An activator protein 1-driven luciferase reporter assay was developed to monitor the activation of the human muscarinic M3 receptor (hM3-R) and evaluate functional potencies of different anticholinergics in Chinese hamster ovary cells. This assay proved to be superior to previously used functional assays [i.e., inositol phosphate accumulation (J Pharmacol Exp Ther 330:660–668, 2009)], thanks to the longer incubation times that allow reaching of pseudoequilibrium for ligands with slower dissociation kinetics, the long-acting muscarinic antagonists. Interestingly, within this system the hM3-R efficiently signaled in an agonist-independent manner. All the antagonists tested were able to inhibit the hM3-R constitutive activity in a concentration-dependent fashion, behaving as full inverse agonists. Curiously, significant differences in potency as antagonists (against carbachol) and inverse agonists were seen for some compounds (N-methyl scopolamine and tiotropium). Given the potential for inverse agonists to cause receptor up-regulation, the effect of chronic exposure to anticholinergics on the expression levels of hM3-R was also tested. Again, significant differences were seen, with some ligands (e.g., tiotropium) producing less than half of the receptor up-regulation caused by other anticholinergics. This study shows that anticholinergics can exhibit differential behaviors, which depend on the pathway investigated, and therefore provides evidence that the molecular mechanism of inverse agonism is likely to be more complex than the stabilization of a single inactive receptor conformation. In addition, differences in the potential of anticholinergics to induce hM3-R up-regulation might have clinical relevance, because many are on the market or in clinical trials as chronic treatment for chronic obstructive pulmonary disease, for example.
Substantial experimental evidence now exists to show that G protein-coupled receptors (GPCRs) can productively couple to G proteins in the absence of agonist to produce a measurable downstream response (Kenakin, 2004; Casarosa et al., 2009), a phenomenon that is termed constitutive activity. To accommodate these empirical observations, the “extended ternary complex” model (Samama et al., 1993) and the more thermodynamically complete “cubic ternary complex model” (Weiss et al., 1996) were developed. Central to these models is the concept that receptors exist in an equilibrium between inactive (R) and active (R*) receptor conformations. The enhanced ability to detect constitutive activity led to the discovery of a unique subclass of ligands, which exert their actions by actively reducing basal receptor activity. This novel pharmacological property, termed “inverse agonism,” has been modeled by assuming that inverse agonists preferentially bind and stabilize the inactive R state of the receptor (Leff, 1995).
According to the simplest interpretation of the two-state receptor model, the constitutively active R* conformation ought to be identical with the agonist-induced AR*, because there is only one single active conformational state. However, there is no a priori reason that this has to be the case. By analogy to ionic channels and enzymes, it is likely that a receptor may possess multiple, distinct active conformations, as supported by increasing evidence (Kenakin, 2003).
Here, we report novel findings on the human muscarinic M3 receptor (hM3-R), which belongs to the seven-transmembrane-containing superfamily of GPCRs and stimulates intracellular inositol trisphosphate production by a Gq/11-mediated mechanism (Wess, 1993). In airway diseases such as chronic obstructive pulmonary disease (COPD) and asthma, the hM3-R, which is highly expressed on airway smooth muscle cells, is an important pharmacological target. Its activation in response to acetylcholine (ACh), which is released from parasympathetic nerve endings, causes bronchoconstricton. Indeed, muscarinic antagonists such as ipratropium or tiotropium are effective bronchodilators that have a particular value in the treatment of COPD, because they block the effects of an increased vagal cholinergic tone (Barnes, 2004).
In the present study, we examined the inhibitory properties of several muscarinic antagonists in a system where both constitutive activity and agonist-induced responses at the hM3R could be observed. Our results support the idea that, despite inducing the same intracellular signaling cascade (here Gq coupling), the constitutively active M3-R* and the agonist-stabilized M3-AR* may represent distinct conformational states of the receptor, which are differently recognized by some antagonists. These findings support a model of GPCR activation in which the assumption of two states (active and inactive) is expanded to include two or more active conformations. In addition, we found evidence that the different anticholinergics have distinct pharmacological behaviors. Therefore, the assumption that inverse agonists operate by stabilizing a common inactive conformation of the receptor may be too simplistic.
Materials and Methods
Chemicals and Reagents.
