Protein kinase C alters the responsiveness of adenylyl cyclases to G protein alpha and betagamma subunits.

The ability of protein kinase C (PKC) to regulate the responsiveness of adenylyl cyclase to different activators was assessed. Membranes prepared from Sf9 cells infected with recombinant baculoviruses encoding either type II or IV adenylyl cyclase were incubated with recombinant PKCalpha (purified from Sf9 cells), and the effects on adenylyl cyclase activity were measured after reconstitution with Gsalpha, Gbetagamma, or forskolin. PKCalpha treatment of type II adenylyl cyclase leads to increases in basal, forskolin-stimulated, and betagamma-stimulated activities and greater sensitivity to stimulation by Gsalpha. Paradoxically, most of the betagamma potentiation of Gsalpha-stimulated activity is eliminated by pretreatment with PKCalpha. By contrast, treatment of type IV adenylyl cyclase with PKCalpha has little effect on the basal, forskolin-stimulated, or betagamma-stimulated activities but markedly reduces the Gsalpha-stimulated and betagamma-potentiated activity of this isoform. These studies demonstrate that protein kinases can alter both the activity of adenylyl cyclase isoforms and their responsiveness to G protein regulation, thereby altering the ability of adenylyl cyclases to integrate signals derived from multiple hormonal inputs.

The ability of protein kinase C (PKC) to regulate the responsiveness of adenylyl cyclase to different activators was assessed. Membranes prepared from Sf9 cells infected with recombinant baculoviruses encoding either type II or IV adenylyl cyclase were incubated with recombinant PKC␣ (purified from Sf9 cells), and the effects on adenylyl cyclase activity were measured after reconstitution with G s␣ , G␤␥, or forskolin. PKC␣ treatment of type II adenylyl cyclase leads to increases in basal, forskolin-stimulated, and ␤␥-stimulated activities and greater sensitivity to stimulation by G s␣ . Paradoxically, most of the ␤␥ potentiation of G s␣ -stimulated activity is eliminated by pretreatment with PKC␣. By contrast, treatment of type IV adenylyl cyclase with PKC␣ has little effect on the basal, forskolin-stimulated, or ␤␥-stimulated activities but markedly reduces the G s␣stimulated and ␤␥-potentiated activity of this isoform. These studies demonstrate that protein kinases can alter both the activity of adenylyl cyclase isoforms and their responsiveness to G protein regulation, thereby altering the ability of adenylyl cyclases to integrate signals derived from multiple hormonal inputs.
Intracellular cAMP concentrations are principally controlled at the level of its synthesis by the hormonal regulation of adenylyl cyclase, the enzyme catalyzing the conversion of ATP to cAMP. Hormonal regulation of adenylyl cyclases is brought about by the receptor-catalyzed activation of heterotrimeric G proteins that in turn regulate the cyclases by the released ␣ or ␤␥ subunits. Molecular genetic approaches have identified nine mammalian isoforms of adenylyl cyclase (types I-IX) (reviewed in Refs. [1][2][3], and studies of these enzymes reveal both common and unique regulatory features. All isoforms tested to date are activated by the GTP-bound form of G s␣ and by forskolin and are inhibited by adenosine analogues termed P-site inhibitors (4). All isoforms of adenylyl cyclase are further regulated by additional inputs in an isoform-specific pattern. For example, increases in intracellular calcium concentrations will inhibit the type V and VI isoforms (5-7) while indirectly activating (via a calmodulin-dependent process) the type I and VIII isoforms (8,9) and inhibiting (by calmodulin kinase) the type III isoform (10).
Regulation of adenylyl cyclase activity by hormone receptors coupled to members of the G i class of G proteins is more complex. The released G i␣ (or G z␣ ) subunits can inhibit the type I, V, and VI isoforms (11,12); G o␣ can also inhibit type I adenylyl cyclase, albeit less potently than G i␣ . The released ␤␥ subunits also regulate adenylyl cyclases in an isoform-specific fashion. Inhibitory effects of ␤␥ subunit complexes on the type I isoform are both evident and more pronounced at lower concentrations than G i␣ inhibition (12)(13)(14). In marked contrast, the type II and IV isoforms are insensitive to inhibition by G i␣ (12) but are activated by ␤␥ subunits in which stimulatory effects are most pronounced in the presence of G s␣ activation (13,15).
