Uptake and metabolism of 3′,5′-cyclic adenosine monophosphate and N,O′-dibutyryl 3′,5′-cyclic adenosine monophosphate in isolated bovine thyroid cells

1.

1. Accumulation of intracellular radioactivity was measured during incubation of isolated bovine thyroid cells with cyclic [³²P]AMP, cyclic [8-³H]AMP and dibutyryl cyclic [8-³H]AMP. With cyclic [³²P]AMP, ³²P cell/medium ratios ranged from 0 to to 0.04 compared to a maximum ³H cell/medium ratio of 0.29 with cyclic [³H]AMP and 0.16 with dibutyryl cyclic [³H]AMP. The excess of intracellular cyclic [³H] over cyclic [³²P]AMP radioactivity was due to extracellular formation of more penetrable dephosphorylated cyclic AMP metabolites which probably served as precursor of intra-cellular cyclic AMP.     

In vivo preparation of (32P) adenosine 3',5'-cyclic monophosphate

A simple method for the preparation of [³²P]adenosine 3′,5′-cyclic monophosphate (cyclic AMP) is described. A culture of Escherichia coli mutant deficient in cyclic AMP receptor protein is incubated with [³²P]orthophosphate of known specific activities (up to 4000 Ci/mole) for several cell doublings. 10¹² cells of this mutant excrete approximately 1.4 μmoles of cyclic AMP/hr. The extracellular cyclic AMP can be purified by adsorption to charcoal, chromatography on an alumina plate, and paper chromatography.     

Studies of kidney plasma membrane adenosine-3′,5′-monophosphate-dependent protein kinase

Porcine kidney cortex was utilized for the preparation of plasma-membrane-enriched and soluble cytoplasmic (cytosol) fractions for the purpose of examining the relative properties of cyclic [³H]AMP receptor and cyclic AMP-dependent protein kinase activities of these preparations. The affinity, specificity and reversibility of cyclic [³H]AMP interaction with renal membrane and cytosol binding sites were indicative of physiological receptors.Binding sites of cytosol and deoxycholate-solubilized membranes were half-saturated at approx. 50nM and 100 nM cyclic [³H]AMP. Native plasma membranes exhibited multiple binding sites which were not saturated up to 1 mM cyclic [³H]AMP. Modification of the cyclic phosphate configuration or 2′-hydroxyl of the ribose moiety of cyclic AMP produced a marked reduction in the effectiveness of the cyclic AMP analogue as a competitor with cyclic [³H]AMP for renal receptors. The cyclic [³H]AMP interaction with membrane and cytosol fractions was reversible and the rate and extent of dissociation of bound cyclic [³H]AMP was temperature dependent. With the plasma-membrane preparation, dissociation of cyclic [³H]AMP was enhanced by ATP or AMP.Assay of both kidney subcellular fractions for protein kinase activity revealed that cyclic AMP enhanced the phosphorylation of protamine, lysine-rich and arginine-rich histones but not casein. The potency and efficacy of activation of renal membrane and cytosol protein kinase by cyclic AMP analogues such as N⁶-butyryl-adenosine cyclic 3′,5′-monophosphate or N⁶,O²-dibutyryl-adenosine cyclic 3′,5′-monophosphate supported the observations on the effectiveness of cyclic AMP analogues as competitors with cyclic [³H]AMP in competitive binding assays.This study suggested that the membrane cyclic [³H]AMP receptors may be closely associated with the membrane-bound catalytic moiety of the cyclic AMP-dependent protein kinase system of porcine kidney.     

Identification and characterization of the adenosine 3′,5′-cyclic monophosphate binding proteins appearing during the development of Dictyostelium discoideum

A photosensitive, radioactive analogue of cyclic adenosine monophosphate, 8-azido-adenosine 3′,5′-[³²P]monophosphate (8-N₃-cyclic AMP), was used to label the cyclic AMP binding proteins of Dictyostelium discoideum. During development cytosolic proteins appear which are specifically labeled by the photoaffinity agent. The proteins are developmentally regulated since they are only found in starved, developing cells. Unlabeled cyclic AMP competes specifically with the labeled analogue for protein binding sites in contrast to unlabeled 5′-AMP which does not compete. A mutant which develops spores but is deficient in stalk cell production produces a different set of cyclic AMP binding proteins from the parent strain.     