MgCl2, carbamoylcholine chloride (carbachol), muscarine chloride, pilocarpine, oxotremorine sesquifumarate, atropine sulfate, pirenzepine dihydrochloride, N-methyl scopolamine bromide, 4-diphenylacetoxy-N-methylpiperidine (4-DAMP), EDTA, NaCl, and HEPES were obtained from Sigma-Aldrich (St. Louis, MO). Acetylcholinesterase from Electrophorus electricus (electric eel) type V-S lyophilized powder, ≥1000 units/mg protein, was obtained from Sigma-Aldrich and reconstituted in distilled water at 1 mg/ml in the presence of 0.1% bovine serum albumin. Ipratropium bromide, tiotropium bromide, aclidinium bromide, and glycopyrrolate bromide were synthesized in the chemical laboratories of Boehringer Ingelheim GmbH (Biberach an der Riss, Germany). The tritration of tiotropium was carried out by RC Tritec AG (Teufen, Switzerland). [3H]Tiotropium was purified by high-performance liquid chromatography (HPLC) on a XBridge (Waters GmbH, Eschborn, Germany) C-8 column, resulting in a radiochemical purity ≥98% and a specific activity of 65 Ci/mmol. All cell culture reagents were purchased from Invitrogen (Carlsbad, CA).
Cell Culture Techniques.
Chinese hamster ovary (CHO) cells stably transfected with the cDNA encoding the hM3-R have been described (Casarosa et al., 2009). CHO-hM3 cells were grown in Ham's F12 medium supplemented with 10% fetal calf serum (FCS) in the presence of the selection agent G418 (400 μg/ml). Cells were maintained at 37°C in humidified air containing 5% CO2.
Transient Transfection for Reporter Gene.
One day in advance, 2 million CHO-hM3 cells were plated in a six-well dish in Ham's F12 medium containing 10% FCS. The next day, cells, typically 50 to 60% confluent, were transiently transfected with the activator protein 1 (AP-1) luciferase reporter gene (PathDetect AP-1 Cis-Reporter Plasmid, Stratagene, La Jolla, CA) using the transfection reagent FuGENE 6 (Roche Diagnostics, Mannheim, Germany) optimized at 2 μg of DNA/6 μl of FuGENE 6 per well, according to the manufacturer's protocol.
Detection of Functional Antagonism with the AP-1 Luciferase Assay.
Thirty hours after transfection of the AP-1 luciferase cDNA, CHO-hM3 cells were resuspended in DMEM-F12 medium without phenol red containing 2% FCS and transferred to 384-white-well plates (PerkinElmer Life and Analytical Sciences, Waltham, MA), where they were stimulated with a range of carbachol concentrations (10−9 to 10−2 M), in the presence or absence of at least seven different concentrations of antagonists. After overnight stimulation (typically 20 h) at 37°C, luciferase activity was quantified with the LucLite Reporter Gene Assay System (PerkinElmer Life and Analytical Sciences) according to the manufacturer's protocol. In brief, 30 μl of substrate solution was added in each well and the plate was further incubated for 30 min in the dark. Afterward, light emission was quantified by using VICTOR2 (PerkinElmer Life and Analytical Sciences), measuring 1 s per well.
Schild plots were created by linear regression in Prism 5.02 (GraphPad Software Inc., San Diego, CA), plotting the dose ratios against the concentration of antagonist to determine pA2 values for antagonists.
Detection of Constitutive Activity and Inverse Agonism with the AP-1 Luciferase Assay.
Thirty hours after transfection of the AP-1 luciferase cDNA, CHO-hM3 cells were resuspended in DMEM-F12 medium without phenol red containing 2% FCS and transferred to 96-white-well plates, where they were incubated with a range of antagonists' concentrations (10−12 to 10−6 M), in the absence of any added agonist.
To rule out the presence of ACh released by the cells, some experiments were performed in the presence of 10 units/ml of acetylcholinesterase from E. electricus (electric eel) type V-S (Sigma-Aldrich). After overnight incubation (typically 20 h) at 37°C, luciferase activity was quantified with the LucLite kit as described above. Data were analyzed with GraphPad Prism 5.02 by nonlinear regression using the equation log(inhibitor) versus response. The best fit between a variable Hill coefficient and a Hill coefficient fixed to unity was determined by using an F test.
Receptor Up-Regulation Studies.
