Abstract
Previous studies on allosteric interactions at muscarinic receptors have often focused on ligand-receptor binding interactions, because ligand binding seemed to reflect functional consequences. The prototypal allosteric agent alcuronium is known to bind with similar affinity to the M2 subtype of muscarinic acetylcholine receptors whether or not the receptors are occupied by the agonist pilocarpine. To determine allosteric modulation of receptor signaling by alcuronium, the effects of pilocarpine were measured in contracting guinea pig left atria and on G-protein coupling in M2-transfected Chinese hamster ovary (CHO) cell membranes. Alcuronium dose-dependently suppressed pilocarpine-induced reduction of isometric contraction force in atria (pIC50, Alc = 5.63) without any effect on the EC50 of pilocarpine, consistent with an allosteric mechanism. In contrast, alcuronium shifted the concentration-effect curve of the agonist oxotremorine M to the right without affecting the maximal effect, in a formally competitive manner (pKA, Alc = 5.54). If pilocarpine remained receptor bound in the presence of alcuronium, this indicates that pilocarpine can no longer act as an agonist. In support of this hypothesis, pilocarpine acted as a competitive antagonist against oxotremorine M in the presence of 10 μM alcuronium. Measuring guanosine 5′-O-(3-[35S]thio)triphosphate ([35S]GTPγS) binding in CHO-M2 membranes yielded similar results. Alcuronium suppressed pilocarpine-induced stimulation of [35S]GTPγS binding (pIC50, Alc = 5.47) without shift in EC50, whereas it competitively shifted the response to oxotremorine M (pKA, Alc = 5.97). [3H]Oxotremorine M binding data corresponded with the functional findings. In conclusion, alcuronium converted the agonist pilocarpine into an antagonist—a novel type of functional allosteric interaction.
The G protein-coupled muscarinic acetylcholine receptors are subject to allosteric modulation (Tuček and Proška, 1995; Ellis, 1997;Christopoulos et al., 1998; Holzgrabe and Mohr, 1998). Initial evidence from experiments with isolated contracting heart preparations demonstrated that gallamine (Clark and Mitchelson, 1976) and alkane-bis-ammonium-type compounds (Lüllmann et al., 1969;Mitchelson, 1975) acted as antagonists with a ceiling effect at high concentrations. Subsequent ligand-receptor binding studies with [3H]N-methylscopolamine ([3H]NMS) supported an allosteric mechanism (Stockton et al., 1983; Jepsen et al., 1988; Choo and Mitchelson, 1989). Allosteric agents retard the dissociation of [3H]NMS, indicating formation of a ternary complex with the radioligand-occupied receptor. In addition, muscarinic allosteric agents interact with unliganded receptors and, thereby, inhibit the association of conventional, orthosteric ligands (Kostenis and Mohr, 1996; Schröter et al., 2000). Hence, allosteric modulation of association and dissociation has opposite effects on equilibrium binding of the orthosteric ligand. Whereas gallamine and many alkane-bis-ammonium agents reduce [3H]NMS binding, alcuronium elevates binding of [3H]NMS to cardiac muscarinic M2 receptors (Tuček et al., 1990) because of pronounced inhibition of [3H]NMS-receptor dissociation (Schröter et al., 2000). These findings prompted a search for allosteric enhancers of acetylcholine binding. Although alcuronium fails to elevate acetylcholine binding at any of the five muscarinic subtypes (Jakubı́k et al., 1997), brucine derivatives with a structure similar to one-half of the alcuronium molecule enhance the binding and action of acetylcholine (Birdsall et al., 1999). Among the five subtypes, acetylcholine-occupied M2 receptors seem to be rather insensitive to this effect of the brucine analogs (Gharagozloo et al., 1999).
