined the effects of PMA on the calcium influx component of the calcium response using the manganese influx technique (32). In these experiments, decreases in fura-2 fluorescence are due to quenching of the dye by Mn 2ϩ , which has entered the cell via a calcium influx pathway (33). As expected, antigen-stimulated Mn 2ϩ influx in cell suspensions was abolished by 100 nM PMA (Fig. 5A). In adherent cells, however, PMA had no effect on manganese influx in response to antigen (Fig. 5B).
Responses to Thapsigargin-Thapsigargin and other inhibitors of the endoplasmic reticulum Ca 2ϩ -ATPase (34) deplete intracellular stores of calcium and activate Ca 2ϩ influx in RBL cells (16,(35)(36)(37), thus bypassing the IP 3 -dependent pathway activated by antigen. If protein kinase C activation by PMA directly inhibits the Ca 2ϩ influx pathway in suspended cells, then PMA should also prevent the thapsigargin-induced Ca 2ϩ influx. However, since the Ca 2ϩ response to antigen is completely abolished by PMA, a more likely possibility is that PMA is inhibiting the Ca 2ϩ responses at, or prior to, the release of  Fig. 3. The maximal change in fluorescence from the pre-stimulation baseline was expressed as a percent of total fluorescence, after correcting for non-fura-2 fluorescence and for leakage of fura-2 from the cells during the experiment. The percent maximal change in fluorescence was then plotted as a fraction of control (percent maximal change in fluorescence without PMA). Data represent the mean and standard deviation of four experiments. Control maximal fluorescence changes were 26.0 Ϯ 5.5% in suspended cells and 26.1 Ϯ 3.3% in adherent cells. The antigen concentration was 1 g/ml.

FIG. 5. PMA inhibits antigen-induced manganese influx in suspended RBL cells, but had little effect in adherent cells.
Suspended (A) and adherent (B) cells were treated with 100 nM PMA 1 min before the addition of 100 M MnCl 2 (Mn). The immediate drop in fluorescence is a result of manganese binding to extracellular fura-2. Two minutes later, the cells were stimulated with 1 g/ml antigen (Ag). Data show fura-2 fluorescence traces from one experiment representative of three. PMA had no effect on fluorescence measurements in unstimulated cells.
Ca 2ϩ from stores. If this is the case then PMA should have no effect on the thapsigargin-induced activation of the calcium influx pathway as monitored by manganese influx, and this is shown in Fig. 6. This is in agreement with recent results obtained by Ali et al. (37), showing that the thapsigargininduced increase in intracellular calcium is not inhibited by PMA. Since activation of protein kinase C by PMA is known to inhibit phospholipase C-␥, it seems likely that PMA is inhibiting the Ca 2ϩ response to antigen by preventing IP 3 production (13,18).
Treatment of RBL cell suspensions with thapsigargin did not induce secretion unless the cells were also treated with PMA (Fig. 7A). This is consistent with work using another endoplasmic reticulum Ca 2ϩ -ATPase inhibitor, cyclopiazonic acid (16). In contrast, thapsigargin alone was able to stimulate secretion in adherent cells, although co-treatment with 50 nM PMA enhanced thapsigargin-induced secretion (Fig. 7B). These data suggest that adherent cells have a constitutive protein kinase C activity that synergizes with thapsigargin to promote secretion.
Effects of Protein Kinase C Inhibitors-If the thapsigargininduced secretion in adherent cells is indeed dependent on constitutive activity of protein kinase C, then the protein kinase C inhibitor GF 109203X should inhibit secretion in response to thapsigargin. This is shown in Fig. 8. PMA-induced down-regulation of protein kinase C in adherent cells (31) also inhibited thapsigargin-induced secretion, with complete inhibition of secretion after 6 h of incubation in 100 nM PMA (Fig.  8). These data support the idea that thapsigargin-induced secretion in adherent RBL cells is dependent upon a constitutive activity of protein kinase C.
Protein kinase C contains two functional domains: a regulatory domain that interacts with the physiological activator diacylglycerol and with PMA, and a catalytic domain that binds ATP and contains the kinase activity. We have shown that adhesion of RBL cells results in a marked loss of sensitivity to PMA as well as an increased activity of protein kinase C, which suggests that the regulatory domain may have been altered in some way. Since calphostin C acts on the the regulatory domain, we predicted that it would be unable to inhibit protein kinase C in adherent cells. Indeed, calphostin C failed to inhibit either antigen-or thapsigargin-induced secretion in adherent cells at concentrations that completely inhibited antigen-induced secretion in cell suspensions (Fig. 9). Since GF 109203X acts on the catalytic domain of protein kinase C (30), this inhibitor should affect suspended and adherent cells similarly, as was shown in Fig. 2.