CHO-hM3 cells, approximately 70 to 80% confluent in 24-well plates, were incubated with the indicated concentrations of antagonists for 30 h. At the end of that period, cells were subjected to an acidic wash to remove surface-bound ligands. Briefly, acidic washes with Ham's F12 medium (with pH set at 2) were allowed to proceed for 30 min, with medium being exchanged every 4 min. Control binding experiments were conducted and showed that this procedure fully removed surface-bound ligands while leaving the receptor surface intact (see Fig. 5A). Afterward, cells were rinsed twice with ice-cold binding buffer (Hanks' balanced salt solution, 10 mM Hepes, and 0.05% bovine serum albumin) to set the pH back to neutral, then receptor expression on the cell surface was monitored by binding of [3H]tiotropium at saturating concentrations (approximately 2 nM). After 1-h incubation, free radioligand was removed by rapid washing of wells with 3 × 1 ml of ice-cold buffer. Cell monolayers were dissolved in 250 μl of 0.1 M NaOH, and radioactivity was determined by liquid scintillation counting in a Tri-Carb beta counter device (PerkinElmer Life and Analytical Sciences).
Chemical Stability Studies.
To analyze their degradation kinetics, compounds were dissolved in 0.1 mol/l KH2PO4 buffer, pH 7.4 at a concentration of approximately 0.35 mg/ml. The solutions were filled in 2-ml HPLC glass vials and placed in a thermostated autosampler that was held at 37°C. HPLC diagrams were collected every 2 h for a total of 20 h. The reversed-phase HPLC [Agilent Technologies (Santa Clara, CA) HP 1100 with diode array detector, equipped with Prontosil C18 ace-EPS column, 125-mm length, 4.6-mm inner diameter] setup consisted in a gradient between eluent A (10 mM ammonium acetate solution, pH 7.0) and eluent B (acetonitrile) with a flow rate of 1 ml/min and quantitative detection at 240 nm. The obtained diagrams were analyzed by using a first-order degradation approach.
All experiments were analyzed by either linear or nonregression analysis with the equations mentioned under the different assay methodologies by using Prism version 5.02 (GraphPad Software Inc.). Individual estimates (either pA2, pIC50, or pEC50 values) were obtained from each experiment and then averaged to provide mean data (± S.E.M.). The statistical significance of differences between data was determined by either Student's t test or one-way analysis of variance with Dunnett's post test for multiple comparisons, as indicated in detail in the legends for Tables 1 and 2.
Selection of an Appropriate Reporter Gene for Monitoring hM3-R Activation.
The hM3-R couples to Gq proteins, resulting in activation of phospholipase C (Wess, 1993). Functional antagonism of a series of muscarinic antagonists was previously characterized in our laboratory by measuring changes in the intracellular concentrations of inositol phosphates (InsPs) (Casarosa et al., 2009). Although having several advantages, such as the ease of performance, sensitivity, and reproducibility, this assay has a major limitation: its stimulation time cannot be prolonged for more than 1 to 2 h. This aspect is of particular importance when considering ligands with slower kinetics of dissociation from the hM3 receptor, the so-called long-acting muscarinic antagonists (LAMAs), e.g., tiotropium (dissociation t1/2 27 h), aclidinium (t1/2 11 h), and glycopyrrolate (t1/2 6 h). LAMAs are unable to reach pseudoequilibrium within this time frame, resulting in deviation of the Schild plot from unity (Casarosa et al., 2009).
To overcome this problem, we intended to set up a reporter gene assay with longer incubation times, using the same CHO-hM3 cell line previously used for the InsP assay, to allow for direct comparison of results. Given the Gq coupling of the hM3-R, efforts concentrated on the firefly luciferase reporter construct under the control of the AP-1 promoter, which typically depends on protein kinase C activation (Angel and Karin, 1991). As expected, the agonist carbachol activated the AP-1-luciferase in a concentration-dependent manner (Figs. 1 and 2). As a control, empty CHO cells were transfected with AP-1. As expected, carbachol was unable to induce AP-1-driven luciferase in the absence of the hM3-R (data not shown).
Subsequently, different parameters, such as cell number and amount of DNA transfected, were optimized to increase assay sensitivity and reproducibility (see Materials and Methods for detailed information). It is noteworthy that different incubation times were tested to ensure reaching of pseudoequilibrium in the presence of LAMAs. Figure 1 shows results obtained with CHO-hM3 cells transfected with the AP-1 reporter and incubated with carbachol in the presence of tiotropium for 5, 8, and 18 h. Increasing the assay incubation to 18 h generated a more robust signal with an improved assay window (the signal-to-background ratio was above 8, compared with less than 2 obtained with shorter incubations). In addition, 18 h was the only time point investigated that delivered a correct determination of a pA2 value through Schild analysis, with a slope of 1.06 (Fig. 1D). Consequently, incubation was allowed to proceed for 20 h in all subsequent luciferase experiments.