Muscarinic allosteric actions depend on the allosteric agent, the orthosteric ligand, and the muscarinic receptor subtype (Lee and El-Fakahany, 1988; Ellis et al., 1991). Jakubı́k et al. (1997)investigated the effects of alcuronium, brucine, and structurally related compounds on the equilibrium binding of 12 agonists in cloned M1-M4 receptors. Among the allosteric agents, alcuronium had, by far, the highest affinity for M2 receptors. When pilocarpine was bound to the M2 receptors, the affinity of alcuronium was even somewhat increased. The factor of cooperativity (Ehlert, 1988a) between alcuronium and pilocarpine, β = 0.37 (Jakubı́k et al., 1997), means a 2.7-fold higher affinity of alcuronium for pilocarpine-occupied M2 receptors compared with free receptors. Therefore, we selected pilocarpine-alcuronium interactions to investigate the functional consequences of ternary complex formation in an intact M2-containing tissue, i.e., contracting guinea pig atria. Alcuronium unexpectedly suppressed the agonist effects of pilocarpine and even caused pilocarpine to behave like a muscarinic antagonist. We confirmed these unique functional properties of a muscarinic allosteric agent at the level of receptor-G protein coupling by measuring pilocarpine stimulation of [35S]GTPγS binding to Chinese hamster ovary (CHO) membranes expressing M2receptors. The combined results demonstrate for the first time that a muscarinic allosteric agent can modulate the intrinsic efficacy of an orthosteric muscarinic receptor ligand.
Materials and Methods
Organ Bath Experiments.
The procedure to measure the actions of muscarinic agents in contracting guinea pig atria has been described previously (Tränkle et al., 1998). Left atria were mounted in organ baths containing 20 ml of oxygenated Tyrode's solution (136.9 mM NaCl, 2.7 mM KCl, 1.8 mM CaCl2, 1.05 mM MgCl2, 11.9 mM NaHCO3, 0.21 mM NaH2PO4, and 5.5 mM dextrose; 95% O2/5% CO2; pH 7.3; 32°C). The preparations were preloaded with 5 millinewton and electrically stimulated via platinum contact electrodes at 3 Hz with rectangular pulses of 5-ms duration. The voltage was 1.5-fold over the threshold of excitation amounting to about 2 V. Contraction force (CF) was recorded isometrically. After an equilibration period of 60 min (CF = 5.5 ± 0.1 millinewton, mean ± standard error,n = 234 atria), a cumulative concentration/effect curve for the negative inotropic action of the agonist under study was recorded. Each agonist concentration was applied for 10 min, a period sufficient to obtain equilibrium effects. Contraction force was expressed as the percentage of the value found at the end of the equilibration period. The agonist was removed from the organ bath over a period of 30 min by replacing the Tyrode's solution with fresh solution every 10 min. Thereafter, the preparation was preincubated with the allosteric test compound for 60 min before another agonist concentration/effect curve was measured in the presence of the allosteric compound. Contraction force was expressed as the percentage of the value at the end of the preincubation period with the allosteric agent. The agonist was washed out using Tyrode's solution still containing the allosteric agent in the concentration as before. After 30 min of washout, the concentration of the allosteric agent was increased, and, after a preincubation phase of 60 min, another concentration/effect curve of the agonist was measured. Modifications of this protocol are mentioned under Results.
Cell Culture and Membrane Preparation.
Cell culture and [35S]GTPγS binding experiments were carried out in a similar way as described by Burford et al. (1995). A CHO cell line stably transfected with the human gene for the muscarinic M2 receptor and wild-type CHO cells were a gift of Prof. Dr. G. Lambrecht (Department of Pharmacology, Biocenter Niederursel, University of Frankfurt/Main, Germany). Cells were cultured at 37°C under humidified air supplemented with 5% CO2 in Ham's F 12 medium containing 10% fetal calf serum, 100 IU/ml penicillin G, 100 μg/ml streptomycin, and 1 mM glutamine. For membrane preparation, cells grown to confluence were treated for 24 h with 5 mM Na-butyrate added to the culture medium before harvesting the cells in a buffer of 10 mM HEPES, 154 mM NaCl, and 0.7 mM EDTA, pH 7.4 at room temperature. The subsequent steps were carried out at 4°C. The cell suspension was centrifuged at 185g (Avanti J25, rotor type JS-7.5; Beckman Instruments, Palo Alto, CA) for 5 min, the supernatant was discarded, the pellet was resuspended in homogenization buffer (10 mM HEPES, 10 mM EDTA, 10 mM NaF, and 10 mM Na2P2O7, pH 7.4), and the first centrifugation step was repeated. The pelleted cells were disrupted using a Polytron homogenizer (PT 10–35; Kinematica AG, Littau, Switzerland; level 7, six bursts of 5-s duration and intervals of 30 s in between), and the resulting homogenate was centrifuged at 40,000g (Avanti J25, rotor type JA 25.50) for 17 min. After discarding the supernatant, the pellet was resuspended in storage buffer (10 mM HEPES and 0.1 mM EDTA, pH 7.4) and centrifuged again. The membranes were resuspended in storage buffer and stored at −80°C.