An additional level of adenylyl cyclase regulation has been suggested by the effects of phorbol esters on adenylyl cyclase activity initially reported by Yoshimasa et al. (16) and more recent reports of phorbol ester effects on expressed adenylyl cyclase isoforms (17)(18)(19)(20)(21)(22)(23). In particular, several groups have demonstrated that phorbol ester treatment of cells exogenously expressing type II adenylyl cyclase results in an increase in adenylyl cyclase activity (18 -21), suggesting that type II adenylyl cyclase may be phosphorylated and stimulated in response to PKC 1 activation. In this report, we examine the effects of purified PKC␣ on the regulatory properties of type II adenylyl cyclase and the other ␤␥-stimulable isoform, type IV. Herein, we demonstrate that in addition to regulating adenylyl cyclase activity, PKC can modulate the responsiveness of adenylyl cyclase to G protein subunits, thereby altering the ability of adenylyl cyclase to integrate signals derived from multiple hormonal inputs.

EXPERIMENTAL PROCEDURES
Sf9 Cell Culture-Procedures for the culture of Sf9 cells and the production, cloning, and amplification of recombinant baculovirus have been outlined by Summers and Smith (24). Baculoviruses encoding type II and type IV adenylyl cyclase have been reported (13,15). Construction of recombinant baculovirus encoding PKC␣ has been described previously (25).
Isolation of Membrane and Purified Recombinant Adenylyl Cyclase Preparations-Sf9 membranes containing individual adenylyl cyclase isoforms were prepared as described (26). The type II adenylyl cyclase isoform was purified from membranes by sequential chromatography on forskolin-Sepharose and lentil lectin-Sepharose columns using published procedures (14).
Purification of PKC␣-PKC␣ was purified from Sf9 cells expressing the recombinant enzyme. Cells (5 ϫ 10 8 cells in 500 ml) were incubated with recombinant baculovirus encoding PKC␣ (multiplicity of infection ϭ 10) and harvested 45-50 h postinfection by centrifugation for 10 min at 1,000 ϫ g. The cell pellet was resuspended in 25 ml of buffer A (20 mM Tris-HCl, pH 7.5, 10 mM EGTA, 2.5 mM EDTA, 1 mM dithiothreitol, and protease inhibitors (1 g/ml leupeptin and lima bean trypsin inhibitor, 2.5 g/ml each of phenylmethylsulfonyl fluoride, 1-chloro-3-tosylamido-7-amino-2-heptanone,andL-1-tosylamido-2-phen- ylethyl ketone, and 2.2 g/ml aprotinin)) and homogenized in a motordriven Teflon homogenizer (20 strokes) at 4°C. The homogenate was centrifuged at 100,000 ϫ g for 1 h at 4°C, and the supernatant was collected and loaded on a Fast Q column (10 ϫ 100 mm; Pharmacia Biotech Inc.). The column was washed with 5 volumes of buffer B (20 mM Tris-HCl (pH 7.5), 1 mM EGTA, 1 mM EDTA, 1 mM dithiothreitol, and protease inhibitors) and eluted with a gradient of 0 -400 mM NaCl in buffer B (10 column volumes) at a rate of 3 ml/min. Fractions containing high PKC activity (as measured by incorporation of [␥-32 P]ATP into histone H3 (27)) were pooled. For some preparations, the recombinant enzyme was further purified by chromatography on a polyacrylamide-immobilized phosphatidylserine column, synthesized, and run as described (28). In all cases, glycerol was added to the final preparations of PKC␣ to a final concentration of 20%, and aliquots of the enzyme were stored at Ϫ70°C.
