Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

It has been postulated that the stimulation of M₃-ACh receptor on the parietal cell membrane results in the activation of phosphatidyl inositol-specific PLC via a GTP binding protein, Gq (Wilkes et al., 1991; Hirschowitz et al., 1995). Histamine is the main stimulant for the acid secretory cascade, and it stimulates H₂ receptors to activate Gs-coupled adenylate cyclase, leading to the production of cAMP. Histamine also elicits an increase [Ca⁺⁺]i in the rabbit (Negulescu et al., 1989), but not the canine, parietal cell (DelValle et al., 1992a). The mechanism by which histamine produces the increase in [Ca⁺⁺]i, e.g., the possible involvement of PLC in the histaminergic stimulus, has not been fully clarified (DelValle et al., 1992b; Wang et al., 1996).

In recent reports, an aminosteroid derivative, U73122, has been frequently used as a PLC inhibitor. When U73122 is used together with its negative control, U73343, any response sensitive to the inhibitor can be considered to be mediated by the activation of membrane-bound PLC. For example, it has been reported that U73122 inhibits receptor-coupled PLC activity in polymorphonuclear neutrophils (Smith et al., 1990; Bleasdale et al., 1990), pituitary cells (Smallridge et al., 1992), neuroblastoma (Thompson et al., 1991), tracheal smooth muscle (Hansen et al., 1995), pancreatic acinar cells (Yule and Williams, 1992), parotid acinar cells (Jorgensen et al., 1995) and fibroblast cell line (Grierson and Meldolesi, 1995). In the present study, our experiments were designed to elucidate the role of PLC in the function of the parietal cell by using U73122 and its negative control, U73343. However, a number of unexpected effects emerged. For example, U73122 did not inhibit the carbachol-induced increase in [Ca⁺⁺]i but augmented the acid secretion induced by any agonist, and the negative control strongly inhibited acid secretion. We were then obliged to clarify their mechanisms of action.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials

U73122 and its negative control, U73343, were obtained from Biomol Research Laboratories, Inc. (Plymouth Meeting, PA) via Funakoshi (Tokyo, Japan). In accordance with the manufacturer's manual, these compounds were dissolved in chloroform and dispensed in aliquots, and the solvent was evaporated by a stream of nitrogen, in order to avoid possible inactivation during storage as a solution. Immediately before use, they were dissolved in dimethylsulfoxide. Fura-2/AM was from Molecular Probes via Wako Pure Chemical (Tokyo, Japan). ¹⁴C-aminopyrine and ³H-phosphatidyl inositol 4,5-bisphosphate were from New England Nuclear (Boston, MA). Other chemicals were all reagent grade and were obtained from Sigma Chemical (St. Louis, MO). Japanese White rabbits were obtained from Shiraishi Inc. (Tokyo, Japan).

Methods

Isolation of rabbit gastric glands and measurement of their functions. Isolated gastric glands were prepared from rabbits by the combination of high-pressure perfusion and collagenase digestion (Berglindh and Öbrink, 1976). Acid secretion of the glands was monitored by measuring the accumulation of a weak base, ¹⁴C-aminopyrine. The stimulation by histamine or dbcAMP was performed at 37°C for 30 min, that by carbachol for 15 min. Stimulation of acid secretion was expressed as the increment of the aminopyrine ratio above the resting value, and the effects of drugs on the agonists were expressed as percentage of the control values. Therefore, when a value lower than resting was obtained with the drug treatment, the result had a negative value.

Purification and enzymatic assay of H⁺,K⁺-ATPase. H⁺,K⁺-ATPase was purified from the microsomes of resting rabbit gastric mucosal homogenate by a sucrose density gradient as described (Hirst and Forte, 1985). Na⁺,K⁺-ATPase was purified from the microsomal fraction of rabbit kidney by a combination of sodium dodesylsulfate treatment and sucrose density gradient (Jorgensen, 1974).

Measurement of phosphodiesterase activity. Phosphodiesterase activity was assayed in a total volume of 0.5 ml containing 0.05 U/ml phosphodiesterase (derived from brain, Sigma), 0.5 U/ml 5-nucleotidase, 50 nM calmodulin and 2 mM cAMP at 30°C for 20 min. The value found when phosphodiesterase was omitted was set as a blank. In some experiments, assays were carried out using partially purified phosphodiesterase from gastric mucosa instead of the brain enzyme.