[35S]GTPγS Binding Experiments.
Membranes (100 μl, final concentration 150 μg of protein/ml) were added to the incubation buffer (900 μl; 10 mM HEPES, 100 mM NaCl, and 10 mM MgCl2, pH 7.4) containing (final concentrations) 0.07 nM [35S]GTPγS (1250 Ci/mmol), 10 μM GDP, and the test compounds at the indicated concentrations. After an incubation period of 60 min at 30°C, the incubation medium was filtered through glass fiber filters (Schleicher & Schüll, Dassel, Germany), and filter-bound radioactivity was measured by liquid scintillation counting. As determined in homologous competition experiments with [3H]N-methylscopolamine ([3H]NMS 0.2 nM, 83.5 Ci/mmol) and increasing concentrations of unlabeled NMS under incubation conditions as described above, the density of M2 receptors amounted to about 2.5 pmol/mg of membrane protein.
[3H]Oxotremorine M Binding Experiments.
M2 receptor-containing membranes from guinea pig hearts that had been prepared as described previously (e.g.,Tränkle et al., 1998) were incubated with 86 Ci/mmol of 1 nM [3H]oxotremorine M in a buffer analogous to that applied by Jakubı́k et al. (1997) but without 0.5 mM GTP (136 mM NaCl, 5 mM KCl, 1 mM MgSO4, 0.2 mM KH2PO4, 0.8 mM Na2HPO4, and 10 mM Na-HEPES, pH 7.4, at 25°C). After time periods appropriate to reach binding equilibrium of up to 3 h, the incubation medium was filtered through glass fiber filters (no. 6, Schleicher & Schüll) followed by two rinses with ice-cold incubation buffer. Filter-bound radioactivity was measured by liquid scintillation counting. Nonspecific binding of [3H]oxotremorine M was determined in the presence of 1 μM atropine and amounted to about 20% of the total binding.
Data Analysis.
Indicated are mean values ± standard error. Unless otherwise indicated concentration/effect curves were fitted to the data by nonlinear regression analysis using the following four-parameter logistic equation
Antagonist-induced parallel curve shifts were quantified by agonist dose-ratios DR = EC50,test compound/EC50,controland analyzed using the equation
Binding constants (KD,KX) were derived from IC50 values according to DeBlasi et al. (1989)and Cheng and Prusoff (1973), respectively. The interaction between [3H]oxotremorine M (L) and the allosteric agent alcuronium (A) was analyzed according to the ternary complex model of allosteric interactions (Ehlert 1988a) using the equation
Combined interactions of [3H]oxotremorine M (L), pilocarpine (X), and alcuronium (A) were analyzed on the basis of the ternary complex model according to Jakubı́k et al. (1997) using the equation
Data analysis, graphical presentations and statistical testing were carried out using the software PRISM version 3.0 and INSTAT version 3.0 (GraphPad, San Diego, CA).
Drugs.
Oxotremorine M iodide was obtained from Research Biochemicals International (Natick, MA). Pilocarpine hydrochloride, gallamine triethiodide, (±)-propranolol, hexamethonium bromide, Ham's F12 medium, fetal calf serum, penicillin G, streptomycin, glutamine, and HEPES were purchased from Sigma-Aldrich Chemie (Steinheim, Germany). Na-butyrate was from Acros Organics (Geel, Belgium). [35S]GTPγS, [3H]oxotremorine M, and [3H]N-methylscopolamine were from PerkinElmer Life Sciences (Homburg, Germany). Alcuronium chloride was generously provided by Hoffmann-La Roche AG (Grenzach Wyhlen, Germany).
Results
Drug Effects on the Contraction Force of Guinea Pig Atria.
