The neurotransmitter 5-hydroxytryptamine (5-HT) mediates a broad range of physiological and pathophysiological responses both centrally and peripherally. High amounts of 5-HT are produced in the gastrointestinal (GI) tract, and 5-HT affects GI motility by signaling through receptors that are present throughout the GI tract (Gershon, 2004; Neal and Bornstein, 2006; Costedio et al., 2007; Gershon and Tack, 2007).

All classes of the extensive 5-HT receptor family, except for the ligand-gated 5-HT₃ receptor, are members of the seven transmembrane-spanning G protein-coupled receptor family. These receptors modulate signal transduction pathways via stimulating or inhibiting adenylyl cyclase, modulating cytosolic calcium concentrations, or activating multiple alternative downstream effectors.

The 5-HT₄ receptor is thought to signal principally, but not exclusively, via G_s-coupled with adenylyl cyclase activation (Bockaert et al., 2004). In the GI tract, 5-HT₄ receptor activation enhances the release of transmitters from excitatory neurons such as cholinergic neurons. Therefore, 5-HT₄ receptor agonists are used to treat bowel conditions that require stimulated propulsive activity (Gershon, 2004; Gershon and Tack, 2007). For 5-HT₄ agonists such as cisapride, mosapride, or tegaserod, this stimulatory activity has been demonstrated in GI preparations isolated from several different species, including the guinea pig, rat, mouse, and human.

Studies with cisapride, a 5-HT₄ agonist with weak 5-HT₃ antagonist and dopamine₂ (D₂) antagonist properties, established a good clinical precedence for treating gastroesophageal reflux disease and other gastric motility disorders. Considering that blocking 5-HT₃ receptors or D₂ receptors accelerates gastric emptying in rats (Zar et al., 1982; Albibi and McCallum, 1983; Costall et al., 1984; Yamano et al., 1997), 5-HT₄ agonism, 5-HT₃ antagonism, and D₂ antagonism may contribute to the GI prokinetic activity of cisapride.

Mosapride is a 5-HT₄ partial agonist with no D₂ antagonist properties that was designed to avoid D₂-mediated side effects (Yoshida et al., 1989). Mosapride increases gastric motility and emptying without signs of autonomic dystonia (Yoshida et al., 1989, 1991; Yoshikawa et al., 1998; Endo et al., 2002; Hiyama et al., 2007). Because of little arrhythmic effects associated with the QT prolongation, mosapride has been used clinically in some countries, including Japan. However, further analysis is needed to conclude that the prokinetic activities of mosapride are attributed only to 5-HT₄ agonism.

Tegaserod, a selective 5-HT₄ partial agonist, was developed for reduced GI motility and transit disorders, and it was approved to treat chronic idiopathic constipation and constipation-dominant irritable bowel syndrome in females in the United States (Camilleri, 2001). Further pharmacological studies have shown that tegaserod is a potent 5-HT_2B receptor antagonist, and they have raised the question regarding the importance of the 5-HT_2B antagonist activity of tegaserod in humans (Beattie et al., 2004). In addition, in 2007 the United States Food and Drug Administration discontinued tegaserod marketing because of cardiovascular safety concerns.

To confirm the role of 5-HT₄ agonism in the gastroprokinetic effects, we identified a potent and selective 5-HT₄ agonist, CJ-033,466, which was originally synthesized at Pfizer Laboratories, and it was specifically exemplified in Pfizer patent applications (WO2003035649 and WO2004026869). In this study, we report the in vitro pharmacological profile of CJ-033,466 and its effects on gastric motility and emptying rates in conscious dogs.

Materials and Methods

Experimental protocols were approved by the Animal Ethics Committee at the Nagoya Laboratories of Pfizer Global Research and Development.

Membrane Preparation

Membranes prepared from cells expressing each human receptor or rat tissue were used in the receptor binding assay. All membranes, except those from Chinese hamster ovary cells expressing human 5-HT_2A receptors (Euroscreen, Brussels, Belgium), 5-HT_2B receptor (Cerep, Celle L'Evescault, France), and 5-HT₇ receptors (PerkinElmer Life and Analytical Sciences, Meriden, CT), were prepared in house. Cells were harvested and homogenized in 50 mM Tris-HCl buffer, pH 7.7, supplemented with protease inhibitors for 5-HT_1A; 50 mM Tris-HCl buffer, pH 7.4, 2 mM MgCl₂, and protease inhibitors for 5-HT_1B and 5-HT_1D; 50 mM HEPES, pH 7.4, supplemented with protease inhibitors for 5-HT_4d; and PBS, pH 7.4, for D₂. Cell suspensions were homogenized, and then they were centrifuged at 40,000g at 4°C. The pellets were resuspended in 50 mM Tris-HCl buffer, pH 7.7, 4 mM CaCl₂, and 10 μM pargyline for 5-HT_1A; 50 mM Tris-HCl buffer, pH 7.4, and 4 mM CaCl₂ for 5-HT_1B and 5-HT_1D; 50 mM HEPES buffer, pH 7.4, for 5-HT_4d; and 20 mM HEPES buffer, pH 7.4, 120 mM NaCl, 1 mM EDTA, and 1 mM EGTA for D₂; and then they were homogenized.