Measurement of agonist-induced changes in [Ca⁺⁺]i in platelets. According to Bleasdale et al. (1990), platelets were obtained from citrated blood of rabbits and suspended in Ca⁺⁺, Mg⁺⁺, BSA-free HBSS (pH 6.5) at 5 × 10⁸ cells/ml. Fura-2 AM (10 µM) was then added, and the platelets were incubated for 30 min at 37°C. After the addition of 1 µM PGE₂, platelets were collected by centrifugation, resuspended in Ca⁺⁺, Mg⁺⁺, BSA-free HBSS (pH 7.4) and transferred to a cuvette containing 1.3 mM CaCl₂ and 0.9 mM MgCl₂ kept at 37°C. Platelet suspensions were excited with dual wavelengths (340 and 380 nm), and the fluorescence emission ratio at 510 nm was recorded using an intracellular calcium analyzer (CAF-110, JASCO Corporation, Tokyo).

Isolation of IP₃-sensitive Ca⁺⁺ store from gastric mucosa and measurement of Ca⁺⁺ release. The procedure was based on the described method for cerebellum (Standerman et al., 1988) with modifications. Rabbit gastric mucosa was homogenized with 120 mM KCl, 10 mM NaCl, 1 mM KH₂PO₄, 0.2 mM MgSO₄, 20 mM HEPES-KOH (pH 7.4) 0.1 mM PMSF and 10 µM pepstatin, and the supernatant of the centrifugation at 20,000 × g for 15 min was incubated with 50 mg/ml of Chelex-100 on ice. After 10 min, the supernatant was transferred to a cuvette containing 2 mM Mg-ATP, 10 mM creatine phosphate, 5 U/ml creatine kinase, 1.2 µM fura-2, 10 µM omeprazole and 5 µg/ml oligomycin, and the Ca⁺⁺ concentration was monitored by CAF-110 as described for platelets. At the start of experiment, 10 µM of CaCl₂ was added, and the decrease in the extravesicular concentration of Ca⁺⁺, as evidenced by the uptake of Ca⁺⁺ by the vesicles, was monitored. When the ratio reached a constant value, test drugs were added to the cuvette.

Measurement of in vitro activity of platelet-soluble phosphoinositide-specific PLC. According to Bleasdale et al. (1990), the platelets were suspended in assay buffer containing 115 mM KCl, 10 mM HEPES/NaOH (pH 6.5), 5 mM KH₂PO₄, 2 mM EGTA, 0.91 mM MgSO₄, 0.1 mM dithiothreitol and 3 µM leupeptin, and were subjected to a freezing-thawing cycle and sonication. The platelet lysate was centrifuged at 105,000 × g for 2 hr, and the supernatant was retained for PLC assays. Assay mixture containing [³H] PIP₂ and the lysate (20 µg protein) was incubated at 37°C for 10 min, and the reaction was terminated with CHCl₃/methanol/HCl, vortexed and centrifuged. Then the radioactivity in the aqueous layer was measured by a liquid scintillation counter.

Measurement of membrane-bound PLC activity. Agonist-stimulated, membrane-bound PLC activity was measured according to the reported method (Hiramatsu et al., 1992). The submandibular glands were isolated from rabbits, minced and homogenized in 50 mM Tris (pH 7.4) by Polytron for 10 min twice. After the centrifugation at 3000 × g for 10 min, the filtrated supernatant was further centrifuged at 40,000 × g for 10 min. The pellet was suspended in 50 mM Tris and centrifuged again and then suspended in Tris. The PLC activity of this membrane fraction was assayed in a total volume of 0.1 ml containing 120 mM KCl, 20 mM NaCl, 1 mM MgSO₄, 10 mM LiCl, 20 mM HEPES/NaOH (pH 7.2), 2 mM ATP, 1 mM CaCl₂, 1 mM EGTA, 0.8 mM deoxycholate, 1 µM GTPS, [³H]PIP₂ and membranes (20 µg protein) at 30°C for 30 min. The subsequent procedures were the same as described above.

