THE JOURNAL OF B~WGICAL CHEYISTRY Vol. 251. No. 6. Isme of March 25. pp. 1913~W20. 1979 Pnntcd in U.S.A. Regulation of Adenylate Cyclase of Neuroblastoma x Glioma Hybrid Cells by a-Adrenergic Receptors I. INHIBITION OF ADENYLATE CYCLASE MEDIATED BY a RECEPTORS* (Received for publication, May 23, 1978) Steven L. Sabol and Marshall Nirenberg From the Laboratorv of Biochemical Genetics. National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Ma&an> 2&.X4 (-)-Norepinephrine and other catecholamines inhibit basal and prostaglandin El-stimulated adenylate cy- clase activities by 35 to 60% in homogenates of NGlOS- 15 neuroblastoma x glioma hybrid cells and markedly reduce adenosine 3':5'-monophosphate levels of intact cells, but do not affect guanosine 3':5'-monophosphate levels. The specificity of the NGlOS-15 receptor for ligands is that of an a receptor, possibly a presynaptic a2 receptor. The inhibition of adenylate cyclase by nor- epinephrine is reversed by a receptor antagonists such as dihydroergotamine or phentolamine, but not by the B receptor antagonist propranolol. The effect of nor- epinephrine on adenylate cyclase activity initially is dependent on GTP; half-maximal inhibition of enzyme activity by norepinephrine is obtained with 0.2 pM GTF'. The inhibition of adenylate cyclase activity by norepi- nephrine is reduced by 10 mu NaF and is abolished by 0.05 mu guanyl-5'-yl imidodiphosphate. Inhibitions of NGlOS-15 adenylate cyclase .mediated by a receptors, opiate receptors, and muscarinic acetylcholine recep- tors are not additive; this suggests that the three spe- cies of receptors can be functionally coupled to the same adenylate cyclase molecules or molecules regu- lating the enzyme. Cellular responses to the endogenous catecholamines nor- epinephrine and epinephrine are mediated by specific receptor molecules, which have been divided into two major classes, a receptors and p receptors, as well as into subclasses, according to the potencies of ligands for the receptors (reviewed in Ref. 1). Typical a receptor responses exhibit the specificity, in order of decreasing potency [epinephrine 2 norepinephrine >> isoproterenol], and are blocked by phentolamine, phen- oxybenxamine, or dihydroergotamine. Typical j3 receptor re- sponses exhibit the specificity [isoproterenol > epinephrine 2 norepinephrine] and are blocked by propranolol. Whereas activation of p receptors usually results in an increase in the activity of adenylate cyclase (EC 4.6.1.1; ATP pyrophosphate-lyase (cyclixing)) and, therefore, an increase in the concentration of adenosine 3':5'-monophosphate (2), the relationship between a receptor activation and adenylate cy- clase is less clearly defined. In certain cells, a receptor acti- vation reduces the intracellular CAMP concentration or re- duces the magnitude of responses elicited by compounds which elevate the CAMP concentration (3-9); thus, Robison et al. (2) proposed that a and j3 receptors may be linked to * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. adenylate cyclase in an opposing manner. However, in cerebral cortical slices, a receptor activation is associated with an increase in CAMP (10-12). In addition, a receptor activation in certain tissues may increase permeability to Ca'+ ions (13- 15) or Cl- ions (16) and/or increase cellular cGMP (13, 14) independently of an alteration of CAMP levels. Clonal NG108-15 hybrid cells,' obtained by fusion of mouse neuroblastoma NlSTG-2 clone (17) with rat glioma clone CGBU-1 (18), possess adrenergic receptors which, in concert with an activating ligand, reduce the PGEl'-dependent in- crease in cellular CAMP (19). The cells also possess PGE, and adenosine receptors that are coupled to the activation of adenylate cyclase and opiate and muscarinic acetylcholine receptors coupled to the inhibition of adenylate cyclase (20- 23). In this report, we show that a receptors are coupled to the inhibition of adenylate cyclase in NG108-15 homogenates and characterize the inhibition. In the accompanying report (24), we show that prolonged a receptor-mediated inhibition of adenylate cyclase in intact NG108-15 cells results in a long lived increase in adenylate cyclase activity. Some of these results have been presented in preliminary form (25). After this work was completed, a receptor-mediated inhibition of adenylate cyclase in human platelet lysates was reported (26). EXPERIMENTAL PROCEDURES Growth of Cells and Preparation of Homogenates-NGIOS-15 cells (subculture 16-22) were grown in Falcon flasks (75 cm' surface area) or Petri dishes (lOO-mm outside diameter, 64 cm' surface area) in 90% DME medium (Grand Island Biological Co., Catalogue No. H- 21) containing 44 rnM NaHC03 and 10% fetal bovine serum (Colorado Serum Co.) supplemented with 0.1 mM hypoxanthine, 1 pM amino- pterin, and 16 PM thymidine in a humidified atmosphere of 90% air, 10% CO, at 365oC. As cultures approached confluency, the medium was changed once or twice daily to maintain the pH between 7.2 and 7.4. Mycoplasma were not detected in the cells or culture medium. For preparation of homogenates, cells from confluent cultures (approximately 15 mg of protein/flask) were harvested 4 to 6 h after the medium was replaced with fresh medium. Cells were washed twice with D2 saline solution (0.17 mru Na2HP04, pH 7.4, 150 mM NaCl, 5.4 mu KCl, 25 mM D-ghCOSe, and 0.2 mM CaClz) and were dissociated with Dl saline solution (D2 solution without CaCl*, pH 6.7) and washed twice with Dl by centrifugation at 250 x g for 5 mm at 25oC. The final washed cell pellet was suspended in 290 mM ' T. Amano, B. Hamprecht, and M. Nirenberg, manuscript in preparation. ' The abbreviations used are: PGE,, prostaglandin E,; PGFP, pros- taglandin Fzo; DME medium, Dulbecco-Vogt modification of Eagle's minimal essential medium; Gpp(NH)p, guanyl-5'-yl imidodiphos- nhate: Ro20-1724. 