Rat cortex (Japan SLC Inc., Shizuoka, Japan) was used as a receptor source for rat 5-HT₃ binding assay. Cortical tissue was homogenized in 50 mM Tris-HCl buffer, pH 7.5, supplemented with protease inhibitors, and then they were centrifuged at 1100g for 10 min at 4°C. The supernatant was further centrifuged at 48,000g for 15 min at 4°C. The pellet was suspended, homogenized in the same buffer, and centrifuged again in the same manner. The resultant supernatant was discarded, and the final pellet was resuspended in 50 mM Tris-HCl buffer, pH 7.5, and homogenized.

Dog striatum and rat brain without the cortex were used for dog and rat 5-HT₄ binding assays, respectively. Dog tissue was homogenized in 25 mM HEPES buffer, pH 7.4, and centrifuged at 48,000g for 15 min at 4°C. The pellets were resuspended, homogenized in 25 mM HEPES buffer, and then centrifuged as described above. The final pellet was resuspended and homogenized in 25 mM HEPES buffer, pH 7.4. Rat tissue was homogenized in 50 mM Tris-HCl buffer, pH 7.5, supplemented with protease inhibitors, and then it was centrifuged at 48,000g for 30 min at 4°C. The pellets were resuspended, homogenized, and centrifuged as described above. The final pellet was resuspended and homogenized in 50 mM Tris-HCl buffer, pH 7.5.

Each membrane fraction was stored at –80°C until use. Protein concentrations were determined using a bicinchoninic acid protein assay kit (Pierce Chemical, Rockford, IL).

Receptor Binding Assay

Radioligand binding assays were conducted at room temperature. Conditions are described in the following order: radioligand and its concentration, reagent for nonspecific binding and its concentration, assay buffer, and incubation time. Human 5-HT_1A: 0.8 nM [³H]8-hydroxy-2-dipropylaminotetralin, 100 μM 5-HT, 50 mM Tris-HCl buffer, pH 7.7, containing 4 mM CaCl₂ and 10 μM pargyline, 60 min; human 5-HT_1B: 5.0 nM [³H]5-HT, 10 μM 5-CT, 50 mM Tris-HCl buffer, pH 7.4, containing 4 mM CaCl₂, 10 μM pargyline, and 0.1% ascorbic acid, 60 min; human 5-HT_1D: 5.0 nM [³H]5-HT, 10 μM 5-CT, 50 mM Tris-HCl buffer, pH 7.4, containing 4 mM CaCl₂, 10 μM pargyline, and 0.1% ascorbic acid, 60 min; human 5-HT_2A: 1.0 nM [³H]ketanserin, 30 μM ketanserin, 50 mM Tris-HCl buffer, pH 7.5, 30 min; human 5-HT_2B: 0.2 nM [¹²⁵I]2,5-dimethoxy-4-iodophenyl-2-aminopropane, 1 μM 2,5-dimethoxy-4-iodophenyl-2-aminopropane, 50 mM Tris-HCl buffer, pH 7.4, 15 min; rat 5-HT₃: 1 nM[³H]BRL-43,694, 50 μM 5-HT, 50 mM Tris-HCl buffer, pH 7.5, 60 min; human 5-HT_4d: 0.2 nM [³H]GR-113,808, 1 μM GR-113,808, 50 mM HEPES buffer, pH 7.4, 60 min; human 5-HT₇: 1 nM[³H]5-CT, 10 μM 5-HT, 50 mM Tris-HCl buffer, pH 7.4, containing 10 mM MgSO₄ and 0.5 mM EDTA, 30 min; human D₂: 0.1 nM [³H]spiperone, 100 nM (+)-butaclamol, 20 mM HEPES buffer, pH 7.4, containing 120 mM NaCl, 1 mM EDTA, and 1 mM EGTA, 60 min; dog 5-HT₄: 0.05 nM [³H]GR-113,808, 30 μM 5-HT, 5 mM HEPES buffer, pH 7.4, 30 min; and rat 5-HT₄: 0.1 nM [³H]GR-113,808, 3 μM 5-HT, 40 mM Tris-HCl buffer, pH 7.5, 30 min.

