Choleragen-stimulated Release of Guanyl Nucleotides from Turkey Erythrocyte Membranes*

Choleragen stimulates adenylate cyclase by ADP ri- bosylating a guanyl nucleotide-binding regulatory protein (G/F). p-Adrenergic hormones also activate the adenylate cyclase of turkey erythrocytes, and it is cur-rently believed that they do so in part by decreasing the affinity of G/F factor for GDP, an effect which is manifested by a hormone-stimulated release of guanyl nucleotides from the membranes. Since choleragen might also activate adenylate cyclase by a similar mechanism, the effect of toxin treatment on the release of guanyl nucleotides from turkey erythrocyte membranes was examined. In the presence of NAD, choleragen was found to stimulate release of guanyl nucleotides from mem- branes which had been preloaded with radiolabeled GTP. No stimulation of release was observed with CAMP or when NAD was replaced by NADP, which does not serve as a substrate for choleragen-catalyzed ADP ribosylation. While either isoproterenol or choler- agen can stimulate release of guanyl nucleotides from the membranes, the amount of guanyl nucleotide re- leased in the presence of both isoproterenol and choleragen was no greater than that released by isoproter- enol alone. Furthermore, when membranes were first treated with choleragen and NAD, the subsequent release of guanyl nucleotides induced by isoproterenol was reduced to -15% of that observed with membranes not treated with the toxin. Therefore, choleragen

' The abbreviations used are: G/F, the GTP-binding protein of the adenylate cyclase complex; App(NH)p, adenosine 5'-(p,yimino)triphosphate, Gpp(NH)p, guanosine 5"(/3,y-imino)triphosphate; EGTA, ethylene glycol bis (p-aminoethyl ether)-N, N,N',N'tetraacetic acid. of adenylate cyclase is active (7)(8)(9). Hydrolysis of the bound GTP by a specific GTPase is thought to result in an inactive G/F-GDP complex (10); reactivation of the cyclase is dependent on dissociation of GDP to permit binding of GTP. Choleragen has been reported to inhibit the hydrolysis of G T P and could, thereby, increase the half-life of the active enzyme-GTP complex (11).
Several hormones are believed to activate adenylate cyclase by binding to receptors which interact with G/F and promote the replacement of bound GDP with GTP. Cassel and Selinger showed that isoproterenol stimulated release of tightly bound [3H]GDP from turkey erythrocyte membranes; the rate of release was correlated with the rate of activation of adenylate cyclase (12). As reported here, we have now found that choleragen in the presence of NAD increases the release of bound guanyl nucleotide from turkey erythrocyte membranes and this nucleotide appears to come from the same pool that is released by isoproterenol.

MATERIALS AND METHODS
App(NH)p, pyruvate kinase, phosphoenolpyruvate (sodium salt), and isoproterenol were purchased from Sigma; Gpp(NH)p was from Boehringer Mannheim; choleragen and dithiothreitol were from Schwarz/Mann; propranolol-HC1 was from Ayerst Laboratories; [a-32P]GTP  Ci/mmol) and [3H]GTP (5-15 Ci/mmol) were from New England Nuclear; polyethyleneimine-cellulose thin layer chromatography sheets were from Brinkmann. Protein was measured by the method of Lowry et al. (13) using bovine serum albumin as standard. Just before addition to membranes, choleragen, 1 mg/ml, was activated by incubation for 10 min at 30 "C in 50 mM glycine buffer, pH 8.0, containing 20 mM dithiothreitol.
Turkey erythrocyte membranes were prepared as described previously (14). The procedure used to determine release of bound guanyl nucleotide from membranes was very similar to that of Cassel and Selinger (12). Membranes were washed twice with 10 volumes of 50 mM Tris-HC1, pH 7.5, containing 0.1 mM EGTA, suspended in 50 mM Tris-HCI, pH 7.5 (hereinafter referred to as buffer), and incubated in buffer containing 0.1 M KCl, 6 mM MgC12, 0.1 mM EGTA, 1 mM dithiothreitol, 0.2 mM App(NH)p, 1 mM phosphoenolpyruvate, and pyruvate kinase, 75 units/ml, (total volume, 1.5 m l ) for 2 min at 37 "C (membrane protein concentration, 2-3 mg/ml). Isoproterenol and [CZ-~'P]GTP or r3H]GTP (-5,000 cpm/pmol), each in 30 p1 of HzO, were then added to final concentrations of 20 and 0.3 PM, respectively, followed after 2 min at 37 "C by propranolol and unlabeled GTP (final concentrations, 10 and 100 p~, respectively). Membranes were sedimented by centrifugation (17,000 X g, 5 rnin), washed three times with eight volumes of ice-cold buffer, and suspended in five volumes of ice-cold buffer containing 0.05 M KCl, 3 mM MgCL, 0.05 rn EGTA, 0.5 mM dithiothreitol, 0.1 mM App(NH)p, 0.5 mM GTP, 0.5 mM phosphoenolpyruvate, and pyruvate kinase, 37 units/ml (-1 mg of membrane protein/ml). After incubation at 37 "C for 10 min, membranes were sedimented by centrifugation and then washed with eight volumes of buffer. For all experiments except that shown in Table 11, the basal (unstimulated) release of guanyl nucleotide was decreased by incubating membranes prior to the experimental period for 1 h at 37 "C in the medium described above (-1 mg of protein/ml) and subsequently washing the membranes with eight volumes of buffer.
To quantify release of the bound nucleotide, samples of membranes prepared in this way (-0.2 mg of protein) were added to buffer