4-(3-butoxv-4-methoxybenzyl)-2-imidasolidinone; EGTA, ethylene glycol bis(@&inoethyl ether)iv,W-tetraacetic acid; WB-4101, 2-([2',6'-dimethoxy]phenoxyethylamino)methylbenzodi- oxan; Hepes, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; car- bachol, carbamylcholine chloride. 1913 1914 Regulation of Adenylate Cyclase by a-Adrenergic Receptors sucrose, 25 mu Tris-HCl, pH 7.5 (4 to 6 mg of protein/ml), and l-ml aliquots were frozen in dry ice and stored at -190oC. Immediately before use analiquot was thawed and the suspension was homoge- nized at 0"C m a ground glass tube with 25 strokes of a Teflon pestle rotating at 1200 rpm. The resultant homogenate was found by phase contrast microscopy to have a ratio of intact cells to nuclei of 0.01 or less. Assays for adenylate cyclase activity were initiated within 5 min after homogenates were prepared. NG10815 particulate fractions were prepared by centrifugation of homogenates at 30,000 x g or 130,000 x g, as indicated in the figure legends, for 15 min. The pellets were washed by dispersion and centrifugation in 290 mM sucrose, 25 mru Tris-HCl, pH 7.5. The final pellets were dispersed in the same buffer (2 mg of protein/ml) and stored at -190oC. These procedures were performed rapidly because a receptor-mediated inhibition of adenylate cyclase was found to decay significantly upon storage of homogenates at O'C for more than 30 min: Protein was measured by the method of Lowry et al. (27) using bovine serum albumin as a standard. Assay of Adenylate Cyclase Activity-Adenylate cyclase activity was routinely determined in lO@pl reaction mixtures containing 30 or 35 mu Tris-HCl (pH 7.5), 5 mM magnesium acetate, 58 or 116 rmu sucrose, 0.25 mM Ro20-1724 (a phosphodiesterase inhibitor), 1 mru [a-32P]ATP (3 to 6 x 10' cpm), 1 mM CAMP, 0.1 mru pargyline hydrochloride to inhibit monoamine oxidase, 20 mM creatine phos- phate, 10 units (67 to 80 pg of protein) of creatine phosphokinaae, 10 l.tM PGE, where indicated, 0.48% ethanol (used as a solvent for PGEl `and Ro20-1724), and unfractionated NG108-15 homogenate protein as indicated (usually 75 to 110 pg of protein/reaction mixture). Unless otherwise indicated, for assays involving catecholamines, reaction mixtures also contained 0.01 mM sodium ascorbate (see preparation of catecholamine stock solutions below). In most experiments sodium ascorbate had little or no effect on adenylate cyclase activity or on the potency of catecholamines as inhibitors of adenylate cyclase. However, for some homogenates prepared during the later stages of the work, ascorbate was found to activate adenylate cyclase. In these cases, indicated in the text, ascorbate was omitted. Reactions were started by addition of homogenate to prewarmed (2 min, 37oC) reaction mixtures. Unless otherwise stated, incubations were at 37oC for 10 mm for basal activity and 5 min for PGE,- stimulated activity. Reactions were stopped by the addition of 0.8 ml of 6.25% trichloroacetic acid at 4oC and 100 pl of [3H]cAMP (10,000 cpm, 39.8 Ci/mmol), and CAMP was purified according to method C of Salomon et al. (28) with a recovery of [3H]cAMP of 70 to 85%. Greater than 90% of the 32P product recovered was shown previously to be CAMP (20). Duplicates usually differed by less than 3%. Less than 0.8% of the 1 mu CAMP in the reaction mixture was degraded by endogenous phosphodiesterase activity in a typical reaction mix- ture incubated for 10 min. Determination of IC, and Apparent Hill Coefficient Values- Adenylate cyclase was determined in the presence of 8 to 12 concen- trations of each adrenergic compound tested. The data were fitted to a four-parameter logistic model (29) using the equation: y= A-D +D 1+ ww where A is the activity in the absence of ligand C is the concentration of ligand resulting in half-maximal inhibition (I&), D is the maxi- malIy inhibited activity at saturating ligand concentration, e.g. at least 0.1 mu (-)-norepinephrine, B is the exponent which is related to the steepness of the dose-response curve and here is termed "apparent Hill coefficient," X is the ligand concentration, and Y is the adenylate cyclase activity. The best values of B and C for each concentration curve with given A and D values were determined by iterative nonlinear least squares analysis using the MLAB program of the PDP-10 computer of the National Institutes of Health Computer Center. The IC, values obtained in this manner agreed well with those obtained by log-probit analysis as well as by inspection of the concentration curves. Assay of CAMP of Intact Cells-Culture medium was removed and 3 ml of cold 5% trichloroacetic acid were added to the cell monolayers. Ceils were washed twice with 5% trichloroacetic acid (1 ml/wash). The trichloroacetic acid extracts and washes were pooled and centri- fuged at 30,000 x p for 20 min. and CAMP was purified (20) and assayed by the method of Gilman (30) in 0.2~ml reaction mixtures each containing 1 pmol of [3H]cAMP and 0.7 pg of partially purified CAMP-dependent protein kinase protein, which bound 0.3 pmol of ["HIcAMP under the conditions used. Cell protein was determined (27) from the NaOH-solubihzed trichloroacetic acid precipitates. Assay of cGMP of Intact Cells-Confluent cultures of NG108-15 cells in 100~mm Petri dishes were incubated in air for 30 min at 37'C in DME medium containing 25 mM Hepes buffer instead of NaHCOs, adjusted to pH 7.4, and adjusted to 340 mosm/kg with NaCl, and supplemented with 0.1 mhr hypoxanthine, 1 pM aminopterin, 16 pM thymidine, and 0.5 mM 