Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Cholinergic neurotransmission at the postganglionic synapses of the parasympathetic nervous system is mediated by the activation of muscarinic receptors. Muscarinic receptor subtypes with distinct but homologous gene sequences have been identified, cloned, and classified as m1, m2, m3, m4, and m5. Structural and pharmacological criteria suggest the presence of at least four subtypes, denoted M₁, M₂, M₃, and M₄, whereas a physiological role for the m5 gene product remains to be identified (Bonner et al., 1988; Buckley et al., 1989). M₁ receptors have been localized to neuronal tissues, including in the brain, and have been implicated in memory circuits. Centrally active muscarinic antagonists can have diverse central nervous system (CNS) side effects, including drowsiness and excitation. M₂ receptors regulate cardiac pacemaker tissues. M₃ receptor-mediated stimulation of smooth muscle can result in diverse effects that include bronchoconstriction, visual accommodation, gastrointestinal peristalsis, and increased detrusor muscle tone in the urinary bladder. The role of the M₄ receptors has not been clearly delineated. The differential localization and functions of these receptor subtypes raise the possibility of designing antagonists that selectively interact with distinct subtypes, thus avoiding the occurrence of adverse effects.

It has been reported that McN-A-343-induced inhibition of twitch contraction in isolated rabbit vas deferens is mediated through the M₁ receptor, and that carbachol-induced bradycardia in isolated right atria is mediated through the M₂ receptor (Eltze, 1988; Lambrecht et al., 1989). It is considered that carbachol-induced contractile responses in isolated rat trachea, ileum, and bladder are mainly mediated through the M₃ receptor. Therefore, these functional assays are frequently used to evaluate the antagonistic activity for muscarinic receptor subtypes. However, smooth muscle M₂ receptors were recently shown to reverse -adrenoceptor-mediated relaxation, which was considered as indirect contraction (Hegde et al., 1997). 4-Diphenylacetoxy-N-methyl piperidine methiodide and para-fluorohexahydrosiladifenidol were popularly used as selective M₃ antagonists in the pharmacological characterization of muscarinic receptor subtypes. Although their M₃ selectivity over M₂ receptors is only 10- to 20-fold in binding assays to cloned human muscarinic receptors, it is believed that the muscarinic M₃ subtype is mainly responsible for muscarinic smooth muscle contraction (Maclagan and Barnes, 1989; Eglen et al., 1996). Because of this limited selectivity of the antagonists between M₂ and M₃ subtypes, their pharmacological discrimination is important.

There are a number of reports indicating that nonselective muscarinic antagonists increase acetylcholine release from isolated human and guinea pig bronchial tissues and enhance bronchoconstriction in certain animals by blocking neuronal M₂ receptors after vagal stimulation (Fryer and Maclagan, 1987; Minette and Barnes, 1988; Kilbinger et al., 1991; Alabaster, 1997). In addition, since M₂ blockade leads to severe adverse effects, such as tachycardia, a major limitation in the systemic use of the classic nonselective muscarinic antagonists is failure to obtain the desired therapeutic responses (Andersson, 1988; Feinberg, 1993). Therefore, M₃ antagonists with high subtype selectivity over M₂ receptors may be of value in anticholinergic therapy.

We designed a new class of 4-acetamidopiperidine derivatives as the lead structure based on the structural features of both the Sicos Nova compound and oxybutynin (Hock, 1967; Kaiser et al., 1993), and obtained compound A, (2R)-N-[1-(6-aminopyridin-2-ylmethyl)piperidin-4-yl]-2-[(1R)-3,3-difluorocyclopentyl]-2-hydroxy-2-phenylacetamide. The present report details the in vitro and in vivo pharmacological profiles of compound A, which is a novel muscarinic receptor antagonist with M₂-sparing antagonistic activity.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

