Search

Stimulation of adenylyl cyclase in the hippocampus is critical for memory formation. However, generation of cAMP signals within an optimal range for memory may require a balance between stimulatory and inhibitory mechanisms. The role of adenylyl cyclase inhibitory mechanisms for memory has not been addressed. One of the mechanisms for inhibition of adenylyl cyclase is through activation of G(i)-coupled receptors, a mechanism that could serve as a constraint on memory formation. Here we report that ablation of G(ialpha1) by gene disruption increases hippocampal adenylyl cyclase activity and enhances LTP in area CA1. Furthermore, gene ablation of G(ialpha1) or antisense oligonucleotide-mediated depletion of G(ialpha1) disrupted hippocampus-dependent memory. We conclude that G(ialpha1) provides a critical mechanism for tonic inhibition of adenylyl cyclase activity in the hippocampus. We hypothesize that loss of G(ialpha1) amplifies the responsiveness of CA1 postsynaptic neurons to stimuli that strengthen synaptic efficacy, thereby diminishing synapse-specific plasticity required for new memory formation.

G proteins are heterotrimeric GTPases that play a key role in signal transduction. The alpha subunit of Gs bound to GTP is capable of activating adenylyl cyclase. The amino acid sequences derived from two X. laevis cDNA clones that apparently code for Gs alpha subunits are 92% identical to those found in the short form of human Gs alpha. Despite this high homology, the X. laevis Gs alpha clones expressed in vitro, yielded a protein that are not able to activate the adenylyl cyclase present in S49 cyc- membranes in contrast with human Gs alpha similarly expressed. This finding suggested that the few amino acid substitutions found in the amphibian subunit are important in defining the functionality of the human Gs alpha. The construction of chimeras composed of different fractions of the cDNAs of the two species was adopted as an approach in determining the regions of the molecule important in its functionality in this assay. Four pairs of chimeras were constructed using reciprocal combinations of the cDNAs coding for human and Xenopus Gs alpha. These eight constructs were expressed in vitro and equivalent amounts of the resulting proteins were assayed in the activation of adenylyl cyclase with GTP gamma and isoproterenol. The results obtained here clearly indicate that the G alpha sequence that extends from amino acid 70 to 140, is important for the functionality of human Gs alpha in activating adenylyl cyclase.

The effect of varying levels of beta 2-adrenergic receptor (beta AR) expression on the capacity of the receptor to activate adenylyl cyclase through regulatory G proteins has been systematically explored in this paper, using differential expression of hamster and human beta AR in L cells. Equations have been developed based on the cycle of G protein activation first proposed by Cassel and Selinger and the mobile receptor model, which assumes that hormone-bound beta AR can stimulate a number of different adenylyl cyclase moieties through the G protein during a single cycle of activation. These equations predict the relationship of receptor number to the EC50 (potency) and Vmax (efficacy) for adenylyl cyclase activation. L cell clones were selected with stable expression of the beta AR over a 2000-fold range of levels (from 5 to 10,000 fmol/mg of membrane protein). Experimentally determined values for the EC50 and the Vmax for epinephrine stimulation of adenylyl cyclase over the entire range of receptor levels were found to be in excellent agreement with predictions of the traditionally accepted models. A method is introduced that allows calculation of beta AR coupling efficiency while taking into account the effect of variable receptor levels. The approach provides a quantitative means for the determination of coupling efficiency of the receptor/G protein/adenylyl cyclase system over wide variations in receptor levels and allows for a rational comparison of coupling efficiencies of wild-type and mutant receptors when receptor levels differ.

The murine receptor for luteinizing hormone (LHR) was cloned and expressed in L cells. This LHR (mature protein of 674 amino acids) is very similar to that of the rat (same length, 36 amino acid differences) but differs significantly more from that of man (673 amino acids, 109 differences). Expression of the murine LHR in L cells led to the appearance of binding sites for human chorionic gonadotropin (hCG) with a Kd of 150 pM and an LH- and hCG-stimulable adenylyl cyclase activity (EC50 = 50-100 pM hCG). Upon labeling pools of phosphoinositides with [3H]myo-inositol, L cells expressing the murine LHR responded to hCG with an increase in their rate of phosphoinositide hydrolysis (EC50 = 2,400 pM hCG). This was accompanied by an increase in intracellular Ca2+ [( Ca2+]i), as determined by the Fura2 method. This increase in [Ca2+]i in response to hCG was dependent on the LHR, for HCG did not affect [Ca2+]i in L cells not expressing the LHR. The effect was not due to the cAMP-forming activity of the LH receptor, for neither forskolin nor prostaglandin E1, which both increase cAMP levels in L cells, had a similar effect in either control or LHR-expressing cells and isoproterenol had no effect in L cells expressing a functionally active hamster beta-adrenergic receptor. The effect was also not due to overexpression of a Gs-coupled receptor, for L cells expressing 8-fold higher levels of the human V2 vasopressin receptor did not mimic the Ca(2+)-mobilizing response of the LH receptor. We conclude that the LH receptor has the capability of activating two intracellular signaling pathways: one leading to stimulation of adenylyl cyclase and resulting in increases in cAMP and a second leading to stimulation of phospholipase C and resulting in formation of inositol phosphates and elevations in [Ca2+]i. These data correlate positively with and provide a mechanistic explanation for previous reports on the ability of hCG to mobilize phosphoinositides and increasing [Ca2+]i in luteal and granulosa cells (e.g. Davis, J. S., West, L. A., and Farese, R. V. (1984) J. Biol. Chem. 259, 15028-15034).