Statistical analysis. Parametric data were expressed as the mean ± S.E.M. Multiple comparisons were analyzed by ANOVA and Dunnett's post-hoc test using a computer program (Super ANOVA, ABACUS Concepts). The level of significance was uniformly set a P < .05, and no further calculation of P value was performed.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

[Ca⁺⁺]i increase in the parietal cell and the effects of U73122 and U73343. When carbachol concentration at higher than 1 µM was added to the glands loaded with fura-2, [Ca⁺⁺]i showed a biphasic increase: a single sharp peak followed by a sustained plateau. Histamine elicited similar pattern of [Ca⁺⁺]i increase, but the overall response tended to be slow. From the dose-response curves for carbachol and histamine, a maximal response was obtained with each agonist at a concentration of 10 µM. Figure 1 depicts the typical pattern of [Ca⁺⁺]i increases by 10 µM carbachol (panel A) and by histamine (panel B). Contrary to expectations, pretreatment with 10 µM U73122 failed to inhibit [Ca⁺⁺]i responses to either carbachol or histamine. When the peak value in the presence of 10 µM U73122 was expressed as a percentage of the control value the responses to 10 µM and 100 µM carbachol were 118.7 ± 14.7% and 112.7 ± 36.9%, respectively, and those to 10 µM and 100 µM histamine were 74.4 ± 25.3% and 109.2 ± 7.9%, respectively (mean ± S.E.M., n = 3-5); thus there was no significant inhibition. In our system, it was already revealed that 2 µM U73122 abolished the [Ca⁺⁺]i response to CCK8 of CHO cells transfected with human CCK_B receptor (Akagi et al., in press, 1997). Therefore, we conclude that PLC in the rabbit parietal cell is apparently resistant to U73122 if the enzyme is practically involved in the [Ca⁺⁺]i increase to carbachol. On the other hand, U73343, the negative control, did not affect the increase in [Ca⁺⁺]i at 10 µM, as expected from previous reports (Bleasdale et al., 1990).

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Effects of U73122 and its negative control, U73343, on [Ca++]i increase induced by carbachol (CCh) or by histamine (Hist) in the rabbit parietal cell. Fura-2-loaded gastric glands were preincubated with dimethylsulfoxide (vehicle; thick line), 10 µM U73122 (fine solid line) or 10 µM U73343 (dotted line) for 10 min and stimulated with 10 µM CCh (panel A) or Hist (panel B) at the point of 60 sec (arrow). Traces are representative of four similar experiments.

Effects of U73122 and U73343 on acid secretion. In the next experiments, we investigated the effects of these compounds on the acid secretion of isolated gastric glands. The glands were stimulated with 100 µM of histamine, carbachol or dbcAMP, and the effects of the inhibitors were expressed as percent of each control. As shown in figure 2, contrary to our expectations, U73122 did not inhibit, but dose-dependently augmented, the acid secretory responses to all agonists tested. The augmentation by the drug was most prominent in the secretion stimulated by dbcAMP; the value at 10 µM U73122 was about 2.4 times the control value. U73122 did not stimulate acid secretion by itself in this concentration range (data not shown). On the other hand, the putative negative control, U73343, dose-dependently inhibited acid secretion irrespective of the stimulant. This was especially evident in the presence of 10 µM U73343, where the value was negative; that is, the aminopyrine ratio was smaller than that of resting.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   U73122 dose-dependently potentiates, but U73343 dose-dependently inhibits, histamine (Hist)-, carbachol (CCh)-, and dbcAMP-stimulated aminopyrine accumulation in gastric glands. Glands were stimulated with histamine (100 µM, 30 min, squares), carbachol (100 µM, 15 min, diamonds) or dbcAMP (100 µM, 30 min, circles) in the presence of the indicated concentration of U73122 (open symbols) or U73343 (closed symbols). Stimulation of acid secretion was expressed as the increment of the aminopyrine (AP) ratio above the resting value, and the effects of drugs on the agonists were expressed as the percentage of the control values. Values are means ± S.E.M. from five separate gland preparations. All the points at 105 M concentration of the drugs are significantly different from control at P < .05.

Effects of U73122 and U73343 on H⁺,K⁺-ATPase activity. Because the increasing effect of U73122 and the inhibitory effect of U73343 were both independent of the stimulant, we postulated that their site of action was on intracellular events beyond receptor activation. Thus we examined their possible effects on the activity of the gastric proton pump, H⁺,K⁺-ATPase. As shown in figure 3, neither 1 nor 10 µM U73122 showed any effect on H⁺,K⁺-ATPase activity. However, U73122 potently inhibited K⁺-pNPPase in the same concentration range. This observation is incomprehensible because U73122, which augments stimulated acid secretion, inhibits the partial reaction of the proton pump without affecting the overall reaction of the enzyme. On the other hand, U73343, which potently inhibited acid secretion, showed no inhibitory effect on either K⁺-pNPPase or H⁺,K⁺-ATPase activity (fig. 3).