Under these conditions, the effects of the different agonists (carbachol, muscarine, oxotremorine, and pilocarpine) were tested (Fig. 2). Each agonist stimulated a concentration-dependent increase in luciferase, with the following pEC50 values: carbachol, 6.5 ± 0.1; muscarine, 6.7 ± 0.1; oxotremorine, 6.9 ± 0.1 and pilocarpine, 6.0 ± 0.1; n ≥ 3. The reporter assay also allowed for a sensitive assessment of agonist intrinsic activity. Compared with carbachol and muscarine, which behaved as full agonists, oxotremorine and pilocarpine exhibited intrinsic activities of 81 ± 3 and 52 ± 2%, respectively (n = 3).
Characterization of Antagonistic Behavior at the hM3-R with the AP-1-Driven Reporter Gene.
To measure the functional antagonism of different anticholinergics, the shifts in the carbachol response curve induced in the presence of different antagonist concentrations were plotted according to Schild analysis (e.g., in Fig. 3 shown for atropine and tiotropium) and pA2 values were determined (Table 1). All anticholinergics tested showed a competitive and surmountable antagonistic behavior (Fig. 3), as indicated by the parallel right shift of the agonist curves in the presence of increasing antagonist concentrations without any significant change in the maximal responses. For all tested antagonists, including LAMAs, the slope of Schild regression obtained with the reporter gene assay was not significantly different from unity over a wide range of antagonists' concentrations. This represents an advantage to previous results obtained with the InsP assay (Casarosa et al., 2009), where fitting the dose ratios over a large range of LAMA concentrations resulted in linear regressions with a slope significantly higher than 1 (slopes were 1.29 ± 0.05; 1.27 ± 0.07, and 1.27 ± 0.05 for tiotropium, aclidinium, and glycopyrrolate, respectively; see Fig. 3B for Schild analysis of tiotropium). These results indicate that, in contrast to the InsP accumulation, the reporter gene assay is ideal for testing antagonists with slower kinetics of receptor dissociation, because its longer incubation time allows the LAMAs to reach pseudoequilibrium and therefore a correct determination of pA2 values.
The pA2 values obtained for each muscarinic antagonist against carbachol in the InsP assay and the AP-1 reporter gene are in close agreement with each other (Table 1), as expected for competitive (i.e., orthosteric, surmountable, and reversible) antagonists inhibiting a common pathway (i.e., Gq coupling). The only exception is represented by aclidinium, which shows a significantly weaker inhibitory potency in the AP-1 luciferase reporter gene assay (pA2 = 9.2) than in the InsP assay (pA2 = 10.0). Given the direct correlation between the decrease in antagonistic potency and the duration of the assay, instability of aclidinium in aqueous solution was investigated as a potential cause. Indeed, kinetic studies addressing the stability of aclidinium showed that its half-life is 3.1 h in buffered solution (pH 7.4) at 37°C (Fig. 4). As a control, chemical degradation of other two muscarinic antagonists that also possess the ester function (ipratropium and NMS) was investigated as well. As shown in Fig. 4, the degradation half-lives are 257 h for NMS and approximately 2300 h for ipratropium. These data, in agreement with the functional readouts, indicate that chemical degradation is not a necessary consequence of the presence of an ester function, but rather a specific characteristic of aclidinium.
Detection of hM3-R Constitutive Activity.
Changes induced by muscarinic antagonists on the basal levels of luciferase expressed in CHO-hM3 cells were investigated as well. In the absence of any added agonist, all tested antagonists decreased AP-1-driven luciferase activity below basal values, suggesting the presence of constitutive activity of the hM3-R in this CHO-hM3 system (Fig. 5). To confirm the M3-mediated constitutive signaling and rule out potential contamination with endogenous ACh, experiments were performed in the presence of acetylcholinesterase from E. electricus (electric eel) type V-S. The addition of as many as 10 units/ml to the cell culture media, which is sufficient to hydrolyze 10 μmol of ACh per min, did not attenuate the basal hM3-mediated signaling nor the observed negative intrinsic activity displayed by muscarinic antagonists in this assay (data not shown). As an additional control, empty CHO cells transfected with AP-1 luciferase were incubated with NMS. The muscarinic antagonist was unable to affect the basal luciferase activity in the absence of the hM3-R (data not shown).