Adenylyl Cyclase Assay-Adenylyl cyclase activity was measured using the procedure described by Smigel (29). All assays were performed at 30°C in a final volume of 100 l, and the final concentration of MgCl 2 and EGTA was 10 and 1 mM, respectively. Before the initiation of the assay, adenylyl cyclase preparations were incubated with PKC␣ for 10 or 20 min at 30°C in a 50-l reaction volume containing 30 mM Na-HEPES (pH 8), 5 mM MgCl 2 , 0.3 mM ATP, 0.4 mM EDTA, 2 mM CaCl 2 , 0.2 mg/ml phosphatidylserine, 1 M phorbol myristic acid, 2 mM dipotassium phosphoenolpyruvate, and 0.067 mg/ml bovine serum albumin. Membrane preparations were assayed for 10 min; purified adenylyl cyclase was assayed for 5 min.
Purification of G Protein Subunits-G protein ␤␥ subunits were purified from bovine brain following procedures by Sternweis and Robishaw (30). Recombinant G s␣ (rG s␣ ) was synthesized in bacteria and purified as described by Lee et al. (31). Protein concentrations were estimated by staining with Amido Black (32). The ␣ subunits were activated by incubation with 50 mM Na-HEPES (pH 8.0), 5 mM MgSO 4 , 1 mM EDTA, 1 mM dithiothreitol, and 400 M GTP␥S at 30°C for 30 min (33); free GTP␥S was removed by gel filtration.
In Vitro Phosphorylation of Type II Adenylyl Cyclase-PKC␣ was incubated at 30°C for 5 min in a 250-l reaction volume consisting of 50 mM Na-HEPES (pH 8), 7 mM MgCl 2 , 0.6 mM EDTA, 3 mM CaCl 2 , 15 M ATP, 3 mM dipotassium phosphoenolpyruvate, 0.1 mg/ml bovine serum albumin, 0.2 mg/ml phosphatidylserine, and 1.5 M phorbol myristic acid. Adenylyl cyclase (1.7 pmol) purified from Sf9 membranes and [␥-32 P]ATP (4,000 cpm/pmol) were then added to the reaction mixture to give a final volume of 350 l and incubated at 30°C for 5 min. The samples were precipitated with trichloroacetic acid, alkylated with N-ethylmaleimide, and subjected to electrophoresis through a 7.5% polyacrylamide gel as described by Taussig et al. (26). Phosphorylated products were visualized by autoradiography using Kodak X-OMAT AR film or by analysis on a Molecular Dynamics 445-SI PhosphorImager.

RESULTS
To examine the effects of PKC on the regulation of type II adenylyl cyclase, we preincubated recombinant PKC␣ (purified from Sf9 cells) with membranes prepared from Sf9 cells infected with a recombinant baculovirus encoding this adenylyl cyclase isoform. Adenylyl cyclase activity was then assayed in the absence or presence of a single concentration of forskolin, rG s␣ , or G protein ␤␥ subunits or a combination of rG s␣ and ␤␥. As shown in Fig. 1A, preincubation of type II adenylyl cyclase with PKC␣ increased basal, forskolin-stimulated, G s␣ -stimulated, and ␤␥-stimulated activities. Surprisingly, PKC␣ pretreatment reduced the activity measured in the presence of rG s␣ ϩ ␤␥ despite the fact that PKC␣ elevated the activity of type II adenylyl cyclase in the presence of either regulator alone.
As shown in Fig. 1B, the effect was saturable with respect to the amount of added PKC␣, as illustrated for the G s␣ -stimulated activity. In addition, the PKC␣ effects can be blocked by bisindolyl maleimide, a potent and specific PKC inhibitor. The boiled PKC␣ preparation had no effect on cyclase activity, ruling out activation due to buffer components (data not shown). Finally, we determined that neither GTP␥S-bound rG s␣ nor ␤␥ is directly phosphorylated by PKC␣, demonstrating that the observed effect is not mediated by the phosphorylation of the cyclase regulators (data not shown).
Similar analysis of the effects of PKC␣ on the type IV isoform (a second ␤␥-stimulated cyclase) revealed quite different results, shown in Fig. 2. Unlike the type II enzyme, treatment of type IV adenylyl cyclase with PKC␣ has little effect on the basal, forskolin-stimulated, or ␤␥-stimulated activities. More strikingly, PKC␣ decreases both the G s␣ -stimulated and G s␣ ϩ ␤␥-stimulated activities of this isoform.