We report the molecular cloning of human gene (S31) containing an open reading frame of 1095 nucleotides, which encodes a protein of 365 amino acids. The encoded protein contains seven hydrophobic putative transmembrane domains considered the hallmark of G protein-coupled receptors. The amino acid sequence shows highest homology to receptors for serotonin (5-hydroxytryptamine). Expression of this receptor in murine Ltk- cells conferred upon these cells the ability to respond to serotonin by inhibition of adenylyl cyclase. No response was observed to isoproterenol, epinephrine, histamine, dopamine or melatonin in the transfected cells. We propose that the human gene S31 encodes a novel serotonin receptor.

Adenylyl cyclases are under positive and negative control by guanine nucleotides and hormones. Stimulatory responses are mediated by a guanine nucleotide- and Mg-binding regulatory component (Ns), a protein that has been purified to homogeneity. Inhibitory responses have been hypothesized to be mediated by an analogous regulatory component (Ni) distinct from Ns, but definitive proof for this is lacking and these effects may result from modulation of Ns activity. Recently, Bordetella pertussis toxin has been shown to ADP-ribosylate a peptide that is not part of Ns, and this coincides with attenuation of hormonal inhibition of adenylyl cyclase. We show here that cyc- S49 cells contain a substrate for ADP-ribosylation by pertussis toxin and that the toxin alters GTP dependent inhibition of cyc- adenyl cyclase activity. As cyc- S49 cells do not contain Ns by several criteria, we conclude that Ni is a distinct and separate regulatory component of adenylyl cyclase.

Pure beta-adrenergic receptors (beta-AR) isolated from guinea pig lung and pure guanine nucleotide binding regulatory protein (NS) of adenylate cyclase isolated from human erythrocytes have been inserted into phospholipid vesicles, resulting in the functional coupling of these two components. The reconstitution of receptor and NS interactions results in the establishment of a guanine nucleotide sensitive state of the receptor that binds agonists with high affinity. Competition curves of isoproterenol for labeled antagonist binding to vesicles containing both beta-AR and NS are biphasic and reveal two affinity states, one of high (approximately 2 nM) and the other of low affinity (approximately 300 nM). In the presence of guanine nucleotides, the competition curves become monophasic and are shifted to a single low-affinity state for the agonist similar to the situation observed in membrane preparations. In addition, the interactions of the receptor and NS lead to the induction of a GTPase activity in NS. The GTPase activity can be stimulated by beta-adrenergic agonists such as isoproterenol (2-5-fold) and is completely blocked by antagonists such as alprenolol in a stereoselective manner. The established hormone responsive activity retains the beta 2-adrenergic specificity conferred by the pure receptor, and similar extents of stimulation (up to 4-fold) are observed with pure receptor from frog erythrocytes, indicating a similar efficiency of coupling between receptors from different species and NS.(ABSTRACT TRUNCATED AT 250 WORDS)

The effects of forskolin on rat liver plasma membrane adenylyl cyclase were studied. The diterpene stimulated the Vmax of the enzyme system with apparent Km values of 3-5 microM. Stimulations were marked both in the absence (20-fold over control) as well as in the presence of various stimulators such as GTP, GuoPP[NH]P, NaF alone or in combination with glucagon. Except with GTP, where stimulations of activities by forskolin and the nucleotide were synergistic (more than additive), stimulations of combinations of GuoPP[NH]P, NaF or glucagon with forskolin were additive. Forskolin did not alter significantly the apparent Km values of the enzyme for MgATP or MnATP or the apparent Ka values (concentrations giving stimulations that are 50% of maximum) for Mg or Mn ions, GTP, GuoPP[NH]P or NaF. Forskolin caused a decrease in the concentration of glucagon required for half-maximal stimulation from 5 microM to 1.5 microM. Except for this effect on the Ka for the glucagon, the only kinetic parameter altered was the Vmax under all conditions tested. Although proteolysis stimulated liver membrane adenylyl cyclase under control conditions, it did not enhance forskolin-stimulated activities. More extensive proteolysis, which resulted in decreased activities in the absence of forskolin, also resulted in reduced forskolin-stimulated activities. 'Uncoupling' of the guanine-nucleotide-binding regulatory component, that mediates guanine nucleotide stimulation by addition of 30 mM MnCl2, did not result in 'uncoupling' of forskolin stimulation. The data indicate that the diterpene forskolin stimulates adenylyl cyclase activity by a novel mechanism that differs from that by which NaF or guanylyl nucleotides affect this membrane-bound system and that the diterpene should be a useful tool with which to explore as yet unrecognized modes of regulation of cyclic AMP production.

A hormone responsive adenylate cyclase has been reconstituted in phosphatidylcholine vesicles from its isolated protein components. The proteins used were the affinity chromatography purified (500-2000-fold) or pure Mr = 64,000 beta-adrenergic receptors (beta AR) isolated from hamster and guinea pig lung membranes, the pure heterotrimeric (Mr: alpha = 42,000; beta = 35,000; gamma approximately equal to 5,000) guanine nucleotide regulatory protein (Ns) isolated from human erythrocyte membranes; and the catalytic unit of the adenylate cyclase (C) solubilized from bovine brain caudate nucleus and resolved from beta AR and Ns by gel filtration. Adenylate cyclase activity in vesicles containing C alone was stimulated by forskolin but not by guanine nucleotides or by the beta-adrenergic agonist isoproterenol. Reconstitution of Ns and C interactions in the lipid vesicles resulted in guanine nucleotide but not beta-adrenergic agonist sensitivity. When beta AR was inserted together with Ns and C into lipid vesicles, the catalytic unit became responsive to beta-adrenergic agonists as well and this stimulation was blocked in a stereoselective manner by the beta-adrenergic antagonist alprenolol. Regulation of adenylate cyclase activity in the reconstituted system by beta-adrenergic agonists, guanine nucleotides, and Mg2+ showed properties similar to those observed in native membranes. The interactions of the various protein components in the reconstituted system were also monitored by GTPase activity. Such activity was observed to occur primarily as a result of receptor-Ns interactions. The results described in this report document the feasibility of studying hormone-responsive adenylate cyclase in a totally reconstituted system which retains the major regulatory properties of the enzyme in its native membrane-bound environment.