Synthesis and release of adenosine 3′: 5′-cyclic monophosphate by Chlamydomonas reinhardtii

One of the labeled compounds synthesized by Chlamydomonas reinhardtii when ³²Pi was supplied was isolated from both the cells and the medium in which the cells had grown. This compound copurified with authentic [8-³H]cAMP by TLC to a constant ratio of ³²P/³H. The compound was degraded by beef heart cyclic nucleotide phosphodiesterase to a product which cochromatographed with authentic 5′AMP, at the same rate as the hydrolysis of authentic cAMP-[³H] to 5′AMP-[³H]. In both cases, 1-Me-3-isoBu-xanthine, a specific inhibitor of the phosphodiesterase, totally blocked the reaction. It is concluded that the compound synthesized by C. reinhardtii was cAMP, 85% of which was released into the medium.     

Catecholamine binding to the α-adrenergic receptors of hamster adipocytes evidence that guanine nucleotides regulate this binding to the α2-receptor subtype

The binding characteristics of the α-component of (?)-[³H]norepinephrine to hamster adipocyte membranes were studied. Binding was rapid, reaching equilibrium in 20 min at 25°C. Dissociation of specific binding by 10 μM phentolamine suggested dissociation from two different sites. The time course of dissociation induced by a 50-fold dilution was unchanged by the addition of norepinephrine, suggesting the absence of cooperative binding sites. [³H]norepinephrine binding was saturable, yielding curvilinear Scatchard plots. Computer modeling of these data further supported the existence of two classes of binding sites, one with high affinity (_D = 23 nM) but low binding capacity (96 fmol/mg protein) and one with low affinity (K_D = 400 nM) but high binding capacity (1000 fmol/mg protein). Adrenergic ligands of competed with [³H]norepinephrine binding in the following order of potency: (?)-norepinephrine>(?)-epinephrine>>(+)-norepinephrine>(?)-isoproterenol. Displacement by the selective α-adrenergic drugs prazosin, clonidine and yohimbine yielded biphasic curves consistent with binding of [³H]norepinephrine to both α₁- (14–22%) and α₂- (78–86%) receptor subtypes. Although Gpp(NH)p failed to alter the binding of [³H]dihydroergocryptine, it severely reduced the binding affinity of (?)-epinephrine, (?)-norepinephrine and the selective α₂-agonist, clonidine. The inhibitory effects of clonidine and of the α-component of (?)-epinephrine on the adrenocorticotropin-stimulated cyclic AMP production in the intact adipocyte were closely correlated with their effects on the binding of both [³H]norepinephrine and [³H]dihydroergocryptine. Conversely, yohimbine but not prazosin markedly antagonised the α-inhibitory effect of norepinephrine on cyclic AMP production. These data led to concluded that [³H]norepinephrine can be successfully used to study the entire α-adrenergic receptor population of hamster fat cells and that the predominant α₂ -receptor subtype exists in two different affinity states for agonists, the proportions of which are modulated by guanine nucleotides.     

β2-Adrenergic regulation of prostaglandin D2 receptor in rabbit platelets

[³H]Prostaglandin D₂ binding to rabbit platelets was increased by about 150% in the presence of β-adrenoceptor agonist, isoproterenol. The isoproterenol-induced potentiation of the [³H]prostaglandin D₂ binding gave a bell-shaped dose-response relationship (maximum response at 3·10⁻⁸ M) in a stereospecific manner. Similar and moderate potentiation was obtained with terbutaline. On the other hand, β-adrenoceptor antagonists such as alprenolol, propranolol and butoxamine (β₂-specific) had no potentiating effect on [³H]prostaglandin D₂ binding; rather, they abolished the isoproterenol-induced increase of [³H]prostaglandin D₂ binding. The β₁-specific antagonist, metoprolol, did not have any effect. Rabbit platelets were found to possess one [³H]prostaglandin D₂ binding site (K_d = 6·10⁻⁷ M, B_max = 787 fmol/mg protein). In the presence of isoproterenol at 3·10⁻⁸ M, B_max was increased with unaltering K_d value. Isoproterenol did not increase [³H]prostaglandin E₁, [³H]prostaglandin E₂ and [³H]prostaglandin F_2α bindings to platelets. The potential effect of isoproterenol was mimicked by forskolin, theophylline, dibutyryl cyclic AMP, prostaglandin E₁ and prostaglandin I₂, but it was abolished by 2′, 5′-dideoxyadenosine, an inhibitor of adenylate cyclase, indicating that elevated level of cyclic AMP may be available for the induction of the increase of [³H]prostaglandin D₂ binding. Prostaglandin D₂-induced cyclic AMP synthesis and antiaggregation activity were also augmented in the presence of isoproterenol. These results suggest a β₂-adrenoceptor-mediated cyclic AMP-dependent mechanism for the regulation of prostaglandin D₂ receptor binding in rabbit platelets.     