Compared with the agonist-induced response, approximately 30% of the total hM3 response corresponds to basal, agonist-independent signaling (Fig. 5A). All of the antagonists tested behaved as full inverse agonists concentration-dependently decreasing the M3-R basal activity to the same minimum (Fig. 5, B and C).
The half-maximal inhibitory concentrations (IC50) of the different anticholinergics necessary to block the hM3-R constitutive signaling are reported in Table 1. When comparing their potencies in inhibiting the agonist-induced response (pA2 values against carbachol) and the constitutive signaling of the hM3-R (pIC50) in the same luciferase assay, the tested anticholinergics revealed different behaviors: whereas most (i.e., atropine, pirenzepine, ipratropium, glycopyrrolate, and DAMP) were equipotent as antagonists and inverse agonists, others (i.e., NMS and tiotropium) showed a significantly higher propensity to inhibit the agonist response than the constitutive activity of the receptor (3.16- and 8.3-fold difference for NMS and tiotropium, respectively). Aclidinium was the only compound to show the opposite trend, i.e., higher potency as inverse agonist; however, because of the compound's inherent instability, large variability was present in the different assays (see, e.g., Fig. 5B), and therefore a large error was associated with the measurements, hampering the reaching of statistical significance.
Taken together, these results suggest that the constitutively active M3-R* and the agonist-stabilized M3-AR*, although activating the same pathway, represent distinct conformational states of the receptor, which are differently recognized by some, but not all, antagonists.
Effects of Chronic Exposure to Inverse Agonists on Receptor Expression Levels.
Next, we assessed whether coincubation with ligands shown to suppress hM3-R constitutive activity could affect receptor expression levels. In these experiments, CHO cells were incubated for 30 h with the different antimuscarinics before M3-R expression at the cell surface was measured with a binding assay with [3H]tiotropium as marker.
Given the slow dissociation kinetics of some of the muscarinic antagonists, the LAMAs (Casarosa et al., 2009), care was taken to develop a washing procedure that would fully remove these compounds bound to the M3-R and allow [3H]tiotropium occupation of the entire cell-surface receptor population. For this control experiment, cells were incubated for 1 h at 4°C (to prevent any change in receptor expression) with the LAMAs (data shown for tiotropium in Fig. 6A), then several buffers set at different pHs were used to extensively wash the cells. As shown in Fig. 6A, complete ligand removal could be obtained by washing with buffers at low pH.
Consequently, after the 30-h incubation at 37°C in the presence of the different concentrations of muscarinic antagonists, cells underwent a 40-min washing procedure, followed by binding with saturating concentrations of [3H]tiotropium, as described in detail under Materials and Methods.
All tested compounds caused concentration-dependent increases in receptor expression levels (Fig. 6, B and C and Table 2), with EC50 values in good agreement with their potency as inverse agonists (Tables 1 and 2). However, the changes in receptor expression differed significantly in the presence of the different ligands. Whereas most antagonists (i.e., atropine, ipratropium, glycopyrrolate, pirenzepine, and aclidinium) caused a significant up-regulation above 40% of basal hM3-R expression levels, a second group of antagonists (tiotropium, 4-DAMP, and NMS) induced less than 20% up-regulation.
As a control, a second cell line was used (CHO-hβ2AR), where the human β2 adrenoceptor is stably transfected as part of the pkCREH mammalian expression vector (i.e., the same vector used for the hM3-R). CHO-hβ2 cells were incubated for 30 h with different muscarinic antagonists (ipratropium, tiotropium, atropine, and NMS; 100 nM), then β2 receptor expression was monitored by radioligand binding (using [3H]CGP12177). No changes in total human β2 receptor expression were seen upon exposure to the different muscarinic antagonists (data not shown), suggesting that ligand-induced up-regulation of the hM3-R takes place through binding and stabilization of the receptor expressed at the cell membrane and not through regulation of mRNA transcription.
Detection of Different hM3-R Active Conformations.