The temporal characteristics of gonadotropin (LH and hCG) and catecholamine interaction with their luteal receptors and activation of adenylyl cyclase were studied in rabbit luteal membranes. studies on hormone-receptor interaction showed that, once bound, [125I]iodo-hCG dissociated from its receptor very slowly. If excess LH was added 30 min after the initiation of [125I]iodo-hCG binding, 85% of the [125I]iodo-hCG bound at 30 min was still bound to the luteal receptors 4.5 h later. The rate of dissociation of [125I]iodo-hCG from its receptor was not altered by 100 microM GTP, 2 mM MgCl2, or GTP plus MgCl2. The slow rate of [125I]iodo-hCG dissociation observed at 30 min was not due to a time-dependent change in the hormone-receptor complex, as the dissociation of [125I]iodo-hCG was equally slow 5 min after the initiation of the binding reaction. Studies on the activation of luteal adenylyl cyclase by LH showed that stimulation by 1 microgram/ml ovine LH (oLH) could be prevented but, once initiated, could not be reversed by antiserum to oLH. This indicates that once bound to the rabbit luteal LH receptor, oLH causes persistent activation of rabbit luteal adenylyl cyclase. In contrast, the activation of luteal adenylyl cyclase by 1 microM isoproterenol could be completely reversed by the addition of 50 nM propranolol 5 min after the initiation of the reaction. The inhibitory effect of the propranolol could be completely overcome by the addition of excess isoproterenol, indicating that catecholamine binding to its luteal beta-receptor is readily reversible. Thus, there appears to be a basic difference in the mechanism by which the gonadotropins and catecholamines interact with their receptors and activate the rabbit luteal adenylyl cyclase.

A large number of hormones and neurotransmitters activate adenylyl cyclase [ATP, pyrophosphate lyase (cyclizing; EC 4.6.1.1.)] catalyzing the formation of cAMP and PPi from ATP in the presence of Mg2+. The cAMP formed is in turn responsible for eliciting the physiological responses of these hormones and neurotransmitters. In addition to hormones and neurotransmitters, fluoride ion, cholera toxin and guanyl nucleotides (GTP and GTP analogs such as GTP gamma S and GMP-P(NH)P) also stimulate adenylyl cyclase activity (Perkins, 1974; Birnbaumer, 1977; Gill, 1977). It has become evident that hormonally-responsive adenylyl cyclase is a multi-component system consisting of at least 3 physically distinct units. The first is the hormone receptor containing a specific site for a given hormone. The second is the catalytic moiety (C component) of adenylyl cyclase bearing the site responsible for catalysis of the cyclizing reaction. The third is the guanyl nucleotide regulatory subunit (G component) which binds guanyl nucleotide. Recently, a GTPase activity has been found to be associated with the G component of adenylyl cyclase (Cassel and Selinger, 1976; Cassel et al., 1977a, b; Lambert et al., 1979). In this review we will present information on the regulation of hormonally-responsive adenylyl cyclases. This is not intended to be a comprehensive review of the literature. Rather, it represents our views on the current status of the regulation of cAMP formation.

Antisera were produced in rabbits against a guanine nucleotide binding protein (N protein), transducin, purified from bovine retina. Antiserum AS/1, which recognized all three subunits (alpha, beta, and gamma) of the holoprotein, was tested for cross-reactivity with the subunits of the adenylyl cyclase [adenylate cyclase; ATP pyrophosphate-lyase (cyclizing), EC 4.6.1.1]-associated stimulatory (Ns) and inhibitory (Ni) N proteins purified from human erythrocytes. As/1 showed strong reactivity against the beta subunits of both Ns and Ni but failed to cross-react with either the alpha or gamma subunits of Ns and Ni. Seven additional antisera against transducin reacted with the beta subunits but not with the alpha or gamma subunits of Ns and Ni. A single antiserum against transducin reacted with the alpha subunit of Ni but not of Ns. Immunostaining of the beta subunits of Ns and Ni was proportional to the amount of beta subunit blotted and to the antiserum concentration. Immunostaining of either human erythrocyte or bovine cerebral cortical plasma membrane proteins with AS/1 showed a single band, comigrating with the beta subunit of transducin; this band was absent in bovine erythrocyte membranes. Estimation of the amount of beta subunit by immunoblotting with AS/1 showed that the beta subunit comprises approximately equal to 2% of bovine cerebral cortical plasma membrane protein, approximately equal to 100-fold more than in human erythrocyte membranes. These findings provide immunochemical evidence for similarities in the beta subunits and differences in the alpha and gamma subunits of this family of N proteins. Antisera against transducin react specifically with the beta subunits of Ns and Ni in crude plasma membranes and, thus, can serve as specific probes for the beta subunit.

cyc- S49 cell membranes contain an adenylyl cyclase activity which is stimulated by forskolin and inhibited by guanine nucleotides and NaF. These inhibitory effects are mediated by an inhibitory guanine nucleotide-binding regulatory component (Ni) affecting the adenylyl cyclase catalytic unit (Hildebrandt, J. D., Sekura, R. D., Codina, J., Iyengar, R., Manclark, C. R., and Birnbaumer, L. (1983) Nature (Lond.) 302, 706-709). Since cyc- S49 cells do not contain a stimulatory guanine nucleotide-binding regulatory component (Ns), these membranes were used to study the requirements and kinetics of activation of Ni in the absence of Ns. Activation of Ni by guanyl-5'-yl imidodiphosphate was time-dependent (i.e. hysteretic) and pseudo-irreversible. Although GTP and guanosine 5'-(beta-thio)diphosphate could prevent the inhibition caused by guanyl-5'-yl imidodiphosphate if added simultaneously with it, they could not reverse the inhibited state induced by previous exposure to guanyl-5'-yl imidodiphosphate. Activation of Ni had an absolute requirement for Mg2+. Unlike the activation of Ns, however, which requires millimolar concentrations of Mg2+ in the absence of hormonal stimulation, activation of Ni requires only micromolar concentrations of the divalent cation. These results support the contention that hormones which activate Ni or Ns do so by altering different parameters of a similar activation mechanism.