The concentration/effect curve for the negative inotropic action of pilocarpine as obtained under control conditions is shown in Fig.1A. The inflection point of the curve was (mean ± standard error, n = 13) pEC50 = 6.10 ± 0.03; the slope factor wasnH = −1.28 ± 0.12; and the maximal effect of pilocarpine was indicated by the bottom plateau of the curve at CFmin = 18 ± 2% of the predrug force of contraction. Repeated measurements of pilocarpine concentration/effect curves in control preparations over intervals of up to 300 min after starting with the first curve revealed that the sensitivity of the preparations to pilocarpine remained stable (data not shown). Alcuronium alone (0.1–100 μM) in the preincubation period of 60 min had no effect on the contraction force of the preparations (data not shown). The time course of the negative inotropic action of pilocarpine seemed to be unchanged under the influence of alcuronium. The pilocarpine concentration/effect curve was hardly affected at 0.1 (not shown) and 0.3 μM alcuronium. At higher concentrations of alcuronium (Fig. 1A), the maximal effect of pilocarpine was reduced (one-way ANOVA, P < 0.0001), whereas the slope factors of the curves (P = 0.298) and the pEC50 values (P = 0.801) were not different from the respective control values.
The concentration/effect relationship for the alcuronium-induced elevation of the lower plateaus CFmin of the pilocarpine curves is depicted in Fig. 1B. For nonlinear regression analysis, the bottom plateau of the curve was fixed at the control level of 18% (see above). The slope factor was not significantly different from unity and was set to nH= 1; the −log inflection point of the curve was at pIP = 5.63 ± 0.14 (n = 34 experiments). The upper plateau of the curve (89 ± 8%) remained slightly below 100%, which may be explained by the spontaneous loss of contraction force over the period of 1 h required to record a cumulative concentration/effect curve of pilocarpine. As mentioned below, when starting immediately with a high concentration of pilocarpine (up to 200 μM), in the presence of 10 μM alcuronium, pilocarpine failed to cause a negative inotropic effect.
The alcuronium-induced attenuation of the maximal effect of pilocarpine was reversible upon washout (data not shown) and was not related to endogenous catecholamines as tested by adding the nonspecific β-adrenoceptor blocking drug propranolol at 1 μM or the ganglionic blocking agent hexamethonium at 100 μM.
We suspected that alcuronium suppressed the intrinsic efficacy of pilocarpine. Thus, pilocarpine should behave like an antagonist against another agonist in the presence of alcuronium. To test this, we chose the agonist oxotremorine M, which is antagonized in a formally competitive fashion by alcuronium in contracting guinea pig atria (Maass et al., 1995). Evidence for the formation of ternary complexes, as reflected by an inhibition of [3H]oxotremorine M dissociation in cardiac membranes, occurred only at high alcuronium concentrations (100 μM) (Maass and Mohr, 1996). As shown in Fig.2A, alcuronium induced a parallel rightward shift of the concentration/effect curve of oxotremorine M (parameters of the control curve pEC50 = 7.92 ± 0.05; nH = −1.56 ± 0.16; CFmin = 7 ± 2%, n = 8). A Schild plot of the data (Fig. 2B) yielded a slope of unity, indicating a formally (but possibly not topologically) competitive antagonism. The −log equilibrium dissociation constant of alcuronium binding to free M2 receptors amounted to pKA = 5.54 ± 0.04,n = 15 and is in line with the value found previously under the same conditions (pA2 = 5.7, Maass et al., 1995). In the next set of experiments, concentration/effect curves of oxotremorine M were recorded in the presence of a fixed concentration of 10 μM alcuronium and stepwise increased concentrations of pilocarpine. In these experiments (Fig.3A), oxotremorine M was first applied alone, then in the presence of alcuronium (preincubation time 1 h), and subsequently in the presence of alcuronium plus one of the indicated concentrations of pilocarpine (preincubation time, 15 min). Alcuronium induced the expected rightward shift of the oxotremorine M curve. In the presence of 10 μM alcuronium, addition of pilocarpine alone even at 200 μM had no negative inotropic effect (data not shown). The concentration/effect curve of oxotremorine M, however, was shifted to the right in a parallel fashion with increasing concentrations of pilocarpine in the presence of alcuronium (Fig. 3A). The antagonistic action of pilocarpine against oxotremorine M was evaluated in a Schild plot using the EC50 of the oxotremorine M curve with alcuronium present and pilocarpine absent as the reference (second curve from the left in Fig. 3A). As shown in Fig.3B, the data points fall on a straight line with a slope of unity. Thus, pilocarpine (in the presence of 10 μM alcuronium) behaved like a competitive antagonist against oxotremorine M. The Schild plot yielded as the −log equilibrium dissociation constant for pilocarpine binding in the presence of 10 μM alcuronium: pK = 6.18 ± 0.06 (n = 11). The concentration-dependent antagonistic action of pilocarpine indicated that pilocarpine could still bind to the receptors in the presence of alcuronium, but its agonistic character was suppressed.