View larger version (28K): [in this window] [in a new window]   Fig. 3.   Effects of U72122 and U73343 on K+-pNPPase and K+-ATPase activities of H+,K+-ATPase purified from rabbit gastric microsomes. K+-pNPPase (left panel) and K+-ATPase (right panel) activities were calculated as the increased value with 20 mM KCl in the presence of 0.1 mM ouabain, and the effects of U72122 (open column) and U73343 (striped column) were expressed as % of vehicle control (dimethylsulfoxide). Values are means ± S.E.M. of three separate experiments. Inhibitions of K+-pNPPase activity by 1 and 10 µM U73122 are both statistically significant (P < .05).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Lineweaver-Burk plot of K+-pNPPase activity vs. K+ concentration. K+-pNPPase activity was measured in the absence () or presence of 0.1 µM () or 1 µM () U73122 under the various KCl concentrations. The experiment was done in duplicate.

Effects of U73122 and U73343 on the PLC activity of rabbit platelets. When it was reported that U73122 was ineffective on the PLC activity in pancreatic  cells (Alter et al., 1994), it was claimed by that the preparation of the drug was inappropriate (Hansen et al., 1995). Thus we checked for reproducibility of the compound in platelets where it was shown to be effective in the original report (Smith et al., 1990; Bleasdale et al., 1990). As depicted in figure 5A, the addition of U46619, a thromboxane A₂ derivative, to platelets loaded with fura-2 elicited an increase in [Ca⁺⁺]i. Pretreatment with 10 µM U73122 potently inhibited the increase, especially the initial phase, which was postulated to be due to the release of intracellular stores.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Effects of U73122 and U73343 on U46619-induced increases in [Ca++]i in platelets (panel A) and on the hydrolysis of PIP2 catalyzed by platelet soluble PLC (panel B). In panel A, platelets were loaded with fura-2 and incubated for 3 min at 37°C with vehicle (dimethylsulfoxide, thick line), with 10 µM U73122 (fine solid line) or with U73343 (dotted line), and the ratio of fluorescence was continuously monitored by an intracellular Ca++ analyzer. Three minutes after the start of measurement (arrow), the thromboxane mimetic U46619 (2 µM) was added. Traces represent three similar experiments. In panel B, platelet-soluble PLC was assayed in the presence of the indicated concentrations of U73122 () or U73343 () as described in "Methods." Values are means ± S.E.M. of four assays.

Effects of U73122 on the carbachol-activated PLC in the rabbit parotid gland. One preparation of membrane-bound PLC from rat parotid glands has been shown by Hiramatsu et al. (1992) to be activated by receptor stimulation with carbachol. We thus examined the PLC activity in a similar preparation. As shown in figure 6, PLC in the membrane fraction was activated by 10 µM to 1 mM carbachol, whereas the absolute values of basal activity and the extent of activation by carbachol varied with different preparations, making it difficult to perform any quantitative analysis. However, it was consistent that no activation by carbachol was observed in the presence of 10 µM U73122 (fig. 6).

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Effects of U73122 on carbachol (CCh)-induced activation of membrane-bound PLC of rabbit parotid gland. The membrane fraction was purified from rabbit submandibular glands, and PLC activity was measured in the presence of various concentrations of carbachol as described in "Methods." Traces are representative of 3 different experiments.

Effects of U73122 on the phosphodiesterase activity. It is well known that phosphodiesterase inhibitors augment acid secretion when stimulated by histamine, AMP or carbachol. Therefore, if U73122 were found to inhibit phosphodiesterase activity, then its activity to increase acid secretion would be understandable. In a brain phosphodiesterase assay, the activities in the presence of 1, 10, and 100 µM U73122 were 98.7 ± 1.3, 97.3 ± 1.5 and 97.9 ± 1.4% of control, respectively (mean ± S.E.M., n = 4). We also used partially purified phosphodiesterase from gastric mucosa to take into consideration tissue-specific subtypes and obtained essentially similar results.