We demonstrated recently that purified preparations of Gs, the stimulatory G protein of adenylyl cyclase, can stabilize Ca2+ channels in inside-out cardiac ventricle membrane patches stimulated prior to excision by the beta-adrenergic agonist isoprenaline or by the dihydropyridine agonist Bay K 8644 and that such preparations of Gs can restore activity to spontaneously inactivated cardiac Ca2+ channels incorporated into planar lipid bilayers (Yatani, A., Codina, J., Reeves, J.P., Birnbaumer, L., and Brown, A.M. (1987) Science 238, 1288-1292). To test whether these effects represented true stimulation and to further identify the G protein responsible, we incorporated skeletal muscle T-tubule membranes into lipid bilayers and studied the response of their Ca2+ channels to G proteins, specifically Gs, and manipulations known to be specific for Gs. In contrast to cardiac channels, incorporated T-tubule Ca2+ channels exhibit stable average activities over prolonged periods of time (up to 20 min at room temperature), allowing assessment of possible effects of G proteins under steady-state assay conditions. We report that exogenously added human erythrocyte GTP gamma S (guanosine 5'-O-(3-thiotriphosphate]-activated Gs (Gs) or its resolved GTP gamma S-activated alpha subunit (alpha s) stimulate T-tubule Ca2+ channels by factors of 2-3 in the presence of Bay K 8644, and of 10-20 in the absence of Bay K 8644 and that they do so in a manner that is independent of concurrent or previous phosphorylation by cAMP-dependent protein kinase. Activation of purified Gs by cholera toxin increases both its adenylyl cyclase stimulatory and its Ca2+ channel stimulatory effects. Ca2+ channels previously stimulated by the combined actions of Bay K 8644 and cAMP-dependent protein kinase still respond to Gs. We conclude that the responses seen are due to Gs rather than a contaminant, that the effect on Ca2+ channel activity is that of a true stimulation, akin to that on adenylyl cyclase, and show that a given G protein may regulate more than one effector system.

The effect of GTP analogs and Mg on the structure of Ns and Ni, the stimulatory and inhibitory regulatory components of adenylyl cyclase, were studied in a comparative manner. Both N proteins, which are alpha beta gamma heterotrimers that differ in their alpha subunits, when exposed to GTP analogs underwent a Mg-dependent conformational change that was not dependent on subunit dissociation. This was seen both as change in sedimentation behavior at 4 degrees C from 4 S to about 3 S and by a property of the new conformation to retain guanine nucleotide tightly bound to it. Warming to 32 degrees C promoted subunit dissociation, each protein giving a mixture of 2 S alpha G and 2 S beta gamma complexes. For both Ns and Ni, these reactions were reversible: 2 S complexes of Ns and Ni associated to 3 S forms on cooling to 4 degrees C, provided the Mg concentration was at or below 10 mM and detergent concentration was below 1%, and the 3 S complexes of these proteins reverted to 4 S forms and released the nucleotide in the cold on chelation of free Mg with EDTA. Reconstitution assays with Ns-deficient membranes from cyc- S49 lymphoma cells revealed that the 3 S form of Ns is a "pre-active" form of the protein. The scheme below summarizes these findings, where G represents a guanine nucleotide. (Formula: see text) Ns and Ni differ in that more Mg is necessary to promote the 4 S to 3 S conversion of Ns than of Ni, and in that both the 2 S to 3 S to 4 S conversions proceeded more readily with Ni than with of Ns. Mg could not be shown to promote subunit dissociation. The above scheme is suggested as a plausible description of the reaction sequence leading from an unactivated to an activated N protein.

We have reported previously that exogenously administered estradiol (E2) results in attenuation of the LH response of rabbit luteal adenylyl cyclase (approximately 50% less activity than control). This was accompanied by a much lesser reduction in the response of the system to the beta-adrenergic agonist isoproterenol (approximately 35% less activity than control). The purpose of the study reported here was to determine if the decreased responsiveness of adenylyl cyclase was the result of altered hormone receptor levels. To this end, hormone receptors were assessed by Scatchard analysis of specific binding. We confirmed that 4-day E2 treatments, which elevated serum E2 levels from 5 to 21 pg/ml. resulted in decreases in LH- and isoproterenol-stimulated adenylyl cyclase activities by 52% and 20%, respectively. In addition, we found that NaF-stimulated activity was also decreased by 20%. Basal adenylyl cyclase activity was unaffected. Upon assessment of the LH and beta-adrenergic receptor levels in luteal membranes, we found that E2 treatment resulted in marked reduction in LH receptor levels to 28% of the control value without changes in the levels of beta-adrenergic receptors. In view of the concomitant changes in the responsiveness of luteal membranes to isoproterenol and NaF, we determined whether E2 treatment affected luteal membrane levels of the stimulatory nucleotide-binding regulatory component (N) of adenylyl cyclase. N component activity was measured using a reconstitution assay that employs the stimulatory N component-deficient cyc- variant of the S49 mouse lymphoma cell line as an acceptor for luteal N component. Using this assay, we found that luteal membrane N component activity was reduced by 20-25% in E2-treated rabbits compared to that in control rabbits. All of the changes noted above were statistically significant. The results uncovered two heretofore unrecognized effects of E2 treatment: 1) loss of LH receptors, and 2) modification of the membrane component responsible for coupling of stimulatory receptors to the catalytic component of adenylyl cyclase.