Drug Effects on M2 Receptor-Mediated G Protein Activation.
We measured the effects of the test compounds on [35S]GTPγS binding in membranes of CHO cells stably transfected with the human M2 receptor gene. To verify that drug effects were mediated via M2 receptors, control experiments were carried out in membranes of wild-type CHO cells; none of the test compounds in high concentrations affected [35S]GTPγS binding (data not shown). In the membranes of CHO cells expressing M2 receptors, pilocarpine stimulated [35S]GTPγS binding concentration dependently (Fig. 4A);Emax was 172 ± 3%, the half-maximal effect occurred at pEC50 = 5.48 ± 0.12 (slope factor nH = 0.68 ± 0.10, n = 3 experiments in up to quadruplicate determinations). The effect of pilocarpine was then measured in the presence of 3 μM (n = 3 experiments in duplicate determinations) and 10 μM (n = 2 experiments in duplicate determinations; Fig. 4A) alcuronium. Alcuronium alone significantly suppressed basal [35S]GTPγS binding (one-way ANOVA, p < 0.0001). In the presence of 3 μM alcuronium, both the inflection point of the pilocarpine curve, pEC50 = 5.43 ± 0.16, and the slope factor of the curve, nH = 0.85 ± 0.23, were unchanged compared with the pilocarpine curve recorded under control conditions (unpaired t test, two-tailed,p > 0.05). The maximal effect with 10−2 M pilocarpine, however, was significantly attenuated by alcuronium (one-way ANOVA, p < 0.0001). In the presence of 10 μM alcuronium, pilocarpine was inactive.
Alcuronium effects on basal [35S]GTPγS binding and on [35S]GTPγS binding in the presence of a high concentration of pilocarpine are depicted in Fig. 4B (n = 2 experiments in triplicate determinations). Basal [35S]GTPγS binding was inhibited by alcuronium half-maximally at pIC50 = 5.86 ± 0.15 (slope factor nH = −0.76 ± 0.16). For comparison, the orthosteric antagonist atropine inhibited basal [35S]GTPγS binding with pIC50 = 9.01 ± 0.10 (nH = −0.66 ± 0.09,n = 2 experiments in triplicate determinations). This result agrees with findings of Hilf and Jakobs (1992): pIC50 8.7. There was no difference between the lower plateaus of the curves for alcuronium and atropine, respectively (Emax, alcuronium = 64 ± 2% andEmax, atropine = 64 ± 1%). The stimulation of [35S]GTPγS binding by 10−2 M pilocarpine was concentration-dependently quenched by alcuronium (pIC50 = 5.47 ± 0.21; slope factor nH = −0.59 ± 0.15) to a minimal level of 57 ± 1%, which is not different from the level in the presence of alcuronium alone (unpaired ttest, two-tailed, p > 0.05). According to a partial F-test, the slope factors both for the effect of alcuronium on basal [35S]GTPγS binding and for alcuronium quenching of the pilocarpine signal were not statistically different from unity (p > 0.05). The slope factor of the atropine curve seemed statistically different from unity (p < 0.05), but the experiments were not designed to study the steepness of the curves (large concentration steps of 1 log unit).