3-isobutyl-1-methylxanthine (a phosphodies- terase inhibitor). Test compounds were then added and dishes were incubated at 37oC for 15 mm. The medium was discarded and cells were suspended in 5 ml of 5% trichloroacetic acid. Intracellular cGMP was purified from the trichloroacetic acid supematants and assayed by radioimmunoassay according to the method of Matsuzawa and Nirenberg (31). The trichloroacetic acid precipitates were assayed for protein (27). Chemicals-The following were kind donations: prostaglandins, Dr. J. E. Pike of Upjohn; Ro20-1724, Dr. H. Sheppard of Hoffmann- La Roche; methoxamine hydrochloride, Burroughs-Wellcome; oxy- metazoline hydrochloride, -Schering; phentolamine hydrochloride, CIBA-GEIGY: naloxone hvdrochloride. Endo: 9.10-dihvdro-a-erso- cryptine, Sandoz; flupher&ine hydrochloride, Squibb; andWB-4101, Ward-Blenkinsop Pharmaceuticals. R-(-)-Norepinephrine hydrochloride, R-(-)-epinephrine bitar- trate, (-)-phenylephrine hydrochloride, (-)-isoproterenol hydrochlo- ride, dopamine hydrochloride, (-)-dopa hydrochloride, 9,10-dihydro- ergotamine, yohimbine hydrochloride, phenoxybenzamine hydrochlo- ride, (&)-propranolol hydrochloride, atropine sulfate, carbamylcho- line chloride, adenosine 5'-triphosphate (prepared from equine muscle or by phosphorylation of adenosine), CAMP-dependent protein kinase (beef heart), and creatine phosphokinase (rabbit muscle) were from Sigma. Pargyline hydrochloride, (+)-dihydroxymandelic acid, and a- methyl-(f)-norepinephrine were from Regis. Other compounds were from the following sources: (+)-norepinephrine bitartrate, Adams; bulbocapnine, K and K; morphine sulfate, Merck; guanosine 5'-tri- phosphate, P-L Laboratories; guanyl-5'-yl imidodiphosphate, ICN; 3- isobutvl-l-methvlxanthine. Aldrich: fa-32P1ATP and TG-"HlcAMP. New England Nuclear. &her compounds were of reagent grade purity. Solutions of adrenergic compounds usually were prepared imme- diately,before use and were stored at -25oC or on ice. Catecholamines were dissolved either in 0.1 mru sodium ascorbate or in 1 mM HCl at 0oC. Stock solutions of dihydroergotamine and dihydroergocryptine were prepared by titrating the free base with HCl (final pH 6.0); these compounds were less soluble at higher pH. RESULTS Effects of Norepinephrine on CAMP and cGMP Levels of Cells and on Adenylate Cyclase Activity-The effects of (-)-norepinephrine and PGEl on intracellular CAMP concen- trations of NG108-15 cells, in the presence or absence of the phosphodiesterase inhibitor Ro20-1724, are shown in Table I (Experiment 1). In the presence of Ro20-1724, 10 pM norepi- nephrine reduced basal and PGE,-stimulated CAMP accu- mulation by celIs to 25%. Cyclic AMP concentrations of cells were low in the absence of the phosphodiesterase inhibitor, and norepinephrine reduced basal cellular CAMP only slightly, if at all, but inhibited PGEi-dependent CAMP accumulation by 93%. These results agree well with those reported previ- ously (19, 32). Substitution of 0.2 mu EGTA for Ca2' ions in the medium did not significantly reduce the a receptor-me- diated decrease in CAMP accumulation; this suggests that the a receptor-mediated response is not dependent on extracellu- lar Ca2+ ions.3 The effects of (-) norepinephrine or PGF2, on intracellular cGMP concentrations of NG108-15 cells are shown in Table I (Experiment 2). Exposure to norepinephrine did not signifi- cantly affect the cGMP concentration of cells, but did decrease the CAMP concentration during the period examined. Expo- sure to PGF2, for 0.5 min, however, elevated the cGMP concentration 5.6-fold as previously found.4 3 R. McGee and M. Nirenberg, unpublished results. ' H. Matsuzawa and M. Nirenberg, manuscript in preparation. Regulation of Adenylate Cyclase by a-Adrenergic Receptors As shown in Fig. 1, 10 PM norepinephrine inhibited basal and PGE,-stimulated adenylate cyclase activities in NGlOS- 15 homogenates 60% and 48% respectively. Reaction rates were linear during the 16min period examined. In other experiments (not shown), norepinephrine inhibited adenylate cyclase for at least 60 min; thus, the effect of norepinephrine does not desensitize rapidly under the conditions used, The rates of basal and PGE,-stimulated CAMP synthesis were proportional to the amount of homogenate protein added in 1915 another; however, similar values were obtained with sepa- rately homogenized portions of the same batch of cells. Receptor Specificity for Adrenergic Ligands-The effects of different ligands and of ligand concentration on basal and PGEL-stimulated adenylate cyclase activities are shown in Fig. 2. Compounds known to activate (Y receptors, such as (-)- norepinephrine, (-)-epinephrine, and dopamine, inhibited basal and PGEL-stimulated adenylate cyclase partially; the maximum inhibitions found were approximately 50 and 35% respectively. The concentrations of these and other ligands required for half-maximal inhibition of adenylate cyclase (It&) are listed in Table II. Clonidine was the most potent inhibitor tested with an I&, of 0.1 pM, but the extent of inhibition of basal activity (24%) was less than that of other inhibitors. Clonidine also partially antagonized the inhibition by norepinephrine (not shown), which suggests that clonidine acts as a mixed agonist-antagonist. The I&,, values for (Y- methyl-(f)-norepinephrine, (-)-norepinephrine, and (-)-epi- nephrine were 0.2, 0.4, and 0.5 pM, respectively, for basal activity, while dopamine and (-)-isoproterenol, a selective fi receptor activator, were less potent inhibitors with IC, values of 5 and 60 w, respectively. (+)-Norepinephrine was 75-fold less potent an inhibitor than (-)-norepinephrine; this indi- cates that the inhibition of adenylate cyclase