Evidence for distinct guanine nucleotide sites in the regulation of the glucagon receptor and of adenylate cyclase activity.

Previous studies have shown that GTP or guanyl-5’-ylimidodiphosphate (Gpp(NH)p) stimulate the activity of hepatic adenylate eyclase and decrease the affinity of the glucagon receptor for the hormone. The studies reported here provide evidence that the effects of the nueleotides on these processes are exerted through functionally and possibly structurally distinct nucleotide sites. The evidence is as follows: (a) the concentration dependency for Gpp(NH)p action on hormone binding is at least 1 order of magnitude higher than for enzyme activation; (b) whereas Gpp(NH)p and GTP are about equipotent on enzyme activation, Gpp(NH)p is considerably less potent than GTP on the hormone binding process; (c) ethylene glycol bis(jSaminoethyl ether)N,N’-tetraacetate decreases the potency of Gpp(NH)p activation of adenylate cyclase but does not alter the potency of Gpp(NH)p action on hormone binding; (d) pretreatment of hepatic membranes with Gpp(NH)p leads to an activated state of adenylate cyclase which persists after washing or addition of GTP. The same pretreatment leads to a low affinity state of the receptor that is readily reversed to a high affinity state by washing the membranes; the high affinity state remains susceptible to GTP action after pretreatment with Gpp(NH)p and washing; (e) treatment of the membranes with phospholipase C (from Bacillus cereus) abolishes the action of Gpp(NH)p on glucagon binding but does not alter the ability of the nucleotide to activate adenylate cyclase. An hypothesis is presented which serves to explain the role of the two guanine nucleotide sites in the overall regulation of adenylate cyclase activation and in the “coupling” reaction between receptor and enzyme.

Previous studies have shown that GTP or guanyl-5'-ylimidodiphosphate (Gpp(NH)p) stimulate the activity of hepatic adenylate eyclase and decrease the affinity of the glucagon receptor for the hormone. The studies reported here provide evidence that the effects of the nueleotides on these processes are exerted through functionally and possibly structurally distinct nucleotide sites. The evidence is as follows: (a) the concentration dependency for Gpp(NH)p action on hormone binding is at least 1 order of magnitude higher than for enzyme activation; (b) whereas Gpp(NH)p and GTP are about equipotent on enzyme activation, Gpp(NH)p is considerably less potent than GTP on the hormone binding process; (c) ethylene glycol bis(jSaminoethyl ether)N,N'-tetraacetate decreases the potency of Gpp(NH)p activation of adenylate cyclase but does not alter the potency of Gpp(NH)p action on hormone binding; (d) pretreatment of hepatic membranes with Gpp(NH)p leads to an activated state of adenylate cyclase which persists after washing or addition of GTP. The same pretreatment leads to a low affinity state of the receptor that is readily reversed to a high affinity state by washing the membranes; the high affinity state remains susceptible to GTP action after pretreatment with Gpp(NH)p and washing; (e) treatment of the membranes with phospholipase C (from Bacillus cereus) abolishes the action of Gpp(NH)p on glucagon binding but does not alter the ability of the nucleotide to activate adenylate cyclase.
An hypothesis is presented which serves to explain the role of the two guanine nucleotide sites in the overall regulation of adenylate cyclase activation and in the "coupling" reaction between receptor and enzyme. It

Guanine
Nucleotide Sites 5943 time course of the assay (4). The pH optimum for these binding studies was 7.0 in contrast to the pH for adenylate cyclase assays. As described by Lin et al. (8), this minimizes the difference in binding properties between native and iodinated species which is particularly accentuated at pH 7.5 to 8.0. However, we found that higher pH did not affect the results obtained although the net binding was lower and the effect of GTP thereby harder to quantitate. Inclusion of adenylate cyclase assay reagents (except for ATP which mimics the action of GTP at high concentrations (1, 2)) also had no effect on the results reported here. Binding reactions were stopped by the addition of 1 ml of cold 20 mM Tris/HCl, pH 7.0, containing 0.2% BSA, followed by immediate filtration under vacuum on oxoid filters (0.45 ~1 which were presoaked for 30 min in 10% BSA (8). The filters were washed two additional times with 1 ml of the same buffer. The filters were counted in a y counter and the error in couxiting was less than 1%. Any variations to this binding assay are described in the figure legends.
The values reported here were obtained by subtracting a background which was determined by carrying out the binding assay in the presence of l,OOO-fold excess of native glucagon (10m6 M). In general, this value was nearly identical to the binding to the filter alone.
Other Methods -The procedures for the pretreatment of liver plasma membranes with Gpp(NH)p, chelators, and digestion of these membranes with phospholipase C (from B. cereus) are described in the figure and table legends.
Analysis for phospholipid hydrolysis was performed as described previously (9). Protein was determined by the method of Lowry et al. (10).

