Functional effects of protein kinase C-mediated phosphorylation of chick heart muscarinic cholinergic receptors.

Muscarinic cholinergic receptors (mAChR) purified from chick heart were phosphorylated by protein kinase C (PKC) and reconstituted with the purified GTP-binding regulatory protein Go. The effects of PKC phosphorylation on the interaction of mAChR with Go were assessed by monitoring for agonist-stimulated guanosine-5'-O-(3-thiotriphosphate) (GTP gamma S) binding to Go, agonist-stimulated GTPase activity of Go, and the capability of Go to induce high affinity agonist binding to mAChR. Both the receptor-stimulated GTP gamma S binding and GTPase activity of Go were markedly diminished as a result of PKC-mediated phosphorylation of the mAChR, whereas the ability of Go to induce high affinity agonist binding to the receptors was unaffected. When mAChR were first reconstituted with Go and then subjected to phosphorylation with PKC, a complete inhibition of the phosphorylation of mAChR by PKC was observed. The inhibitory effect of Go on mAChR phosphorylation was concentration-dependent and was prevented by the presence of GTP gamma S in the reaction mixtures. Taken together, these results indicate that the phosphorylation of mAChR by PKC modulates receptor/G-protein interactions and that the ability of the receptors to act as substrates for PKC may be regulated by receptor/G-protein interactions.

the receptors to transduce signals. Desensitization appears to involve decreases in: the affinity of the receptors for agonist (11)(12)(13), the efficiency of receptor/G-protein interactions (4, 14), and/or the number of functional cell surface receptors (15). Activation of protein kinases and phosphorylation of the mAChR seem to be involved in the desensitization process (11)(12)(13)(14)(15)(16)(17)(18)(19). Approaches to define the role of receptor phosphorylation in mAChR regulation have been reported in several studies. It has been shown in intact cell studies that agonist activation of chick and porcine heart mAChR leads to phosphorylation of the mAChR on serine and threonine residues to a stoichiometry of 3-5 mol of phosphate/mol of receptor (14,17,18). In in vitro studies it has been shown that purified mAChR from cardiac tissues, reconstituted into phospholipid vesicles, are excellent substrates for phosphorylation by a padrenergic receptor kinase (19) and a similar or related kinase (20). These reactions occur in a strictly agonist-dependent manner. In addition, purified mAChR from chick heart (21) and porcine cerebrum (22) are also excellent substrates for protein kinase C (PKC). Notably, the phosphorylation of the mAChR that is catalyzed i n vitro by PKC occurs in an agonistindependent manner (21,22), which is in marked contrast to the @-adrenergic receptor kinase-catalyzed events (18,20). The functional effects of mAChR phosphorylation have not been fully defined.
In this work we have studied the functional consequences of phosphorylation of chick heart mAChR by PKC by analyzing the interactions of the receptors with the G-protein Go. In addition, we have also determined the ability of Go to modify the ability of the receptors to serve as substrates for PKC.

MATERIALS AND METHODS
Purification of d C h R and G-proteins-Muscarinic receptors were purified from chick heart ventricles using the procedure of Haga and Haga (23), as described by Kwatra and Hosey (17), and were reconstituted into lipid vesicles as described previously (18,24). The specific activity of the purified receptors was -1 nmol of [3H]quinuclidinyl benzilate (QNB) bound/mg of protein (17). G-proteins were purified from calf brain according to the procedure described by Sternweis and Robishaw (25), which resulted in the purification of a mixture of heterotrimeric Go and Gi proteins, in a ratio of 4:1, respectively. Go was further purified from the G,/Gi mixture by the use of a Mono-Q column (Pharmacia LKB Biotechnology Inc.) as described by Katada et al. (26). Prior to mixing with reconstituted receptors, the amount of Go was determined by an [36S]GTPyS binding assay (25). In the text, the amount of Go used for experiments refers to the amount assayed before reconstitution.
Purification of Protein Kinase C-Protein kinase C was purified from chick brain according to the procedure of Woodgett and Hunter (27). This procedure results in the isolation of a doublet of 78-80 kDa (27) which probably contains the a, j3, and y isoforms of protein kinase C (28). The PKC preparation was homogeneous as assessed by silver staining after SDS-gel electrophoresis, was highly dependent on calcium, and was insensitive to cyclic nucleotides and calmodulin.