[Ca⁺⁺]i increase elicited by U73122. We investigated the direct effects of U73122 on [Ca⁺⁺]i in the parietal cell. At 1 µM, U73122 failed to elicit an increase in [Ca⁺⁺]i, whereas Ca⁺⁺ transients were observed with a 10 µM application (fig. 7). As shown before, we confirmed that U73122 did not affect the [Ca⁺⁺]i response to carbachol. On the other hand, U73343, the negative control, did not show any effect on [Ca⁺⁺]i up to 10 µM (data not shown).

View larger version (15K): [in this window] [in a new window]   Fig. 7.   U73122 by itself elicits an increase in [Ca++]i in rabbit parietal cells. Fura-2-loaded gastric glands were treated with 10 µM U73122 (solid line, at the first arrow), followed by 100 µM carbachol (CCh; second arrow). The [Ca++]i response to 100 µM carbachol alone (dotted line) is superimposed as a comparison. Traces are representative of three separate experiments.

View larger version (19K): [in this window] [in a new window]   Fig. 8.   Ca++ release from Ca++ store membranes from rabbit gastric mucosa. The fraction containing Ca++ store membranes was brought up to 2 mM Mg-ATP, 10 mM creatine phosphate, 5 U/ml creatine kinase, 1.2 µM fura-2, 10 µM omeprazole and 5 µg/ml oligomycin, and Ca++ concentration was monitored by CAF-110 as described in "Methods." Before the starting of the experiment, 10 µM of CaCl2 was added, and the decrease in extravesicular concentration of Ca++ resulting from the uptake of Ca++ by the vesicles was monitored until the ratio reached a constant value. In panel A, 0.8 µM IP3 was repeatedly added to the cuvette, which resulted in relatively constant releases of Ca++ and sensitivity to 200 µg/ml heparin. In panel B, U73122 up to 10 µM failed to induce the release of Ca++, whereas a small and slow release of Ca++ was induced by 100 µM. Ca++ release by IP3 was not affected by U73122. Data are representative of at least three separate measurements with essentially same results.

Acid secretion in the digitonin-treated gastric glands. In order to clarify the inhibitory mechanism of U73343 on acid secretion, we investigated its effects on the ¹⁴C-aminopyrine accumulation of digitonin-treated gastric glands. As shown in figure 9, the aminopyrine ratio of permeabilized glands was increased by the addition of 1 mM ATP, and a further increase was observed with the inclusion of the potassium ionophore valinomycin (10 µg/ml). When 0.1 to 10 µM U73343 was added in the presence of ATP and valinomycin, a dose-dependent inhibition of aminopyrine accumulation was observed (fig. 9). The inhibition curve was similar to that of intact glands stimulated by the secretagogues shown in figure 2.

View larger version (19K): [in this window] [in a new window]   Fig. 9.   U73343 dose-dependently inhibits ATP-dependent aminopyrine accumulation in digitonin-permeabilized gastric glands. Permeabilized glands were suspended in a buffer containing 150 mM KCl, 10 mM HEPES (pH 7.4), 1 mM MgSO4 and 5 µg/ml oligomycin. 14C-aminopyrine accumulation was measured in the absence of 1 mM ATP (open column), in the presence of 1 mM ATP alone (hatched column) and in the presence of 1 mM ATP and 10 µg/ml valinomycin (filled column). The effects of U73122, 0.1 to 10 µM, were estimated in the latter preparation. Values are mean ± S.E.M. of four separate gland preparations.

Effects of U73343 on the proton transport by isolated gastric vesicles. On the basis of the experiments showing that U73343 inhibited ATP-dependent acid secretion in digitonin-permeabilized glands, we considered the most feasible site of action of U73343 to be the proton pump. However, we had already found that U73343 itself was not a proton pump inhibitor (fig. 3). Therefore, the possibility remained that there was a direct action on the ionic movement. In order to assess its action on proton transport, we measured H⁺,K⁺-ATPase-dependent proton transport by purified gastric vesicles using the acridine orange quenching technique. As shown in figure 10A, a rapid quenching of fluorescence was observed after the addition of the membranes followed by 20 µg/ml valinomycin in the presence of ATP and KCl, an H⁺,K⁺-ATPase-dependent proton transport. When maximal quenching was attained, 1, 3 and 10 µM U73343 were sequentially added. A recovery of the fluorescence, or a dissipation of the proton gradient, was observed at 3 µM or higher. The recovery rate was slower than that by nigericin, a cation exchange ionophore, added at the end of the experiment but was much faster than that by the inhibition of H⁺,K⁺-ATPase by the proton pump inhibitor omeprazole or by the depletion of ATP with glucose plus hexokinase (data not shown). This suggests that U73343 was working as a protonophore. As shown in figure 10B, U73122 did not have such an effect.