To gain further insight into the effect of PKC␣ on G s␣ -stimulated activities, we examined the effect of PKC treatment on dose-dependent stimulation by rG s␣ . Inspection of the doseresponse curve in Fig. 3A reveals that PKC-treated type II adenylyl cyclase is more responsive to stimulation by rG s␣ than the nontreated enzyme; no change in the maximum G s␣ -stimulated activity was observed. By contrast, PKC treatment of type IV adenylyl cyclase reduces its responsiveness to G s␣ . PKC␣ does not affect the basal type IV cyclase activity; how- ever, the G s␣ -stimulated activity is inhibited over the entire range of rG s␣ concentrations tested (Fig. 3B). We could not determine if maximal G s␣ -stimulated activity was also reduced because we were unable to reach saturating rG s␣ concentrations for the kinase-treated type IV enzyme.
We next tested the effect of PKC␣ on the dose-dependent stimulation of type II adenylyl cyclase by ␤␥ in both the absence and presence of G s␣ stimulation. In the absence of rG s␣ , ␤␥ stimulation of type II adenylyl cyclase was very modest, requiring high nanomolar to micromolar concentrations of ␤␥ to see substantial stimulation. Nevertheless, pretreatment with PKC␣ resulted in an augmentation of the ␤␥ response (Fig. 4A), and in the presence of high (300 nM) ␤␥ concentrations, activities were more than additive with the enhanced basal activity and therefore not due solely to increases in the basal activity of the enzyme. Upon the addition of rG s␣ to the assay (in which ␤␥ effects were magnified), two striking observations were apparent (Fig. 4B). First, the addition of rG s␣ caused a leftward shift in the ␤␥ dose-response curve (compare the curves marked with E in Fig. 4, A and B), suggesting that G s␣ either allosterically increases the apparent affinity of the cyclase for ␤␥ or exposes a second high affinity ␤␥ site on the enzyme. Second, pretreatment of the type II cyclase with PKC␣ eliminated superstimulation by ␤␥ by removing this high affinity interaction (Fig. 4B).
Like type II adenylyl cyclase, type IV is stimulated by ␤␥ in the presence of rG s␣ . In contrast to the type II enzyme, we observed no stimulation of type IV adenylyl cyclase at the ␤␥ concentrations tested when assayed in the absence of rG s␣ , and pretreatment of this cyclase with PKC␣ had little effect when assayed under these conditions (Fig. 2). However, as shown in Fig. 5, PKC treatment of type IV membranes blocked the ability of ␤␥ to superstimulate type IV adenylyl cyclase in the presence of rG s␣ . Because PKC treatment reduced stimulation of type IV adenylyl cyclase by G s␣ alone, we examined whether the PKC treatment decreased G s␣ stimulation of type IV adenylyl cyclase below a threshold activity necessary to observe ␤␥ superstimulation. At reduced rG s␣ concentrations (Fig. 5, Ⅺ) that give comparable cyclase activity (in the absence of ␤␥) to that of the PKC-treated enzyme assayed at high G s␣ concentration (q), ␤␥ superstimulation was observed for the nontreated sample but not for the kinase-treated samples. When higher concentrations of G s␣ (3 M) were tested, still no ␤␥ superstimulation was observed in the presence of PKC, even though ␤␥ superstimulation was still evident in the absence of PKC at rG s␣ concentrations 3 orders of magnitude lower (3 nM), in which G s␣ stimulation of type IV was not measurable (data not shown). These results suggest that PKC treatment of the type IV adenylyl cyclase causes a block of ␤␥ effects in addition to reducing the sensitivity to G s␣ , thereby altering the ability of this cyclase to integrate coincidental hormonal inputs from G sand G i /G o -mediated pathways.