Phosphorylation of mAChR by Protein Kinase C-The phosphoryl-ation of the purified mAChR by protein kinase C was carried out as previously described (21). Ligand Binding Assays with Reconstituted mAChR and Go-Reconstitution of phosphorylated and nonphosphorylated mAChR with Go was performed as described by Haga et al. (24) using a ratio of 250:l of G,:receptor. After reconstitution, 50-60% of mAChR was recovered routinely, as assessed by QNB binding. In the text, the indicated amount of mAChR used refers to the amount of receptors measured by ligand binding after reconstitution, unless otherwise specified. Prior to reconstitution with Go, the phosphorylated and nonphosphorylated mAChR (which were also subjected to the conditions of the phosphorylation reaction, but without protein kinase C) were chromatographed on a Sephadex G-50 column (2 ml) in order to remove the ATP and other reagents of the phosphorylation reactions. Ligand binding to the reconstituted mAChR was carried out with the antagonist [3H]QNB (Amersham Corp.), and agonist affinity was determined using varying concentrations of the agonist carbachol in competition studies as previously described (24,29). Briefly, the assay GTPase Assays-The reconstitution of the receptor with Go was performed as described above for the GTP-yS binding assays. GTPase activity was assayed essentially according to the procedure described by The reaction mixtures were incubated at 30 "C for 45 min, and were stopped by addition of 10 pl of cold 50% trichloroacetic acid with immediate chilling on ice. The mixtures were centrifuged for 10 min at 2,500 rpm, and then 90 pl of the supernatant was removed and assayed for inorganic [32P]phosphate as previously described (31).

Effect of Protein Kinase C Phosphorylation of Chick Heart mAChR on the Ability of the Receptors to Stimulate GTPyS
Binding to Go-Chick heart muscarinic receptors were purified, reconstituted into lipid vesicles, and phosphorylated by PKC as previously described (21). Phosphorylation occurred to an extent of 4-5 mol of P/mol of receptor and was unaf- stimulated the binding of the nucleotide to Go; -70% of the total GTPyS binding after 30 min of reaction was due to agonist-stimulated binding. This effect was prevented by the presence of atropine in the reaction mixture. In contrast, the phosphorylated receptors (Fig. 1B) exhibited a diminished capacity for agonist-stimulated binding of GTPyS to Go. With the phosphorylated receptors a 35-40% decrease of the carbachol-stimulated GTPyS binding to Go was observed relative to the nonphosphorylated receptors. These results suggest that the phosphorylation of the mAChR by PKC may impede the interaction of the receptor with Go, which may result in a decrease in agonist-stimulated binding of GTPyS to the Gprotein.

Effect of Phosphorylation of the mAChR by Protein Kinase C on the Ability of the Receptors to Modulate the GTPase
Actiuity of Go-To further determine the effect of PKC phosphorylation of mAChR on mAChR/G, interactions, we studied the ability of the receptors to stimulate the GTPase activity of Go in an agonist-dependent manner. Nonphos-phorylated receptors were used as controls. As shown in Fig.  2, the presence of mAChR plus carbachol in the reaction mixtures stimulated the catalytic activity of Go by -13-fold. This effect was prevented by atropine. In contrast, the phosphorylated receptors stimulated GTPase activity to a significantly smaller degree. Only -45-50% of the Pi released in the control was observed in the presence of carbachol and the phosphorylated mAChR. These results are consistent with those obtained with the GTPyS binding assays and indicate that phosphorylation of the receptors by PKC affects mAChR coupling to Go and perturbs the capacity of the agonistactivated mAChR to stimulate GTPyS binding to Go and the GTPase activity of Go.
Effect of Protein Kinase C Phosphorylation of rnAChR on the Ability of Go to Induce High Affinity Agonist Binding to mAChR-The capacity of Go to induce high affinity agonist binding to the phosphorylated and nonphosphorylated mAChR was also studied. Phosphorylated and nonphosphorylated mAChR were reconstituted with Go using a proportion of 1 pmol of receptor/250 pmol of G-protein (24). Agonist affinity was determined in competition studies as described previously (21)(22)(23)(24)(25)(26)(27)(28)(29). As shown in Fig. 3, in the absence of Go, both the phosphorylated and the nonphosphorylated receptors exhibited a single low affinity state for the agonist carbachol with a Kd of -110 p~. In the presence of Go, both the nonphosphorylated and the phosphorylated receptors also exhibited a high affinity state with a Kd of -30-40 nM for carbachol and the dose-response curves were best fit with a two-state model. The percentage of mAChR exhibiting high affinity for carbachol was not different between the phosphorylated and the nonphosphorylated receptors under the conditions tested (Fig. 3) These results indicate that the phosphorylation of mAChR by PKC may not affect the ability of Go to induce high affinity agonist binding to the receptor.
