P2Y2 Purinergic and M3 Muscarinic Acetylcholine Receptors Activate Different Phospholipase C-β Isoforms That Are Uniquely Susceptible to Protein Kinase C-dependent Phosphorylation and Inactivation*

Activation of phospholipase C-β (PLC-β) by G protein-coupled receptors typically results in rapid but transient second messenger generation. Although PLC-β deactivation may contribute to the transient nature of this response, the mechanisms governing PLC-β deactivation are poorly characterized. We investigated the involvement of protein kinase C (PKC) in the termination of PLC-β activation induced by endogenous P2Y2 purinergic receptors and transfected M3 muscarinic acetylcholine receptors (mAChR) in Chinese hamster ovary cells. Activation of P2Y2 receptors causes Gαq/11 to associate with PLC-β3, whereas M3mAChR activation causes Gαq/11 to associate with both PLC-β1 and PLC-β3 in these cells. Phosphorylation of PLC-β3, but not PLC-β1, is induced by activating either P2Y2receptors or M3 mAChR. We demonstrate that PKC rather than protein kinase A mediates the G protein-coupled receptor-induced phosphorylation of PLC-β3. The PKC-mediated phosphorylation of PLC-β3 diminishes the interaction of Gαq/11 with PLC-β3, thereby contributing to the termination PLC-β3 activity. These findings indicate that the distinct temporal profiles of PLC activation by P2Y2 receptors and mAChR may arise from the differential activation of PLC-β1 and PLC-β3 by the receptors, coupled with a selective PKC-mediated negative feedback mechanism that targets PLC-β3 but not PLC-β1.

tion make it difficult to determine unambiguously where PKC acts to limit PLC-␤ activity.
The purpose of this study was to determine the role of PKC in the poorly characterized termination of GPCR-stimulated PLC activity. We investigated PLC-␤ activity regulated by endogenous P2Y 2 purinergic receptors (25) and transfected M 3 muscarinic acetylcholine receptors (mAChR) (11) expressed in Chinese hamster ovary (CHO) cells. It was previously reported that P2Y 2 receptors, like most GPCRs, transiently activate PLC-␤ (25), whereas M 3 mAChR activate PLC-␤ for periods longer than 10 min (6,9,11). We found that P2Y 2 receptors functionally couple only to PLC-␤3, whereas M 3 mAChR functionally couple to both PLC-␤1 and PLC-␤3 in CHO cells. Activating P2Y 2 receptors or M 3 mAChR induces the phosphorylation of PLC-␤3, but not PLC-␤1, in these cells. We demonstrate that PKC, rather than PKA, mediates the GPCRinduced phosphorylation of PLC-␤3. Thus, PLC-␤3 phosphorylation does not result from cross-regulation by adenylyl cyclase acting on PLC signaling. Our results indicate that PKC participates in a negative feedback loop to inhibit the interaction of G␣ q/11 with PLC-␤3, limiting the duration of PLC-␤3 activity. This PKC-mediated negative regulation of PLC-␤3, coupled with the differential activation of PLC-␤1 and -␤3 by M 3 mAChR and P2Y 2 receptors, offers a mechanism for the distinct temporal profiles of PLC activation by these two receptors.
Co-immunoprecipitation of PLC-␤ with G␣ q/11 -Co-immunoprecipitation of PLC-␤1 or PLC-␤3 with G␣ q has been described previously (26). CHO cells were incubated for 2 h in serum-free medium and exposed to agonists for 1 min in the presence of 1 mM dithiobis[succinimidyl propionate] to induce protein cross-linking and subsequently lysed in immunoprecipitation buffer described above. The lysate was centrifuged at 13,000 g for 10 min, and the resulting supernatant was rotated (90 min, 4°C) with G␣ q/ll antibody and protein A-agarose beads (Life Technologies, Inc.). After washing three times in lysis buffer, the immunoprecipitates were eluted with sample buffer (30 min, 4°C), subjected to SDS-PAGE, and electrophoretically transferred to PVDF membranes. The PVDF membranes were blocked and probed by ECL-Western blotting as described below using antibodies to G␣ q/ll , PLC-␤1, or PLC-␤3.