Here, we describe the setup of a readout system, an AP-1-driven reporter gene, which allows the detection of both agonist-induced and constitutive activity for the hM3-R. Pharmacological characterization of a series of anticholinergics indicates disparate pharmacological behavior upon receptor interaction. Although all of the antagonists studied here can be classified as full inverse agonists based on their ability to decrease agonist-independent activation of the AP-1 transcription factor, we found interesting differences in the pharmacology of these compounds when comparing their respective potencies as neutral antagonists (i.e., their pA2 values against carbachol) and inverse agonists (Table 1). Our interpretation of these data implies the presence of different active hM3-R conformations, either stabilized by agonists such as carbachol and muscarine (M3-AR*) or the agonist-free active conformation itself (M3-R*). Importantly, antagonists tested in these assays are orthosteric to each other and the agonist carbachol (Casarosa et al., 2009), ruling out the probe dependence seen with allosteric antagonists as a potential explanation for these pharmacological discrepancies.
The present finding that M3-R* and M3-AR* represent different conformations, which are differently recognized by some antagonists, is in general agreement with the emerging concept of multiple active states for a given receptor. Observations supporting this novel concept have been reported for many GPCRs (Kenakin, 2003). However, most studies have focused on inversion of agonists' potencies or efficacies when looking at different signaling pathways mediated by the same receptor, supporting the idea that agonists select among distinct receptor conformations to activate different downstream effectors, as seen e.g., with GPCRs showing pleiotropism in G protein coupling (Kenakin, 2003). Instead, our current results, i.e., the pharmacological differences observed with some antagonists in blocking agonist-induced and constitutive activation of the AP-1 transcription factor, suggest that for the hM3-R multiple active conformations exist for the regulation of the same signaling pathway, namely Gq coupling. These findings are in agreement with biophysical studies obtained with β2 adrenoceptors fluorescently labeled at Cys265, a region that is sensitive to receptor conformational changes (Ghanouni et al., 2001). Fluorescence lifetime spectroscopy showed that different β-agonists produced different arrays of receptor conformations, consistent with ligand-selective active states. Instead of a single receptor active state for G protein activation, there is an ensemble of microconformations (Kenakin, 2002) that are all capable of producing the same pharmacological effect (in this case, Gq protein activation) but have different overall tertiary conformations.
Detection of constitutive activity also allowed for pharmacological differentiation among the tested anticholinergics: although for most compounds pA2 values and inverse agonist potencies were in close agreement, NMS and tiotropium showed a significantly lower potency as inverse agonists (without affecting their negative intrinsic activity). These data suggest that the different anticholinergics stabilize different conformations and possibly that they achieve their inverse agonistic effect via distinct mechanisms of action. Classically, inverse agonists are envisioned as ligands that stabilize the inactive R conformation (Leff, 1995). One alternative mode of inverse agonist action has been proposed (Strange, 2002): it does not involve a redistribution of receptor states but one in which the inverse agonist switches the active conformation of the receptor to an inactive state that still retains the ability to bind and sequester G proteins but is unable to activate them. A species of receptor that is inactive but still binds G protein is one of the key features of the cubic ternary complex model (Weiss et al., 1996), which distinguishes it from the extended ternary complex model.
Differences in Antagonist-Mediated hM3-R Up-Regulation.
It is generally assumed that inverse agonists might cause receptor up-regulation: their inhibition of the spontaneous formation of receptor active states and consequent phosphorylation and internalization, coupled with normal receptor synthesis and expression, would lead to an increased surface density of receptor (Milligan and Bond, 1997). However, most studies analyzing the effects of inverse agonists on receptor up-regulation were performed with CAM receptor mutants (Zeng et al., 2003; Dowling et al., 2006). Unfortunately, interpreting results obtained with CAM receptors is not always straightforward: the creation of a CAM can diminish stabilizing constraints within the receptor, leading to an inherently unstable receptor that is more susceptible to destabilization and/or proteolytic degradation (Milligan and Bond, 1997). Consequently, the expression level of the mutant is increased by any ligand, either agonist or antagonist, regardless of its efficacy, as shown, e.g., with a CAM chimeric M3-R (Zeng et al., 2003). Here instead, we tested the ability of the different anticholinergics to up-regulate the wild-type hM3-R, ruling out potential artifacts. Again, different profiles emerged: although most compounds induced a similar level of receptor up-regulation (40–50% versus untreated) with a good correlation between their potency as inverse agonists and EC50 values for receptor up-regulation, a second set of compounds (NMS, tiotropium, and DAMP) caused a significantly reduced hM3-R up-regulation (less than 20%). Differences in up-regulation do not relate to differences in ligands' negative intrinsic activity, because all tested anthicholinergics behaved as full inverse agonists. However, a correlation between induction of up-regulation and differential propensity to behave as inverse agonists seems plausible, as two of the three compounds that are causing a less pronounced M3-R up-regulation (tiotropium and NMS) show a significant difference in their potencies as neutral antagonists (pA2 values) and inverse agonists (Table 1). The different ability to induce receptor up-regulation supports a model where the anticholinergics stabilize distinct M3-R conformations, all equally effective in blocking G protein signaling (though possibly through different mechanisms), but differently able to affect receptor internalization and its steady-state expression at the cell membrane. This is in agreement with a model where internalization and G protein activation are mediated by distinct receptor conformations, as suggested by the ability of antagonists or inverse agonists to elicit internalization (Azzi et al., 2003).