Washed membrane particles from pig Graafian follicles larger than 6 mm in diameter were found to contain an adenylyl cyclase system that is highly responsive to luteinizing hormone (LH). Under the assay conditions used (1.5 mM ATP, 5 mM MgCl2, 1.4 mM EDTA, and 1.0 mM ethylene glycol bis(beta-aminoethyl)-N,N'-tetraacetic acid, pH 7.0) LH stimulated enzyme activity up to 7-fold. Half-maximal activation was obtained between 3 and 6 X 10(-10) M of added LH, provided the reaction was carried out under equilibrium conditions. In contrast to findings in large follicles, adenylyl cyclase activity was found to be unresponsive to LH in small immature follicles (1 to 2 mm in diameter) and partially responsive to LH (2- to 3-fold stimulation over basal) in follicles of medium size (3 to 5 mm in diameter). Washed membrane particles from follicles larger than 6 mm were found to contain also an ATP- and Mg2+-dependent process which, upon incubation at 30 degrees, causes the adenylyl cyclase system to lose its susceptibility to respond to LH. Because this loss of LH-stimulated activity was not accompanied by a concomitant loss of basal activity, it was concluded to be the expression of a desensitization reaction. Desensitization was almost undetectable at 5 mM total added MgCl2 and required, to be readily detectable, addition of MgCl2 concentrations in excess of 10 mM. LH was found to stimulate desensitization in the presence of 1 mM cyclic adenosine 3':5'-monophosphate by diminishing the requirement for MgCl2. Loss of LH-stimulated adenylyl cyclase activity was readily detectable at 5.0 mM total added MgCl2 in the presence of 10 mug per ml of LH. Half-maximal stimulation of desensitization by LH was found to occur at about 25 X 10(-10) M, i.e. at about 4 to 6 times higher concentrations than needed for half-maximal stimulation of adenylyl cyclase activity. Half-maximal LH-stimulated desensitization was obtained with 0.5 to 0.7 mM ATP in the incubation. Neither GTP, nor adenyl-5'-yl imidodiphosphate, the nonphosphorylating imidodiphosphate analogue of ATP, supported the desensitization reaction, suggesting that this process may be mediated via phosphorylation of one or more of the components of the adenylyl cyclase system. LH-stimulated desensitization was detected also in washed membrane particles derived from dissected rabbit Graafian follicles, whose adenylyl cyclase becomes desensitized in vivo within minutes after human chorionic gonadotropin injection (Hunzicker-Dunn, M., and Birnbaumer, L., Endocrinology, in press), but not in membrane particles from corpora lutea of pregnant or pseudopregnant rabbits, whose adenylyl cyclase becomes desensitized in vivo only several hours or days after human chorionic gonadotropin injection (Hunzicker-Dunn, M., and Birnbaumer, L., Endocrinology, in press). It is suggested that both actions of LH (stimulation of adenylyl cyclase and stimulation of desensitization) are receptor-mediated, and that neither of them is mediated by cyclic adenosine 3':5'-monophosphate...

Effects of glucagon and guanyl nucleotides on the rat liver plasma membrane adenylyl cyclase were studied. It was established that: 1) glucagon stimulates the fully guanyl-5'-yl imidodiphosphate (GMP-P(NH)P)-activated enzyme between 20 and 70%, provided a guanyl nucleotide is present in the assay; 2) glucagon has no effect on adenylyl cyclase activity in membranes activated fully by GMP-P(NH)P and then washed free of nucleotides. It is concluded that occupancy of the guanyl nucleotide binding site that activates the catalytic moiety of the system is not sufficient to promote hormone-receptor coupling to adenylyl cyclase and that occupancy of a second site by guanyl nucleotides is essential to effect stimulation of adenylyl cyclase by the glucagon-receptor complex. The data presented raise the question whether the guanyl nucleotide site that promotes coupling is distinct from the guanyl nucleotide site that modulates binding of glucagon to receptor and whether the occupancy of the guanyl nucleotide site associated with the catalytic moiety is necessary for coupling.

Transient and steady state kinetics of the interaction of liver plasma membrane adenylyl cyclase with GTP and the GTP analogues guanyl-5'-yl imidodiphosphate (GMP-P(NH)P), guanyl-5'-yl diphosphonate (GMP-P(CH2)P), and guanosine 5'-(3-O-thio)triphosphate (GTP gamma S) was studied before and after treatment of membranes with preactivated cholera toxin. (a) In control experiments, GTP stimulated the enzyme partially and without a noticeable lag, while GTP analogues stimulated to varying degrees (GTP gamma S approximately GMP-P(NH)P greater than GMP-P(CH2)P greater than or equal to GTP) and with lag periods that varied with the nucleotide (GMP-P(NH)P approximately GMP-P(CH2)P greater than GTP gamma S). (b) In toxin-treated membranes, transient kinetics and degree of activation by GTP analogues were unaltered, while activation by GTP remained rapid and became as effective as the most effective of the analogues. (c) Toxin treatment had no effect on the concentration-effect curves for GTP analogues but resulted in a 5- to 9-fold lowering of the apparent K alpha with which GTP stimulates the system. These results indicate that the degree of stimulation of the liver adenylyl cyclase by GTP and its analogues is independent of the susceptibility of the beta-gamma bond of the guanosine triphosphate to be hydrolyzed and that lags in progress curves are not due to slow dissociation of "resident" GDP molecules. The time transients of the interaction of the rat liver adenylyl cyclase with combinations of GTP and GMP-P(NH)P were studied and found to be of the competitive type regardless of the time and sequence of addition of the two nucleotides. The results are consistent with GMP-P(NH)P interacting with and dissociating from the system very slowly and not leading to the formation of an irreversibly activated state of the enzyme. The data obtained on regulation of the basic (no hormone added) liver adenylyl cyclase by guanyl nucleotides are interpreted in terms of a two-state enzyme model (Birnbaumer, L., Bearer, C.F., and Iyengar, R. (1980) J. Biol. Chem. 255, 3552-3557) and the results are discussed in the light of current views of regulation of adenylyl cyclases and involvement of a GTPase. Our experiments are consistent with, but not proof for, presence of a cholera toxin-inhibited GTPase in intimate association with the liver adenylyl cyclase.