View larger version (20K): [in this window] [in a new window]   Fig. 10.   U73343, but not U73122, cancels the proton gradient in the gastric vesicles formed by H+,K+-ATPase. Purified gastric microsomal membranes (50 µg/ml) were added (memb.) to the cuvette containing 125 mM KCl, 50 mM sucrose, 2.5 mM N-Tris(hydroxymethyl) methyl-2-aminoethane sulfonic acid (pH 6.8), 0.4 mM MgATP, 25 µM EDTA and 1.5 µM acridine orange, and the fluorescence changes (excitation, 493 nm; emission, 540 ± 6 nm band-pass filter) were continuously monitored with a spectrophotofluorometer. Subsequent addition of 20 µM valinomycin (val.) caused a rapid fluorescent quenching, which indicates that the membranes were producing a H+,K+-ATPase-dependent proton gradient. Addition of U73343, 1 to 10 µM, gradually recovered the fluorescence, or canceled the formed proton gradient in the left panel, whereas U73122, 3 and 10 µM, did not show such an effect. Note that 10 µM nigericin (nig.), a cation-exchanging ionophore, produced a rapid recovery of the fluorescence at the end of experiments. Traces are representative of at least three separate recordings with essentially identical results.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The PLC inhibitor most frequently used in recent research is U73122. Although its precise mechanism of action has not yet been elucidated, there have been many reports in which receptor-coupled PLC was inhibited by this compound in various cell types, e.g., polymorphonuclear neutrophils (Smith et al., 1990; Bleasdale et al., 1990), platelets (Bleasdale et al., 1990), pituitary cells (Smallridge et al., 1992), neuroblastoma (Thompson et al., 1991), tracheal smooth muscle (Hansen et al., 1995), pancreatic acinar cells (Yule and Williams, 1992) and fibroblast cell line (Grierson and Meldolesi, 1995). We also observed that U73122 abolished the [Ca⁺⁺]i response to CCK8 of CHO cells transfected with human CCK_B receptor (Akagi et al., in press, 1997). We thus applied this PLC inhibitor to isolated rabbit gastric glands. No inhibition was observed in the [Ca⁺⁺] transients elicited by carbachol or histamine, contrary to the consensus that the Ca⁺⁺ release from intracellular stores, at least by the activation of M₃ receptor in the parietal cell, is mediated by the activation of PLC. Moreover, 1 to 10 µM U73122 dose-dependently augmented the acid secretion of isolated rabbit gastric glands stimulated by the agonists tested. These results suggested that U73122 was not a PLC inhibitor, at least in the rabbit parietal cell.

Before drawing this conclusion, we had to examine a technical point. One report showed that U73122 was not acting as a PLC inhibitor in rat Langerhans islets or in a  cell line (Alter et al., 1994). It has been suggested that U73122 lost its activity in this study because of the inappropriate preparation of the drug (Hansen et al., 1995). Although we were very careful about this in the present study, we could not ignore this possibility, so we examined the effects of U73122 on the functional activity of rabbit platelets. Because this drug was originally reported to be specific and effective in human polymorphonuclear neutrophils and platelets (Smith et al., 1990; Bleasdale et al., 1990), these confirmatory experiments using rabbit platelets also worked to exclude the possibility of species differences. We found that 10 µM U73122 inhibited the U46619-induced [Ca⁺⁺]i increase in platelets. The activity of soluble PLC from platelets was resistant to U73122 up to 50 µM and was completely inhibited by 100 µM of the compound. There was an apparent discrepancy between the inhibitory activity in vivo and in vitro; 10 µM U73122, which was enough to inhibit platelet Ca⁺⁺ transient, was ineffective on the PLC assay using the soluble enzyme. This phenomenon was also observed in the original report (Bleasdale et al., 1990). The authors reported that the IC₅₀ value of U73122 for platelet soluble PLC with PIP₂ as substrate was 40 µM, whereas that for Ca⁺⁺ transient was 1 µM. The IC₅₀ value of another PLC inhibitor, neomycin, was 4 µM for the former and 1 mM for the latter. Thus U73122 is specific for membrane-bound, receptor-coupled PLC, but neomycin is not. It should be difficult to assess the effect of an inhibitor on the membrane-bound PLC mediating Ca⁺⁺-transient by its effect on the activity of soluble PLC. There has been no report that U73122 has its effect on PLC; rather, some have suggested that it acts on the interaction site between G protein and PLC (Thompson et al., 1991). Therefore, it should be necessary to assess the effects on the membrane-bound, receptor-mediated type of PLC.