Membrane fractions were incubated with tritiated radioligands in the absence or presence of serially diluted compounds. Except for the 5-HT_4d binding assay, the reaction was terminated by filtering the reaction mixture onto a GF/B filter (Whatman, Maidstone, UK). The quantity of radioactivity captured on the filter was measured by liquid scintillation counting. For the 5-HT_4d binding assay, wheat germ agglutinin-coated SPA beads (1 mg/well; GE Healthcare, Little Chalfont, Buckinghamshire, UK) were used, and the receptor-bound radioactivity was quantified by MicroBeta plate counting (PerkinElmer Life and Analytical Sciences, Waltham, MA). The K_i values were determined using the Cheng-Prusoff equation (Cheng and Prusoff, 1973).

Functional Assay

Human 5-HT_4d-transfected HEK293 cells (h5-HT_4d/HEK293) were grown at 37°C and 5% CO₂ in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum, 20 mM HEPES, 200 μg/ml hygromycin B, 100 units/ml penicillin, and 100 μg/ml streptomycin. Cells were passaged every 4 to 5 days with PBS/1 mM EDTA.

To measure intracellular cAMP production, h5-HT_4d/HEK293 cells were grown to 60 to 80% confluence. The medium was changed to DMEM containing 10% dialyzed fetal calf serum, and the cells were incubated overnight to remove endogenous 5-HT. On the day of the assay, samples were prepared in 96-well plates. To determine the maximal elevation of cAMP, 1 μM 5-HT was used. The cells were harvested with PBS/1 mM EDTA, centrifuged, and washed with PBS. The cell pellet was resuspended in DMEM supplemented with 20 mM HEPES, 10 μM pargyline, and 1 mM 3-isobutyl-1-methylxanthine at a concentration of 1.6 × 10⁵ cells/ml, and they were incubated for 15 min at room temperature. The reaction was initiated by adding cells to the plates (2 × 10³ cells/well). After a 15-min incubation at room temperature, 1% Triton X-100 was added to stop the reaction. After a 30-min lysis at room temperature, a cAMP-XL665 conjugate was added to the lysate followed by an anti-cAMP-cryptate conjugate. After a 60-min incubation at room temperature, samples were measured on a 1420 ARVOsx multilabel counter (PerkinElmer Life and Analytical Sciences; excitation, 320 nm; emission, 665 nm/620 nm; delay time, 50 μs; and window time, 400 μs).

Data were analyzed using the ratio of fluorescence intensity at 620 and 665 nm for each well. Enhanced cAMP production elicited by each compound was normalized to the amount of cAMP elevated by 1 μM 5-HT. Curve-fitting analysis was performed with Prism version 3.02 (GraphPad Software Inc., San Diego, CA).

Rat Tunica Muscularis Mucosae Assay

An additional functional assay for 5-HT₄ agonism was conducted using a rat tunica muscularis mucosae (TMM) tissue preparation. Male CD IGS rats were sacrificed by isoflurane inhalation, and a 2-cm segment of the intrathoracic esophagus was excised and placed in Krebs' solution (119 mM NaCl, 4.7 mM KCl, 25 mM NaHCO₃, 0.6 mM MgSO₄, 1.2 mM KH₂PO₄, 11.1 mM glucose, and 1.3 mM CaCl₂, pH 7.4). To obtain 5-HT₄ receptor-mediated relaxation without neurogenic and complicated contractions, the external muscularis propria containing the plexus of the esophagus was carefully removed to isolate the TMM as described by Baxter et al. (1991). In this preparation, 5-HT₄ agonists induce relaxation through intracellular cAMP production subsequent to direct activation of 5-HT₄ receptors located on the smooth muscle cells. The strips were suspended in a 10-ml organ bath containing Krebs' solution at 37°C with 95% O₂ and 5% CO₂ aeration under 0.5-g tension, and they were equilibrated for 15 min. In all assays, 3 μM indomethacin, 1 μM ketanserin, and 120 μM ascorbic acid were included in the Krebs' solution. Concentration-effect curves (CECs) were obtained after contracting the rat TMM with 1 μM carbachol. Responses were measured isometrically using the TB-612T transducer (Nihon Kohden, Tokyo, Japan) coupled to a PowerLab data acquisition system (ADInstruments, Colorado Springs, CO). Two CECs were constructed per tissue, the first CEC for 5-HT and the second CEC for CJ-033,466. Involvement of 5-HT₄ receptors was confirmed by preincubating tissues with 1 μM SB-203,186 (McLean and Coupar, 1995), a selective 5-HT₄ receptor antagonist, which was added to the bath 5 min before the addition of carbachol. Data for the CECs are expressed as the means ± S.E.M. Curve fits were analyzed by a sigmoidal dose-response (variable slope) analysis with nonlinear regression in Prism (GraphPad Software Inc.). The pEC₅₀ values were calculated based on the molar concentration of the test compound that produced 50% of the maximal response. The E_max value was calculated relative to the maximal response to 5-HT.