Dose-Response Effects of GTP and Gpp(NH)p on Y-glucagon Binding
and Adenylate Cyclase Activity- Fig.  1 illustrates the concentration dependency for Gpp(NH)p activation of adenylate cyclase and for the effects of this nucleotide on the binding of Y-glucagon to its receptor. Half-maximal concentration for activation* was 0.4 PM, whereas the half-maximal concentration required for effects on glucagon binding was about ZO-fold higher (7 FM). By contrast potency for the effects of GTP on these processes was essentially identical (0.2 pM) as seen in Fig. 2. These differences in the potency of Gpp(NH)p and GTP on hormone binding and enzyme activation gave the first indication that the guanine nucleotide sites responsible for these effects on the receptor and enzyme may be functionally distinct.
Effects of Chelators on Actions of Gpp (NH)p - Spiegel et al. (11) have recently shown with the turkey erythrocyte adenylate cyclase system that EDTA inhibits the ability of Gpp(NH)p to activate adenylate cyclase during pretreatment. We have found with the hepatie adenylate cyclase system that this apparent inhibitory effect of chelators (EGTA or EDTA) results from a shift in the concentration of Gpp(NH)p required for activation during pretreatment. This is illustrated with EGTA in Fig. 38 where an approximately 7-fold shift to the right was observed in the presence of chelator. In contrast, when the effects of Gpp(NH)p on 'Y-glucagon binding to its receptor were assayed in the presence of chelators, essentially no change was observed for the potency of Gpp(NH)p action on hormone binding to the receptor. It should be emphasized that the effects of Gpp(NH)p on activity and binding were assayed in the presence of 30 pM App(NH)p to minimize the hydrolysis of Gpp(NH)p by nucleotide pyrophosphohydrolases present in the hepatic plasma membrane (12). If the effects of the chelators were also simply to protect against enzymatic degradation of the nucleotide, one would have expected a marked leftward * In these studies we were concerned only with the concentration of Gpp(NH)p needed for one-half maximal response. The relation of the true K., to this value may be quite complex due to considerations such as the irreversible nature of the activation by this nucleotide. shift in potency of Gpp(NH)p on both processes in contrast to the results we obtained.
As discussed by others, the effects of chelators on the adenylate cyclase system are complex (5, 13 duced by Gpp(NH)p is reversible. These findings can be explained by tight binding of Gpp(NH)p to the regulatory site associated with adenylate cyclase in contrast to weak, reversible binding of Gpp(NH)p to the regulatory site associated with the receptor. The latter is in accord with the considerably lower potency of Gpp(NH)p on the binding relative to the activation process.
Effects of Phospholipase C -Previous studies on the hepatic adenylate cyclase system showed that treatment of the membranes with phospholipase C from Bacillus cereus resulted in a state of receptor which displayed lower affinity for the hormone and in which the effects of GTP on hormone binding were abolished (15,16). It was of interest therefore, to examine the effects of phospholipase C digestion on the two guanine nucleotide effects. In the experiments described in Fig. 4, '29glucagon was incubated with control or phospholipase Ctreated membranes until equilibrium binding had been attained (about 12 min). This was followed by addition of unlabeled glucagon in the absence or presence of Gpp(NH)p, and then the time course of dissociation of bound labeled hormone was followed. In Panel A of Fig. 4, it can be seen that Gpp(NH)p enhanced the rate of dissociation of the labeled hormone from control membranes incubated with three different concentrations of 12SI-glucagon. Under identical incubation conditions, no effects of Gpp(NH)p were observed on dissocia-3 It should be noted that the lack-of effect of GTP at higher concentrations of the hormone is consistent with the fact that GTP alters the affinity of the receptor for glucagon but not the concentration of receptors that bind the hormone (2, 10).  Table II show that phospholipase C treatment resulted in decreased activities relative to control membranes for basal, The procedure for phospholipase C treatment of membranes was described in the legend to Table II. After this treatment, the membranes were suspended to a protein concentration of 0.2 mg/ml in 20 mM TrislHCl, pH 7.5, containing 0.2% BSA, and incubated in 7-ml aliquots with 1.5, 4.5, or 9 x lo-lo M 'Yglucagon (240,000 cpm/pmol) at 30". After 10 and 12 min (A-A), l- Liver membranes were thawed, resuspended to 2 mg/ml in 10 mM TrislHCl, pH 7.5, containing 1 mM dithiothreitol, and centrifuged at 27,000 x g for 15 min. The pelleted membranes were then suspended in 10 mM Tris/HCl, pH 7.5, containing 1 mM dithiothreitol and 1 mM CaCl, to a protein concentration of 0.5 mglml, incubated at 25" for 5 min with phospholipase C (12 pg/ml of membrane protein), and then centrifuged at 27,000 x g for 15 min. The pelleted membranes were resuspended in 10 rnM Tris/HCl, pH 7.5, containing 1 mM dithiothreitol, and recentrifuged at 27,000 x g for 15 min prior to resuspension in the same buffer to assay enzymatic activity. The concentration of Gpp(NH)p used in each assay was 10e5 M, while the concentration of glucagon was lOme M and that of GTP was 10m5 M. All enzymatic activities are expressed as picomoles per 10 min per mg of protein. GTP-, and Gpp(NH)p-stimulated activities and in complete loss of hormone response. These changes required hydrolysis in excess of 40% of the membrane lipids (data not shown). It can be seen that under those conditions in which a loss of guanine nucleotide action on the receptor took place the fold stimulation of adenylate cyclase activity by either GTP or Gpp(NH)p remained the same as that seen in control membranes. The differential effect of phospholipase digestion on receptor regulation and enzymatic activation by Gpp(NH)p again suggests that functionally distinct sites were involved in these two processes, although we cannot eliminate the possibility that these differential effects may be due, in part, to an alteration in the hormone receptor site.