We measured the adenylyl cyclase (AC) activity in dissected CL and the responsiveness of the AC system to LH, FSH, and prostaglandin (PG)E1 at different times following the administration of high doses of hCG or hLH to pseudopregnant and pregnant rats and rabbits. In rabbits, ovulatory doses of hCG promoted desensitization of the AC system in both CL of pregnancy and CL of pseudopregnancy (PSP), but at varying rates. At least a 50% decline in the LH-stimulated AC system was demonstrable 2 h after the hCG injection in CL obtained during PSP and the first 18 days of pregnancy. However, after day 21, AC activity was unaltered at 2 or 24 h after hCG injection, necessitating as much as 72 h for the AC system to become desensitized to LH. It seems that CL in the last third of pregnancy are afforded partial protection from the desensitizing effects of hCG. This protective effect was found not to be conferred upon follicles contained in ovaries after day 21 of pregnancy or upon newly, hCG-induced 3-day-old CL in 24-day pregnant rabbit ovaries. hCG-induced desensitization of CL adenylyl cyclase in rabbits was prevented neither by cauterization of tertiary follicles not by the continued administration of estradiol-17beta (1.5 mug SC twice daily), suggesting that this effect of hCG is due to a direct interaction with the CL, and not due to interruption of the follicular estrogen supply. In rats, the injection of an ovulatory dose of hCG (50 IU SC into prepubertal rats; 50 IU ip plus 50 IU SC into mature rats) also induced desensitization of the AC system in ovaries of superovulated prepubertal rats and in CL of pseudopregnant and pregnant rats. Desensitization of the AC system was not detectable at 2 h, was 30% of total by 6 h, and was complete at 24 h after hCG injection. Both regression of the CL and desensitization of the AC system in CL are induced only by doses of hCG which are ovulatory and not subovulatory. Desensitization of AC appears to precede functional luteolysis, at least in the pseudopreganant rabbit. Thus, the apparent close association between hCG-induced luteolysis and the desensitization of the adenylyl cyclase system in CL would suggest that desensitization may be a marker for luteal regression.

The subunit composition of the Ns and Ni, the human erythrocyte stimulatory and inhibitory regulatory proteins of adenylyl cyclase, respectively, were analyzed by a sodium dodecyl sulfate-containing discontinuous urea and polyacrylamide gradient gel electrophoresis system designed for the study of low molecular weight polypeptides. This system disclosed that these proteins, in addition to their known alpha and beta subunits, contain an additional small peptide of apparent molecular weight of 5,000 (5K). This "5K peptide" is also present in preparations of another protein which we termed "40K protein" on the basis of its hydrodynamic behavior and whose primary protein constituent is the Mr 35,000 beta subunit of the above regulatory proteins. Analyzing Ni, the 5K peptide was functionally related to the protein by showing that its apparent Stokes radius changes from 5.9 to 5.1 nm after treatment with guanyl-5'-yl imidodiphosphate and magnesium in parallel with the alpha and beta subunits. These data are interpreted as evidence for the existence of a third subunit associated with the regulatory proteins of adenylyl cyclase. We call this subunit gamma and propose a minimum subunit structure for these proteins of the alpha beta gamma type.

The final step in a scheme for the purification of the guanine nucleotide- and Mg2+-binding stimulatory regulatory component (Ns) of adenylyl cyclase [adenylate cyclase; ATP pyrophosphate-lyase (cyclizing), EC 4.6.1.1] from human erythrocyte membranes involves chromatography over hydroxylapatite (HAP) which yields two fractions. The first fraction (HAP I) contains predominantly two peptides that, upon sodium dodecyl sulfate/polyacrylamide gel electrophoresis, migrate with Mr values of 39,000 and 35,000. The second fraction (HAP II) contains predominantly Ns formed of two peptides of Mr 42,000 and 35,000. The HAP I, Mr 39,000 peptide is shown to be a substrate for the ADP-ribosylating toxin of Bordetella pertussis (pertussis toxin). Upon sucrose density gradient centrifugation, both the Mr 39,000 and the Mr 35,000 peptides of HAP I migrate at about 4 S. Treatment of HAP I with guanine nucleotide and Mg2+ prior to centrifugation results in a coordinated change in the migration of both peptides to 2 S. It is postulated that HAP I contains an alpha beta heterodimeric protein composed of an alpha subunit of Mr 39,000 and a beta subunit of Mr 35,000. Further, this protein dissociates under the influence of guanine nucleotides and Mg2+ into its individual alpha and beta subunits. Because previous studies have shown that treatment of cells and cell membranes with pertussis toxin results in attenuation of the effects of hormones that inhibit adenylyl cyclase activity, and because this effect correlates with the ADP-ribosylation of a Mr approximately equal to 40,000 peptide, we believe that we have purified a guanine nucleotide- and Mg2+-binding inhibitory regulatory component of adenylyl cyclases--i.e., the Ni.