The measurement of membrane-bound enzymes, especially their activation by agonists, is extremely difficult, but one available preparation has been reported by Hiramatsu et al. (1992), who successfully activated membrane-bound PLC obtained from rat parotid glands with carbachol. We first examined the activity in the membrane fraction obtained from rabbit parotid glands and found receptor-mediated activation of PLC by carbachol from 1 µM to 1 mM, although the absolute activity and the extent of activation varied among preparations. We found that 10 µM U73122 consistently abolished the activation by carbachol, which suggests that U73122 is an effective inhibitor on the membrane-bound PLC in the parotid glands. These results are consistent with the observation that the drug inhibited the Ca⁺⁺ transient in this tissue (Jorgensen et al., 1995).

Using the same isolation technique, we tried to measure membrane-bound PLC activity in the rabbit gastric mucosa, but we were unsuccessful in observing any consistent activation by carbachol in vitro. Therefore, the reason for the resistance of the putative gastric PLC to U73122 remains uncertain. The following possibilities exist: 1) There are different subtypes of PLC in different tissues or cells with respect to their sensitivity to U73122. 2) U73122 is metabolized or inactivated within the parietal cell. 3) The increase in [Ca⁺⁺]i by activation of the M₃ receptor is not mediated by PLC. In order to exclude the third possibility, which is rather absurd, we applied neomycin, another PLC inhibitor, on the Ca⁺⁺ transient in the parietal cell elicited by histamine or carbachol. We observed no effect of neomycin up to 1 mM but achieved a complete inhibition of the responses to both histamine and carbachol at 10 mM. This suggests that both histamine and carbachol involve a PLC-mediated [Ca⁺⁺]i increase. However, it was considered a nonspecific effect, because 10 mM neomycin abolished acid secretion measured by ¹⁴C-aminopyrine accumulation in the isolated glands stimulated by any agonist, including dbcAMP (unpublished observations). Thus the pharmacological analysis of acid secretion by using the so-called PLC inhibitors is practically impossible at present.

It was an unexpected observation that U73122 augmented the acid secretory responses. Because the effect was most prominent in the acid secretion stimulated by dbcAMP and was independent of the secretagogue, it was proposed that the site of action was on some step beyond receptor activation and production of cAMP. U73122 does not seem to augment acid secretion via the elevation of intracellular cAMP, because it lacks an inhibitory effect on phosphodiesterase. This was supported by the original report (Bleasdale et al., 1990) that U73122 never increased either basal or PGI₂-stimulated cAMP contents in platelets. U73122 does not seem to act on the final step of acid secretion either, because it did not augment H⁺,K⁺-ATPase activity in vitro. The most reasonable possibility comes from its action to increase [Ca⁺⁺]i as shown in figure 7. The fact that the augmentation by U73122 was most prominent on the secretion stimulated by dbcAMP is understandable, considering the potentiating interaction between cAMP and [Ca⁺⁺]i increase within the parietal cell (Negulescu et al., 1989). Recently, it was reported that U73122 could increase [Ca⁺⁺]i in rabbit pancreatic acinar cells (Jin et al., 1994) and cultured neuronal cells (Willeums et al., 1994). In the present experiment using isolated Ca⁺⁺ stores, no direct release of Ca⁺⁺ was observed with U73122 up to 10 µM. This suggests that U73122 itself is not an intracellular Ca⁺⁺ releaser but that it indirectly induces Ca⁺⁺ release, e.g., by its partial agonistic effect on PLC. However, we cannot entirely rule out the possibility of a direct effect on the store, considering the weak Ca⁺⁺-releasing effect of 100 µM U73122, because the responsiveness or the sensitivity of the isolated Ca⁺⁺ stores might be reduced during the isolation procedure.