Physiological Relevance of hM3-R Constitutive Activity.
Presently, the physiological importance of constitutive activity for GPCRs is unclear. Accumulating evidence for its potential role comes from studies showing constitutive signaling in cells endogenously expressing the receptor of interest (de Ligt et al., 2000). However, evidence of GPCR constitutive activity in vivo is still scarce, with few excellent studies detailing inverse agonism at GPCRs expressed in vivo in native systems (e.g., Morisset et al., 2000; De Deurwaerdère et al., 2004). In addition, somatic receptor mutations leading to CAM receptors are a causal factor in certain diseases, e.g., retinitis pigmentosa, congenital night blindness, and familial hyperthyroidism (de Ligt et al., 2000). The extent to which GPCRs are constitutively active depends on the level of receptor and G-protein expression levels: in general, in efficiently coupled tissues with large receptor reserves, this may be an important factor in the pharmacological control of tissue function (Kenakin, 2004).
The physiological role of hM3-R constitutive signaling is currently unknown. There is, however, evidence that an increased cholinergic tone may be an important feature of chronic pulmonary diseases such as COPD (Barnes, 2004a,b). This increased cholinergic tone might, at least in part, result from an increase in the constitutive signaling of hM-3Rs, as suggested by recent studies indicating an up-regulation of M3-R in COPD patients, compared with healthy individuals (Profita et al., 2005, 2009). In addition, increasing the expression levels of effector proteins influences the basal signaling of GPCRs: for example, artificial overexpression of Gαq proteins greatly enhances the constitutive activity of hM3-R in a cellular model (Burstein et al., 1995). To this end, it has been shown that proinflammatory stimuli known to play a role in the pathogenesis of COPD, such as tumor necrosis factor α and cigarette smoke, increase the expression of Gαq and Rho proteins (Hotta et al., 1999; Chiba et al., 2005), which are known signaling partners for the hM3-R. If indeed constitutive activity was operative in this pathology, inverse agonists would be uniquely valuable in the retardation of disease progression. An important appendix to the use of inverse agonists in the clinical practice relates to their potential to induce receptor up-regulation and tolerance upon chronic treatment. Indeed, studies have indicated that the up-regulation of muscarinic receptors as a consequence of chronic atropine administration in animals (Chevalier et al., 1991; Wall et al., 1992). In particular, chronic exposure to atropine resulted in an up-regulation of M3-R that was associated with enhanced airway smooth muscle contraction in rabbits (Witt-Enderby et al., 1995), suggesting that the phenomenon might be of relevance with chronic treatment in clinical use.
Negative efficacy is a molecular property of anticholinergics, which might be desirable in clinical use for COPD, but attention should be paid to their potential for receptor up-regulation, leading to tolerance and possibly a rebound effect upon antagonist withdrawal. Importantly, our results indicate that it is possible to dissociate inverse agonistic properties from potential to induce up-regulation, as it has been seen, for example, with tiotropium. In agreement with these results, recent data from the UPLIFT (Understanding Potential Long-Term Impacts on Function with Tiotropium) study involving almost 6000 COPD patients indicated no loss of tiotropium bronchodilatory effects throughout the trial, which lasted 4 years, ruling out tolerance to this drug in the clinical setting (Tashkin et al., 2008).
- Received October 30, 2009.
- Accepted December 23, 2009.
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
- G protein-coupled receptor
- human muscarinic receptor
- long-acting muscarinic antagonist
- chronic obstructive pulmonary disease
- Chinese hamster ovary
- activator protein 1
- N-methyl scopolamine
- fetal calf serum
- high-performance liquid chromatography
- inositol phosphate
- constitutively active mutant
- Copyright © 2010 by The American Society for Pharmacology and Experimental Therapeutics