In Vitro Studies

Binding Affinity for Human and Rat Muscarinic Receptor Subtypes. In competition studies, specific binding of [³H]N-methylscopolamine (NMS; New England Nuclear, Boston, MA) was determined using membranes from Chinese hamster ovary (CHO) cells expressing cloned human m1, m2, m3, m4, or m5 receptors (Receptor Biology, Baltimore, MD), rat m1 or m3 receptors (American Type Culture Collection, Manassas, VA), and rat heart tissue. These CHO cells expressing cloned rat m1 or m3 receptor, and rat heart tissue were homogenized in 3 volumes of 50 mM Tris-HCl (pH 7.4)-1 mM EDTA containing 20% sucrose with Polytron PT-10. The homogenates were centrifuged at 10,000g for 30 min at 4°C. The supernatants were centrifuged at 100,000g for 60 min at 4°C. The pellets were suspended in 50 mM Tris-HCl (pH 7.4)-5 mM MgCl₂ and centrifuged at 100,000g for 60 min at 4°C. The pellets were resuspended in the above-mentioned buffer (25 mg/ml for CHO cells expressing cloned rat m1 or m3, and 50 mg/ml for rat heart tissue) and stored at 80°C as membrane preparations. In the binding assay, the membrane preparations were incubated with 0.19 to 0.2 nM [³H]NMS in 50 mM Tris-HCl, 10 mM MgCl₂, and 1 mM EDTA (pH 7.4) for 2 h at room temperature. Final protein concentrations were 22 µg/ml (human m1), 70 µg/ml (human m2), 54 µg/ml (human m3), 20 µg/ml (human m4), 116 µg/ml (human m5), 481 µg/ml (rat m1 and m3), and 2500 µg/ml (rat heart). Assays were performed in a total volume of 500 µl. Nonspecific binding was measured in the presence of 1 µM NMS and it was less than 2% of total binding. Free and membrane-bound [³H]NMS was separated by filtration over glass filters (UniFilter-GF/C; Packard Instruments, Meriden, CT) using a cell harvester (Filtermate 196; Packard Instruments). Radioactivity was counted by a liquid scintillation counter (TopCount; Packard Instruments).

Antagonism Activity in Rat Right Atria and Trachea and in Rabbit Vas Deferens. Male Sprague-Dawley rats (270-470 g; Charles River Japan, Yokohama, Japan) were exsanguinated. The right atria and trachea were isolated. The trachea was prepared free of serosal connective tissue and single open ring preparations were obtained. The right atria and trachea were placed in 20- and 5-ml organ baths containing modified Krebs-Henseleit solution maintained at 32°C, were continuously aerated with 95% O₂ and 5% CO₂, and were connected to isometric transducers (TB-651T; Nihon Kohden, Tokyo, Japan) with sutures. Mechanical responses were recorded isometrically by a multichannel polygraph (RMP-6018; Nihon Kohden). Following stabilization of the right atria and the trachea with an initial tension of 0.5 and 1 g, respectively, cumulative concentration-response curves to carbachol were obtained before and after addition of the test compounds. For the right atria assay, the beating rate was measured and the responses were expressed as a percentage of basal beating rate. For the trachea assay, carbachol-induced contraction was measured and the responses were expressed as a percentage of the contractile response to carbachol (100 µM).

In Vivo Studies

Bronchoconstriction Assays. Acetylcholine-induced bronchoconstriction in anesthetized rats. Male Sprague-Dawley rats (390-490 g) were anesthetized with urethane (1 g/kg) and -chloralose (50 mg/kg) injected intraperitoneally. A tracheal cannula was inserted and the animals were ventilated (6 ml/kg, 90 strokes/min; model 683, Harvard Apparatus, South Natick, MA) with room air. A cannula was inserted into the jugular vein for drug administration. The animals were paralyzed with succinylcholine (10 mg/kg s.c.) and placed in the Plethysmograph-Box (PLYAN-model; Buxco Electronics, Sharon, CT). Airflow was measured by placing a pneumotachograph, which was connected to a differential pressure transducer (DP45-14; Validyne, Northridge, CA), between the tracheal cannula and the ventilator. Transpulmonary pressure was measured with a differential pressure transducer (DP45-28; Validyne): one port was attached to a sidearm adapter of the Plethysmograph-Box and the other port was left open to the atmosphere. Airflow and transpulmonary pressure were monitored by a Pulmonary Mechanics Analyzer (model 6; Buxco Electronics), which calculated lung resistance (R_L) and dynamic compliance (C_dyn) on a breath-by-breath basis.