We next sought to determine whether the regulatory changes induced by PKC treatment of membranes were due to the phosphorylation of adenylyl cyclase by PKC. Using procedures that we previously developed, we purified recombinant type II adenylyl cyclase from Sf9 cells and examined the effects of PKC on the activity of this purified cyclase preparation. Preincubation of purified type II cyclase with purified PKC resulted in an increase in G s␣ -stimulated activity that was blocked by the addition of bisindolyl maleimide, a specific PKC inhibitor (Fig.  6A). However, the magnitude of this response was consistently much smaller than that observed in membrane preparations (compare Figs. 6A and 2A), and similarly small effects of PKC on purified type II adenylyl cyclase were seen when forskolinstimulated activity was assayed (data not shown). This loss of responsiveness to PKC is not due to the purification process; rather, it is evident upon removal of the type II adenylyl cyclase from the membrane upon detergent solubilization. As shown in Fig. 6B, the addition of ␤-dodecyl maltoside (the detergent used to solubilize and purify the recombinant cyclase) to the PKCtreated membranes resulted in the loss of the enhanced forskolin response of the type II adenylyl cyclase. Similar results were obtained when the order of treatment was reversed (detergent addition followed by PKC treatment), when other cyclase regulators were used, or when other detergents were tested (data not shown). 2 The observation that the effects of PKC on type II adenylyl cyclase activity (measured in membranes) are lost after detergent solubilization suggests that either: 1) the effects of PKC we measured may not be due to a direct phosphorylation of the cyclase by PKC but may result from phosphorylation of some accessory component present in the membranes; or 2) the conformational state of the cyclase in membranes is important to observe effects of phosphorylation by PKC. This second possibility is not unique to the type II adenylyl cyclase isoform. For example, Kawabe et al. (22) have shown that in intact cells or with purified enzyme, type V adenylyl cyclase can be phosphorylated (and its activity augmented) by PKC activation, whereas no effect is observed in membranes treated with PKC. In addition, calmodulin stimulation of type I adenylyl cyclase can be dramatically affected by detergents or phospholipid composition of the surrounding bilayer (34). 3 These results highlight the importance of the interaction between adenylyl cyclase and its membrane environment for the regulation of adenylyl cyclase activity, either for conformational restraint or protein-lipid and/or protein-protein interactions.
Clearly, type II adenylyl cyclase is phosphorylated in response to PKC activation. When coexpressed in Sf9 cells with PKC␣, phosphorylation of type II adenylyl cyclase was detected in response to activation of the kinase by phorbol esters (21). In addition, as shown in Fig. 7, PKC can directly phosphorylate type II adenylyl cyclase; incubation of purified type II adenylyl cyclase with purified recombinant PKC␣ in the presence of [␥-32 P]ATP resulted in a dose-dependent increase in phosphorylation of the cyclase that could be blocked by PKC inhibitors. However, in these experiments, we could maximally phosphorylate the enzyme to a stoichiometry of only 0.2. Similar stoichiometry was measured when type II adenylyl cyclase was affinity-purified from in vitro phosphorylated membrane preparations. In addition, reconstituting the purified cyclase into lipid vesicles of varying phospholipid compositions did not improve the incorporation of phosphate into the molecule. In summary, because the effects of PKC on type II adenylyl cyclase activity are drastically reduced in detergent solution and low stoichiometry of phosphorylation is observed with purified type II adenylyl cyclase, we are unable to attribute this phosphorylation to the changes in cyclase activity and responsiveness to G protein subunits and cannot rule out a possible role of an accessory component present in the Sf9 cell membranes that is required for the response of the cyclase to PKC. DISCUSSION Adenylyl cyclase activity, and ultimately intracellular cAMP concentrations, can be regulated by a number of hormonal inputs coupled to heterotrimeric G protein-mediated pathways. Hormone receptors coupled to the G s and G i subclass of G proteins can directly activate or inhibit adenylyl cyclase activity through the interactions of the released ␣ and ␤␥ subunits with the cyclase. Another level of regulation is provided by the cross-talk of distinct G protein-mediated signaling pathways 2 Y. Ishikawa, personal communication. 3  with adenylyl cyclase isoforms. For example, hormonal activation of a G q -coupled receptor leading to the activation of PKC (via a process involving phospholipase C activation, diacylglycerol and IP 3 production, and released intracellular calcium) can have effects on the activities of different adenylyl cyclase isoforms (17)(18)(19)(20)(21)(22)(23). Our studies demonstrate that in addition to regulating adenylyl cyclase activity itself, PKC can modulate the ability of the adenylyl cyclase isoforms to respond to different G protein subunits.