Histamine H-1 receptors on rat peritoneal mast cells

In order to characterize the receptor subtype involved in histamine stimulation of increased cyclic AMP levels in rat mast cells with consequent impairment of anaphylactically induced mediator release, the binding of the H-1 receptor antagonist [³H] pyrilamine to mast cells was examined. Pyrilamine bound rapidly, in a saturable and reversible fashion, and with increased binding at 4°C as compared with 21°C and 37°C. [³H] Pyrilamine binding was displaced by H-1 antagonists (tripelnnamine > yrilamine ≧ iphenhydramine) > histamine > the H-2 antagonist, cimetidine. H-1 agonists displaced pyrilamine binding less efficiently than histamine but better than H-2 agonists. Rat mast cells have a single homogeneous population of low affinity (K_D = 222 ± 33 nM) H-1 receptors with a Bmax of 9.7 ± 2.3 pm/10⁶ mast cells and 5.4 ± 0.92 × 10⁶ binding sites per mast cell. Thus, the mast cell has an H-1 type histamine receptor which is probably involved in histamine-induced cyclic AMP increases.     

Effect of hexoses and mannoheptulose on cyclic AMP accumulation and insulin secretion in rat pancreatic islets

The effects of various sugars on the simultaneous release of insulin and accumulation of cyclic AMP were studied in collagenase isolated rat pancreatic islets. d-Glucose stimulated the formation of cyclic AMP at 3 and 60 min of incubation, whether measured by a label incorporation technique, or by the protein kinase binding assay of Gilman. Only d-glucose and d-mannose were able to stimulate insulin release and cyclic [³H]AMP accumulation in the absence of other substrate. d-fructose had a stimulatory effect in the presence of 3.3 mM d-glucose only at a high concentration (38.8 mM), and enhanced the effects of 8.3 mM glucose when added at the concentration of 8.3 mM. d-Galactose was effective only together with 8.3 mM d-glucose. The order of potency of these hexoses, both regarding insulin secretion and cyclic [³H]AMP accumulation, was glucose-mannose-fructose-galactose.l-Glucose and 3-O-methylglucose had no effects at 60 min when incubated together with 8.3 mM d-glucose, whereas at 3 min, 3-O-methylglucose induced a small stimulation of the cyclic [³H]AMP response.d-mannoheptulose and d-glucosamine inhibited the insulin and cyclic [³H]-AMP responses to 27.7 mM glucose. Mannoheptulose suppressed completely the glucose effect on cyclic nucleotide accumulation within 90 s.Although under all incubation conditions, the threshold stimulatory or inhibitory concentration of a given agent was identical for insulin release and cyclic [³H]AMP accumulation, these two variables showed quantitative differences in incubations of 60 min, the magnitude of the changes in insulin secretion being larger than that for the cyclic nucleotide. It is suggested that modulation of islet cyclic AMP level is an important step in the transmission of the effect of various sugars on insulin release; however, glucose and possibly other sugars may also enhance insulin release by additional mechanisms not involving the adenylate cyclase-cyclic AMP system of the β-cell.     

Characterization of an Adenylate Cyclase-Linked Serotonin (5-HT1) Receptor in a Neuroblastoma × Brain Explant Hybrid Cell Line (NCB-20)