Methacholine provocation test in anesthetized dogs. The methacholine provocation test by the oscillation method, a procedure used to measure airway responsiveness in humans, has been used to assess the bronchodilatory activity of several bronchodilators (Takishima et al., 1981; Aizawa et al., 1996). By modifying the clinical methodology, we evaluated the bronchodilatory activity of muscarinic antagonists in dogs (Sato et al., 1999). Beagle dogs weighing 10.0 to 12.6 kg were anesthetized with sodium pentobarbital (30 mg/kg i.v.). A cuffed endotracheal tube (7.5 mm i.d.) was inserted into the trachea and connected to an astograph (TCK-6100H; Chest, Tokyo, Japan) during spontaneous respiration. An astograph is an apparatus for a measuring airway hyperresponsiveness with a system that delivers serially increasing doses of methacholine from 12 identical nebulizers. Each nebulizer is connected to the main tube between the mouthpiece and the flowmeter. The air compressor can be switched from one nebulizer to the next at set time intervals. In this way, 12 kinds of aerosol can be sequentially delivered. Simultaneously, the methacholine dose-response curve on respiratory resistance (Rrs), which is defined as the flow resistance of the total respiratory system, including the lungs, airways, and chest wall, was obtained by the forced 3-Hz oscillation method. After baseline recording of Rrs, normal saline (0.9% NaCl) was inhaled through the first nebulizer for 1 min. This was followed by 1-min inhalations of methacholine through the next nebulizer in doubling doses from 5 × 10⁷ M (0.078 mg/ml) to 2.5 × 10⁴ M (40 mg/ml) without any intervals. When Rrs had become twice the initial value, salbutamol aerosol (1 mg/ml) was given for 2 min through the nebulizer to achieve bronchodilation. The methacholine provocation dose was evaluated as the cumulative methacholine unit causing a doubling of basal Rrs. One methacholine unit was expressed as 1-min inhalation at 1 mg/ml. Compound A was administered orally by gavage in the conscious state. Four, 8, 12, or 24 h later, the methacholine provocation test was conducted. Bronchodilatory activity for each dose of compound A was expressed as shifts in the cumulative methacholine units between the nontreated control and the test drugs.

Bradycardia Assays. Male Sprague-Dawley rats (306-490 g) were used. The animals were anesthetized with urethane (1 g/kg) and -chloralose (25 mg/kg) injected intraperitoneally. The trachea, carotid artery, and jugular vein were cannulated after a midline neck incision. The animals were pretreated with succinylcholine (10 mg/kg s.c.) and then ventilated (6 ml/kg, 90 strokes/min; model 683, Harvard Apparatus) with room air. Heart rate was integrated from the blood pressure signal. Bradycardia was induced by intravenous administration of acetylcholine (10 µg/kg) delivered into the jugular vein at 5-min intervals. Once three similar responses were obtained, test drugs were administered 5 min before the subsequent acetylcholine challenge.

Tremor Assays. Four- to 5-week-old male ICR mice (24.8-31.2 g; Charles River Japan) were used. The test drugs were administered intravenously 5 min before i.p. injection of oxotremorine (0.4 mg/kg). After treatment with oxotremorine, mice were scored as having or not having tremor during a 15-min observation period. No apparent effects or slight effects (creeping movements) were graded as zero (protected animal); intermittent or constant whole body tremor was given a value of 1.

Brain Concentration Assays (Brain Penetrability). Test drugs were intravenously injected into male Sprague-Dawley rats. Fifteen minutes, 1 h, or 2 h later, blood samples were collected from the inferior vena cava with a heparinized syringe under light ether anesthesia and plasma was obtained by centrifugation and stored at 80°C for the high performance liquid chromatography assay with quadrupole tandem mass spectrometry. The brain was quickly removed, weighed, and homogenized in 4 volumes of ice-cold 50 mM Tris-HCl, 10 mM MgCl₂, and 1 mM EDTA, pH 7.4. Fifty microliters of the brain homogenates or the plasma sample was mixed with acetonitrile/methanol (50:50, 100 µl). After centrifugation (8000g, 4°C, 5 min), the supernatants were filtered over a Samprep C02-LH (Millipore, Bedford, MA). The plasma and brain concentration of the test drugs was determined by quadrupole tandem mass spectrometry. The K_p values were calculated as the ratio of brain level to plasma level. Brain penetration was expressed as a percentage of the total amount of administered compound.

Ex Vivo [³H]Pirenzepine Binding Assays for Rat Cerebral Cortex. Test drugs were intravenously injected into male Sprague-Dawley rats. The rats were sacrificed at 15 min after dosing. The brain was quickly removed and the cerebral cortex was rapidly dissected. The cerebral cortex was weighed and homogenized in 10 volumes of ice-cold 50 mM Tris-HCl, 10 mM MgCl₂, and 1 mM EDTA, pH 7.4. The inhibition of binding of [³H]pirenzepine (New England Nuclear) binding to the M₁ receptor in 20 µl of cerebral cortex membranes was determined by adding 2 nM [³H]pirenzepine in 500-µl total volume of the above-mentioned buffer. After a 2-h incubation at 25°C, free and membrane bound [³H]pirenzepine were separated by filtration over glass filters (UniFilter-GF/C; Packard Instruments) using cell harvester (Filtermate 196; Packard Instruments). The radioactivity was counted by a liquid scintillation counter (TopCount; Packard Instruments). Nonspecific binding was measured in the presence of 10 µM atropine.