Translocation and Tyrosine Phosphorylation of Protein Kinase C Isozymes from Suspended and Adherent Cells-The difference in the protein kinase C response to PMA of suspended versus adherent cells might be mediated by differential membrane translocation of protein kinase C isozymes. We therefore examined the distribution of isozymes in membrane and cytosolic fractions from suspended and adherent cells. Table I shows the relative amounts of each protein kinase C isozyme present in the membrane fraction from resting and antigen-stimulated cells, in the presence or absence of PMA. Prior to stimulation, the membrane-associated protein kinase C-␣ was 3-fold higher in adherent cells, with 6.4% of this isozyme present in the membrane fraction of adherent cells as compared to 2.1% for suspended cells. A difference was also observed for protein kinase C-⑀, where adherent cells had less enzyme in the membrane fraction than suspended cells. The membrane-association of the ␤ and ␦ isozymes (5-6% and 25-30%, respectively) was the same in adherent and suspended cells, prior to stimulation.
Activation of suspended and adherent cells by antigen also revealed a difference in the ability of the calcium-dependent protein kinase C-␣ and -␤ isozymes to translocate (Table I).
There was a 5-fold increase in membrane-associated protein kinase C-␣ in both adherent and suspended cells in response to antigen, but again the extent of translocation was 3-fold higher in adherent cells. Antigen stimulation caused a 2-3-fold increase in membrane-associated protein kinase C-␤ in suspended cells, whereas a 6-fold increase was seen with adherent cells. In contrast, no difference between suspended and adherent cells was observed for translocation of protein kinase C-␦ and -⑀ in response to antigen. Treatment of cells for 3 min with 50 nM PMA (a concentration that effectively inhibited 75% of the secretory response of suspended cells) resulted in translocation to the membrane of all isozymes except . Although the extent of this translocation varied between isoforms (see Table I), no statistically significant differences were seen between adherent and suspended cells. However, the differential distribution of protein kinase C-␣ and -␤ in adherent and suspended cells that was seen in response to antigen appeared to be maintained in PMA-treated cells (Table I). No large differences were observed for protein kinase C-␦ and -⑀ in response to antigen, since both of these isozymes were already localized to the membrane by PMA treatment (Table I). An additional experiment using 100 nM PMA showed a similar trend, although the PMA alone induced a more substantial translocation of isozymes and so additional translocation in response to antigen was not as great (data not shown). Thapsigargin-induced elevation of intracellular calcium in adherent and suspended cells did not affect membrane association of any of the isozymes except for protein kinase C-⑀, which increased from 7.6 Ϯ 4.6% to 21.3 Ϯ 6.1% in adherent cells (n ϭ 3).
Tyrosine phosphorylation of the individual protein kinase C isozymes was assessed by immunoprecipitation of the individual isozymes and immunoblotting of the resolved proteins with antibody to phosphotyrosine. Only protein kinase C-␦, which was previously shown to be tyrosine-phosphorylated (27), was tyrosine-phosphorylated in response to antigen or PMA. In resting cells a trace amount of tyrosine phosphorylation of protein kinase C-␦ was also noted. However, in all cases the state of tyrosine phosphorylation of protein kinase C-␦ from adherent and suspended cells was similar (data not shown).

DISCUSSION
In the past, there have been discrepancies in the literature describing the effects of PMA on antigen-stimulated secretion in RBL cells. Pecht and colleagues (12,15) found that PMA potentiates secretion at low concentrations (Ͻ15 nM) and inhibits secretion at higher concentrations. However, Beaven's laboratory (13,14) has shown that PMA has no effect on anti- gen-induced secretion. One difference between these two sets of experiments is that Pecht's group worked with cell suspensions while Beaven's group worked with adherent cells. Our results clearly demonstrate that adherent RBL cells are markedly resistant to PMA. Only in suspended cells did PMA inhibit antigen-induced increases in intracellular [Ca 2ϩ ] i (Figs. 3 and 4) and calcium influx (Fig. 5), and have complex effects on secretion (Fig. 1). Thus, we propose that the earlier discrepancies were due to differences between adherent and suspended cells. In the past, results obtained using cells in suspension have often been compared with other data obtained using adherent cells. Our findings highlight the importance of making all measurements under the same experimental conditions.