The best understood system for transduction of extracellular messages into intracellular signals is the hormone receptor-coupled adenylate cyclase. In such systems receptors are functionally coupled to the enzyme by two special proteins, termed the stimulatory and inhibitory guanine nucleotide regulatory proteins (Ns and Ni, respectively). These proteins, thought to mediate, respectively, stimulatory and inhibitory influences on the adenylate cyclase, are members of a larger class of heterotrimeric guanine nucleotide regulatory proteins involved in membrane signal transduction. We have studied the interactions of the various components of the adenylate cyclase system by co-reconstituting pure beta-adrenergic receptors, pure Ns and Ni, and functionally resolved preparations of the catalyst in phospholipid vesicles. In the absence of Ni, beta-adrenergic receptor/Ns-mediated catecholamine stimulation of the enzyme is relatively modest (approximately 1.3-fold). Surprisingly, however, when Ni is also present, stimulation increases dramatically (up to 7-8-fold) because of a greater suppression of basal relative to agonist-stimulated enzyme activity. Thus, Ni may actually be required for maximal agonist stimulation as well as for inhibition of the adenylate cyclase.

Catecholamine-induced stimulation of adenylyl cyclase in ventricular membranes of kitten and Xenopus laevis was antagonized competitively by carazolol. Apparent equilibrium constants ranging between 100-170 pM were estimated for the beta-adrenoceptor-carazolol complex.

We injected hCG into superovulated rats on the seventh day of pseudopregnancy and confirmed previous findings that this results in both homologous desensitization of luteal adenylyl cyclase (loss of responsiveness to LH) and heterologous desensitization of the same adenylyl cyclase system (partial loss of responsiveness to catecholamines), and that these changes are associated with the loss of available unoccupied LH receptors (down-regulation) but not with any discernible loss of beta-adrenergic receptors. We tested the hypothesis that the heterologous component of the above changes might be due to alterations in the function of the nucleotide-binding N component of adenylyl cyclase that intervenes between receptors and catalytic units of adenylyl cyclase. This was done by assessing N component activity in reconstitution assays that measured the capacity of luteal N to mediate, in cyc- S49 lymphoma membranes, stimulation of adenylyl cyclase independently by the guanine nucleotide guanylyl imidodiphosphate, by NaF, or by the lymphoma membrane beta-adrenergic receptor. By all of these modes of assay, heterologous desensitization of luteal adenylyl cyclase to beta-adrenergic stimulation was found to be associated with a proportionally similar decrease in N component activity. This change in N component activity could be due to either quantitative or qualitative alterations. It is speculated that if the change is of a qualitative nature, the alteration may be a cAMP-dependent phosphorylation reaction of one of the subunits of the N component.

Membranes were obtained from Xenopus laevis oocytes after removal of follicular cells by collagenase treatment. [32P]ADP-ribosylation with pertussis toxin showed them to contain a single Mr = 40000 substrate for this toxin that co-migrates on sodium dodecylsufate-polyacrylamide gel electrophoresis with pure human erythrocyte Ni, the inhibitory regulatory component of adenylyl cyclase. [32P]ADP-ribosylation of oocyte membranes with cholera toxin also showed presence of a single substrate but of Mr = 42000. These results indicate, that the adenylyl cyclase system of oocytes, like that of somatic cells and unlike that of spermatozoids, contains the catalytic unit C and both of the known regulatory N components. The possible susceptibility to pertussis toxin of the guanine nucleotide-dependent inhibition of oocyte adenylyl cyclase by progesterone was investigated. This action of progesterone is mediated by a membrane bound receptor as opposed to a receptor of cytosolic or nuclear localization. However, the inhibitory effect of progesterone was unaffected by pertussis toxin, even though the oocyte membrane Ni was fully ADP-ribosylated with pertussis toxin, as revealed by lack of further [32P]ADP-ribosylation on subsequent re-incubation with pertussis toxin. These results indicate that the action of progesterone, in spite of being nucleotide-dependent, is either not mediated by Ni, suggesting the existence of an additional nucleotide regulatory component, or if mediated by Ni, involves a mode of regulation of this coupling protein that is different from that by which all other inhibitory hormones act on adenylyl cyclase.

We have reported in recent studies that exogenous estradiol (E2) suppresses luteal LH-responsive adenylyl cyclase activity in pseudopregnant rabbits. The purpose of the present study was to determine whether this suppression is reversible. High or low level E2-filled Silastic capsules or empty capsules were sc implanted in day 5 pseudopregnant rabbits. On day 8 of pseudopregnancy, the high level E2 implants were either sham replaced, replaced with low level E2 implants, or replaced with empty capsules. The low level E2 implants and empty capsules were sham replaced. Animals from each of the five resulting groups were killed on days 9-12 of pseudopregnancy (1, 2, 3, and 4 days postimplant manipulation). As previously reported using an intermittent injection protocol, exogenous E2 had little effect upon serum progesterone concentrations. Both high and low level E2 implants suppressed the luteal LH-responsive adenylyl cyclase, but the suppression due to the low level E2 implants was not as great as that for the high level E2 implants. Within 24 h of switching from high to low level E2 implants, LH-responsive adenylyl cyclase activity increased from the level found for animals with high level E2 throughout to that found for animals with low level E2 throughout. Total withdrawal of exogenous E2 resulted in a precipitous fall in serum progesterone concentrations, as predicted by previous studies. However, within 4 days of withdrawal, both serum progesterone and luteal LH-responsive adenylyl cyclase activity had returned to control values. E2 implants also suppressed serum LH concentrations and follicular LH-responsive adenylyl cyclase activities. Both of these effects were reversed within 24-48 h after implant withdrawal. We conclude, therefore, that effects of exogenous E2 are reversible and that the previously reported E2-induced dependency upon exogenous E2 is related to the experimental protocol used.