Expression of Results. Values are expressed as mean ± S.E. unless otherwise noted. Statistical analyses were performed by ANOVA and post hoc multiple comparison was performed with the modified t test (Dunnett's) for the bronchoconstriction assay and Fisher's exact test for the tremor assay. Estimates of inhibitory potency for acetylcholine-induced bronchoconstriction (increases in R_L) and acetylcholine-induced bradycardia (decreases in heart rate) were expressed as ED₅₀ [50% inhibitory dose (mg/kg i.v.) of antagonist against vehicle-treated control response]. The ED₅₀ values were calculated from dose-response (% inhibition) curves using least-squares linear regression analysis. The dose-response curves were obtained on the basis of the percentage of inhibition at each log-dose level compared with the vehicle-treated control group. For methacholine provocation test, statistical analyses were performed between the nontreated control and the test drugs with paired Student's t test. p values less than 0.05 were considered significant.

Materials. Compound A (Fig. 1) was synthesized at the Tsukuba Research Institute of Banyu Pharmaceutical Co., Ltd. as described in WO 9805641. Darifenacin and revatropate were synthesized as described in Pfizer's published patent application. [³H]NMS and [³H]pirenzepine were purchased from New England Nuclear. All other reagents were purchased from Sigma Chemical Co. (St. Louis, MO).

View larger version (8K): [in this window] [in a new window]   Fig. 1.   Chemical structure.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

In Vitro Studies

Binding Affinity for Human and Rat Muscarinic Receptor Subtypes. In competition binding assays, compound A inhibited [³H]NMS binding to cloned human muscarinic m1, m2, m3, m4, and m5 receptors expressed in CHO cells with K_i values of 1.5 ± 0.06, 540 ± 20, 2.8 ± 0.03, 15 ± 0.58, and 7.7 ± 0.26 nM, respectively (Fig. 2; Table 1). Its affinity for human m3 receptors was 190-fold greater than that for human m2 receptors. The selectivity of compound A for the human m3 over the human m2 receptor was much greater than that of darifenacin (56-fold). Atropine and ipratropium nonselectively inhibited [³H]NMS binding to all of the muscarinic receptor subtypes. In addition, the binding affinities of compound A to the rat muscarinic receptor subtypes were in agreement with those obtained in the binding assays using cloned human receptors (Table 2). In the saturation binding study, Scatchard analysis indicated that compound A reduced the apparent affinity of [³H]NMS but caused no change in B_max (Fig. 3), and the resulting K_i values were consistent with those obtained in the competition assays. These results indicate that compound A binds reversibly and competitively with high affinity to muscarinic M₃ receptors, and has extremely high selectivity for the M₃ over the M₂ receptor.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Displacement of [3H]NMS binding to cloned human m1 (left), m2 (center), and m3 (right) receptors expressed in CHO cells. [3H]NMS binding was measured in the absence and presence of compound A (), darifenacin (), ipratropium (), or atropine () as described under Experimental Procedures. Each data point is presented as the mean ± S.D. of three experiments.

View this table: [in this window] [in a new window]   TABLE 1 Binding affinities of compound A and reference drugs to cloned human muscarinic receptor subtypes The Ki values were calculated by the method of Cheng and Prusoff (1973) and are presented as the mean ± S.E. of three experiments. The values in parentheses indicate the selectivity of binding affinity against cloned human m3 receptor and the value of Hill slope, respectively.

View this table: [in this window] [in a new window]   TABLE 2 Binding affinities of compound A to rat muscarinic receptor subtypes The Ki values were calculated by the method of Cheng and Prusoff (1973) and are presented as the mean ± S.E. of three experiments. The values in parentheses indicate the selectivity of binding affinity against cloned rat m3 receptor and the value of Hill slope, respectively.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   [3H]NMS saturation binding to cloned human m3 receptors expressed in CHO cells in the presence or absence of 10 nM compound A. Kd and Bmax values were as follows: human m3 control (), 81 and 64.2 pM; human m3 + 10 nM compound A (), 507 and 66.0 pM, respectively. The Ki values for compound A to m3 receptors were 1.9 ± 0.3 nM. The Ki value for compound A was calculated from the following relationship: Ki = Kd[Kd'  Kd] × [compound A], where Kd is the control value and Kd' is the value in the presence of 10 nM compound A. Data are presented as the mean ± S.D. of four experiments.