Clonal cell line NCB-20 (a hybrid of mouse neuroblastoma N18TG2 and Chinese hamster 18-day embryonic brain expiant) expressed both high- (K_D 180 nM) and low-affinity (>3000 nM) binding sites for [³H]serotonin (5-HT) which were absent from the parent neuroblastoma. The low-affinity binding site was eliminated by 1 μM spiperone. The order of drug potency for inhibition of high-affinity [³H]5-HT binding was consistent with a 5-HT₁ receptor (5,6 - dihydroxytryptamine = 5-HT = methysergide = 5-methoxytryptamine > cyproheptadine = clozapine = mianserin > spiperone > dopamine = morphine = ketanserin = norepinephrine). [³H]5-HT binding was inhibited by guanine nucleotides (e.g., GTP and Gpp(NH)p), whereas antagonist binding was not; as-corbate was also inhibitory. A 30-min exposure of cells to 1—2 μM 5-HT or other agonists produced a three- to fivefold stimulation of cyclic AMP levels. The order of potency for 5-HT agonist stimulation of basal cyclic AMP levels and 5-HT antagonist reversal of agonist-stimulated levels was the same as the order of drug potency for inhibition of high-affinity [³H]5-HT binding, suggesting linkage of the 5-HT₁ receptor to adenylate cyclase in NCB-20 cells.     

Hormone action at the membrane level VI. Binding of (—)-[3H]dihydroalprenolol and (±)-[3H]isoproterenol to turkey erythrocytes and correlation with adenylate cyclase activity

The binding of (±)-[³H]isoproterenol and (—)-[³H]dihydroalprenolol to intact turkey erythrocytes was studied using a rapid centrifugation technique. The binding of both ligands is rapid, dissociable, stereospecific and inhibited by (—)-propranolol. The total number of isoproterenol binding sites is 2800 sites/ cell. This consists of a low and high affinity site both of which show stereospecific binding. The high affinity isoproterenol site has a K_d of 15.5—19.5 nM and has 600 sites/cell. The low affinity isoproterenol site has a K_d of 195 nM and has 2200 sites/cell. The binding of (—)-[³H]dihydroalprenolol shows one type of site with a K_d of 7.8 nM and has 2500 sites/cell. The agonists epinephrine, norepinephrine, soterenol and p-hydroxyphenylisoproterenol which were tested by competition for binding showed a 6—25-fold greater affinity for the high affinity site determined by (±)-[³H]isoproterenol as compared to the (—)-[³H]dihydroalprenolol binding site. However, the antagonists propranolol, practolol and metrapolol showed similar affinities for the binding sites as determined by competition of binding of either labeled isoproterenol or dihydroalprenolol. These studies indicate that isoproterenol binding can recognize two independent stereospecific β-adrenergic receptors or can recognize two different conformational states of a single receptor. Provisional calculations are made on the turnover number of adenylate cyclase under physiological conditions using intact erythrocytes. The turnover number is 4000 molecules of cyclic AMP/10 min per high affinity receptor.     

Characteristics of the catecholamine and histamine receptor sites mediating accumulation of cyclic adenosine 3',5'-monophosphate in guinea pig brain

—Five areas of guinea pig brain were examined to determine the properties of the receptor sites mediating increases in [³H]adenosine 3′,5′-monophosphate (cyclic AMP). Both epinephrine and histamine were effective in causing increases in cyclic AMP in slices derived from cerebral cortex, hippocampus or amygdala, but not in diencephalon or brainstem. Stimulation of slices of cerebral cortex by either epinephrine or histamine resulted in a small, but reproducible, decrease in specific radioactivity of the [³H]-cyclic AMP produced, as did stimulation of the hippocampus by epinephrine. The catecholamine receptor was an α-adrenergic receptor in all three areas where epinephrine was effective; α-adrenergic stimulation, but not β-adrenergic stimulation, increased levels of [³H]-cyclic AMP. Furthermore, α-, but not β-adrenergic blocking agents, prevented the epinephrine- induced increase of both [³H]- and total cyclic AMP in cerebral cortex and hippocampus. Only antihistaminic agents were capable of antagonizing the histamine-induced increase of both [³H]- and total cyclic AMP in these two brain areas. The catecholamine receptor in the amygdala also appeared to be an α-adrenergic receptor. The effects of histamine and epinephrine together were far greater than the sum of effects of either hormone alone in both cerebral cortex and hippocampus.     