is dependent on a stereospecific interaction. (-)-Phenylephrine, methoxamine, and oxymetazoline, which are potent Q receptor activators in other systems, were relatively weak inhibitors of adenylate cyclase (I& 9 to 80 PM). Serotonin, an OL receptor activator in some systems, did not reduce intracellular CAMP levels4 or inhibit adenylate cyclase. A precursor of norepinephrine (dopa) and a metabolite (3,4dihydroxymandelic acid) did not affect adenylate cyclase activity. Thus, the receptor exhibits the specificity of an a: receptor. The potency order [cr-meth- the range 25 to 150 pg of protein/reaction mixture (not shown). The extent of maximum inhibition of adenylate cyclase by norepinephrine varied somewhat from one batch of cells to TABLE I Effects of norepinephrine on intracellular CAMP and cGMP of NG108-15 cells Experiment 1: duplicate Petri dishes (60~mm diameter), each with 3 mg of cell protein, were incubated in an atmosphere of 10% COz, 90% air with 5 ml of growth medium/dish without serum, supple- mented with 0.1 mu ascorbic acid and 0.1 mM pargyline, with or without 250 PM Ro20-1724, for 20 min at 37% Ten micromolar PGE, and/or 10 PM (-)-norepinephrine was then added where indicated. Dishes were incubated for an additional 10 min at 37oC then intra- cellular CAMP was determined. Experiment 2: triplicate Petri dishes (100~mm diameter), each with 10 mg of cell protein, were incubated as described under "Experimental Procedures." Then the following compounds were added where indicated: 1 PM HCl (control), 10 FM (-krorepinephrine and 1 pM HCl, or 10 pM PGFh. Cultures were incubated at 37oC for the times indicated, then intracellular cGMP and, where indicated, CAMP were determined. Experiment No. and Additions Picomoles Picomoles Min- cGMP/mg UtC?S cAMP/mg protein f SE. pro;; * 1. CAMP: With Ro20-1724 None 10 160 + 4 Norepinephrine 10 38 f 10 PGE, 10 4100 + 490 PGE, + norepinephrine 10 1040 f 57 CAMP: Without phosphodiesterase inhibitor None 10 12 + 3 Norepinephrine 10 10 + 1 PGE, 10 79Ort67 PGE, + norepinephrine 10 52 + 11 2. cGMP Control 0 Control 0.5 Control 2.0 27 zt 2 Norepinephrine 0.25 Norepinephrine 0.5 Norepinephrine 1.0 Norenineohrine 2.0 16 f 1 0.57 + 0.12 0.53 f 0.09 0.47 + 0.03 0.51 +- 0.04 0.55 f 0.15 0.58 f 0.16 0.36 f 0.04 PGF2, - 0.5 p I PGE, 3.0 * 0.20 3500 -2000 r . 55 - 1500 4 0 - lOOcg d - 500 5 a FIG. 1. Effect of norepinephrine (NE) (10 pM) on the rate of ["`PIcAMP formation in an NG108-15 homogenate. Aliquots (50 ~1) were withdrawn at the indicated times from a 690~~1 reaction mixture containing 735 pg of homogenate protein. A, basal rate; B. rate in the presence of 10 PM PGE,. -1 is] FIG. 2. Inhibition of (A) basal and (B) PGE,-stimulated adenylate cyclase activity by adrenergic compounds. Each reaction contained 100 pg of homogenate protein and one of the following compounds at the concentrations indicated: (-)-norepinephrine ((-)-NE), (-)-epi- nephrine bitartrate @PI, bitartrate concentration adjusted with so- dium salt to 0.1 mM in all tubes), dopamine (DA), (-)-phenylephrine (PE), (+)-norepinephrine ((+)-NE), and (-)-isoproterenol (ISO). One hundred per cent corresponds to the following specific activities (picomoles of [32P]cAMP/min/mg of protein) for the compounds listed: (A) 11.1, 10.6, 10.4, 10.3, 9.0, 10.5, respectively; and (RI 142, 167, 144, 191, 167, and 155, respectively. 1916 Regulation of Adenylate Cyclase by a-Adrenergic Receptors TABLE II Concentrations of adrenergic compounds required for inhibition of NGIOB-15 adenylate cyclase activity Adenylate cyclase activity of homogenates (average 105 ng of protein/reaction mixture) was assayed with or without 10 PM PGE, in the presence of 10 to 12 concentrations of each compound in duplicate reaction mixtures. Half-maximal inhibitions (I& -L SE.) were found from the resultant concentration curves as described under "Experi- mental Procedures." Numbers in parentheses refer to the number of experiments averaged. Otherwise, values listed refer to the most reliable experiment. IC%l Compound Basal PCE, PM Clonidine 0.1" a-Methyl-(i)-norepineph- 0.2 0.1 rine (-)-Norepinephrine 0.4 + 0.1 (9) 0.6 f 0.2 (2) (-)-Epinephrine 0.5 f 0.03 (3) 0.3 Dopamine 5 20 Oxymetaxoline 9 (-)-Phenylephrine 20 30 (+)-Norepinephrine 30 160 (-)-Isoproterenol 60 30 (f)-Methoxamine 80 70 (-)-Dopa >100* (-)-3,4-Dihydroxyman- >lOOb delic acid Serotonin >lOOh >lOCP " Maximum inhibition less than that for catecholamines. h Little or no effect at 100 PM, the highest concentration tested. ylnorepinephrine > norepinephrine >> phenylephrine > meth- oxamine] suggests that NGlOB-15 o receptors resemble pre- synaptic os receptors more than postsynaptic o1 receptors of smooth muscle (33-35). Average apparent Hill coefficients obtained from the de- pendence of inhibition of basal and PGE]-stimulated adenyl- ate cyclaae on ligand concentration were as follows: (-)-nor- epinepbrine, 0.71 -t 0.04 (n. = 9) and 0.91 f 0.16 (n = 2), respectively; and (-)-epinephrine, 0.72 -+ 0.11 (n = 3) and 0.67 -+ 0.16 (n = 2), respectively. The average apparent Hill coef- ficient for other ligands was 0.8. These results suggest either heterogeneity of receptors or adenylate cyclase or negative cooperativity in ligand-receptor interactions and/or in the functional coupling of the [ligand-receptor] complex with adenylate cyclase. Therefore, the ICM values (Table I) do not necessarily approximate dissociation constants for the [ligand . receptor] complexes. The inhibition of adenylate cyclase by norepinephrine was blocked by the reversible OL receptor antagonists dihydroer- gotamine and phentolamine as well as the irreversible cr receptor antagonist phenoxybenzamine, but not by the p