Antagonistic Activity in Rabbit Vas Deferens and Rat Right Atria and Trachea. Compound A showed concentration-dependent inhibition of McN-A-343-induced reductions in the twitch response to electrical field stimulation in rabbit vas deferens, carbachol-induced bradycardia in rat right atria, and carbachol-induced contraction in rat trachea, with K_B values of 11, 650, and 1.2 nM, respectively (Fig. 4; Table 3). The antagonistic effect of compound A in isolated tracheal tissue was 9.2- and 540-fold more potent than that in the vas deferens and cardiac tissue. Although the antagonistic effect of compound A in the trachea was 5 times less potent than that of darifenacin (K_B value of 0.22 nM, Table 3), the in vitro selectivity of compound A for tracheal over cardiac tissue (540-fold) was much greater than that of darifenacin (68-fold) and revatropate (10-fold). On the other hand, the in vitro selectivity of compound A for tracheal over vas deferens tissue (9.2-fold) was lower than that of darifenacin (15-fold). Revatropate in vas deferens tissues was 29-fold more potent than that in the tracheal tissues.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Functional antagonism by compound A (, vehicle; , 0.03 µM; , 0.1 µM; , 0.3 µM; , 3 µM; , 6 µM; , 10 µM) for rabbit vas deferens (left), rat atria (center), and trachea (right). For rabbit vas deferens assay, the dose ratio of EC50 values with or without the antagonist was 6.9 at 0.03 µM, 15.9 at 0.1 µM, and 29.6 at 0.3 µM, respectively. For rat atria assay, the dose ratio of EC50 values with or without the antagonist was 6.3 at 3 µM, 10.0 at 3 µM, and 21.4 at 10 µM, respectively. For rat trachea assay, the dose ratio of EC50 values with or without the antagonist was 32.5 at 0.03 µM, 106.0 at 0.1 µM, and 240.0 at 0.3 µM, respectively.

View this table: [in this window] [in a new window]   TABLE 3 Antagonism potencies (KB values) of compound A and reference drugs in isolated rabbit vas deferens, rat right atria, and rat trachea For rabbit vas deferens assay, contractions induced by electrical field stimulation (0.5 ms, 30 V, 0.05 Hz) were measured in the presence of yohimbine at 1 µM. The cumulative concentration-response curves to McN-A-343 were obtained before and after addition of test drugs. For rat right atria assay, the beating rate was measured and the cumulative concentration-response curves to carbachol were obtained before and after addition of test drugs. For rat trachea assay, carbachol-induced contractions were measured and the cumulative concentration-response curves to carbachol were obtained before and after addition of test drugs. The KB values were calculated from the following equation: KB = [C]/(concentration ratio  1), where concentration ratio is the ratio of EC50 values with or without the antagonist and [C] is the concentration of the antagonist. The values in parentheses indicate the selectivity of antagonist potency (KB) against carbachol-induced contractions in isolated rat trachea. The KB values are presented as the mean ± S.E. of 3 to 12 preparations.

In Vivo Studies

Bronchoconstriction Assays. Acetylcholine-induced bronchoconstriction in anesthetized rats. Intravenous administration of acetylcholine (50 µg/kg i.v.) produced a transient increase in R_L) and a decrease in C_dyn that peaked within 30 s after the injection and returned to the baseline within 2 to 3 min. Pretreatment with i.v. and p.o. compound A dose dependently inhibited the acetylcholine-induced increase in R_L (Fig. 5). The ED₅₀ values of compound A were 0.022 mg/kg for the i.v. dose and 0.24 mg/kg for the p.o. dose. In the time course study, pretreatment with compound A at 3 mg/kg p.o. induced 86.6 ± 2.3, 83.2 ± 6.2, 76.5 ± 9.1, 64.9 ± 6.0, and 60.7 ± 6.7% inhibition at 1, 2, 4, 6, and 8 h postdosing, respectively. In the same animals, the decrease in C_dyn, which was obtained simultaneously with the measurement of R_L, was also inhibited by compound A (data not shown).