Synthesis and Release of Cyclic Adenosine 3':5'-Monophosphate by Ochromonas malhamensis

The chrysophycean alga, Ochromonas malhamensis Pringsheim, was shown to synthesize cyclic adenosine 3′:5′-monophosphate (cAMP) and to release it into the culture medium. Cells contained 3 to 3,000 picomoles per gram fresh weight; medium contained up to 20 times the amount in the cells. Putative [³²P]cAMP was purified from cultures supplied [³²P]phosphate. The compound was identified as [³²P]cAMP by co-chromatography with authentic cAMP through 10 serial steps; by chemical deamination at the same rate as authentic cAMP, to a ³²P compound with the chromatographic behavior of cIMP; and by its conversion through the action of cyclic nucleotide phosphodiesterase to a ³²P compound with the chromatographic behavior of 5′-AMP. A two-step procedure involving chromatography on alumina and on Dowex 50 purified the unlabeled compound from cells or medium sufficiently for it to be assayable by competitive inhibition of binding of [³H]cAMP to cAMP-binding protein (Gilman assay) or by stimulation of cAMP-dependent protein kinase. The activity was destroyed by cyclic nucleotide phosphodiesterase with the same kinetics as authentic cAMP, provided that an endogenous inhibitor of the phosphodiesterase was first removed by an additional purification step.     

Cyclic AMP synthesis in Xenopus laevis oocytes inhibition by progesterone

[α-³²P] ATP was microinjected into Xenopus oocyte and neosynthesized cyclic AMP was isolated. Cholera toxin inhibited progesterone-induced maturation and stimulated after 3 h of preincubation the amount of neosynthesized cyclic AMP. Progesterone decreased the neosynthesis of cyclic AMP during the first hour following addition of the hormone.     

Changing pattern of cyclic AMP-binding proteins during germination of Mucor racemosus sporangiospores.

Interation of cyclic AMP with a profoundly changing pattern of specific binding proteins was shown during aerobic germination of sporangiospores from the fungus Mucor racemosus. 32P-labeled 8-azido-cycli AMP, an analogue of cyclic AMP that forms a covalent linkage with the binding proteins under u.v. light, was used as the ligand. Binding proteins carrying this photoaffinity label were separated by polyacrylamide-gel electrophoresis and identified by radioautography. Equibiltrium dissociation constants (Kd) and binding-response curves in the presence of competing nucleotides were identical for both 8-azido-cyclic [32P]AMP and cyclic [3H]AMP. A quantitative binding assay with both 8-azido-cyclic [32P]AMP and cyclic [3H]AMP over the time course of sporangiospore germination indicated a parallel relationship between cyclic AMP-binding capacity and the intracellular concentrations of cyclic AMP reported in a previous study [Paznokas & Sypherd (1975) J. Bacteriol. 124, 134--139]. Both of these parameters attained transient high values at a time of development when addition of exogenous cyclic AMP prevents hyphal-germ-tube emergence. The measured Kd values did not change during sport germination.     

Cyclic adenosine 3',5'-monophosphate phosphodiesterase from Mucor rouxii: regulation of enzyme activity by phosphorylation and dephosphorylation.

Crude preparations of cyclic adenosine 3′, 5′-monophosphate phosphodiesterase were activated 1.5 to 2 fold by incubation with ATP, Mg²⁺ and cyclic AMP in a reaction which was both, time and temperature dependent. Cyclic AMP phosphodiesterase remained in an activated state upon filtration of the enzymatic preparation through Sephadex G-25 and ion-exchange chromatography. Activation of the enzyme in the presence of [γ ³²P]ATP resulted in a significant amount of [³²P] protein-bound radioactivity. Reversible deactivation of cyclic AMP phosphodiesterase was enhanced by Mg²⁺ and was accompanied by the release of [³²P] protein bound radioactivity. The evidence is consistent with a mechanism for controlling cyclic AMP phosphodiesterase through phosphorylation-dephosphorylation sequence.     

Adenosine 3',5'-monophosphate receptor proteins in HeLa cells

The adenosine 3',5'-monophosphate receptor proteins of HeLa cells have been characterized. Using the Millipore filter assay, in the presence of 5'AMP and a phosphodiesterase inhibitor, specific [3H]cyclic AMP binding was detected in cytosol and in a nuclear-free particulate fraction, but not in nuclei. Both preparations exhibited biphasic Scatchard plots. 8-Azido[32P]cyclic AMP was used as a photoaffinity probe to covalently link ligand with receptor proteins. Proteins were then separated on denaturing gels and analyzed by autoradiography. The cytosol exhibited four specific binding proteins, with molecular weights of 46 000, 50 000, 52 000 and approx. 120 000. The 50 000/52 000 doublet could not be interconverted by phosphorylation-dephosphorylation reactions. On DEAE-cellulose, the 50 000-dalton protein eluted with peak II cyclic AMP-dependent protein kinase. The other proteins eluted with Peak I and with a binding peak not associated with kinase activity. Only the 50 000 protein was precipitated by type II protein kinase antibody from bovine heart. In the particulate fraction, the 120 000 protein was not detectable, but 8-azido[32P]cyclic AMP treatment revealed the other three proteins, with a relative increase in the 50 000-dalton protein. The results suggest that HeLa cells have four binding proteins which can associate with catalytic subunit and that the Peak I enzyme is heterogeneous, consisting of several distinct regulatory subunits.     