receptor antagonist propranolol (Fig. 3). The (Y antagonist property of phentolamine and phenoxybenzamine could be demonstrated only in the presence of 10 to 50 pM naloxone, a specific opiate receptor antagonist. This is because phentol- amine or phenoxybenzamine also were weak activators of the opiate receptor and thereby inhibited NG108-15 adenylate cyclase (half-maximal inhibition at 2 pM or 5 pM, respectively). These compounds are known to interact with rat brain opiate receptors (36). The apparent dissociation constants (Kc,,,) of various re- ceptor antagonists in reversing norepinephrine-dependent in- hibition of adenylate cyclase are shown in Table III. The specificity of the receptor for antagonists is consistent with that of an a receptor, with dihydroergocryptine, diiydroer- gotamine, and yohimbine the most potent tested (Khpp 0.005 to 0.07 w). The Knapp of phentolamine (0.2 PM) was higher *4 , : II. &, 2 EJIHYDROERG~TAMINE P F. F~TOPRANOLOL ' `--F-F-e -& *5Nmo4N& rvmLTARITY FIG. 3. Effect of adrenergic receptor antagonists on the norepi- nephrine-dependent inhibition of basal NG108-15 adenylate cyclase. Reaction mixtures without (0) or with (0) 10 PM (-)-norepinephrine (NE) contained 108 gg of homogenate protein and receptor antagonist at concentrations indicated. TABLE III Apparent dissociation constants of receptor antagonists in blocking norepinephrine-dependent inhibition of adenylate cyclase activity Basal adenylate cyclase activity was assayed with or without 10 PM (-)-norepinephrine (NE) in the presence of different concentra- tions of antagonist. The concentrations (EC!,) resulting in half-max- imal reversal of norepinephrine inhibition were used to calculate the apparent dissociation constants Km,, of the antagonist from the equation (37) Khpp = EC&(1 + [NE]/Ks&, where K,,, is the apparent dissociation constant for norepinephrine-a receptor binding, assumed here to be 0.4 PM (Table I). Compound KlhPD PM 9,10-Dihydro-a-ergocryptine" 0.605 9,10-Dihydroergotamineb 0.01 Yohiibine 0.07 Phentolamine' 0.2 Phenoxybenxamine'sd 0.2 WB-4101 0.2 Fluphenaxine 0.2 (f)-Propranolol ZlOO Bulbocapnine >loo' Atropine >lOO' Naloxone >lOO " Mixed agonist-antagonist effect noted. b Mixed agonist-antagonist effect noted in some but not all experi- ments `Naloxone (50 PM) added to reaction mixtures to block opiate receptor activation. ' Irreversible antagonist. ' Little or no effect at 100 pM, the highest concentration tested. than that found in other systems (about 0.01 w), but is identical with that found for the reversal of cy-adrenergic inhibition of platelet adenylate cyclase (38). Fluphenazine, a potent antagonist of dopamine receptors in the central nerv- ous system (39), blocked NGlOB-15 receptors with a Khpp of 0.2 JLM, which is consistent with the (Y antagonist property of Regulation of Adenylate Cyclase by a-Adrenergic Receptors 1917 phenothiazines. The receptor antagonists propranolol (p re- ceptors), bulbocapnine (dopamine receptors), atropine (mus- carinic acetylcholine receptors), and naloxone (opiate recep- tors) had little or no effect on norepinephrine-dependent inhibition of adenylate cyclase at concentrations of 100 PM or less. Responses of Parent Cell Lines of the Hybrid Line NGI08- 15-As shown in Table IV, adenylate cyclase activity in ho- mogenates of CGBU-1 rat glioma cells was stimulated by norepinephrine; however, when the p receptors of these cells were blocked by propranolol, norepinephrine had no effect on adenylate cyclase activity. However, adenylate cyclase of N18TG-2 mouse neuroblastoma cells was inhibited by nor- epinephrine, but slightly less than that of the hybrid cells. This suggests that expression of a receptors in the hybrid cells is a property derived from the neuroblastoma parent. Effect of GTP-Hormonal stimulation of adenylate cyclase has been shown to require low concentrations of GTP (see Ref. 40 for review). To determine whether (Y receptor-me- diated inhibition of NG108-15 basal adenylate cyclase requires GTP, a washed particulate fraction was assayed with or with- out GTP in a system in which the ATP concentration was reduced to 0.1 mM and ATP synthesized by phosphorylation of adenosine was used to reduce the level of guanine nucleotide contaminants. In the absence of added GTP (Fig. 4A), nor- epinephrine did not inhibit the initial rate of CAMP synthesis between 0 and 4 min, but norepinephrine-dependent inhibi- tion slowly appeared during further incubation. However, in the presence of 1 pM GTP (Panel B), the lag in the initial rate was abolished and norepinephrine inhibited adenylate cyclase activity approximately 45% at each time tested between 2 and 15 min. The relationship between GTP concentration and the reaction rates between 0 and 4 min and between 4 and 8 min are shown in Panels C and D, respectively. Between 0 and 4 min, three effects of GTP on adenylate cyclase can be seen: (a) stimulation of the initial rate (half-maximal stimulation at 5 x lo-" M GTP); (6) enhancement of norepinephrine-depend- ent inhibition from 6% to 44% (half-maximal effect at 3 x lo-' M GTP); and (c) inhibition of activity in the absence or presence of norepinephrine (>lO-" M GTP). Between 4 and 8 min, GTP did not stimulate the rate of CAMP synthesis but increased norepinephrine-dependent inhibition from 17% to 46% (half-maximal effect at 3 x lo-' M GTP) and inhibited activity above lo-" M GTP. These results indicate that inhi- bition of adenylate cyclase by norepinephrine is dependent upon GTP. TABLE IV Effect of norepinephrine on adenylate cyclase activity of neuroblastoma x glioma hybrid NG108-15 cells and parental cell lines Adenylate cyclase reaction mixtures contained, where indicated, homogenate protein from CGBU-1 cells (subculture 24, 170 pg), NlSTG-`2 (subculture 12, 146 pg), or NGlOS-15 cells (190 pg) and 10 PM (-)-norepinephrine, 10 j.4~ PGE,, or 20 11sr (fhxopranoloL as indicated. The activities listed are means of duplicate determinations. Parental cells Additions Neuroblas- Rat glioma Mouse neu- toma x CGBU-1 roblastoma glioma hy- N18TG-2 brid NG108. 