View larger version (12K): [in this window] [in a new window]   Fig. 5.   Effect of compound A on acetylcholine-induced bronchoconstriction in rats. Compound A was administered i.v. () 5 min or orally () 1 h before acetylcholine (50 µg/kg i.v.) challenge, respectively. Data are presented as the mean ± S.E. of five animals. **p < 0.01 and ***p < 0.001, ANOVA followed by Dunnett's test versus vehicle.

Methacholine provocation test in anesthetized dogs. In the methacholine provocation test, compound A (0.1-1 mg/kg p.o.) dose dependently shifted the methacholine Rrs response curves to the right in anesthetized dogs (Fig. 6). Compound A at 0.1 mg/kg caused 12.8-, 7.8-, and 3.8-fold shifts of the cumulative methacholine units at 4, 8, and 12 h after administration, respectively. The inhibitory activity of compound A (0.3 and 1 mg/kg p.o.) on aerosolized methacholine-induced bronchoconstriction was present at 24 h.

View larger version (12K): [in this window] [in a new window]   Fig. 6.   Time course for the inhibitory effects of orally administered compound A in the dog methacholine provocation test. Compound A at 0.1, 0.3, and 1 mg/kg (p.o.) was administered at 8, 12, and 24 h before the methacholine provocation test. Shifts in the methacholine provocation dose were calculated as the drug-treated methacholine units divided by the nontreated control units in each animal. Data are presented as the mean ± S.E. of three animals. *p < 0.05, **p < 0.01, and ***p < 0.001, paired Student's t test.

Bradycardia Assays. In anesthetized rats, i.v. injections of acetylcholine (10 µg/kg) caused reproducible and transient decreases in heart rate. Compound A at i.v. doses as large as 3 mg/kg did not affect acetylcholine-induced bradycardia; however, partial inhibition was observed at 10 mg/kg i.v. As summarized in Table 4, the ED₅₀ value for compound A was approximately 10 mg/kg. In contrast, the ED₅₀ value for darifenacin (ED₅₀ = 0.18 mg/kg i.v.) is much lower than that of compound A. Compound A displayed substantial in vivo bronchoselectivity versus cardioselectivity (450-fold), and the bronchoselectivity was superior to that of darifenacin (11-fold).

Tremor Assays. In all vehicle-treated mice, i.p. injections of oxotremorine (0.4 mg/kg) produced intermittent or constant whole body tremor graded as score 1. Intravenous pretreatment with scopolamine completely inhibited the oxotremorine-induced tremor at a dose of 0.1 mg/kg, but NMS did not inhibit the tremor response even at the 30-fold higher dosage (3 mg/kg i.v.). Compound A was also without effect even when administered at a dose of 10 mg/kg i.v. (Table 5).

Brain Concentration Assays (Brain Penetrability). When administered to rats at a dose of 3 mg/kg i.v., the brain penetration rate (% penetration) and the K_p of compound A were lower than those obtained with darifenacin or scopolamine (Table 6). The time course study confirmed the time-dependent decreases in brain and plasma concentration after i.v. administration of compound A (brain concentration: 67 ± 2.4 ng/g at 15 min, 42 ± 1.8 ng/g at 1 h, 25 ± 2.6 ng/g at 2 h; plasma concentration: 510 ± 25 ng/ml at 15 min, 190 ± 17 ng/ml at 1 h, 85 ± 12 ng/ml at 2 h, n = 3, respectively). These results indicate that the brain penetrability of compound A is significantly less than that of darifenacin or scopolamine in rats.

Ex Vivo [³H]Pirenzepine Binding Assays in Rat Cerebral Cortex. Intravenous administration of scopolamine at a dose of 0.3 mg/kg inhibited almost completely the specific [³H]pirenzepine binding in rat cerebral cortex (81.9 ± 1.7%, n = 3). Compound A showed only slight inhibition even at an intravenous dose of 3 mg/kg (12.9 ± 2.2%, n = 3). The inhibitory activity was almost similar to that of NMS at 3 mg/kg i.v. (8.1 ± 6.4%, n = 3).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