Binding of [ 35 S]GTP␥S to G␣ q/11 -Binding of [ 35 S]GTP␥S to G␣ q/11 was determined by modification of a method described previously (26). CHO cells were washed in phosphate-buffered saline, suspended in reaction buffer (50 mM HEPES, 100 mM NaCl, 6 mM MgCl 2 , 2 mM EDTA, 10 M GDP, 150 nM GTP␥S, pH 7.4), and subjected to a Ϫ70°C freeze/thaw cycle to disrupt cell membranes. The freeze/thawed cells were incubated for 15 min (37°C), with 30 nM [ 35 S]GTP␥S in the absence or presence of agonists for 10 min (37°C). The samples were solubilized by rocking in lysis buffer (50 mM HEPES, 150 mM NaCl, 20 mM MgCl 2 , 100 M GDP, 100 M GTP, 0.5% Nonidet P-40, 1 mM phenylmethylsulfonyl fluoride, 200 g/ml leupeptin, pH 7.4). The samples were centrifuged (13,000 ϫ g, 10 min, 4°C), and the resulting supernatants were incubated (1.5 h, 4°C) with G␣ q/ll antibody and protein A-agarose. The immunoprecipitates bound to protein A-agarose were washed two times in lysis buffer, resuspended in distilled water, and transferred to scintillation vials containing Ultima-Gold scintillation fluid (Packard Bioscience). The amounts of [ 35 S]GTP␥S bound to the immunoprecipitated G␣ q/ll were determined by liquid scintillation counting using an LS-6000 ␤-counter.
Measurement of G␣ q/11 -stimulated PLC Activity in Membranes-CHO cells were labeled for 24 h in Ham's F-12 medium containing 5% fetal bovine serum and 5 Ci/ml [ 3 H]myo-inositol. The cells were washed in phosphate-buffered saline, suspended in reaction buffer (described above), and subjected to a Ϫ70°C freeze/thaw cycle to disrupt cell membranes. Membranes were challenged with GTP␥S for 10 min, 37°C, after which the supernatant was removed for determination of released inositol phosphates by liquid scintillation counting as described previously (9,27).
Measurement of PLA 2 Activity-Arachidonic acid release was measured as an indictor of phospholipase A 2 activity as described previously (28). CHO cells were plated in 24-well culture plates at a density of 2.5 ϫ 10 5 cells/ml in Ham's F-12 medium containing 5% fetal bovine serum. After incubation for 48 h, the medium was supplemented with 0.2 mCi/ml [ 3 H]arachidonic acid, and the cells were incubated overnight. The cells were rinsed three times in Ham's F-12 medium, followed by a 1-h incubation in Ham's F-12 medium containing 5% fetal bovine serum. Cells were exposed to carbachol for 15 min, and an aliquot of the culture supernatant was subjected to liquid scintillation counting to determine [ 3 H]arachidonic acid content.

RESULTS
To determine which PLC isoforms are expressed by CHO-m3 cells, PLC-␤1 and -␤3 were immunoprecipitated from 35 S-labeled CHO cells, submitted to SDS-PAGE, and visualized by autoradiography (Fig. 1A). By this method we found that significant levels of PLC-␤1 and -␤3 are expressed by these cells. In contrast, neither PLC-␤2 nor -␤4 could be detected in immunoprecipitates or immunoblots of the CHO-m3 cells (Fig. 1B).
Confusion exists regarding the kinases mediating the GPCRstimulated phosphorylation of PLC-␤. Several studies have  1 and 3). The immunoprecipitates were subjected to SDS-PAGE followed by autoradiography. The efficiency of immunoprecipitation was determined by comparing the immunoreactivity of 5% of lysis material with the total recovered in immunoprecipitation (data not shown). B, lysates prepared from CHO cells (lanes 1 and 3), U937 cells (lane 2), or mouse brain tissues (lane 4) were probed by ECL-Western blotting using antibodies to PLC-␤2 (lanes 1 and 2) or PLC-␤4 (lanes 3 and 4). Results are representative of three independent experiments that produced similar results.