15 None Norepinephrine PGE, PGE, + norepinephrine Propranolol Propranolol + norepi- nephrine pm01 CAMPlminlmgprotein 43 5.5 7.7 113 4.1 4.6 48 97 128 116 76 94 43 43 FIG. 4. Effect of GTP on inhibition of adenylate cyclase by nor- epinepbrine. Reaction mixtures (400 ~1) contained 0.1 mM [a-"`P]ATP (unlabeled ATP prepared by phosphorylation of adenosine), 129 c~% of protein of NG108-15 particulate fraction (30,000 x g, 15 mm), and, as indicated, 100 pM norepinephrine (NE) and 0 to 10 PM GTP. A and B, adenylate cyclase activity in the absence or presence of 1 F GTP, respectively, as a function of time. C and D, average specific activities between 0 and 4.17 min or 4.17 and 8.17 min, respectively, in the absence or presence of 100 pM norepinephrine and per cent inhibition by norepinephrine (dashed lines). Effects of Sodium Fluoride and Gpp(NH)p-Sodium fluo- ride (reviewed in Ref. 41) and guanyl-5'-yl imidodiphosphate (42) activate adenylate cyclase from many sources, but the fully activated enzyme then appears to be unresponsive to hormones which usually increase its activity. NaF or Gpp(NH)p activate NGlOS-15 adenylate cyclase and concom- itantly reduce or abolish the responsiveness of the enzyme to inhibition by opiates or activation by PGE, (20,21). As shown in Fig. 5, 10 mM NaF or 50 pM Gpp(NH)p increased activity of NG108-15 adenylate cyclase to 3.4 or 2.8 times, respectively, the basal activity. Whereas 10 pM norepinephrine decreased the basal activity by 41%, norepinephrine reduced the activity of the fluoride-stimulated enzyme by only 11% and did not reduce the activity of the Gpp(NH)p-stimulated enzyme. These results show that NaF and Gpp(NH)p reduce the effectiveness of norepinephrine as an inhibitor of adenylate cyclase and suggest that these compounds affect either inter- action of the [norepinephrine- a receptor] complex with ade- nylate cyclase and/or the binding of norepinephrine to the receptor. Nonadditiuity of Adenylate Cyclase Inhibition Mediated by Three Species of Receptors-a Receptors, opiate receptors, and muscarinic acetylcholine receptors of NG108-15 cells each mediate only a partial inhibition of adenylate cyclase. Thus, the question arises as to whether each species of receptor can become functionally coupled to the same population or to distinct populations of adenylate cyclase molecules. To inves- tigate this possibility, enzyme activity was determined in the presence of activators of each species of receptor (norepineph- rine, morphine, and carbachol, respectively), separately, and in various combinations (Table V). Each ligand was tested at a concentration which resulted in at least 90% of the maximum inhibition obtained with a saturating concentration of ligand as determined by the dependence of inhibition on the concen- tration of norepinephrine (Fig. 2), morphine (20), or car- bachol.5 The maximum inhibitions by norepinephrine or mor- ' S. K. Sharma and M. Nirenberg, manuscript in preparation. 1918 Regulation of Adenylate Cyclase by a-Adrenergic Receptors 15 20 25 30 MINUTES Fro. 5. Effect of 50 pM guanyl-5'-yl imidodiphosphate and 10 mM sodium fluoride on adenylate cyclase inhibition by 10 pM (-)-norepi- nephrine (NE). Each reaction mixture (700 pl) contained 875 gg of NGlO&15 homogenate protein. Aliquota of 50 ~1 were withdrawn at indicated times. Symbols and specific activities (picomoles of CAMP/ min/mg of protein) calculated from linear portions of time courses are as follows: none (O), 10.0; norepinephrine (O), 5.9; NaF (m), 34.2; NaF plus norepinephrine (Cl), 30.4; Gpp(NH)p (A), 28.1; and Gpp(NH)p plus norepinephrine (A), 28.1. phine are similar and exceed that of carbachol alone. The extent of inhibition by norepinephrine or morphine alone (about 46%) was similar to that obtained with combinations of two or three Iigands. To show that the three ligands employed activate different species of receptors, each ligand was tested in the presence of dihydroergotamine, naloxone, or atropine to block a receptors, opiate receptors, or muscarinic receptors, respectively. Inhibition of adenylate cyclase by each receptor activator was prevented only by the antagonist spe- cific for the appropriate receptor. It should be noted that the maximum inhibitions of PGEr- stimulated CAMP accumulation in intact NG108-15 cells by norepinephrine, morphine, or acetylcholine alone are in the range 75 to 90% (Refs. 19, 22, and 23, Table I), considerably greater than the maximum inhibitions of adenylate cyclase found in homogenates. These results suggest that at least three or four species of receptors can be functionally coupled to the same population of adenylate cyclase molecules or molecules regulating the enzyme. DISCUSSION The results demonstrate that the adrenergic receptors of NG108-15 cells exhibit the specificity for ligands characteristic of o receptors. Activation of the (Y receptors results in inhibi- tion of basal and PGE1-stimulated adenylate cyclase. Inhibi- tion by norepinephrine is dependent on GTP and is partially or completely abolished by NaF or Gpp(NH)p. The results suggest that inhibitory (Y receptors and stimulator-y PGEi receptors interact with the same population of adenylate cyclase molecules because the decrease in CAMP formed due to norepinephrine is greater in the presence of PGEl than in the