Opposite effects of acute and repeated administration of desmethylimipramine on adrenergic responsiveness in rat pineal gland.

The effect of acute and repeated desmethylimipramine (DMI) treatment on catecholamine-stimulated production of adenosine 3', 5'-monophosphate (cyclic AMP) in rat pineal gland was studied $\text{in}$$\text{vivo}$. In rats exposed to continuous illumination, the administration of isoproterenol (2μmol/kg) to control animals produced a marked increase in the concentration of cyclic AMP in pineal gland. In contrast, norepinephrine (2μmol/kg) failed to increase the levels of cyclic AMP. After acute treatment with DMI (single injection, 38μmol/kg, i. p.), the isoproterenol-induced rise in cyclic AMP was not significantly different from that measured in control animals. However, acute DMI treatment did allow a significant elevation in the concentration of cyclic AMP in pineal gland in response to norepinephrine. In rats given nine injections of DMI (38μmol/kg, i.p., twice daily) neither isoproterenol nor norepinephrine caused a significant increase in the concentration of cyclic AMP in pineal glands. Although acute treatment with DMI had no significant effect on [³H] dihydroalprenolol binding, chronic treatment with DMI significantly reduced [³H] dihydroalprenolol binding in the pineal gland. The results of this study suggest that while a single administration of DMI can enhance adrenergic responses elicited by norepinephrine, chronic administration of DMI leads to compensatory decreases in receptor density and adrenergic responsiveness.     

Cyclic AMP,cyclic GMP,and bean rust uredospore germlings

Ten millimolar cyclic AMP (cAMP) or cyclic GMP (cGMP) induced bean rust uredospore germlings to undergo one round of mitosis and to form septa, processes normally associated with appressorium formation. To assess the possibility of cyclic nucleotide regulation of bean rust development, we used an 8-azido-[³²P]cAMP photoaffinity probe to identify three cyclic nucleotide binding peptides. The peptides bound either cAMP or cGMP. The phosphorylation of one peptide in uredospore germling extracts by [γ-³²P]ATP was stimulated by either 1 μM cAMP or cGMP, but only in the presence of 10 mM Na₂MoO₄, a phosphatase inhibitor. Uredospores contain about 1500 and 23 pmol cAMP and cGMP/g dry wt, respectively, as determined by radiobinding assays.     

Two types of cyclic GMP binding site associated with the cyclic AMP-dependent protein kinase from lymphocytes

Low- and high-affinity binding sites for cyclic GMP were found to be associated with the cyclic AMP-dependent protein kinase (ATP: protein phosphotransferase, EC 2.7.1.37) from human tonsillar lymphocytes, but neither of them was identical with the cyclic AMP binding site.The enzyme activated by cyclic GMP phosphorylated the same site of calf thymus H2b histone as the cyclic AMP activated enzyme; however, more complex kinetics of activation were found with cyclic GMP.Two classes of cyclic GMP binding site were demonstrated by kinetic analysis of cyclic [³H]GMP binding in the enzyme preparations eluted by 0.1 M potassium phosphate (pH 7.0) from DEAE cellulose. The high-affinity cyclic GMP binding site (K_d about 44 · 10^?8 M belonged to some complex form of the protein kinase, as evidenced by the mutual inhibition of cyclic AMP binding and high affinity cyclic GMP binding. However, the high-affinity cyclic GMP binding site disappeared on Sephadex G-100 gel chromatography of the enzyme preparation, whereas the cyclic AMP binding activity was recovered quantitively as separate fractions. The low-affinity cyclic GMP binding site (K_d 2–5 · 10^?6 M) was demonstrated by the inhibitory effect of 10^?5 M cyclic GMP on cyclic AMP binding in each cyclic AMP binding fraction obtained by gel chromatography. However, cyclic AMP did not inhibit the binding of cyclic GMP to the low-affinity binding site.