Choleragen-stimulated Release of Nucleotides from Membranes
To identify the nucleotide released by choleragen, membranes loaded with [a-32P]GTP and washed as described above were incubated at 37 "C in the medium used for assay of nucleotide release containing 2 m GDP and 2 m M NAD with or without choleragen, 250 pg/ml (-0.5 mg of membrane protein in 1 ml). After 1 h, the samples were cooled to 0 "C and centrifuged at 17,000 X g for 15 min. Samples (100 pl) of supernatants were applied to polyethyleneiminecellulose thin layer sheets. Guanyl nucleotides were separated by chromatography with 1.2 M LiC1. Membranes incubated with choleragen released 3,110 and those incubated without, 2,380 cpm/mg of protein/h. Of the radioactivity recovered from chromatograms of choleragen-treated and control samples, respectively, 13 and 12% comigrated with GTP, 27 and 36% with GMP, and 60 and 52% with GDP. From these data, it was calculated that 85% of the labeled nucleotide released by choleragen was GDP.

RESULTS AND DISCUSSION
Membranes that had bound radiolabeled GTP in the presence of isoproterenol and had then been extensively washed as described under "Materials and Methods" released radiolabeled guanyl nucleotide at a relatively constant rate during incubation at 37 "C in the release assay medium; 40 to 50% of the bound nucleotide was released in 1 h. As reported by Cassel and Selinger (12), addition of isoproterenol produced an immediate increase in the rate of release, which returned to the basal rate in 2 to 3 min. Addition of choleragen and NAD also accelerated release but to a lesser degree (Fig. 1). In this case, the rate of release did not return to the basal rate for almost 90 min.
Addition of choleragen or NAD alone did not alter the release of guanyl nucleotide, whereas, in the experiment shown in Table I, when both were present release was increased from 0.33 to 0.45 pmol/mg of protein/h. When NADP, which does not serve as a substrate for choleragen (15), was substituted for NAD, no stimulation of release was observed (Table I) GTP were prepared and incubated at 37 "C for the indicated times for assay of guanyl nucleotide release as described under "Materials and Methods" with the further addition of 2 m M NAD. In one experiment, samples were incubated with or without 50 p~ isoproterenol and with or without choleragen, 250 pg/ml, in another. The increment in nucleotide released induced by isoproterenol (0) or choleragen (0) is shown. Release in the absence of additions was 0.16 pmol/mg of protein/3 min in the experiment in which isoproterenol-stimulated release was measured and 1.00 pmol/ mg of protein/3 h when choleragen-stimulated release was measured.

TABLE I Effects of chozeragen andpyridine nucleotides on release of guanyl nucleotides from turkey erythrocyte membranes
Membranes were incubated with additions as indicated for 60 min and guanyl nucleotide release was determined as described under "Materials and Methods." At the beginning of the incubation, bound nucleotide was 0.82 pmol/mg of protein.

Choleragen-stimulated Release
of Nucleotides from Membranes ment of turkey erythrocyte membranes with choleragen decreases the time lag normally observed for activation of adenylate cyclase by the poorly hydrolyzable GTP analog, Therefore, choleragen may activate adenylate cyclase, at least in part, by a mechanism similar to that of P-adrenergic hormones; that is, both choleragen and P-adrenergic hormones may decrease the affinity of G/F for guanyl nucleotides, thus increasing the rate of release of GDP and freeing the site for subsequent binding of the activating ligand, GTP. GPPWWP.