indicated roles for both PKC (17,18) and PKA (19). It is more likely that PKC, rather than PKA, acts in a classical negative feedback loop to mediate rapid PLC phosphorylation and deactivation. The involvement of PKA may indicate cross-regulation, which is unlikely to be a mechanism for rapid PLC phosphorylation and deactivation, because G␣ q/11 -coupled receptors do not generally activate adenylyl cyclase. The ability of PKA to mediate agonist-induced phosphorylation of PLC-␤3 was tested using the highly specific PKA antagonist, Rp-CAMPS. Concentrations of Rp-CAMPS well above those known to completely block PKA activation (28,29) had no effect on the carbacholinduced phosphorylation of PLC-␤3 at (Fig. 4). This finding indicates that PKA does not participate in PLC-␤3 phosphorylation induced by the M 3 mAChR. The inability of PKA to mediate carbachol-induced PLC-␤3 phosphorylation is not due to an inability of PKA to phosphorylate PLC-␤3, because activation of PKA with forskolin induces PLC-␤3 phosphorylation (Fig. 4).
In contrast to the effects of inactivating PKA, inactivation of PKC with the specific antagonist BIM significantly inhibited the carbachol-or ATP-induced phosphorylation of PLC-␤3 (Fig.  4A). This finding indicates that PKC participates in the agonist-induced phosphorylation of PLC-␤3. Further evidence that PKC mediates PLC-␤3 phosphorylation is provided by our finding that PLC-␤3 is phosphorylated when PKC is directly acti-vated by PMA (Figs. 3 and 4). PMA typically increased PLC-␤3 phosphorylation 3-4-fold above basal levels. Interestingly, activation of PKC does not induce the phosphorylation of PLC-␤1 (Fig. 4B). This finding suggests that the lack of agonist-induced phosphorylation of PLC-␤1 is due to an inability of PLC-␤1 to serve as a substrate for PKC-mediated phosphorylation in these cells.
The mechanism by which PKC inhibits PLC-␤ is poorly characterized. We found that treatment with PMA diminishes GTP␥S-stimulated inositol polyphosphate generation in the CHO-m3 cells (Fig. 5). This result indicates that PKC stimulation diminishes the ability of GTP␥S to activate PLC-␤. This event may occur because PKC stimulation diminishes the ability of activated G␣ q/11 proteins to associate with PLC-␤. To investigate this possibility, we determined whether PKC stimulation alters the co-precipitation of PLC-␤ with G␣ q/11 in agonist-treated cells (Fig. 6). Carbachol induces G␣ q/11 to associate with PLC-␤1 (Fig. 6A, lane 3) and PLC-␤3 (Fig. 6B, lane 3). Treatment with PMA completely abolishes the carbachol-induced association of G␣ q/11 with PLC-␤3 (Fig. 6B, lane 6). This result suggests that the PKC-mediated phosphorylation of PLC-␤3 diminishes the interaction of G␣ q/11 with the phospholipase. This response is not restricted to signaling by the muscarinic receptor, because treatment with PMA also diminishes the ATP-induced association of G␣ q/11 with PLC-␤3 (Fig. 6B,  lane 5). Treatment with PMA does not alter the agonist-induced association of G␣ q/11 with PLC-␤1 (Fig. 6A, lane 6), consistent with our finding that PMA has no effect on PLC-␤1 phosphorylation. To determine whether PKC activation alters other aspects of receptor signaling, the effects of PMA on the agonist-induced activation of G␣ q/11 were investigated. Activation of P2Y 2 receptors stimulates an approximately 500% increase in G␣ q/11 activation, as indicated by the binding of [ 35 S]GTP␥S to G␣ q/11 (Fig. 7A). This response is unaffected by pretreatment with PMA (Fig. 7A). Similarly, the M 3 mAChR-stimulated increase in G␣ q/11 activation is unaffected by PMA pretreatment (Fig.  7A). These findings indicate that the GPCR-dependent activation of G␣ q/11 is intact in PMA-treated cells. Consistent with this result, we found that PMA treatment actually enhances the ability of the muscarinic receptor to activate phospholipase A 2 , as indicated by arachidonic acid release (Fig. 7B).