Measurement of Gastric Antral Motility in Conscious Dogs

Preparation of Animals. Five healthy male beagles (Oriental Yeast Co., Ltd., Tokyo, Japan) weighing 9 to 15 kg, were anesthetized with isoflurane, and the abdominal cavity was opened under aseptic conditions. Extraluminal force transducers (IS-12S; Star Medical, Tokyo, Japan) were sutured onto the selomuscular layer of the gastric antrum 3 cm proximal to the pyloric ring, according to the method of Itoh et al. (1977). The transducer's lead wires were taken out of the abdominal cavity through a skin incision between the scapulae. After surgery, the dogs were placed in protective jackets and housed in individual cages. Gastric motility was recorded starting at least 2 weeks after surgery.

Experimental Procedures. After an overnight fast, the dogs were placed in a shielded room, and gastric motility was recorded in the fasted state. Gastric motility was measured with a telemetry system (GTS-800; Star Medical), and data were acquired on a personal computer with acquisition software (Eight Star; Star Medical). After confirming the incidence of an interdigestive migrating complex (IMC, a typical motility pattern of the upper gastrointestinal tract in the fasted state) at regular intervals, CJ-033,466, cisapride, or vehicle (0.1% methyl cellulose) was administered orally, and gastric motility was recorded for 8 h.

To measure postprandial gastric motility, the dogs were fed 100 g of solid meal (DS-A, Oriental Yeast Co., Ltd.) after confirmation of IMC. The dogs ingested it within 10 min. Two hours after feeding, CJ-033,466, cisapride, or vehicle was administered orally. The postprandial motility pattern lasted at least 5 h after administration; therefore, motility was recorded for this duration.

To determine whether CJ-033,466 affects the 5-HT₄ receptor, SB-203,186, a selective 5-HT₄ receptor antagonist (1 mg/kg) or saline (vehicle for SB-203,186) was injected intravenously into dogs in the postprandial state (2 h after feeding with 100 g of solid meal). Immediately after, CJ-033,466 (0.1 mg/kg) or 0.1% MC (vehicle for CJ-033,466) was administered orally, and gastric motility was observed for 2 h.

Data Analysis. To measure the gastric motility quantitatively, motor indexes that represent areas of contractions were calculated. The signals from the force transducer were acquired on a personal computer and analyzed by processing software (Analyze II; Star Medical). Because the motility patterns between the fasted and postprandial states were dissimilar, different calculation formulae were used for each state. In the fasted state, the areas surrounded by the contraction curve and the baseline were determined every 2 h after administration. For standardization, the calculated areas were divided by the IMC peak height before administration, and they were used as the motor index (modified from a previous report; Sato et al., 2000). In the postprandial state, the areas for every 1-h period after administration were calculated, expressed as a percentage of the 1-h period before administration, and they were used as the postprandial motor index (Mine et al., 1997). In the antagonist study, we determined the influence of SB-203,186 by the sum of postprandial motor indexes for 0 to 1 and 1 to 2 h after administration. Results are presented as the means ± S.E.M. Statistical analysis was performed with Dunnett's or Bonferroni's multiple comparison test.

Measurement of Gastric Emptying Rates in Conscious Dogs

Test Meal Preparation. The test meal was prepared according to a previous report (Sato et al., 2000) with minor modifications. Namely, 10 ml/kg caloric semisolid meal (OKUNOS-A; 14.2% carbohydrate, 5.1% protein, 2.7% fat, 1.02 kcal/ml; Horika-foods, Niigata, Japan) was thoroughly mixed with 10 mg/kg acetaminophen.