absence of PGEi. Thus, norepinephrine can regulate basal adenylate cyclase activity and the sensitivity and magnitude of enzyme responses to a ligand for another species of receptor that activates the enzyme. The response of the enzyme to an activating ligand may be shifted reversibly from a subsensitive to a supersensitive state by altering the relative concentrations of activating and inhibiting ligands. In addition, the results suggest that three species of NG108-15 receptors that mediate TABLE V Lack of additivity of NG108-I5 adenylate cyclase inhibitions mediated by a, opiate, and muscarinic acetylcholine receptors Basal adenylate cyclase activity was assayed with 116 pg of NGlOB- 15 homogenate protein and where indicated: 10 PM (-)-norepineph- rine (NE), 10 PM morphine, 59 w carbamylcholine chloride (car- bachol), 10 CM dihydroergotamine, 10 w naloxone, or 10 pM The results are means of duplicate determinations. atropine. Additions None NE Morphine Carbachol NE + morphine Morphine + carbachol NE + carbachol NE + morphine + carbachol Dihydroergotamine NE + dihydroergotamine Morphine + dihydroergotamine Carbachol + dihydroergotamine Naloxone NE + naloxone Morphine + naloxone Carbachol + naloxone Atropine NE + atropine Morphine + atropine Carbachol + atropine Dihydroergotamine + naloxone + atropine NE + morphine + carbachol + dihydroergotamine + naloxone + atropine Picomoles cAMP/min/mg s of control protein 9.2 100 4.9 53 5.0 54 6.7 73 4.7 51 5.0 54 5.2 56 5.2 56 8.4 100 7.7 92 5.6 67 6.4 76 9.1 100 5.2 57 9.9 109 7.0 77 8.5 100 5.2 61 4.9 58 9.0 106 9.7 100 8.4 87 inhibition of adenylate cyclase (a receptors, muscarinic ace- tylcholine receptors, and opiate receptors) also interact with the same population of adenylate cyclase molecules and/or regulatory molecules because the inhibitions mediated by the three species of receptors are not additive. At least two, and possibly more, species of (Y receptors have been distinguished on the basis of differences in the relative potencies of ligands. It has been proposed that postsynaptic o1 receptors mediate excitatory responses such as smooth muscle contraction, while presynaptic ci2 receptors of axons that release norepinephrine mediate the feedback inhibition of norepinephrine release at synapses. and perhaps other in- hibitory processes (reviewed in Refs. 33 to 35). The specificity of NGlOS-15 o receptors for ligands resembles that of the presynaptic eq receptor because the potency of cw-methyl- norepinephrine is high and the potencies of methoxamine and phenylephrine are low relative to that. of norepinephrine; in addition, the potency of the antagonist yohimbine is high relative to that of phentolamine (Tables II and III). However, clonidine acts as a mixed agonist-antagonist with respect to NG108-15 adenylate cyclase activity, an effect characteristic of postsynaptic lyl receptors (43). In addition, norepinephrine did not reduce basal or serotonin-stimulated acetylcholine release from NG108-15 cells6 These results show that the a receptors of NG108-15 cells have some properties attributed to presynaptic op receptors and at least one property attrib- uted to postsynaptic LY, receptors. Additional studies, that will be reported elsewhere, show that NG108-15 membranes con- 6 S. Wilson and M. Nirenberg, unpublished results. Regulation of Adenylate Cyclase by a-Adrenergic Receptors 1919 tain saturable sites which bind [3H]dihydroergocryptine with high affinity* and that the 3H-ligand is displaced by ligands that are known to interact selectively with (Y receptors7 Human platelets, the only other cell type in which (Y recep- tor-mediated inhibition of adenylate cyclase has been une- quivocally demonstrated (26, 38,44), possess a receptors that resemble NG108-15 receptors, insofar as, for example, yo- himbine is more potent than phentolamine. It is tempting to speculate that different subclasses of a receptors may mediate different biochemical responses, and that one subclass, possi- bly CQ, may mediate inhibition of adenylate cyclase in a variety of cell types. It would be of interest to determine whether activation of CQ, muscarinic acetylcholine, or opiate receptors of presynaptic nerve terminals results in a reduction of intra- cellular CAMP, and if so, whether decreases in CAMP concen- tration are required for the inhibition of neurotransmitter release elicited by ligands interacting with these receptors (reviewed in Ref. 34). Micromolar concentrations of GTP are required for (Y re- ceptor-mediated inhibition of NG108-15 adenylate cyclase. GTP stimulates the initial rate of adenylate cyclase activity, and norepinephrine `inhibits the GTP-dependent increase in enzyme activity. However, the GTP concentration resulting in maximal stimulation of adenylate cyclase activity (0.2 pM) is less than that required for maximal inhibition by norepi- nephrine (1 PM). Higher concentrations of GTP are inhibitory. These results suggest that GTP has multiple effects on basal adenylate cyclase activity. A GTP requirement for ar receptor- mediated inhibition of human platelet adenylate cyclase has been reported (45), and a similar GTP requirement for opiate receptor-mediated inhibition of NG108-15 adenylate cyclase by morphine has been found." The demonstration that GTP is required for receptor-mediated inhibition of adenylate cy- clase must be considered in the context of models that attempt to describe the role of GTP in receptor-mediated activation of adenylate cyclase (46-49). We suggest that GTP is required for coupling of receptors to adenylate cyclase and that this process may result in either activation or inhibition of the enzyme, depending on the species of receptor. In a