Results with guinea pig atria suggested that alcuronium affects G protein activation by oxotremorine M and pilocarpine differently. The concentration/effect curves of oxotremorine M in the absence and in the presence of alcuronium are depicted in Fig.5A. The control curve was characterized by pEC50 = 7.32 ± 0.11 andEmax = 165 ± 2% (nH = 0.76 ± 0.12,n = 3 experiments in 2-fold determinations). Alcuronium at 10 μM (n = 3 in 2-fold), 100 μM (n = 1 in 6-fold), and 1000 μM (n = 1 in 4-fold), induced increasing shifts of the oxotremorine M concentration/effect curve. The maximal effect of oxotremorine M was not affected by alcuronium compared with the control curve (F-test,p > 0.05). A Schild plot of the curve shifts is depicted in Fig. 5B. The data points fall on a straight line with a slope not different from unity, which indicates a formally competitive interaction between alcuronium and oxotremorine M. The pKA for alcuronium binding to free M2 receptors amounted to 5.97 ± 0.04 (n = 5). We used oxotremorine M to check whether pilocarpine was still bound under the influence of alcuronium: The effect of oxotremorine M was measured in the presence of a combination of 10 μM alcuronium plus 100 μM pilocarpine (n = 2 experiments in duplicate determinations). Compared with 10 μM alcuronium alone, the presence of pilocarpine induced a further rightward shift of the oxotremorine M concentration/effect relationship (asterisks in Fig. 5A). Curve fitting to the data points (not shown) suggested a lowered maximum; the corresponding experiments in guinea pig atria had not revealed a diminution of the maximal effect of oxotremorine M in the presence of alcuronium plus pilocarpine (Fig.3A). In any case, under both experimental conditions, the combination experiments demonstrate that pilocarpine remained bound in the presence of alcuronium.
Drug Interactions with Labeled Agonist Binding to M2Receptors.
As mentioned before, the binding experiments ofJakubı́k et al. (1997) pointed to a modest positive cooperativity between pilocarpine and alcuronium (β = 0.37), whereas our functional experiments did not give evidence for an increased potency of pilocarpine in the presence of alcuronium and vice versa. We carried out radioligand binding experiments that were designed in a stepwise approach analogous to that of Jakubı́k et al. (1997) with the following main modifications. We used the agonist [3H]oxotremorine M instead of the antagonist [3H]N-methylscopolamine, and we omitted 0.5 mM GTP from the assay medium. First, the displacement of [3H]oxotremorine by unlabeled oxotremorine M (Fig. 6A) yielded the −log equilibrium dissociation constant of [3H]oxotremorine M binding, pKD = 8.82 ± 0.28. Second, the curve for the competition between [3H]oxotremorine and pilocarpine (Fig. 6A) gave the binding constant of pilocarpine, pKX = 6.77 ± 0.05. Third, the reduction of [3H]oxotremorine M binding by increasing concentrations of alcuronium (Fig. 6B) yielded (eq. 3) the binding constant of alcuronium at free M2receptors, pKA = 6.09 ± 0.06, and its factor of cooperativity with [3H]oxotremorine M, α = 64 ± 10. In accordance, alcuronium has a low affinity for oxotremorine M-occupied M2 receptors (p[αKA] = 4.28) as seen previously in [3H]oxotremorine M dissociation experiments (Maass and Mohr, 1996). Fourth, parallel experiments carried out in the presence of a fixed concentration of 0.3 μM pilocarpine (Fig. 6B) yielded (eq. 4) the factor of cooperativity between alcuronium and pilocarpine, β = 1.12 ± 0.25. Thus, under the present conditions, there is a nearly neutral cooperativity between alcuronium and pilocarpine, i.e., the affinity of alcuronium for pilocarpine-occupied receptors (p[βKA] = 6.04) is almost the same as in free receptors and vice versa. The parameters characterizing the actions of the test compounds in the three applied assays are compiled in Tables1 and 2.
Discussion
The main finding of the present study is that the allosteric modulator alcuronium converts the agonist pilocarpine into an antagonist in intact myocardial tissues. Muscarinic receptors are linked to myocardial contraction force by a complex signaling pathway. To gain insight into the events on the molecular level, we also studied ligand effects on G protein activation in membranes of CHO cells expressing M2 muscarinic receptors. In both assays, the allosteric agent alcuronium did not affect the potency of the muscarinic agonist pilocarpine but reduced its efficacy. In contrast, with oxotremorine M as the agonist, alcuronium had no effect on the efficacy but reduced the potency. Because both organ bath and in vitro assays yielded corresponding results, we conclude that effects observed in the contracting guinea pig atria seem to result from alcuronium modulation of agonist-mediated G protein activation. Because alcuronium had no effect on G protein activity in wild-type CHO cells but only in CHO cells expressing M2 receptors, the site of action of alcuronium seems to be the receptor protein.