Effect of Go on the Phosphorylation of rnAChR by Protein  interactions, it was also of interest to determine if Go could modify the ability of mAChR to serve as substrates for PKC. Purified and reconstituted mAChR were subjected to phosphorylation by protein kinase C in the absence and in the presence of Go, plus and minus the agonist carbachol. As shown in the autoradiogram of Fig. 4A, phosphorylation of the mAChR in the absence of Go (lunes 1 and 2) results in a major phosphorylated band at -80 kDa which corresponds to the chick heart mAChR. The stoichiometry of phosphorylation was -5 mol of phosphate/mol of receptor (21). The phosphorylation reaction was unaffected by agonist ( l u n e 2) as reported previously (21). In contrast, the presence of Go ( 5 pmol of Go/l pmol of mAChR) in the reaction mixture markedly inhibited the ability of protein kinase C to phosphorylate the reconstituted mAChR (lane 5). The stoichiometry of phosphorylation was reduced to -1.2 mol of P/mol of receptors. This inhibition of mAChR phosphorylation by Go occurred in the absence or presence of agonist (lane 6). Go was also subjected to phosphorylation by protein kinase C (lunes 3 and 4 ) , but no significant phosphorylation of Go was observed under these conditions. Only a faint band at -78 kDa was observed which corresponds to autophosphorylated PKC.
We further studied the effects of various concentrations of Go on the PKC-mediated phosphorylation of mAChR. Puri-  7 pmol) (lanes 1,2,5, and 6) or purified Go (lanes 3 and 4 ) were electrophoresed on SDS gels and visualized by autoradiography. fied mAChR were reconstituted with different ratios of Go and subjected to phosphorylation by PKC. As shown in Fig.  5 (upper panel) the Go-mediated inhibition of mAChR phosphorylation by PKC was concentration-dependent. The extent of inhibition was quantified by determining the stoichiometries of phosphorylation (Fig. 5 , lower panel). When mAChR were reconstituted with Go in a ratio of 10 pmol of Go/pmol of mAChR, only -0.8 mol of phosphate/mol of receptor was incorporated. This effect of Go was prevented by the presence of GTPyS (100 PM) in the reaction mixture (Fig.  5, lane 5 in upper panel and asterisk in lower panel). This latter result was predicted if the inhibitory effect of Go was due to an interaction of mAChR with Go.

DISCUSSION
In previous work we reported that mAChR purified from chick heart and reconstituted into phospholipid vesicles can serve as effective substrates for PKC (21). In contrast to results obtained in in vitro phosphorylation studies using padrenergic receptor kinase (19), phosphorylation of the chick heart mAChR by PKC was unaffected by agonist occupancy. Similar results were obtained by Haga et al. (22) using mAChR purified from porcine brain. The functional consequences of mAChR phosphorylation by PKC in mAChR signaling were not fully elucidated in our previous study; however, several studies have reported that PKC may play an important role in the regulation of mAChR function (reviewed in Ref. 32). It has been shown that the exposure of neuroblastoma cells to phorbol esters, well known activators of PKC, results in a decrease in the number of mAChR on the cell surface (15,32, 33) and attenuation in receptor-mediated stimulation of phosphatidylinositol hydrolysis ( 4 3 4 ) and cGMP production (35). The inhibition of these stimulatory effects has been proposed to be associated with a feedback regulation of mAChR by PKC phosphorylation (36), but direct evidence for this has not been demonstrated. In this work we present evidence that phosphorylation of mAChR by PKC may directly affect the coupling of the receptors to Go, which results in a decrease in mAChR-mediated stimulation of GTPyS binding and the catalytic activity of the G-protein.
The interaction of mAChR with Go has been analyzed in several studies (24, 30, 37). It has been shown that agonist interaction with mAChR effectively stimulates the binding of GTPyS to Go as well as the hydrolysis of GTP by Go in reconstituted vesicles containing purified mAChR and Go (30, 37). Similar results were obtained in this work under control conditions using nonphosphorylated mAChR. When these functions were monitored in reconstituted vesicles containing mAChR phosphorylated by PKC, a -35-50% inhibition was observed compared with the nonphosphorylated receptors. These results suggest that the phosphorylation of mAChR by PKC can perturb receptor/(;, coupling and diminish the transducing capacity of the receptor. It has not yet been demonstrated if similar events occur in intact cells. However, mAChR-mediated stimulation of phosphoinositide turnover may result in an increase in the intracellular concentration of diacylglycerol and Ca2+ and a consequent activation of PKC (36). Thus PKC may regulate mAChR function by phosphorylating the receptors, resulting in altered receptor/ G-protein coupling.