Summer rabbits appeared to be less sensitive to the presence of exogenous estradiol or to the withdrawal of exogenous estradiol than were Winter rabbits. Suppression of the luteinizing hormone(LH)-stimulated adenylyl cyclase of corpora lutea by estradiol-filled silastic capsules, which resulted in serum estradiol concentrations approximately 3.8 X control (high level), appeared to be greater in Winter rabbits than in Summer rabbits. Both high level and low level (2.1 X control) estradiol implants suppressed follicular LH-stimulated adenylyl cyclase in Winter rabbits but neither size capsule had an effect in Summer rabbits. Serum LH concentrations were equally low in Winter rabbits using either high or low level estradiol implants, while only the high level implants caused a decrease in serum LH in Summer rabbits. Withdrawal of the exogenous estradiol caused a precipitous fall in serum progesterone concentrations in Winter rabbits; estradiol withdrawal in Summer rabbits was without effect.

GDP and GTP regulation of receptor-mediated stimulation of adenylyl cyclases in membranes of S49 murine lymphoma cells (S49), NS-20 murine neuroblastoma cells (NS-20), rabbit corpora lutea (CL), and turkey erythrocytes were studied under assay conditions which minimized conversion of added GTP to GDP and of added GDP to GTP. Hormonal stimulation in all systems required guanine nucleotide addition. In the presence of GTP, adenylyl cyclase activity in S49, NS-20, and CL was stimulated respectively by isoproterenol and prostaglandin E1 (PGE1), by PGE1 and the adenosine analog, phenylisopropyladenosine, and by PGE1 and isoproterenol, with the first of the listed stimulants eliciting higher activities than the second. Activity in turkey erythrocyte membranes was stimulated by isoproterenol. GDP was partially effective in promoting hormonal stimulation, being able to sustain stimulation by isoproterenol and PGE1 in S49 cell membranes and by PGE1 in CL membranes. In NS-20 membranes, both GDP and guanosine-5'-O-(2-thiodiphosphate) (GDP beta S) were inhibitory on basal activity, yet promoted limited but significant stimulation by PGE1. In turkey erythrocytes, stimulation by isoproterenol could not be elicited with GDP or GDP beta S. Thus, although less effective than GTP in promoting hormonal stimulation of several adenylyl cyclase systems, GDP was clearly not inactive. Concentration effect curves for active hormone in the presence of GDP had higher apparent Ka values than in the presence of GTP. In spite of differences between the effects of GTP and GDP on hormonal stimulation of adenylyl cyclase activities, GTP and GDP affected equally well isoproterenol binding, regardless of whether or not its receptor could be shown to stimulate adenylyl cyclase in the presence of GDP. Determination of transphosphorylation of GDP to GTP showed that at saturating concentrations, the proportion of GDP converted to GTP is negligible and unaffected by hormonal stimulation. Concentrations giving 50% inhibition were determined for GTP- and GDP-mediated inhibition of guanyl-5'-yl imidodiphosphate stimulation in the absence and presence of stimulatory hormones. In all four systems studied, GTP and GDP interacted with about equal potency and hormonal stimulation was not accompanied by a selective decrease in affinity for GDP. One way to explain all of the results obtained is to view hormonally sensitive adenylyl cyclase systems as two-state enzymes whose activities are regulated by GTP and GDP through an allosteric site related to the catalytic moiety, and receptors as entities that are inactive and hence unable to couple unless occupied by hormones and activated by any guanine nucleotide through a distinct receptor-related process.

In vitro studies were conducted to evaluate the possible mechanisms of LH-induced desensitization of the rabbit follicular adenylyl cyclase (AC) system. We tested the effects of cAMP, dibutyryl cAMP, and inhibitors of various cellular functions on LH-stimulated AC activity as well as the reversibility of AC desensitization. Refractoriness of the AC to LH was induced by a1 1- or 2-h incubation of Graafian follicles with 10 microgram/ml LH. We fund that the initial 60-min phase of AC desensitization to LH in Graafian follicles was not prevented by a 60-min preincubation of follicles with 11 microM puromycin, 30 microM cycloheximide, 8 microM actinomycin D, 5 microgram/ml cytochalasin B, 50 microM colchicine, or 1 or 10 mM trinitrophenol or by a 90-min preincubation of follicles with 50 microM colchicine. We evaluated the effects of cAMP and dibutyryl cAMP on LH-stimulated AC activity by incubating Graafian follicles with these nucleotides for 30-min to 4 h. While LH-stimulated AC activity was not significantly reduced in follicles which had been incubated 30-min or 1-h with either nucleotide, 2 h incubations resulted in significant reductions in LH-stimulated AC activity, and 4-h incubations promoted a complete refractoriness of the LH-stimulable AC. cAMP also caused desensitization of the FSH-stimulable AC in 4-h incubations, but not in the 1-h incubations. Lastly, once the follicular AC was desensitized to LH, neither 4-guanylyl imidodiphosphate nor ATP could reverse desensitization. These results indicate that AC desensitization in rabbit Graafian follicles is biphasic event. The initial 60-min phase is not mediated by cAMP, RNA, or protein synthetic events, by energy-requiring events inhibited by trinitrophenol, or by microtubule- or micro-filament-associated processes. A secondary phase occurs within 2-h and appears to be mediated, at least in part, by cAMP.