The distinct effects of alcuronium on pilocarpine and oxotremorine M likely result from structural differences in the respective agonist/receptor-complexes. With respect to the binding of these agonists, the differential interaction of alcuronium with oxotremorine M- and pilocarpine-occupied M2 receptors has been demonstrated by Jakubı́k et al. (1997) in radiolabeled antagonist binding experiments. Likewise, the radiolabeled agonist binding experiments of the present study indicate that the affinity of alcuronium for pilocarpine-occupied receptors was markedly higher than for oxotremorine M-occupied receptors (Table 2). Alcuronium binding to pilocarpine-occupied receptors takes place at concentrations that are close to the concentrations in which alcuronium binds to free M2 receptors. The slight divergence between the cooperativity factors found by Jakubı́k et al. (1997) and in this study for the interaction of alcuronium and pilocarpine (β = 0.37 and β = 1.12, respectively) may be accounted for by the different assay conditions. Furthermore, Jakubı́k et al. (1997), using the antagonist [3H]N-methylscopolamine in the presence of GTP, found identical binding affinity for pilocarpine and oxotremorine M (Table 1), whereas we found ∼100-fold lower binding affinity of pilocarpine compared with oxotremorine M in the present study, using [3H]oxotremorine M in the absence of GTP (Table 1). Remarkably, the functional experiments in the contracting guinea pig atria and the [35S]GTPγS binding experiments (Table 1) yielded a similar ratio. Thus, the [3H]agonist binding assay applied here seems to mimic the conditions of the functional experiments. Taken together, the findings of the binding and functional assays (Table 2) are compatible with the notion that alcuronium, in terms of the ternary complex model of allosteric interactions, shows nearly neutral cooperativity with pilocarpine and pronounced negative cooperativity with oxotremorine M.
In line with the binding data, the pilocarpine concentration/effect curves for the negative inotropic effect and for the G protein activation (Figs. 1A and 4A) remain in the same concentration range when alcuronium is present. In both sets of experiments a shift of the oxotremorine M concentration/effect curve by pilocarpine (in the presence of alcuronium) provides strong evidence that pilocarpine actually is bound to the receptor and acts as an antagonist (Figs. 3and 5A asterisks). The alcuronium-induced reduction of the maximal effect of pilocarpine, therefore, indicates that the formation of ternary complexes is paralleled by a loss of pilocarpine's intrinsic efficacy to induce G protein activation.
Loss of agonist efficacy by formation of ternary complexes has not yet been described in muscarinic receptors. When self-limiting antagonistic actions were observed in organ-bath experiments with allosteric agents such as gallamine and alkane-bis-ammonium-type compounds and the agonists carbachol, acetylcholine, and oxotremorine (e.g.,Lüllmann et al., 1969; Mitchelson, 1975; Clark and Mitchelson, 1976; Tränkle et al., 1998), the formation of ternary complexes was the pivotal event but the maximal agonist effects were maintained. Furthermore, allosteric augmentation of the action of acetylcholine (Birdsall et al., 1999), oxotremorine M, and other agonists (Doležal and Tuček, 1998) has been observed without an impairment of the maximal agonist response. Enhanced agonist binding depended on the formation of ternary complexes, without changes in agonist efficacy. In a report on gallamine inhibition of rat myocardial adenylate cyclase by oxotremorine M and the partial agonistN-methyl-N-(1-methyl-4-pyrrolidino)-2-butynyl acetamide, Ehlert (1988b) observed reduction of the maximal effect ofN-methyl-N-(1-methyl-4-pyrrolidino)-2-butynyl acetamide and raised the possibility that gallamine may have influenced the intrinsic efficacy of the partial agonist.