We focused our efforts to make muscarinic receptor antagonists with M₂-sparing antagonistic activity. Compound A displays high selectivity for the M₃ over the M₂ receptor in rat as well as human binding assays. In rat isolated tissue assay, compound A has 540-fold in vitro selectivity for trachea over the cardiac muscarinic receptors. Furthermore, the results of in vivo rat bronchoconstriction and bradycardia assays indicate that compound A almost completely inhibits acetylcholine-induced bronchoconstriction at a dose of 0.1 mg/kg (i.v.) and that has approximately 450-fold in vivo selectivity for bronchoconstriction over bradycardia. These findings suggest that compound A is a highly selective M₃ antagonist in in vitro and in vivo functional assays as well as in radioligand binding assays, and that cholinergic contraction responses in airway smooth muscle are mainly mediated through the M₃, but not M₂, receptors. The net physiological role of the indirect contraction mediated by smooth muscle M₂ receptors is unknown in detail. Then, there remains some question whether blockade of M₂ and/or M₃ muscarinic receptors is the more rational approach for cholinergically evoked smooth muscle contractions, especially, under the pathological conditions. Although nonselective muscarinic antagonists have proven clinical efficacy in the treatment of obstructive airway diseases, gastrointestinal disorders and urinary dysfunction, a limitation in the systemic use of such compounds is failure to obtain desired therapeutic responses without concomitant side effects such as tachycardia, which is mediated by the cardiac M₂ receptor (Feinberg, 1993; Hendrix and Robinson, 1997). Thus, muscarinic M₃ receptor antagonists with high selectivity for the M₃ over the M₂ receptor such as compound A are anticipated to have minimal cardiac effects, and may be of value in anticholinergic therapy. In addition, recent studies suggest the existence of prejunctional M₂ receptors located on airway parasympathetic nerve terminals of various species, including human (Jacoby et al., 1993). The prejunctional M₂ receptors function as autoreceptors to inhibit further acetylcholine release (Kilbinger et al., 1991). Blockade of the prejunctional inhibitory receptors by both nonselective muscarinic antagonists such as ipratropium and M₂-selective antagonists, such as methoctramine, enhance the bronchoconstriction induced by vagal nerve stimulation in guinea pigs, and thereby weaken the functional blockade of postjunctional M₃ receptors in airway smooth muscle (Fryer and Maclagan, 1987). In addition, we recently demonstrated that the blockade of airway neuronal M₂ receptors stimulates SO₂-induced mucus hypersecretion in a rat bronchitis model (Aoki et al., 1999). Furthermore, M₂ receptors also have been identified in the bladder, and it appears that they have functions at least in part as inhibitory autoreceptors there as well. It has been suggested that M₃ antagonists with far greater selectivity for M₃ over inhibitory M₂ autoreceptors might lead to an ideal anticholinergic with greater efficacy than the existing nonselective antagonists.

Compound A has slightly higher affinity for M₁ compared with M₃ receptors. Since it had been suggested that McN-A-343-induced inhibition of twitch contraction in isolated rabbit vas deferens was mediated through M₁ receptors (Eltze, 1988), it was believed that the McN-A-343-induced response was useful as the functional assay for M₁ receptors. However, in our functional assay, compound A is more 9-fold less potent in rabbit vas deferens than rat trachea. This difference in potencies is similar to that for darifenacin (15-fold), although in binding assay darifenacin has 6.5-fold lower affinity for M₁ receptors. Interestingly, other investigators speculate that the presynaptic M₄ receptors would be presented in rabbit vas deferens (Sagrada et al., 1994). In our binding assay, compound A and darifenacin also has 5.4- and 10-fold lower affinity for M₄ compared with M₃ receptors. This selectivity is well consistent with the selectivity of rabbit vas deferens over rat trachea. Therefore, one possible consideration is that the McN-A-343-induced response may be complementarily induced by the both M₁ and M₄ receptors. The multiplicity of presynaptic muscarinic receptor subtypes in rabbit vas deferens remains to be clarified.

Since darifenacin was reported as a selective M₃ antagonist (Alabaster, 1997), we compared in vitro and in vivo selectivity of compound A with that of darifenacin. However, our parallel studies with darifenacin showed this drug to be less selective in in vivo functional assays. Alabaster's (1997) report suggests that darifenacin was able to differentiate between M₃ receptors in different tissues. When compared with atropine, darifenacin showed a degree of selectivity for guinea pig ileum relative to other smooth muscle preparation such as trachea. The molecular mechanism for this selectivity is not understood, since only one M₃ receptor has been identified from molecular sequence studies. Numerous functional studies indicate that the airways of several species, including rats and humans, contain at least three muscarinic receptor subtypes: M₁, M₂, and M₃ receptors. Therefore, it is considered that the coexistence of muscarinic receptor subtypes may modulate the potency of the inhibitory effects on muscarinic agonist-induced contraction and affect the in vivo selectivity for bronchial over cardiac tissue. Additionally, other investigators found some differences with regard to the affinities of antagonists to cloned versus endogenous muscarinic receptors (Eglen et al., 1996). These differences may arise from variations in experimental conditions such as the level of membrane receptor glycosylation or lipid concentration. Of course, we cannot rule out the possibility of the influence of tissue distribution after i.v. treatment with darifenacin.