As diagrammed in Fig. 8, the type II and IV adenylyl cyclases, isoforms that exhibit almost identical regulatory patterns with respect to their responses to G protein subunits, are affected quite differently by PKC. Both adenylyl cyclases can integrate hormonal inputs from G s -, G i -, and G q -coupled receptors that are mediated by G s␣ , ␤␥, and PKC, respectively.
Hormonal activation of PKC␣ causes an increased responsiveness of type II adenylyl cyclase to G s␣ , leading to an enhanced G s␣ -stimulated activity as outlined in Fig. 8A. PKC␣ also enhances ␤␥ stimulation of type II adenylyl cyclase, although these effects require high nanomolar to micromolar concentrations of ␤␥, a concentration more than 2 orders of magnitude higher than that required to see ␤␥ effects on other adenylyl cyclases, phospholipases, or ion channels (13,35,36). In contrast, lower concentrations (EC 50 ϭ 5-15 nM) of ␤␥ are required to see superstimulation of type II adenylyl cyclase in the presence of G s␣ , and therefore, augmentation of ␤␥ stimulation by PKC␣ or G s␣ seems to occur by distinct mechanisms. Our surprising observation that PKC␣ augments ␤␥ stimulation but blocks the ␤␥ superstimulation of type II adenylyl cyclase (in the presence of G s␣ ) may be due to an inability of G s␣ to expose the high affinity ␤␥ site on type II cyclase after PKC treatment.
Type IV regulation is modulated very differently by PKC␣, as diagrammed in Fig. 8B. PKC␣ activation leads to a reduced responsiveness of type IV adenylyl cyclase to G s␣ , resulting in decreased G s␣ -stimulated activity. This is the first demonstration of an inhibitory effect of PKC on cloned adenylyl cyclase activity and may serve to explain the inhibitory effects of phorbol esters on the G s␣ -stimulated adenylyl cyclase activity observed in some mammalian cell lines (37,38). In addition to the effects on G s␣ activation, PKC␣ blocks the superstimulation of type IV adenylyl cyclase by ␤␥ subunits.
Adenylyl cyclase activity can be regulated by multiple hormonal inputs that are mediated by the direct interaction of the FIG. 6. Effect of PKC on the activity of purified and detergentsolubilized type II adenylyl cyclase. A, recombinant type II adenylyl cyclase (0.16 pmol) purified from Sf9 cells was incubated in the absence (f) or presence (p) of PKC␣; bisindolyl maleimide, a PKC inhibitor, was added as indicated. Adenylyl cyclase activity was then assayed in the presence of 100 nM rG s␣ . Assays were performed in duplicate (bars, S.D.), and the results are representative of three experiments. B, membranes (5 g) prepared from Sf9 cells expressing type II adenylyl cyclase were incubated with carrier (f) or 0.03 unit of PKC␣ (p) for 10 min at 30°C. The membranes were then treated with 0.03% dodecyl ␤-maltoside (ϩDetergent) or water (ϪDetergent) at 0°C. Adenylyl cyclase activity was then assayed in the presence of 50 M forskolin. Assays were performed in duplicate (bars, S.D.), and the results are representative of three experiments. cyclase with G protein subunits or via G protein-regulated pathways leading to the release of intracellular calcium and the subsequent activation of PKC. The conditional stimulation of adenylyl cyclase activity by ␤␥ subunits (in the presence of G s␣ ) serves as an example of the temporal integration of G sand G i -mediated hormonal signals by type II and IV adenylyl cyclases (13,15,39). PKC differentially affects the ability of type II adenylyl cyclase to respond to G s and/or G i activation, rendering the enzyme more sensitive to stimulation by G s␣ yet insensitive to superstimulation by ␤␥ released by G i or G o activation. Integration of these same signals derived from hormonal activation of G s (G s␣ ), G i (␤␥), and G q (PKC) is seen for the type IV isoform; however, in the case of this cyclase, the activation of PKC reduces both the G s␣ and ␤␥ stimulation of the enzyme. Our current study suggests that an added level of adenylyl cyclase regulation is imparted by the integration of G q -mediated pathways (leading to the activation of PKC) that can alter the inherent activities of adenylyl cyclases, but more importantly, that modulate regulation by G protein subunits.