Pilocarpine is known as a partial agonist in M2receptors, whereas oxotremorine M is a full agonist (McKinney et al., 1991; Vogel et al., 1997). In line with this, pilocarpine reduced the contraction force to CFmin = 18 ± 2%, whereas the maximal effect of oxotremorine M (CFmin = 7 ± 2%) was more pronounced (p < 0.002, unpaired t test). In the [35S]GTPγS binding experiments, the maximal effect of pilocarpine was equal to that of oxotremorine M, but this may be related to the high level of M2receptor expression in the transfected CHO cells. Pilocarpine's sensitivity to allosteric modulation of intrinsic efficacy may be accentuated by its partial agonist character. Alcuronium alone had an inverse agonist action in this study similar to that of other muscarinic allosteric agents, e.g., the alkane-bis-ammonium agent hexane-1,6-bis(dimethyl-3′-phthalimidopropyl-ammonium bromide) (Hilf and Jakobs, 1992). Also in cardiomyocytes, alcuronium and gallamine induced an inverse agonist action on G protein activation (Jakubı́k et al., 1996) similar to that seen with conventional muscarinic antagonists (Jakubı́k et al., 1995). Taken together, these prototypal allosteric agents suppress basal G protein activity as inverse agonists and, thus, seem to stabilize preferentially inactive conformations of the muscarinic M2 receptor. This negative intrinsic activity could have caused the conversion of pilocarpine into an antagonist. As depicted in Fig. 4, alcuronium suppressed the efficacy of pilocarpine to stimulate G proteins (Fig.4B, upper curve) in the same concentration range in which it reduced spontaneous receptor activity (Fig. 4B, lower curve). This finding suggests that alcuronium's binding site, which is located in the entrance of the ligand binding cavity of the receptor protein (Ellis and Seidenberg, 2000; Buller et al., 2002), has a similar conformation in free and pilocarpine-occupied receptors. One may speculate that pilocarpine-mediated receptor activation goes along with a conformationally malleable receptor state, whereas the rigid alcuronium molecule, upon binding to the allosteric site, may imprint an inactive, more rigid conformation on the pilocarpine/receptor complexes. It is intriguing to speculate that muscarinic allosteric modulators could also induce the reverse action, i.e., preferential allosteric stabilization of active agonist/receptor complexes, thereby, increasing the intrinsic efficacy of an agonist. In case of adenosine A1 receptors, the 2-amino-3-benzoylthiophene derivative PD 81,723 allosterically enhances agonist action and promotes constitutive receptor activity (e.g., Bruns and Fergus, 1990;Kollias-Baker et al., 1997).
The allosteric modulation of muscarinic receptor signaling described here differs from a topological and mechanistic point of view from the findings that Litschig et al. (1999) made with another G protein-coupled receptor, i.e., the metabotropic glutamate receptor hmGluR1b. A noncompetitive antagonist was found to inhibit receptor signaling without affecting glutamate binding. In that case, the antagonist did not influence basal receptor signaling activity, and the antagonist action was seen with a variety of agonists. The metabotropic glutamate receptors belong to a subfamily of G protein-coupled receptors that is characterized by a long N-terminal amino acid chain, which contains the agonist binding domain. Possibly, the noncompetitive antagonist binds between the extracellular glutamate binding domain and the transmembrane domain mediating signal transduction (Litschig et al., 1999). Muscarinic acetylcholine receptors differ such that the agonist binding domain resides in the transmembrane region of the ligand binding cavity and the allosteric site is located above that place, i.e., in the entrance of the ligand binding pocket. Thus, the present study demonstrates another, novel type of interference with G protein-coupled receptor signaling.
A major goal is to identify allosteric agents that modulate agonist and antagonist effects at specific muscarinic receptor subtypes. Here, we demonstrate for the first time that the intrinsic efficacy of muscarinic agonists may be subject to allosteric modulation at the M2 subtype. Future studies will give more insight in how far this action depends on the type of allosteric agent, the type of orthosteric ligand, and the subtype of muscarinic receptor.
Acknowledgments
We thank Iris Witten and Frauke Mörschel (University of Bonn) for skillful technical assistance.
Footnotes
-
The work was supported by grants provided by the Deutsche Forschungsgemeinschaft (to K.M.) and by the German Academic Exchange Service, DAAD (to N.E.).
- Abbreviations:
- ANOVA
- analysis of variance
- M2 receptor
- M2 subtype of muscarinic acetylcholine receptor
- CF
- contraction force
- CHO cells
- Chinese hamster ovary cells
- GTPγS
- guanosine-γ-thiotriphosphate
- NMS
- N-methylscopolamine
- Received November 26, 2001.
- Accepted January 25, 2002.
- The American Society for Pharmacology and Experimental Therapeutics