The importance of the adhesion process in the modification of cellular activities such as differentiation and proliferation has been recognized in many cell types (38,39). It is thus not very surprising that other aspects of the cellular response should also be affected by cell adhesion. One of the events following RBL cell adhesion is tyrosine kinase activation (2,3,19). Although similar results have not yet been reported for serine/ threonine kinases, we suggest that protein kinase C itself might be activated, either directly or indirectly, during or after adhesion. Since thapsigargin only induced secretion from RBL cell suspensions if PMA was also present (Fig. 7A), we used thapsigargin to test whether protein kinase C is constitutively active in adherent cells. If the enzyme is activated with cell adhesion, thapsigargin should induce secretion in adherent cells without the need for PMA co-treatment and this was indeed the case (Fig. 7B). Similar results were obtained with the calcium ionophore A23187 (data not shown). This secretion was inhibited by the protein kinase C inhibitor GF 109203X and by down-regulation of protein kinase C (Fig. 8), suggesting that the thapsigargin-induced secretion in adherent cells is indeed dependent upon a constitutive protein kinase C activity.
We have attempted to identify how this increased level of protein kinase C activity might be achieved, and which isozymes are involved. An attractive hypothesis is that adhesion causes an alteration in protein kinase C that affects the function of the regulatory domain of the enzyme. Since the binding site for PMA is on the regulatory domain, this could also explain the loss of sensitivity of adherent cells to PMA as well as the apparent increase in protein kinase C activity in adherent cells. If the regulatory domain is altered, protein kinase C would still be sensitive to GF 109203X, because this inhibitor acts on the catalytic site of protein kinase C (30). However, the protein kinase C inhibitor calphostin C should not inhibit the altered protein kinase C, because it acts on the regulatory domain (40). Indeed, calphostin C was unable to inhibit thapsigargin-induced secretion in adherent cells at concentrations that inhibited antigen-stimulated secretion in cell suspensions (Fig. 9).
One mechanism by which this alteration might be accomplished is by the proteolytic cleavage of the regulatory domain of protein kinase C from the catalytic domain, leaving a constitutively active protein kinase M fragment (41,42). However, this would not explain the potentiation of thapsigargin-induced secretion by PMA that is still seen in adherent cells (Fig. 7B). Another possibility is that the function of the regulatory domain is altered by phosphorylation. This is supported by data suggesting that several of the protein kinase C isozymes can become phosphorylated (27,31) and that tyrosine phosphorylation of protein kinase C-␦ occurs on the regulatory domain (43). However, we failed to detect any differences between adherent and suspended cells in the tyrosine phosphorylation of any of the protein kinase C isozymes. One mechanism suggested by the ability of thapsigargin to stimulate secretion in adherent cells without PMA treatment is that protein kinase C is activated in adherent cells when intracellular Ca 2ϩ is increased, even without diacylglycerol stimulation. However, in response to stimulation with thapsigargin, we did not detect membrane translocation of any of the protein kinase C isozymes except for protein kinase C-⑀ in adherent cells. The ability of thapsigargin to induce membrane translocation of protein kinase C-⑀ has been described previously in GH 4 C 1 rat pituitary cells (44); since protein kinase C-⑀ is not calcium-dependent, this effect may be an indirect consequence of the thapsigargin-induced increase in intracellular Ca 2ϩ (44).
Another possibility is that cell adhesion may activate the kinase by inducing translocation of protein kinase C to the plasma membrane in a manner similar to activation by PMA or antigen. Our experiments suggest that protein kinase C-␣ and -⑀ are indeed differentially distributed in adherent versus suspended cells, with greater membrane translocation of protein kinase C-␣ in adherent cells, and of protein kinase C-⑀ in suspended cells (Table I). We also observed that with antigen stimulation the calcium-dependent protein kinase C-␣ and -␤ isozymes were translocated to the membrane to a greater extent in adherent than in suspended cells (Table I). Protein kinase C-␤ is able to reconstitute antigen-induced secretion in permeabilized cells (31), while protein kinase C-␣ and -⑀ have been shown to inhibit of phospholipase C-␥, thus preventing IP 3 production and the release of calcium from stores (18). a Values were derived from suspended (S) or adherent (A) cells. b The membrane-associated isozyme was expressed as a percentage of the sum of the isozyme in the membrane and cytosolic fractions, as determined by densitometry. Protein kinase Cdid not translocate under any conditions studied and was therefore not included in the Thus, both potentiating and inhibitory protein kinase C isozymes show differential distribution in adherent and suspended cells.
In conclusion, we have shown that following adhesion, RBL cells lose sensitivity to PMA and display a constitutive activity of protein kinase C, perhaps because the regulatory domain of protein kinase C has been altered in some way. Since mature mucosal mast cells reside in tissues, adherent cells should be more representative of mast cells in vivo. It is possible that the increase in protein kinase C activity represents a regulatory mechanism which allows mature, adherent mast cells to achieve greater sensitivity to intracellular Ca 2ϩ , thus leading to full physiological activation. Activation of protein kinase C when mast cells adhere may therefore be an important link between physiological stimulus and cell response in mast cells and perhaps in other cell types as well.