The methacholine provocation test by the oscillation method, a procedure used to measure airway responsiveness in humans, has been used to assess the activity of several bronchodilators (Takishima et al., 1981; Aizawa et al., 1996). By modifying the clinical methodology, we recently established a simple and reproducible assay system to evaluate the bronchodilatory activity of muscarinic antagonists in dogs (Sato et al., 1999). In this dog methacholine provocation test, orally administered compound A at 0.1 to 1 mg/kg induces a potent and long-lasting bronchodilation. In our preliminary pharmacokinetic study, plasma C_max and area under the plasma-concentration time curve from time 0 to 10 h after dosing rose linearly, whereas plasma t_1/2 and oral bioavailability did not change following oral administration at this dose range. At 0.1 mg/kg (p.o.), plasma C_max and t_1/2 were 45 ± 10 ng/ml (101 ± 22.5 nM) and 8.2 ± 1.4 h, respectively. These pharmacokinetic profiles support that the in vivo bronchodilatory activity of compound A in dog methacholine provocation tests is nearly compatible with its in vitro M₃ antagonistic potency. Anticholinergics have been used in the treatment of obstructive airway diseases. Subsequent development of the quaternary derivatives of atropine such as ipratropium bromide resulted in the potential clinical use of muscarinic antagonists as aerosol agents. Although they are better tolerated than parenterally administered atropine, the quaternary derivatives may be not ideal anticholinergic bronchodilators because of their short duration and nonselective profile for muscarinic receptor subtypes. Compound A (0.1-1 mg/kg p.o.) displays long-lasting bronchodilation, and has high selectivity for M₃ receptors over M₂ receptors. Further study is needed to evaluate the tolerability of compound A, including anticholinergic adverse effects such as mydriasis and dry mouth.

Blockade of the central muscarinic receptors impairs performance on memory tasks and causes an amnesia-like syndrome in rodents, primates, and humans (Bartus et al., 1987; Broks et al., 1988; Preston et al., 1988). Recently, it was reported that oxybutynin, which has potent antimuscarinic activity, caused a decrease in the retention time of the passive avoidance response (Sugiyama et al., 1999). Thus, the CNS activity of compound A was estimated in three rodent assays. The first was the antitremor assay of compound A. In this assay, the tremor response is induced by oxotremorine, a central active muscarinic agonist (Shannon et al., 1994). In conscious mice, no evidence of inhibitory effects of compound A on oxotremorine-induced tremor was obtained even at a dose of 10 mg/kg i.v. Oyasu et al. (1994) reported that oxybutynin inhibited tremor in mice, and that the K_p until 4 h after i.v. dosing with labeled 1 mg/kg oxybutynin ranged from 1.58 to 3.70 in rats. The second assay was the brain penetrability of compound A: brain and plasma levels of compound A were measured after i.v. injection, and the K_p and brain penetration rate (% penetration) were less than those of reference drugs such as darifenacin and scopolamine. The third assay was the inhibition of compound A on ex vivo labeled-pirenzepine, selective M₁ antagonist, binding in rat cerebral cortex: inhibition potency of compound A was measured after i.v. injection, and its potency (% inhibition) was almost similar to that of NMS. Compound A has high affinity for M₁ receptors. M₁ receptors have been implicated in memory function. Taken together, these results in CNS assays strongly suggest that compound A serves less antagonistic activity for central M₁ receptors due to its less brain penetrability, and that compound A is a less brain-penetrant M₃ antagonist than is darifenacin or scopolamine in rats.

In conclusion, compound A has high selectivity for M₃ receptors over M₂ receptors, displays potent, oral M₃ antagonistic activity, and does not significantly inhibit central muscarinic receptors because of low brain penetration. Thus, compound A is anticipated to have fewer cardiac or CNS side effects, and may be an excellent pharmacological tool to explore the therapeutic utility of an M₂-sparing muscarinic antagonist.