number of cell types, an increase in intracellular cGMP accompanies a reduction of the CAMP level (50). a Receptor activation elevates the cGMP level in some tissues (13, 14) but has no effect on the cGMP level of NG108-15 cells (Table I). Thus, a receptor-mediated decreases in the CAMP level are not necessarily associated with alterations in the cGMP level. Clonal NGlOS-15 cells are relatively homogeneous and eas- ily obtained in large quantities. These cells generate action potentials and synthesize, store, and secrete acetylcholine and form synapses with striated muscle cells. The cells, therefore, can be used as a model system for studies on a receptor properties and possible functions, including those that may regulate tram+synaptic communication. In the accompanying report (24), we show that a receptors of NG108-15 cells mediate long term as well as short term effects on adenylate cyclase activity. Acknowledgments-We thank Deborah Carper for assistance with the cGMP determinations, and Drs. Robert Ruffolo and James Ken- imer for helpful discussions. REFERENCES 1. Triggle, D. J., and Triggle, C. R. (1976) Chemical Pharmacology ' S. Ayukawa and M. Nirenberg, manuscript in preparation. ' D. Wilkening, S. Sabol. and M. Nirenberg, submitted for publi- cation. of the Synapse, Chap. 3, pp. 233-430, Academic Press, New York 2. Robison, G. A., Butcher, R. W., and Sutherland, E. W. (1967) Ann. N. Y. Acad. Sci. 139.703-723 3. Turtle, J. R., and Kipnis, D: M. (1967) Biochem. Biophvs. Res. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. Commun. 28,797~802 - I Abe, K., Robison, G. A., Liddle, G. W., Butcher, R. W., Nicholson, W. E., and Baird, C. E. (1969) Endocrinology 85,674-682 Moskowitz, J., Harwood, J. P., Reid, W. D., and Krishna, G. (1971) Biochim. Biophys. Acta 230,279-285 Triner, L., Vulhemoz, Y., Verosky, M., and Nahas, G. G. (1970) Life Sci. 9, 707-712 Kaliner, M. A., Orange, R. P., Koopman, W. J., Austen, K. F., and Laraia, P. J. (1971) B&him. Biophys. Acta 252, X0-164 Field. M.. Sheerin. H. E.. Henderson. A.. and Smith. P. L. (1975) Am. J:Physiol.229,86-92 Handler. J. S.. Bensinzer. R.. and Orloff. J. (1968) Am. J. Phvsiol. 215, 1024-1031 - I ~ Chasin, M., Rivkin, I., Mamrak, F., Samaniego, S. G., and Hess, S. M. (1971) J. Biol. Chem. 246.3037-3041 Perkins, J. P., and Moore, M. M. (1973) J. Pharmacol. Exp. Ther. 185,371-378 Schultz, J., and Daly, J. W. (1973) J. Neurochem. 21,573-579 Schultz, G., and Hardman, J. G. (1975) Adu. Cyclic Nucleotide Res. 5,339-351 Butcher, F. R., Rudich, L., Emler, C., and Nemerovski, M. (1976) Mol. Pharmacol. 12,862-870 Sehnger, Z., Esmerl, S., and Schramm, M. (1974) Proc. Natl. Acad. Sci. 7% S. A. 71, 128-131 Bulbring, E., and Szurszewski, J. H. (1974) Proc. R. Sot. Lond. B Biol. Sci. 185, 225-262 Minna, J., Glazer, D., and Nirenberg, M. (1972) Nature New Biol. 235,225-231 Amano, T., Hamprecht, B., and Kemper, W. (1974) Exp. Cell Res. 85, 399-408 Traber, J., Reiser, G., Fischer, K., and Hamprecht, B. (1975) FEBS Lett. 52. 327-332 Sharma, S. K., Nirenberg, M., and Klee, W. A. (1975) Proc. Natl. Acad. Sci. U. S. A. 72,590-594 Sharma, S. K., Klee, W. A., and Nirenberg, M. (1977) Proc. Natl. Acad. Sci. U. S. A. 74,3365-3369 Traber, J., Fischer, K., Buchen, C., and Hamprecht, B. (1975) Nature 255,558-560 Traber, J., Fischer, K., Latzin, S., and Hamprecht, B. (1975) Nature 253, 120-122 Sabol, S. L., and Nirenberg, M. (1979) J. Biol. Chem. 254,1921- 1926 Sabol, S. L., and Nirenberg, M. (1977) Fed. Proc. 36, 736 Jakobs, K. H., Saur, W., and Schultz, G. (1976) J. Cyclic Nucleo- tide Res. 2,381-392 Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275 Salomon, Y., Londos, C., and Rodbell, M. (1974) Anal. Biochem. 58, 541-548 Rodbard, D., and Frazier, G. R. (1975) Methods Enzymol. 37B, 3-22 Gilman, A. G. (1970) Proc. Natl. Acad. Sci. U. S. A. 67.305-312 Matauzawa, H., and Nirenberg, M. (1975) Proc. N&l. Acad. Sci. U. S. A. 72,3472-3476 Hamprecht, B., and Schultz, J. (1973) Hoppe-Seyler's Z. Physiol. Chem. 354, 1633-1641 Langer, S. Z. (1974) Biochem. Pharmacol. 23, 1793-1806 Starke, K. (1977) Rec. Physiol. Biochem. Pharmacol. 77, 1-124 Berthelsen. S.. and Pettinner. W. A. (1977) Life Sci. 21. 595-606 Cicero, T. b., `Wilcox, C. g., `and Meyer, E. k. (1974) `Biochem. Pharmacol. 23.2349-2352 Cheng, Y.-C., and Prusoff, W. H. (1973) Biochem. Pharmacol. 22, 3099-3108 38. Newman, K. D., Williams, L. T., Bishopric, N. H., and Lefkowitz, R. J. (1978) J. Clin. Znuest. 61, 395402 39. Clement-Cormier. Y. C.. Kebabian. J. W.. Petzold. G. L.. and Greengard, P. (1974) Proc. Natl. Acad. Sci. U. S. A. 71, 1113- 1117 40. Rodbell, M., Lin, M. C.. Salomon, Y., Londos, C., Harwood, J. P., Martin, B. R., RendeIl, M., and Berman, M. (1975) Adu. Cyclic Nucleotide Res. 5.3-29 41. Perkins, J. P. (1973) Adu. Cyclic Nucleotide Res. 3, l-64 42. Londos, C., Salomon. Y., Lin, M. C., Harwood, J. P.. Schramm. 1920 Regulation of Adenylate Cyclase by a-Adrenergic Receptors M., Wolff, J., and Rodbell, M. (1974) Proc. Natl. Acad. Sci. U. 7207-7213 S. A. 71,3087-3090 47. Lad, P. M., Welton, A. F., and Rodbell, M. (1977) J. Biol. Chem. 43. Starke, K., Montel, H., Gayk, W., and Merker, R. (1974) Naunyn- 252, 5942-5946 Schmiedeberg's Arch. Pharmacol. 285, 133-150 48. Levitzki, A. (1977) Biochem. Biophys. Res. Commun. 74, 1154- 44. Kafka, M. S., Tallman, J. F., and Smith, C. C. (1977) Life Sci. 21, 1159 1429-1438 49. Blume. A. J.. and Foster. C. J. (1976) J. Biol. Chem. 251. 3399- 45. Jakobs, K. H., Saur, W., and Schultz, G. (1978) FEBS Lett. 85, 3404' 167-170 50. Goldberg, N. D., O'Dea, R. F., and Haddox, M. K. (1973) Adv. 46. Williams, L. T., and Lefkowitz, R. J. (1977) J. Biol. Chem. 252, Cyclic Nucleotide Res. 3, 155-223