Experimental Procedures. Four male beagles (Oriental Yeast Co., Ltd.), weighing 9 to 11 kg, were used. The dogs were fasted overnight, and then they were fed the test meal (10 ml/kg), which was ingested within 5 min. Venous blood samples were withdrawn every 15 min up to 90 min. To induce a gastroparesis model, clonidine was injected s.c. 30 min before feeding. CJ-033,466 (0.03 mg/kg), cisapride (1 mg/kg), or vehicle (0.1% methylcellulose) was administered orally just before the clonidine injection. Plasma acetaminophen concentrations were determined by liquid chromatography-tandem mass spectrometry using [D7] 4-acetamindophenol as the internal standard. Acetaminophen is not absorbed through the stomach but quickly through the small intestine; thus, gastric emptying is expressed as the increase in plasma acetaminophen concentrations. First a dose finding study of clonidine was performed (n = 3), and then the effect of pretreatment of CJ-033,466 on the clonidine-induced gastroparesis was evaluated (n = 4).

Data Analysis. Time-plasma acetaminophen concentration profiles were compared by repeated measures analysis of variance followed by Dunnett's multiple comparison test.

Compounds

CJ-033,466, cisapride [(±)-4-amino-5-chloro-N-[1-[3-(4-fluorophenoxy)-propyl]-3-methoxy-4-piperidinyl]-2-methoxybenzamide], mosapride [(±)-4-amino-5-chloro-2-ethoxy-N-[[4-(4-fluorobenzyl)-2-morpholinyl]-methyl]-benzamide], and tegaserod [2-((5-methoxy-1H-indol-3-yl)-methylene)-N-pentylhydrazinecarboximidamide] were synthesized in Pfizer Japan Inc. (Tokyo, Japan). SB-203,186 was purchased from Sigma-Aldrich (St. Louis, MO).

Results

The data generated in the binding and functional assays with CJ-033,466 (Fig. 1) are summarized in Tables 1, 2, 3 together with comparative data for cisapride, mosapride, and tegaserod.

In Vitro Pharmacological Profiling Study. In the receptor binding assay, CJ-033,466 displaced [³H]GR-113,808 binding to membrane fractions of human 5-HT_4d-transfectants, with a K_i of 1.26 nM. Likewise, CJ-033,466 exhibited high affinity for the rat and dog 5-HT₄ receptors, with K_i values of 0.27 and 0.21 nM, respectively. CJ-033,466 showed greater than 1000-fold selectivity for 5-HT_4d over 5-HT_1A, 5-HT_1B, 5-HT_1D, 5-HT_2A, 5-HT_2B, 5-HT₃, 5-HT₇, and D₂ receptors. Furthermore, CJ-033,466 was tested at more than 50 receptors by radioligand binding assay, which was performed by Cerep. CJ-033,466 at a concentration of 1 μM did not inhibit specific bindings at all the other receptors by more than 50% (data not shown). The assays studied include receptors for acetylcholine (M₁, M₂, and M₃, muscarinic nonselective, nicotinic neuronal, muscle-type), adenosine (A₁ and A₂), adrenaline (α₁, α₂, β₁, and β₂), angiotensin-II (AT₁ and AT₂), benzodiazepine, dopamine (D₁, D₃, and D₄), γ-aminobutyric acid (nonselective), glutamate (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid, kainate, and N-methyl-d-aspartate), glycine (strychnine-insensitive), histamine (H₁ and H₂), motilin, neurokinin (NK₁ and NK₂), opiate (μ, nonselective), sigma, glucocorticoid, estradiol, thyroid hormone, and vasopressin, ion channels for calcium (L-type dihydropyridine site, diltiazem site, verapamil site, and N-type), potassium (ATP-dependent, voltage-dependent, and Ca²⁺-dependent), sodium and chloride, transporters (norepinephrine, dopamine, GABA, choline, and 5-HT) and enzymes (phospholipase C, monoamine oxidase, thyrosine hydroxylase, and acetylcholine esterase).

Cisapride demonstrated affinity for the 5-HT₄ receptors, with a K_i value of 117 nM, but it displayed higher affinities for 5-HT_2A and D₂, with K_i values of 3.1 and 17.4 nM, respectively. Mosapride showed moderate affinity for 5-HT_2A, 5-HT_2B, and D₂ receptors, with K_i values of 1890, 330, and 823 nM, respectively, as well as its binding affinity for 5-HT₄ receptors, with a K_i value of 347 nM. Tegaserod exhibited relatively high affinities for the 5-HT_1A, 5-HT_1B, 5-HT_1D, 5-HT_2A, and 5-HT_2B receptors, with K_i values of 18.7, 30, 19, 68, and 9.1 nM, respectively, which were similar to the affinity for 5-HT₄ receptors, with a K_i value of 17.2 nM.