The preceding results indicate that the GPCR-mediated activation of PKC stimulates PLC-␤3 phosphorylation, resulting in reduced PLC-␤3 activity because of diminished interactions of the phosphorylated PLC-␤3 with G␣ q/11 . This model predicts that the activation of PLC-␤3 by GPCR should be enhanced when PKC activity is inhibited and diminished when PKC activity is stimulated. To test this prediction, we measured the effects of pharmacologically altering PKC activity on GPCRmediated Ca 2ϩ mobilization, which is an indicator of PLC-␤ activity. We found that inactivation of PKC with BIM enhances Ca 2ϩ mobilization induced by ATP (Fig. 8A) and carbachol (Fig.  8B). Conversely, activation of PKC with PMA diminishes Ca 2ϩ mobilization induced by these agonists (Fig. 8). These findings support the model that the PKC-mediated phosphorylation of PLC-␤3 diminishes the activity of the phospholipase. DISCUSSION The purpose of this investigation was to characterize the involvement of PLC-␤ phosphorylation in the termination of GPCR-dependent PLC activation. Activation of PLC-␤ by P2Y 2 receptors and M 3 mAChR was investigated because these GPCRs are known to stimulate transient and sustained PLC-␤ activity, respectively. Our results indicate that activation of these receptors induces the PKC-mediated phosphorylation of PLC-␤3. This phosphorylation event may deactivate PLC-␤3 by inhibiting the association of G␣ q/11 with the phospholipase. In contrast, PLC-␤1 is not phosphorylated by PKC and thus is not deactivated by this mechanism. Our results suggest that the activation of PLC-␤1 by M 3 mAChR, but not by P2Y 2 receptors, may contribute to the sustained PLC-␤ activity induced by M 3 mAChR activation. This model is depicted in Fig. 9.
The activation of PLC-␤ by GPCRs typically surges and declines rapidly (6 -8). The GPCR-induced activation of PLC-␤ is usually detectable in just a few seconds and frequently peaks around 10 -20 s before falling either to the base line or to a plateau level above the base line. These changes in PLC-␤ activity are mirrored by similar changes in [Ca 2ϩ ] i concentration. Negative feedback inhibition of PLC-␤ by PKC is one mechanism that has been proposed for the rapid termination of phosphatidylinositol 4,5-bisphosphate hydrolysis and [Ca 2ϩ ] i mobilization (4,17,18). However, investigations addressing the effects of PKC activation on the GPCR-activated PLC-␤ cascade have been limited by the inability to conclusively pinpoint the site of PKC action.
Because of the large number and complexity of components involved, PKC could act at many sites in the GPCR signaling pathway to limit PLC-␤ activity. Direct activation of PKC with PMA is known to induce the phosphorylation of M 3 mAChR, with undetermined effects on M 3 mAChR function (12). This finding, along with demonstrations that some other GPCRs can be phosphorylated by PKC (22,23), led to the speculation that PKC activation inhibits events at the receptor level. However, we found that activation of PKC with PMA does not alter the ability of M 3 mAChR or P2Y 2 receptors to activate G␣ q/11 . This result indicates that PKC inhibits a post-receptor event. Consistent with this conclusion, we found that PKC activation inhibits IP 3 generation that is induced by directly activating G proteins with GTP␥S, without involving GPCRs. These findings indicate that PKC inhibits the GPCR signaling cascade only at a point after the activation of G␣ q/11 .
Our studies show that PKC activation induces the phosphorylation of PLC-␤3 and diminishes the ability of PLC-␤3 to interact with G␣ q/11 . In contrast, PKC activation does not induce the phosphorylation of PLC-␤1, nor does it affect the ability of PLC-␤1 to associate with G␣ q/11 . These findings suggest that PKC inhibits PLC-␤3 activity by phosphorylating PLC-␤3 and diminishing the ability of PLC-␤3 to interact with G␣ q/11 . Although PKC activation may inhibit other events following PLC-␤ activation, our results indicate that the PKC- mediated inhibition of G␣ q/11 and PLC-␤3 interactions is one way that the GPCR signaling cascade can be deactivated.