DISCUSSION
This study provides substantial evidence that guanine nucleotides modify the states of the receptor and adenylate cyclase through functionally distinct sites. In addition to the results presented here, there is other evidence which supports this view and which suggests, furthermore, that these sites reside on structurally different binding components. Such evidence includes previous findings (1, 2) that GDP is equipotent with GTP in glucagon receptor regulation but is a potent competitive inhibitor of the activation process (4). These findings suggest different structural requirements for guanine nucleotides at the two sites. More recently, we have shown that after Lubrol solubilization of the adenylate cyclase complex, the nucleotide-sensitive receptor can be resolved from the nucleotide-activatable catalytic unit. Given evidence for functionally and probably structurally distinct nucleotide sites involved in the regulation of receptor and enzyme activity, what role do these sites play in the overall regulation of enzyme activity and, most importantly, in the so-called "coupling" reaction between receptor and enzyme? De Haen (17) and Cuatrecasas et al. (181 have summarized evidence suggesting that the hormone receptors and adenylate cyclase are separate molecular entities and have proposed that the receptors and enzyme need not be present in stoichiometric amounts in the membrane or even normally juxtaposed within the plane of the membrane. Orly and Schramm (19) have shown recently that catecholamine /3 receptors can be transferred separate from adenylate cyclase and in a functional state between membranes of different cell types. Their findings provide strong evidence for the separateness of catalytic and receptor units. What factors are involved Evidence for Distinct Guanine Nucleotide Sites that promote interaction between these units? It has been proposed (20) that binding of the hormone to a receptor state having a favored coupling conformation is a prime factor in the coupling reaction, and a role for GTP was suggested in this process. We have shown in previous studies (21) that GTP regulates the conformation of the glucagon receptor by changing the receptor state from one having a high affinity (K,, = 2 ml and which displays essentially irreversible binding characteristics to at least two other states of the receptor. Approximately 90% of the receptor sites are converted to low affinity (Kd = 10 to 20 nM), whereas the remaining sites have considerably higher affinity.
Since GTP regulates, through independent sites, both the glucagon receptor and adenylate cyclase, we suggest that the following hypothesis, presented schematically below, may explain the formation of high and low affinity states of the receptor, provides the basis of the coupling reaction, and places in focus the role of GTP and glucagon in the overall regulation of enzyme activity. In the "uncoupled" Equilibrium (I) the hormone receptor (Rl exists in two states. The R,, ("desensitized") form has a high affinity for hormone (H) and low affinity for nucleotide. The converse holds for the R, ("sensitized") form. Similarly, the catalytic unit (E) exists in two forms in the uncoupled Equilibrium (II). E, is inactive and has low affinity for guanine nucleotides, whereas E, is active and has higher affinity for the nucleotides. In the simplest case, only the E2 and R, forms are conformationally compatible for coupling. Such linkage gives rise to the third "coupled" state (III) of the receptor CR,) and catalytic units and is accompanied by an increase in the affinity for the hormone (IV). According to this hypothesis, the bulk of the low affinity receptor states induced by GTP are relevant to the coupling process and can be considered as the favored precursor states in the process.
The advantage of a scheme of uncoupled equilibria is the fact that not only the effective concentrations of hormone and nucleotide but also the effective stoichiometry and configuration of the macromolecular components are important in the overall regulation. Such an hypothesis may explain two of our earlier observations. First, activation of adenylate cyclase by glucagon displays a hyperbolic function of glucagon receptor occupation (41. Second, GTP causes a rightward shift in the glucagon binding curve (2), but a leftward shift in the dose-response curve for hormonal activation of the enzyme (4). The latter can be explained readily by an excess of receptors over the catalytic unit. The hypothesis also predicts that, in the absence of GTP, increasing hormone concentrations would shift the equilibrium to the "uncoupled" high affinity form of the receptor and concomitant shifts toward the E, or inactive state of the enzyme. This may be the basis of the "desensitization" phenomenon that has been reported recently for the catecholamine-sensitive adenylate cyclase system in frog erythrocytes (22).