It has been shown that prolonged exposure of neuroblastoma cells to phorbol ester also results in a decrease in the affinity of mAChR for agonist (38). In order to investigate the possibility that this might be due to the phosphorylation of mAChR by PKC, we studied the ability of Go to induce high affinity agonist binding to the phosphorylated and nonphosphorylated receptors. A ratio of 1 pmol of mAChR/250 pmol of Go was used in the reconstitution process in order to facilitate detection of the effects of the G-protein (37). Surprisingly, both the phosphorylated and the nonphosphorylated receptors showed a similar percentage of high affinity binding (48 and 53%, respectively) and a similar affinity for the agonist carbachol (& of -30-40 nM). These apparently different effects of phosphorylation on receptor/(;-protein interactions may be explained in several ways. First, the phosphorylation of mAChR by PKC may affect receptormediated activation of Go without affecting the ability of Go t o induce high affinity agonist binding to the receptors. This possibility seems plausible in light of studies with the padrenergic receptors which suggest that different domains are involved in receptor-mediated activation of GB and effects of G. on receptor affinity (51,52). A second interpretation of the differential effects of mAChR phosphorylation in receptor/(;, interaction is that the presence of a third component like arrestin or P-arrestin may be necessary to observe certain effects of phosphorylation (39,53). Third, the possibility that the relatively high proportion of G,:receptor used for the ligand binding assays (uersus the lower proportion used in GTPase activity and GTPyS binding assay) could have masked an effect of the phosphorylation of mAChR by PKC to modify Go-induced high affinity agonist binding. On the other hand, the results we obtained are consistent with those reported by Haga et al. (22) who assessed effects of PKC phosphorylation on mAChR purified from porcine brain. In contrast, previous results from our group (21), from studies in which we used a mixture of GJG,, showed a positive effect of the PKC phosphorylation of mAChR on the G-protein induced high affinity state of the receptors. This may have been due to the mixture of G-proteins used or some differences in the conditions of the reactions. The mixture of Gi/G, used in our previous study (21) was only -70% pure as assessed by silver staining after SDS-gel electrophoresis and we cannot discount the possibility that impurities in the mixture may have contributed to the difference in results. Furthermore the previous study of the effects of the mixture of Gi/Go on high affinity agonist binding was performed in the presence of 0.1% Lubrol plus 0.1% cholate, whereas the present studies were performed in the absence of cholate. We have not yet tested whether these factors contributed to the results, nor do we know how phosphorylation of the receptor affects interactions with purified isoforms of Gi. We plan to address these issues in future studies.
In view of the ability of PKC to modify mAChR/G, interactions, it is interesting that Go can inhibit the phosphorylation of the receptors by PKC. This inhibition caused by Go was reversed by the presence of GTPyS in the reaction mixture but was unaffected by the presence of agonist. Agonists are thought to stabilize receptor/G-protein interactions, whereas the binding of GTPyS to Go should result in an uncoupling of the receptors from the G-protein and a dissociation of the G-protein trimer into its a and By subunits. The G-protein-mediated inhibition may be due to the interaction of the third cytoplasmic loop of the receptor with the G-protein, where the potential sites for PKC phosphorylation are located. This inhibitory effect of Go on mAChR phosphorylation has important implications for the regulation of the receptors by PKC in intact cells, as it suggests that only receptors dissociated from G-proteins may serve as substrates for PKC.
In summary, the present results indicate that the phosphorylation of chick heart mAChR by PKC interferes with the capacity of the receptors to stimulate GTPyS binding and GTPase activity of the G-protein Go without affecting the ability of Go to induce high affinity agonist binding state of the receptors. The presence of Go in the reaction mixtures results in an inhibition of the phosphorylation of mAChR by PKC. The inhibitory effect of Go was completely suppressed by the nucleotide GTPyS, indicating that it was specifically due to the presence of the G-protein in the reconstituted system.
A question that remains to be clarified is: how many subtypes of mAChR(s) undergo phosphorylation by PKC? Molecular cloning studies have demonstrated that there are at least five distinct subtypes of mAChR (ml-m5) (40-47). Studies with several neuronal cells suggest that ml, m3, and m4 subtypes may be regulated by PKC. It has been shown that mammalian m2 mAChR purified from cardiac tissue are apparently not substrates for PKC (22,50). For the present studies we have used purified chick heart mAChR. So far, two subtypes have been reported to be present in chick heart, m2 and m4 (48,49), with the m2 receptor being expressed at much higher levels than the m4 receptor (49). If the protein levels of the two receptors reflect the amount of mRNA expressed, then the chick heart contains predominantly m2 receptors. Using different expression systems which will allow us to obtain significant amounts of expressed subtypes of mAChR with relative ease, we expect to address the question of which subtypes of mAChR are regulated by PKC in the near future.