The potentiating effect of U73122 on stimulated aminopyrine accumulation is qualitatively similar to that observed by Tsunoda et al. (1993), who noted the potentiation of histamine-stimulated aminopyrine accumulation by a tyrosine kinase inhibitor, genistein. In contrast to the present results, genistein failed to elicit any potentiation of carbachol- or dbcAMP-stimulated response. The mechanism of potentiation by genistein appears to be different from that by U73122, because the action of genistein was not related to [Ca⁺⁺]i mobilization (Tsunoda et al., 1993). The present results appear to be more similar to those observed by Levine et al. (1990, 1991), who noted the potentiation of histamine- or dbcAMP-stimulated aminopyrine accumulation by nonsteroidal anti-inflammatory drugs and postulated its mechanism to be the entry of extracellular Ca⁺⁺.

Another riddle of U73122 was its effects on H⁺,K⁺-ATPase reactions. U73122 strongly inhibited K⁺-pNPPase activity, which has been considered to be a partial reaction of H⁺,K⁺-ATPase, while having no effect on the overall H⁺,K⁺-ATPase activity. The lack of an effect on H⁺,K⁺-ATPase activity was not surprising, because U73122 did not inhibit but rather augmented acid secretion; however, the apparent dissociation of K⁺-pNPPase from H⁺,K⁺-ATPase activity was controversial. Because the compound has steroidal structure, similar to cardiac glycosides, we became interested in the phenomenon in connection with the inhibition of Na⁺,K⁺-ATPase by cardiac glycosides. Like its effect on H⁺,K⁺-ATPase, U73122 was found to be ineffective on the overall Na⁺,K⁺-ATPase activity, but K⁺-pNPPase activity of Na⁺,K⁺-ATPase was inhibited by the compound, although the inhibition was rather weak. The mechanism of the apparent dissociation of the activity is unclear at present, but a similar example was reported by others (Ray and Nandi, 1986), who found that spermine inhibited K⁺-pNPPase competitively with K⁺ but did not affect H⁺,K⁺-ATPase activity. The effect of U73122 on K⁺-pNPPase activity has not clarified any physiological function, but it might be a useful tool to analyze the reaction mechanisms of H⁺,K⁺-ATPase and Na⁺,K⁺-ATPase.

U73343, the negative control of U73122, dose-dependently inhibited acid secretion stimulated by all agonists tested without affecting the [Ca⁺⁺]i increase induced by carbachol or histamine. The effect of 10 µM U73343 was almost equivalent to that of the same concentration of a proton pump inhibitor, omeprazole, such that the aminopyrine ratio decreased to a value lower than resting (Hirschowitz et al., 1995). The aminopyrine accumulation of digitonin-permeabilized glands stimulated by KCl, ATP and valinomycin was also inhibited by 1 to 10 µM U73343 proved its inhibitory effect to be independent of the second messenger. Although this mode of action was similar to that of omeprazole (Hersey et al., 1989), the compound itself did not show any inhibitory activity on the hydrolysis of ATP or pNPP, which suggests that it is not a proton pump inhibitor. On the other hand, U73343 at higher than 1 µM dissipated the proton gradient formed by H⁺,K⁺-ATPase in gastric vesicles. Thus we conclude that U73343 inhibited acid secretion by working as a protonophore. This is somewhat surprising, because U73122 lacks protonophore activity, and the structural difference between U73343 and U73122 is only one double bond in the side chain. A similar effect was reported by Mamiya et al. (1993), where KN-93, an inhibitor of calmodulin-dependent protein kinase II, inhibited aminopyrine accumulation by its protonophore effect, whereas a similar inhibitor, KN-62, did not show such an effect. If the protonophore acts on mitochondria, it might uncouple them to induce various physiological responses in any cell type. Therefore, it is possible that this putative negative-control compound produces unexpected responses in variety of tissues other than gastric glands.

In summary, U73122 did not alter the agonist-induced increase in [Ca⁺⁺]i, but it did by itself induce an increase in [Ca⁺⁺]i in rabbit parietal cells. The drug was also effective in potentiating acid secretion produced by several agonists, especially activators of the cAMP-dependent protein kinase pathway, probably via a potentiating effect through increased [Ca⁺⁺]i. U73122 did not inhibit H⁺,K⁺-ATPase activity or ATP-mediated proton uptake, but it inhibited K⁺-pNPPase activity, a putative partial reaction of the H⁺,K⁺-ATPase. U73343, the negative-control compound for U37122, inhibited acid secretion stimulated by any agonist, apparently working as a protonophore. Care should be taken when interpreting experiments using these drugs in other tissues. It is obviously necessary to develop more specific and convenient pharmacological tools for analyzing the role of the PLC pathway in the intracellular events of acid secretion.