In the human 5-HT_4d transfectants, CJ-033,466 significantly stimulated intracellular cAMP production, with an EC₅₀ value of 0.927 nM and E_max value of 54% relative to 5-HT (Table 3). CJ-033,466 also produced concentration-dependent relaxations of the rat esophageal TMM, with a mean EC₅₀ value of 1.3 nM and E_max value of 58% (Fig. 2). The relaxation elicited by CJ-033,466 was antagonized completely by 1 μM SB-203,186, a selective 5-HT₄ receptor antagonist, whereas SB-203,186 alone did not show any effects on this preparation. Based on these findings, it is possible that CJ-033,466 works as a partial agonist for both human and rat 5-HT₄ receptors.

Cisapride demonstrated weaker agonist activity in the cAMP and TMM assays, with EC₅₀ values of 140 nM (E_max, 106%) and 49 nM (E_max, 57%), respectively. Mosapride also showed a weak agonistic activity, with an EC₅₀ value of 983 nM (E_max 52%), whereas tegaserod demonstrated a potent activity, with an EC₅₀ of 6.7 nM (E_max, 113%) in the cAMP assay.

Effects of CJ-033,466 and Cisapride on Gastric Antral Motility in the Fasted State.Figure 3A shows typical tracings of the effects of CJ-033,466 and cisapride on gastric antral motility when they were administered immediately after termination of phase III period of the IMC. The vehicle did not induce any changes in the incidence of contractile activity. CJ-033,466 (0.01–0.3 mg/kg p.o.) stimulated gastric antral motility by an increased frequency of contractions. The amplitude of CJ-033,466-induced contractions reached the level of natural phase III contractions. Cisapride (3 mg/kg p.o.) also stimulated gastric antral motility in a similar manner. The motor index of gastric antral motility was increased by CJ-033,466 (0.01–0.3 mg/kg p.o.) in a dose-dependent manner, and it reached a statistically significant difference at a 0.03-mg/kg dose (Fig. 3B). The increased motor activity at a 0.3-mg/kg dose lasted 6 to 8 h after administration. Increases in the motor index through the stimulatory effect of cisapride (p.o.) on gastric antral motility reached statistical significance at a 1-mg/kg dose, and but it lasted 6 to 8 h only at a 3-mg/kg dose.

Effects of CJ-033,466 and Cisapride on Gastric Antral Motility in the Postprandial State. In the postprandial state, gastric antral motility is characterized by regular, low-level contractile activity. This pattern of motility became stable 1 h after feeding, and it was maintained for at least 6 h thereafter. CJ-033,466 (0.01–0.3 mg/kg p.o.) increased the amplitude of contractile activity without affecting its frequency, and cisapride (3 mg/kg p.o.) stimulated postprandial gastric antral motility in a similar manner (Fig. 4A). Figure 4B shows the effects of CJ-033,466 (0.01–0.3 mg/kg p.o.) and cisapride (0.3–3 mg/kg p.o.) on the postprandial motor index. CJ-033,466 dose-dependently increased the motor index, and it reached statistical significance at a 0.03-mg/kg dose, whereas cisapride induced a statistically significant response at 1 mg/kg.

Effect of SB-203,186 on the CJ-033,466-Induced Stimulation of Gastric Antral Motility in the Postprandial State. To evaluate the involvement of 5-HT₄ receptors, we investigated the effect of pretreatment with SB-203,186, a selective 5-HT₄ receptor antagonist, on CJ-033,466-induced stimulation of postprandial gastric antral motility. SB-203,186 did not affect the basal antral motility in the postprandial state when given alone. In contrast, when SB-203,186 (1 mg/kg i.v.) was administered just before CJ-033,466 (0.1 mg/kg p.o.), it completely blocked the CJ-033,466-induced stimulation of antral motility and increase in the postprandial motor index (Fig. 5, A and B).

Effects of CJ-033,466 and Cisapride on Gastric Emptying Rate in Clonidine-Induced Gastroparesis in Conscious Dogs. Gastric emptying, expressed as an elevated plasma acetaminophen concentration, proceeded rapidly after the test meal was ingested, and it reached maximal levels at 75 min. Clonidine (5–10 μg/kg s.c.) pretreatment dose-dependently delayed gastric emptying (Fig. 6A). The delay in gastric emptying induced by clonidine at a 10-μg/kg dose was significant. Therefore, we used this clonidine dose as a gastroparesis model. Next, the effects of CJ-033,466 and cisapride on the gastroparesis model were investigated. Oral administration of CJ-033,466 (0.03 mg/kg) significantly restored the delayed gastric emptying induced by clonidine (10 μg/kg s.c.) to normal levels. Cisapride (1 mg/kg p.o.) equally restored delayed gastric emptying (Fig. 6B).