It is intriguing that PKC activation induces phosphorylation of PLC-␤3 but not PLC-␤1. The reason for the lack of PLC-␤1 phosphorylation is unknown. It is unlikely that any differences in the intracellular compartmentalization of PLC-␤1 and PLC-␤3 contribute to their different phosphorylation profiles in CHO-m3 cells, because PMA probably has equal access to different intracellular sites. It is conceivable that PLC-␤1, but not PLC-␤3, is associated intracellularly with protein partners that protect it from PKC-mediated phosphorylation, although evidence for this possibility is lacking.
Activation of PKA, rather than PKC, has been found to induce PLC-␤3 phosphorylation in some systems (19). This PKA-mediated phosphorylation of PLC-␤3 would represent cross-regulation between GPCRs that signal through adenylyl cyclase-and PLC-␤-mediated pathways and not negative feedback control. This distinction is critical because PKC-mediated negative feedback can be a mechanism for the termination of PLC activity. In contrast, PKA-mediated cross-regulation is unlikely to be a mechanism for the termination of PLC activity because G␣ q/11 -coupled GPCRs do not generally activate adenylyl cyclase. We demonstrated that PLC-␤3 phosphorylation induced by P2Y 2 receptors or M 3 mAChR is not mediated by the action of PKA nor by an increase in cAMP. Agents that elevate cAMP do not affect PLC-␤ activation by these receptors. Similarly, it was previously reported that signaling by the plateletactivating factor receptor, which activates PLC-␤ via G␣ q/11 and induces PLC-␤3 phosphorylation, is insensitive to the action of forskolin or other PKA-activating agents (30). However, we did find that agents that increase cAMP can stimulate modest increases in PLC-␤3 phosphorylation (typically 1.5fold). It is possible that PKA regulates PLC-␤ activity only when PLC-␤ is activated by specific GPCR-mediated signals that differ from those generated by M 3 mAChR or P2Y 2 receptor stimulation in CHO cells. This possibility is supported by a previous report that PKA activation alters signaling by the formyl peptide receptor (30), which activates PLC-␤ via G␣ i rather than via G␣ q/11 . The differential coupling of M 3 mAChR and P2Y 2 receptors to PLC-␤1 and PLC-␤3 was somewhat surprising. Whereas the stimulation of M 3 mAChR induced the interaction of G␣ q/11 with PLC-␤1 and PLC-␤3, the stimulation of P2Y 2 receptors only induced the interaction of G␣ q/11 with PLC-␤3. These differences may occur because greater levels of G␣ q/11 are activated by stimulating M 3 mAChR than by stimulating P2Y 2 receptors in these cells. Stimulation of P2Y 2 receptors may not activate enough G␣ q/11 to allow the detectable association of PLC-␤1 with G␣ q/11 in our assays. Alternatively, it is possible that PLC-␤1 is located in unique intracellular sites that make it more susceptible to activation by M 3 mAChR than by P2Y 2 receptors.
Our results suggest an explanation for the ability of the M 3 mAChR to induce sustained PLC-␤ activation. Sustained PLC-␤ activity is induced by stimulating endogenous M 3 mAChR in cultured neuronal and lung cancer cell lines (6,9), indicating that sustained activation of PLC-␤ is not simply a result of nonphysiological signaling by transfected mAChR.
Instead, our results suggest that sustained PLC-␤ activation may be induced by M 3 mAChR because these receptors activate PLC-␤1, which is not susceptible to PKC-dependent inhibition. In contrast, only transient PLC-␤ activation may be induced by P2Y 2 receptors because these receptors activate only PLC-␤3, which is susceptible to PKC-mediated inhibition. We found that inactivation of PKC with BIM converts P2Y 2 receptormediated [Ca 2ϩ ] i mobilization from a transient event to a sustained event. According to our model, treatment with BIM diminishes the PKC-mediated inhibition of PLC-␤3, resulting in prolonged activation of PLC-␤3 and sustained [Ca 2ϩ ] i mobilization after P2Y 2 receptor activation. In addition to this PKCmediated regulation of PLC-␤, other negative and positive feedback events undoubtedly contribute to the different abilities of the M 3 mAChR and P2Y 2 receptor to regulate PLC-␤ activity. However, the activation of specific PLC-␤ isoforms that have different susceptibilities to PKC-mediated inhibition provides an intriguing mechanism for the temporal regulation of PLC-␤ activity by GPCRs.