Discussion

It is known that several 5-HT receptor subtypes and D₂ receptor are expressed in the GI tract, and that they are involved in GI motility (Hoyer and Martin, 1997; Taniyama et al., 2000). Gastroprokinetic agents that have a high affinity for the 5-HT and D₂ receptors have been developed to treat bowel disorders. However, because of the low selectivity of these agents, it is not fully understood which receptor mediates the gastroprokinetic efficacy. For example, cisapride, which acts as a 5-HT₄ partial agonist, has been reported to enhance gastric motility by also blocking D₂ receptors (Schuurkes et al., 1985; Yoshida, 1999). Likewise, tegaserod, which was originally identified as a 5-HT₄ agonist, was discovered to be a 5-HT_1A receptor agonist and a potent 5-HT_2B receptor antagonist (Kahrilas et al., 2000; Beattie et al., 2004). Mosapride was reported as a partial 5-HT₄ agonist free of D₂ antagonist properties, and it was designed to avoid D₂-mediated side effects (Yoshida et al., 1989). Our radioligand binding assays revealed that cisapride has high affinity for 5-HT_2A and D₂ receptors comparable with 5-HT₄ receptors, as tegaserod does for 5-HT_1A, 5-HT_1B, 5-HT_1D, 5-HT_2A, and 5-HT_2B receptors. In recent studies, it has been indicated that 5-HT_2A and 5-HT_2B receptors, not only the 5-HT₄ receptor, modulate gastric emptying in a cooperative way (Borman et al., 2002; McCullough et al., 2006; Komada and Yano, 2007). In the light of these findings, the mechanism of action of cisapride or tegaserod in GI tract remains to be fully established. Mosapride showed moderate affinity for 5-HT₄ receptors, with comparable affinity for the 5-HT_2A, 5-HT_2B, and D₂ receptors. This result suggests that mosapride lacks selectivity for the 5-HT₄ receptors, which is inconsistent with a previous report (Yoshida et al., 1989). Considering that 5-HT_1A agonism and D₂ and 5-HT₃ antagonism may also affect GI motility, these agents could not seem to function as prokinetics only by activating 5-HT₄ receptors.

Through our search for more selective and potent 5-HT₄ agonists, CJ-033,466 was found to have high affinity for human 5-HT₄ receptors, with greater than 1000-fold selectivity for 5-HT₄ receptors over other 5-HT receptors and D₂ receptors in vitro. In addition, it has been shown that CJ-033,466 has less activity for human ether-a-go-go-related gene channel than cisapride and mosapride (Toga et al., 2007). This compound will be a useful pharmacological tool to investigate the physiological role of 5-HT₄ receptors with lack of QT prolongation and other risks.

We confirmed that CJ-033,466 has agonistic activity for 5-HT₄ receptors not only in cAMP assay using human 5-HT_4d transfectants but also in TMM assay using rat tissues in consideration of the existence of various variants of 5-HT₄ receptors in GI tract. CJ-033,466 and cisapride showed a partial (54% relative to 5-HT) and full agonism (106%), respectively, in the cAMP assay. In contrast, both compounds showed partial agonistic activities in the rat TMM assay. The reasons why cisapride exhibited full agonism in the cAMP assay are not fully elucidated, but artificial events in the stable transfectants, such as G protein-coupling efficiency and receptor stabilization in a ligand-specific conformation, might be responsible for discrepancies between the two assays (Bach et al., 2001; Banères et al., 2005). Under more physiological or pathophysiological conditions such as human tissue preparations of urinary bladder detrusor muscle, cisapride is reported to act as a partial agonist (Chapple et al., 2004). Our results also showed that both CJ-033,466 and cisapride had partial agonism using rat TMM. Based on these findings, it is possible that CJ-033,466, like cisapride, works as a partial agonist of human 5-HT₄ receptors. Among the other 5-HT₄ agonists tested in the cAMP assay, mosapride showed a very weak agonistic activity, and tegaserod was confirmed to be a potent 5-HT₄ agonist, but it also inhibited 5-HT₁, 5-HT_2A, and 5-HT_2B at similar concentrations, indicating that tegaserod is not selective for 5-HT₄. Therefore, we used CJ-033,466 as a 5-HT₄ partial agonist, which has high selectivity over the other relevant receptors, and we investigated the effects of specific 5-HT₄ activation on GI motility using a dog in vivo model.

In the dog model, we investigated the stimulatory effects of CJ-033,466 and cisapride on gastric antral motility in both the fasted and postprandial states. Before the study, pharmacokinetic parameters of CJ-033,466 in dogs were determined as follows: bioavailability, 24%; volume of distribution, 23 ± 2 l/kg; t_1/2, 2.4 ± 0.2 h; and unbound fraction, 40% (n = 3; Y. Nakai and T. Matsuura, unpublished data). Although CJ-033,466 has a relatively high volume of distribution value, its bioavailability and t_1/2 values are comparable with and unbound fraction is higher than those of cisapride. Based on these parameters, we evaluated the effects of CJ-033,466 in comparison with cisapride by oral administration. CJ-033,466 and cisapride significantly stimulated gastric motility in both of the states at the same dose ranges. The in vivo potency of CJ-033,466 was 30 times greater than cisapride, and it was consistent with the in vitro potency. The manner in which CJ-033,466 stimulated gastric antral motility was different between the fasted and postprandial states. CJ-033,466 increased the contraction frequency in the fasted state, and it stimulated the amplitude of contractile activity in the postprandial state. These manners of action were identical to those of cisapride. These data suggest that although CJ-033,466 and cisapride use the same mechanism to stimulate gastric antral motility, CJ-033,466 is more potent and effective than cisapride.

It has been shown that 5-HT₄ receptors are located in the myenteric plexus and muscle layers of the stomach and colon in guinea pigs and humans (Sakurai-Yamashita et al., 1999). Gastric antral muscles in dogs have excitatory neuronal 5-HT₄ receptors (Prins et al., 2001), and activation of these 5-HT₄ receptors promotes contractile activity of the gastric antrum through enhanced cholinergic transmission in the myenteric plexus (Taniyama et al., 2000). Cisapride and mosapride enhanced gastric antral motility through 5-HT₄ receptor activation in the postprandial state in conscious dogs (Yoshida et al., 1991; Mine et al., 1997; Sato et al., 2000). In the present study, CJ-033,466-stimulated gastric motility in the postprandial state was completely blocked by pretreatment with SB-203,186, a selective 5-HT₄ receptor antagonist. Our data are consistent with previous reports, and they strongly suggest that 5-HT₄ receptors play an essential role in the stimulatory effect of CJ-033,466 on gastric antral motility in conscious dogs. In contrast, SB-203,186 alone did not affect basal gastric antral motility in the postprandial state, suggesting that 5-HT₄ receptors play a regulatory role in GI motility.

Stimulation of contractile activity in the gastric antrum does not always lead to enhanced gastric emptying, because physiologically coordinated gastroduodenal contraction is necessary for gastric emptying. Therefore, we needed to confirm the effect of CJ-033,466 on gastric emptying to demonstrate the potential usefulness as a gastroprokinetic agent. Acetaminophen, a marker for gastric emptying rates, was mixed with a semisolid caloric meal, and then it was fed to dogs. This drug is not absorbed from the stomach, but it is rapidly absorbed from the small intestine, and there is a good correlation between acetaminophen absorption rates and gastric emptying rates of the test meal (Heading et al., 1973). Oral administration of CJ-033,466 and cisapride increased the gastric emptying rates in the dog gastroparesis model with equal efficacy at 0.03- and 1-mg/kg doses, respectively, which were the minimal doses that provided a significant effect in the gastric motility study. These results clearly indicate that CJ-033,466 and cisapride are useful gastroprokinetic agents. It is reported that blocking the D₂ receptor improves gastroduodenal coordination (Schuurkes and Van Nueten, 1984). However, our results suggest that activation of 5-HT₄ receptors alone is sufficient to stimulate well coordinated gastric motility. In addition, it has been reported that human 5-HT₄ receptor is easily desensitized by an exposure to 5-HT (Mialet et al., 2003), and generally, receptor desensitization may limit therapeutic efficiency of drugs that act as an agonist on it. Present data revealed that CJ-033,466 is a partial agonist for 5-HT₄ receptors. Thus, CJ-033,466 is expected to stimulate physiological gastric motility with reduced or absent propensity to elicit tachyphylaxis and desensitization of 5-HT₄ receptors, and it is of interest to prove this concept by a repeated administration study in vivo.

In conclusion, CJ-033,466 is an orally active, partial agonist for 5-HT₄ receptors with greater than 1000-fold selectivity over relative receptors. CJ-033,466 produces a significant gastroprokinetic effect that is 30 times more potent than cisapride. Therefore, CJ-033,466 is a highly selective agonist for 5-HT₄ receptors that does not effect D₂ and other 5-HT receptors, and it is a potential therapy for gastroparesis patients. From a clinical perspective, CJ-033,466 is expected to provide better gastroprokinetics with reduced side effects due to its high potency and selectivity.